Hypofractionated radiotherapy for early (operable) breast ... · Hypofractionated radiotherapy for early (operable) breast cancer: technical document Part 1 - Hypofractionated Radiotherapy

$Page 1: Hypofractionated radiotherapy for early (operable) breast ... · Hypofractionated radiotherapy for early (operable) breast cancer: technical document Part 1 - Hypofractionated Radiotherapy$
Hypofractionated radiotherapy for

early (operable) breast cancer:

technical document

Part 1 - Hypofractionated Radiotherapy Systematic

Review Update 2014

Part 2 - Overview of Left vs Right sided tumour

evidence October 2014

Part 3 - Hypofractionated Radiotherapy Systematic

Review 2011

November 2014

Hypofractionated radiotherapy for the

treatment of early breast cancer

An updated systematic review

August 2014

Hypofractionated radiotherapy for the treatment of early breast cancer: an updated systematic review, was

prepared and produced by:

Cancer Australia

Locked Bag 3 Strawberry Hills NSW 2012 Australia

Tel: +61 2 9357 9400 Fax: +61 2 9357 9477

Website: www.canceraustralia.gov.au

© Cancer Australia (2014)

ISBN Online: 978-1-74127-266-6

Copyright statement:

Internet sites

This work is copyright. You may download, display, print and reproduce the whole or part of this work in unaltered

form for your own personal use or, if you are part of an organisation, for internal use within your organisation, but only

if you or your organisation do not use the reproduction for any commercial purpose and retain this copyright notice

and all disclaimer notices as part of that reproduction. Apart from rights to use as permitted by the Copyright Act

1968 or allowed by this copyright notice, all other rights are reserved and you are not allowed to reproduce the

whole or any part of this work in any way (electronic or otherwise) without first being given the specific written

permission from Cancer Australia to do so. Requests and inquiries concerning reproduction and rights are to be sent

to the Publications and Copyright contact officer, Cancer Australia, Locked Bag 3, Strawberry Hills, NSW 2012

Copies of Hypofractionated radiotherapy for the treatment of early breast cancer: an updated systematic review

can be downloaded from the Cancer Australia website: www.canceraustralia.gov.au or ordered by telephone: 1800

624 973.

Recommended citation

Cancer Australia. Hypofractionated radiotherapy for the treatment of early breast cancer: an

updated systematic review. Cancer Australia, Surry Hills, NSW, 2014.

http://www.canceraustralia.gov.au/

Hypofractionated radiotherapy for the treatment of early breast cancer iii

Contents

Acknowledgments ............................................................................................................................... v

Executive summary ............................................................................................................................ vii

1 Background .............................................................................................................................. 1

1.1 Breast cancer in Australia ...............................................................................................1

1.2 Use of radiotherapy for the treatment of early breast cancer ...............................1

1.3 Existing Cancer Australia clinical practice guidelines ..............................................2

2 Methods .................................................................................................................................... 3

2.1 Inclusion criteria ................................................................................................................3

2.2 Literature search ..............................................................................................................3

2.3 Exclusion criteria ...............................................................................................................4

2.4 Data extraction ................................................................................................................5

2.5 Quality assessment ..........................................................................................................5

3 Results ........................................................................................................................................ 7

3.1 International guidelines ..................................................................................................7

3.2 Systematic reviews ...........................................................................................................8

3.3 Randomised controlled trials .........................................................................................8

3.4 Ongoing trials ................................................................................................................. 16

4 Discussion ............................................................................................................................... 18

5 Conclusion .............................................................................................................................. 21

Appendix A Contributors.................................................................................................................... 22

Appendix B Literature databases searched ................................................................................... 23

Appendix C Search strategy ............................................................................................................. 24

Appendix D Guideline and clinical trial sites searched ................................................................ 25

Appendix E Flowchart of inclusion/exclusion ................................................................................. 26

Abbreviations ...................................................................................................................................... 27

References ........................................................................................................................................... 28

iv Hypofractionated radiotherapy for the treatment of early breast cancer

Tables

Table 1 Study characteristics of included RCTs from new publications .................................... 1

Table 2 Patient characteristics of included RCTs from new publications ................................. 2

Table 3 Study characteristics of conference abstracts ................................................................ 2

Table 4 All-cause mortality reported in START A and START B trials (2013)17 .............................. 3

Table 5 Causes of death in START A and START B trials (2013)17 .................................................. 3

Table 6 Any breast cancer-related events reported in START A and START B trials

(2013)17 ..................................................................................................................................... 4

Table 7 Cumulative incidence-free estimates reported by Spooner et al (2012)19 ................ 5

Table 8 START A and START B relapses17 ........................................................................................... 6

Table 9 Physician-assessed normal tissue effects by fraction schedule in START A

(2013)17 ................................................................................................................................... 11

Table 10 Physician-assessed normal tissue effects by fraction schedule in START B

(2013)17 ................................................................................................................................... 12

Table 11 Incidence of other late adverse effects according to fractionation

schedule in START (2013)17 .................................................................................................. 13

Table 12 Change in photographic breast appearance at 2 years by fractionation

schedule20 ............................................................................................................................. 14

Table 13 Acute skin reactions during treatment by fractionation schedule in UK FAST

trial (2011)20 ........................................................................................................................... 15

Hypofractionated radiotherapy for the treatment of early breast cancer v

Acknowledgments

Funding

The development of this report was funded by the Australian Government through Cancer

Australia.

Contributors

Cancer Australia gratefully acknowledges the support of the many individuals and groups

who contributed to the development of this report. See Appendix A

Hypofractionated radiotherapy for the treatment of early breast cancer: an updated

systematic review was developed with input from an expert multidisciplinary Working Group

with the following members:

Dr Marie-Frances Burke (Chair) Radiation Oncologist

Ms Jan Rice Breast Care Nurse

Dr Patsy Soon Breast Surgeon

Dr Kirsty Stuart Radiation Oncologist

Ms Bronwyn Wells Consumer Representative


Hypofractionated radiotherapy for the treatment of early breast cancer vii

Executive summary

Cancer Australia published clinical practice guidelines on the use of hypofractionated

radiotherapy for early (operable) breast cancer in November 2011. The guidelines were

based on a Cancer Australia systematic review which included available evidence from

randomised controlled trials (RCTs) published between January 2001 and March 2010. The

RCT evidence in the first systematic review came from five trials: RMH/GOC, Canadian trial,

two Standardisation of Breast Radiotherapy Trials (START A and START B) and Spooner. At that

time full publications were available for the first four of these, whilst the Spooner trial was

described in abstract form only.

Since the publication of the 2011 Cancer Australia clinical practice guidelines, further

evidence on hypofractionated radiotherapy for the treatment of early breast cancer has

been published, including 10 year follow-up results from the two START trials. The current

systematic review was undertaken to identify new and updated evidence on the use of

hypofractionated radiotherapy for the treatment of early breast cancer following surgery

and to support the update of Cancer Australia’s 2011 clinical practice guidelines.

An updated literature search of electronic databases was undertaken in November 2013

(search span from January 2010 to November 2013) to identify relevant literature published

since the 2011 systematic review search dates. The primary search was limited to randomised

controlled trials conducted in humans published in the English language.

The literature search identified one new RCT (UK FAST 2011) and updated full publications

from trials included in the first systematic review: the START A, START B trials (Haviland 2013)

and the Spooner trial (Spooner 2012). Three further RCTs were identified that have only been

published as conference abstracts to date. The updated search identified three international

guidelines , but no systematic reviews. One meta-analysis (of START A, START B, and their pilot

trial RMH/GOC) was identified in the search (Haviland 2013).

Standard radiotherapy is typically delivered over a period of five to six weeks using a

standard 2 Gy (Gray) radiation dose per fraction, in 25 to 30 fractions, to a total dose of 50 to

60 Gy.1 Hypofractionated radiotherapy involves fewer, larger-dose radiation treatments that

are usually delivered over a shorter time period compared to standard radiotherapy

regimens.2 The total dose of radiation used in a course of hypofractionated radiotherapy is

reduced to compensate for the increased toxicity of larger daily fractions.

With regard to overall survival, all RCTs have demonstrated that hypofractionated

radiotherapy is equivalent or superior to conventionally fractionated radiotherapy. The START

B trial reported a statistically significantly higher overall survival rate and significantly longer

disease-free survival in the hypofractionated radiotherapy arm compared with the standard

radiotherapy arm. The START A trial reported no significant difference in all-cause 10 year

mortality rates between the hypofractionated radiotherapy schedules and standard

radiotherapy. Spooner et al (2012) reported no significant difference between

hypofractionated radiotherapy and standard radiotherapy schedules for 2, 5, 10 and 15 year

overall survival estimates. The Spooner trial also reported similar 2, 5, 10 and 15 year relapse-

free survival estimates and cumulative incidence-free estimates between the

hypofractionated radiotherapy and standard radiotherapy regimens.

viii Hypofractionated radiotherapy for the treatment of early breast cancer

Local relapse rates were similar between hypofractionated radiotherapy regimens and

standard radiotherapy for START A, START B and the Spooner 2012 trials. START A and START B

also reported no significant difference in 10 year local-regional relapse rates between

hypofractionated radiotherapy schedules and standard radiotherapy.

START B reported significantly fewer distant relapses at 10 years in patients receiving

hypofractionated radiotherapy (40 Gy in 15 fractions over 3 weeks) compared with patients

receiving standard radiotherapy.

Moderate or marked breast shrinkage, telangiectasia, and breast oedema were

experienced significantly less ofen in patients who received hypofractionated radiotherapy

regimens of 39 Gy in 13 fractions over 5 weeks in the START A trial and 40 Gy in 15 fractions

over 3 weeks in the START B trial compared with patients receiving standard radiotherapy.

The UK FAST trial reported increased risk for mild or marked change in 2 year photographic

breast appearance for patients in the hypofractionated radiotherapy group (30 Gy in 5

once weekly fractions of 6.0 Gy) compared with conventionally fractionated radiotherapy,

and compared with the hypofractionated radiotherapy schedule of 28.5 Gy in 5 once

weekly fractions of 5.7 Gy. Results were comparable between the 28.5 Gy schedule and 50

Gy. Three-year rates of physician-assessed moderate/marked adverse effects in the breast

were also significantly higher in the hypofractionated radiotherapy group of 30 Gy in 5 once

weekly fractions of 6.0 Gy compared with the standard radiotherapy group and the

hypofractionated radiotherapy group of 28.5 Gy in 5 once weekly fractions of 5.7 Gy. The

rates were similar between the 28.5 Gy and 50 Gy groups

Hypofractionated radiotherapy for the treatment of early breast cancer 1

1 Background

1.1 Breast cancer in Australia

In 2009, breast cancer was the most common cancer in Australian women (excluding non-

melanoma skin cancer), accounting for 27.4 per cent of all new cancers in women.3 In 2014

it is estimated that 15,800 women will be diagnosed with breast cancer in Australia.4 At the

end of 2008, it was estimated that there were 159,325 Australian women alive who had been

diagnosed with breast cancer in the previous 27 years, including 57,327 women diagnosed in

the previous 5 years.5

1.2 Use of radiotherapy for the treatment of early breast cancer

Early breast cancer is defined as tumours not more than five centimetres in diameter, with

either impalpable or palpable but not fixed lymph nodes and with no evidence of distant

metastases.6 Breast conserving surgery (BCS) plus radiotherapy is a standard alternative to

mastectomy for eligible patients with stage I or II invasive breast cancer.2 A meta-analysis by

the Early Breast Cancer Trialists Group (EBCTG) of 10,801 women from 17 randomised trials of

radiotherapy vs. no radiotherapy following breast conserving surgery reported that following

BCS, radiotherapy to the conserved breast halves recurrence rates and reduces breast

cancer death rate by about a sixth.7

Standard radiotherapy is typically delivered over a period of five to six weeks using a

standard 2 Gy (Gray) radiation dose per fraction (treatment), in 25 to 30 fractions, to a total

dose of 50 to 60 Gy.1 Hypofractionated radiotherapy has been demonstrated to have

comparable outcomes to standard radiotherapy in recent randomised controlled trials

(RCTs).2,8 Hypofractionated radiotherapy involves fewer, larger-dose radiation treatments

that are usually delivered over a shorter time period compared to standard radiotherapy

regimens.2 As each daily fraction is larger than a conventional daily fraction and the

radiation therapy is delivered in a shorter time period; the total delivered dose of radiation is

reduced, to account for the more intense treatment. Based on radiobiological principles,

calculations using the number of fractions (time), total dose, dose per fraction and the

alpha-beta ratio, indicate the hypofractionated biological effective dose is often similar to a

conventional dose.

Sensitivity of tissues to radiation fraction size is described by the α/β ratio. Lower α/β values

indicate greater sensitivity to fraction size than higher α/β values.9 It has been hypothesised

that breast cancer is as sensitive to fraction size as normal breast tissue with a low α/β value,

and confirmation from the RMH/GOC trial of hypofractionated radiotherapy would indicate

that fewer, larger fractions are as effective as conventional 2 Gy fractions.9

The National Breast Cancer Audit reported that in Australia most breast cancer patients

receive radiotherapy following BCS (86%) and 71% of patients who undergo mastectomy

with large tumours (>5cm) receive radiotherapy.10

2 Hypofractionated radiotherapy for the treatment of early breast cancer

1.3 Existing Cancer Australia clinical practice guidelines


radiotherapy for early (operable) breast cancer in November 2011.11 The guidelines were


randomised controlled trials published between January 2001 and March 2010. The 2011

clinical practice guideline included the following:

In women with early breast cancer who require post-operative whole breast radiotherapy

and for whom hypofractionated radiotherapy is being considered, women should be

informed of the potential benefits and risks, and potential side effects and adverse events

of hypofractionated radiotherapy and conventionally fractionated radiotherapy.

Hypofractionated radiotherapy can be offered as a suitable alternative to conventionally

fractionated radiotherapy for women:

aged 50 years and over

with pathological stage T1-2, N0, M0

with low or intermediate histologic grade breast cancer

who have undergone breast conserving surgery

with clear surgical margins

There is insufficient evidence to make a recommendation for or against the use of

hypofractionated radiotherapy for women:

aged less than 50 years

with pathologic stage T3+ and/or N1+ tumour

with high histologic grade breast cancer

who are treated with total mastectomy

who receive chemotherapy and/or targeted biological therapies

Optimal schedule:

Recommended hypofractionated schedules for whole breast radiotherapy, based on

current evidence are:

42.5 Gy in 16 fractions given at the rate of one fraction per day, 5 fractions per

week over 22 days

40 Gy in 15 fractions given at the rate of one fraction per day, 5 fractions per

week over 21 days

Adverse events and toxicity

When selecting an appropriate radiotherapy schedule, consideration should be given to

the possibility of adverse events including early acute reactions and late toxic effects


2 Methods

The current systematic review addresses the following research question:

1. What is the effectiveness of hypofractionated radiotherapy compared to

conventionally fractionated radiotherapy for the treatment of early breast cancer?

2.1 Inclusion criteria

Participants

Women with early (invasive) breast cancer treated with surgery (BCS or mastectomy).

Intervention/comparison

Hypofractionated radiotherapy following surgery compared with either:

a) Standard radiotherapy following surgery

b) Other regimens of radiotherapy following surgery.

Outcome measures

Outcome measures of interest were:

Overall survival

Disease-free survival

Recurrence

Adverse events

Toxicity

Cosmetic outcomes

Quality of Life (QoL).

2.2 Literature search

A systematic literature search was conducted in November 2013 to identify relevant studies

which addressed the inclusion criteria. The search was conducted using several databases

(see Appendix B), including:


Medline

Embase

Pubmed

Cochrane Library.

The search strategy, based on the previous systematic review, used combined key terms

which described early (invasive) breast cancer and hypofractionated radiotherapy (see

Appendix C). The primary search was limited to randomised controlled trials (RCTs)

conducted in humans which were published from January 2010 to November 2013 in the

English language.

After the removal of duplicates and the addition of further citations identified via cascade

searching, a total of 384 unique citations remained. The titles and abstracts of these citations

were assessed independently by two reviewers to determine eligibility for the current review

based on the criteria described previously. Ineligible studies were classified using the

exclusion criteria below. For citations which provided insufficient information to assess

eligibility, the full text was retrieved for assessment, by the same two reviewers.

In addition to the above databases, guidelines and clinical trial websites were searched for

relevant information. Specific international guideline organisations were searched as well as

the National Guidelines Clearinghouse and the Guidelines International Network (GIN)

guideline database. Clinical trial sites searched included clinical trials.gov (USA) and

controlled trials.com (UK). Further information on sites can be found in Appendix D.

The following websites were searched from January 2010 to November 2013 to identify

recently published abstracts about hypofractionated radiotherapy for early breast cancer:

American Society of Clinical Oncology (ASCO)

San Antonio Breast Cancer Symposium (SABCS)

2.3 Exclusion criteria

Papers were excluded if they met any of the following criteria:

Inappropriate study type: studies other than randomised controlled trials

Inappropriate population: studies in a population other than as defined in the

inclusion criteria

Inappropriate interventions: studies not investigating hypofractionated radiotherapy

regimens as defined in the inclusion criteria

Inappropriate outcomes: studies not reporting on one or more the outcomes defined

in the inclusion criteria

Not published in the English language

Published prior to 2010.

Based on these criteria, 299 articles were excluded. The full texts of the remaining 85 citations

were retrieved and assessed to identify which met the inclusion criteria for the review. After

full text assessment, nine citations were identified as eligible for the current review (see

Appendix E).


2.4 Data extraction

Data extraction was performed by two reviewers and verified by both to ensure accuracy.

Descriptive data extracted from these studies included characteristics of the patient

population, study interventions and primary outcomes.

Outcome data extracted from the studies included overall survival, local recurrence,

adverse events, toxicity, cosmetic outcomes and quality of life.

2.5 Quality assessment

Primary studies included in the systematic review were critically appraised according to

criteria previously published by the National Health and Medical Research Council (NHMRC).

The following questions were considered:

Was an appropriate method used for treatment assignment?

Was there control of selection bias after treatment assignment (such as intention to

treat analysis, minimal patients lost to follow-up)

Was the study blinded?

Was there standardised outcome assessment (if blinding was not possible)?

Were groups well matched at baseline?

Was the study powered to detect a difference in primary outcome?

The quality of individual studies was rated as good (i.e., low risk of bias), fair (moderate risk of

bias) or poor (high risk of bias). The methodology used in the current systematic review was

based on the NHMRC checklist for appraising the quality of intervention studies, the box

below outlines the criteria used to assess the included studies.

NHMRC Checklist for appraising the quality of studies of Interventions*

1. Method of treatment assignment

a. Correct, blinded randomisation method described

OR randomised, double-blind method stated

AND group similarity documented

b. Blinding and randomisation stated but method not described

OR suspect technique (e.g. allocation by drawing from an envelope)

c. Randomisation claimed but not described and

investigator not blinded

d. Randomisation not mentioned

2. Control of selection bias after treatment assignment

a. Intention to treat analysis AND full follow-up

b. Intention to treat analysis AND <15%

loss to follow-up

c. Analysis by treatment received only OR no mention of withdrawals

d. Analysis by treatment received

AND no mention of withdrawals

OR more than 15% withdrawals/loss-to-follow-up/post-randomisation


exclusions

3. Blinding

a. Blinding of outcome assessor

AND patient and care giver

b. Blinding of outcome assessor

OR patient and care giver

c. Blinding not done

4. Outcome assessment (if blinding was not possible)

a. All patients had standardised assessment

b. No standardised assessment OR not mentioned

*Source: modified from I Chalmers, Cochrane Handbook; available on the Cochrane Library

CDROM


3 Results

3.1 International guidelines

Four relevant international guidelines were identified in the previous Cancer Australia

systematic review:

American Society for Radiation Oncology (ASTRO) 2010: guidelines on fractionation

for whole breast irradiation

New Zealand Guidelines Group (NZGG) 2009: guidelines for management of early

breast cancer

National Institute of Clinical Excellence (NICE) 2009: guidelines for early and locally

advanced breast cancer

Scottish Intercollegiate Guidelines Network (SIGN) 2005: management of breast

cancer in women

This updated systematic review identified a further three relevant clinical practice

recommendations from the European Society of Medical Oncology (ESMO) (2013),12 German

Society of Radiation Oncology (DEGRO) (2013),13 and Nice-Saint-Paul de Vence guidelines

(2013).14 No updates of the four guidelines identified in the previous systematic were

identified.

The ESMO guidelines for diagnosis, treatment and follow-up of primary breast cancer,

published in November 2013, included the following recommendation regarding

hypofractionated radiotherapy:

Shorter fractionation schemes (e.g. 15-16 fractions with 2.5-2.67 Gy single dose) have

shown similar effectiveness and comparable side-effects [Level of evidence I, Grade

A].12

DEGRO published practical guidelines in 2013 on radiotherapy for breast cancer which

updated guidelines published in 2007. The guideline included the following recommendation

for hypofractionated radiotherapy:

In elderly patients with tumours <5cm and without locoregional lymph node disease,

who do not receive chemotherapy, as an alternative to normofractionated whole

breast radiotherapy, hypofractionated schedules (e.g. 5 x 2.666 Gy/week up to 40

Gy) may be considered (Level of evidence 1a, Grade B).13

The guideline also included the following conclusion from the DEGRO panel:

Normofractionated whole breast irradiation plus sequential boost remains standard.13

Hypofractionated whole breast irradiation with single doses up to 2.7 Gy in 15-16

fractions to total doses of 40-42 Gy is an option for older women with pT1-2 pN0

tumours who need no chemotherapy. The additional use of a sequential boost is

possible.13


Hypofractionated whole breast irradiation plus boost either by simultaneously

integrated boost or by hypofractionated sequential application is discouraged

outside clinical trials.13

The French expert review board of Nice-Saint-Paul de Vence published the 4th edition of the

adjuvant radiotherapy Nice-Saint-Paul de Vence guidelines.14 The guideline included the

following indications regarding hypofractionation and whole breast irradiation:

Eligibility for hypofractionated whole breast irradiation was defined by the most

representative breast cancer patients enrolled in clinical trials (START A, START B,

Canadian trial), which are: patients older than 50 years with invasive ductal

carcinoma, pT1, pN0, histological grade I/II, HR+, HER2-, having complete surgical

excision (experts’ agreements).14

Hypofractionated schemes are as follows: total dose= 42.4 Gy/15 fractions/ 3 weeks;

or total dose= 41.6 Gy/13 fractions/5 weeks; or total dose= 40 Gy/15 fractions/3 weeks

(Level of evidence I, grade A).14

The value of a boost to the lumpectomy cavity has not been established (Level of

evidence 1, grade A) and should not be recommended but rather could be left to

the discretion of the treating physician.14

The indication for a hypofractionated schedule (± boost to the lumpectomy site) in

breast cancer patients who do not belong to the subgroup mentioned above, should

be left to the discretion of the treating physician (experts’ agreements).14 The

treatment planning should keep the dose homogeneity index (less than 10%), as well

as the heart and lung exposure as low as possible.

3.2 Systematic reviews

Two published systematic reviews were identified by the first Cancer Australia review: James

2008 and Kalogeridi 2009. No new systematic reviews were identified in the updated

systematic search for the current review. A meta-analysis of locoregional recurrence rates

from three of the included RCTs (START A, START B and RMH/GOC) was identified (Haviland

2013). The objective of this post hoc sub-group analysis was to explore whether the following

characteristics modify response to hypofractionated radiotherapy: patient age, breast size,

tumour grade, axillary node status, type of surgery, cytotoxic chemotherapy, tumour bed

boost radiotherapy, and lymphatic radiotherapy.

3.3 Randomised controlled trials

In total, six RCTs were identified, all of which compared one or more hypofractionated

radiotherapy regimen to a standard radiotherapy regimen of 50 Gy in 25 fractions over 5

weeks for the treatment of early breast cancer. Of these six RCTs, five were previously

identified in the first systematic review, and one is newly published. In addition to the newly

published trial (UK FAST), updated publications were identified for START A and B, and for

Spooner. An additional three RCTs published as conference abstracts were identified,

although these provide limited information regarding study design and results.


It should be noted that the current review provides a detailed description of the newly

identified evidence only. This relates to four out of the six RCTs in the evidence base.

However, to provide context for the derivation of the current recommendations, the

accompanying guidelines include a consolidated summary of information for all six RCTs.

Study/Author Publication(s) in first Cancer Australia

systematic review (2011)

New publication(s) in current Cancer

Australia systematic review

RMH/GOC

Owen 2006; Yarnold 2005 Haviland 2010

Canadian

Whelan 2010; Whelan 2002 -

START A

Bentzen 2008a; Hopwood 2010 Haviland 2013; Haviland 2010

START B

Bentzen 2008b; Hopwood 2010 Haviland 2013; Haviland 2010

Spooner Spooner 2008 (conference abstract only) Spooner 2012

UK FAST Identified as an ongoing clinical trial FAST 2011

Study characteristics

The four RCTs with new, full publications included patients with early invasive breast cancer

(stage I-II, T1-3a, N0-1, M0). The studies examined a range of hypofractionated radiotherapy

regimens, including:

39 Gy in 13 fractions over 35 days(RMH/GOC trial9,15 and START A16,17)

40 Gy in 15 fractions over 21 days (START B17,18 and Spooner trial19)

41.6 Gy in 13 fractions over 35 days (START A16,17)

30 Gy in 5 fractions over 5 weeks (UK FAST20)

28.5 Gy in 5 fractions over 5 weeks (UK FAST20)

The START A trial tested two hypofractionated radiotherapy regimens. The START B, Spooner

trial and UK FAST trial each tested one hypofractionated radiotherapy regimen. In all trials,

the conventional radiotherapy regimen used as a comparator was 50 Gy in 25 fractions,

delivered over 5 weeks.

All studies compared conventional radiotherapy and hypofractionated radiotherapy

following surgery. Two trials included women who had undergone breast conserving surgery

only (Spooner trial, UK FAST trial). Two trials included women who had undergone breast

conserving surgery or mastectomy (START A and START B).

Median follow up ranged from 37.3 months to 16.9 years.

Study characteristics of the four included RCTs for which there are new publications are

presented in table 1 and patient characteristics are presented in table 2. A more detailed

discussion of the design and patient characteristics of these studies is presented below, by


study. Study characteristics of the RMH/GOC and Canadian trials are available in the 2011

systematic review in table 17. Note: a consolidated summary of study and patient

characteristics from all six RCTs included in the evidence base (i.e., those listed below plus

RMH/GOC and Canadian) is included in the accompanying guidelines document.


Table 1 Study characteristics of included RCTs from new publications

Author,

Year,

Location

Patients (N) Median

follow-up

Intervention Comparator Tumour bed

boost

Outcomes

START A ,

2013

UK17

Moderate

risk of bias

2236

9.3yrs

41.6 Gy in 13

Fractions over 5

weeks (n=750) OR

39Gy in 13

fractions over 5

weeks (n=737)

50 Gy in 25 fractions

over 5 weeks (n=749)

Both START trials

permitted

prescription of a

sequential tumour

bed boost dose of

10 Gy in five fractions

BCS: 1512 (61%)

Primary: local-regional

relapse and late

normal tissue effects.

Secondary: local

relapse, distant

relapse, DFS, OS

START B

2013

UK

Moderate

risk of bias

2215 9.9yrs

40 Gy in 15

fractions over 3

weeks (n=1110)



Both START trials

permitted

prescription of a

sequential tumour

bed boost dose of

10 Gy in five fractions

BCS: 868 (43%)

Spooner,

2012

UK19

Moderate

risk of bias

707

n=358*

(radiotherapy)

n=349 (no

radiotherapy)

16.9yrs

40 Gy in 15

fractions over 3

weeks (n=181)

Supplementary boost

of direct 10-14 MeV

electron field of 15

Gy in five daily

fractions



Supplementary boost

of direct 10-14 MeV

electron field of 15 Gy

in five daily fractions

All irradiated patients:

358 (100%)

Primary: locoregional

relapse rate 5 years.

Secondary: survival

and locoregional

tumour control

n=707 randomised to:

radiotherapy (n=358) or no radiotherapy


(n=349)

UK FAST trial,

2011

UK20

Moderate

risk of bias

915 3.1yrs

30 Gy in 5 once

weekly fractions of 6

Gy over 5 weeks

(n=308) or

28.5 Gy in 5 once

weekly fractions of

5.7 Gy

over 5 weeks

(n=305)

50 Gy in 25 fractions of

2 Gy over 5 weeks

(n=302)

No boost Primary: change in

photographic breast

appearance.

Secondary: radiation-

induced changes in

the breast, and local

tumour control

*Group of interest. Abbreviations: DFS=Disease Free Survival; Gy=Gray; OS=overall survival.

Table 2 Patient characteristics of included RCTs from new publications

Study/Author Patient inclusion criteria Demographics Prior Surgery Prior treatment

START A17

Women with operable invasive breast cancer (pT1-

3a, pN0-1, M0) requiring radiotherapy after surgery

(BCS or mastectomy, with clear tumour margins ≥1

mm)Age > 18 yrs.

Median age: 57 (25-85)yrs.

Tumour size <2cm: 1138 (51%)

Positive lymph nodes: 643 (29%)

Tumour grade 1 or 2: 1572 (70%)

BCS: 1900 (85%) Adjuvant chemo: 793

(35%)

Tamoxifen: 1758 (79%)

Lymphatic radiotherapy:

318 (14%)

START B17

Women with operable invasive breast cancer (pT1-

3a, pN0-1, M0) requiring radiotherapy after surgery

(BCS or mastectomy, with clear tumour margins ≥1

mm)Age > 18 yrs.


Tumour size <2cm: 1412 (64%)

Positive lymph nodes: 504 (23%)

Tumour grade 1 or 2: 1667 (75%)

BCS: 2038 (92%) Adjuvant chemo: 491

(22%)

Tamoxifen: 1928 (87%)

Lymphatic radiotherapy:

161 (7.3%)

Spooner,

201219

Patients with early clinical stage I and II breast

cancer

Women were eligible if they had histologically

proven adenocarcinoma of the breast that had

been completely surgically removed resulting in a

cosmetically satisfactory breast; clinical tumour

measurement had to be less than 5 cm with no

clinically palpable axillary nodes and no evidence

of systemic disease. Age not included in inclusion

criteria


Menopausal status: pre- 87 (24%),

post- 237 (66%), peri- 34 (10%)

Infiltrating ductal carcinoma: 75%

Tumour grade 1 or 2: 39 (17%) or

134 (57%)

Tumour size: 2.0 (0.12-8.0)cm

BCS: 358 (100%) Tamoxifen: 358 (100%)


UK FAST trial20

Women with early stage breast cancer (invasive

carcinoma) and favourable prognostic features

including age > 50 yrs., BCS, pathological tumour

size <3.0 cm, complete microscopic resection of

tumour and negative axillary node status

established by appropriate surgical staging.

Median age: 63.2 (50-88)yrs.

Tumour size: 1.3 (0.05-3.0)cm

Tumour grade 1 or 2: 50Gy- 94

(31%) or 176 (58%), 30Gy- 113

(37%) or 159 (52%), 28.5- 102 (33%)

or 168 (55%)

BCS with complete

microscopic

resection of tumour

Tamoxifen: 50Gy- 227

(75%), 30Gy- 243 (79%),

28.5Gy- 224 (73%)

No chemotherapy

(exclusion criteria)

Abbreviations: BCS=breast conserving surgery



UK Standardisation of Breast Radiotherapy (START) trials

The START trials began in 1998, on the basis of results of the UK START pilot study (the

RMH/GOC trial) that assessed two lower dose regimens (39 Gy and 42.9 Gy) of a 13 fraction

regimen delivered over 5 weeks compared with 50 Gy in 25 fractions over 5 weeks.17 The

START trials were two randomised, unmasked trials of women recruited between 1999 and

2002, from UK radiotherapy centres; START A and START B.17

A total of 2236 early breast cancer patients were included in START A. Patients were

randomly assigned to either 50 Gy in 25 fractions over 5 weeks (control group) (n=749) or 41.6

Gy in 13 fractions over 5 weeks (n=750) or 39 Gy in 13 fractions over 5 weeks (n=737) after

complete excision (BCS or mastectomy).17 START A was designed as a superiority trial, with a

target sample size of 2,000 providing 80% power to detect a difference of 5% between the

control group and each hypofractionated regimen (two-sided alpha=0.05).

START B randomised 2215 early breast cancer patients to either 50 Gy in 25 fractions over 5

weeks (control group) (n=1110) or 40 Gy in 15 fractions over 3 weeks (n=1105) after complete

excision (BCS or mastectomy).17 START B was designed as a non-inferiority trial, with a target

sample size of 1,840 providing 95% power to exclude an increase of 5% in the primary

outcome for the control group versus the hypofractionated regimen (one-side alpha=0.025).

Both START trials permitted a sequential tumour bed boost dose of 10 Gy in 5 fractions. This

was required to be planned before randomisation.17

Haviland et al (2013) reported 10 year follow-up results of both START trials for local-regional

relapse, late normal tissue effects, local relapse, distant relapse, disease-free survival and

overall survival.17 The authors also reported post-hoc subgroup meta-analysis, which

compared the combined hypofractionated radiotherapy regimens of START A and B and the

pilot trial versus the control groups for local-regional relapse and the incidence of any

moderate or marked physician-assessed normal tissue effects in the breast (shrinkage,

induration, oedema, or telangiectasia). This updated search also identified a 2010 meta-

analysis of the START trials by Haviland et al (2010) (published as a letter to the editor),

however this was for 5-year follow-up and therefore the Haviland 2013 meta-analysis of 10-

year follow-up is reported in this systematic review.

The START trials were previously included in the 2011 Cancer Australia systematic review.

Spooner 2012

Spooner et al (2012) reported results of a UK based RCT which compared immediate

radiotherapy or delayed salvage treatment (no radiotherapy) following BCS (mastectomy

was excluded).19 Patients receiving radiotherapy were further randomised to either standard

radiotherapy (50 Gy in 25 fractions over 5 weeks) or hypofractionated radiotherapy (40 Gy in

15 fractions over 3 weeks). All irradiated patients received a supplementary tumour bed

boost of 15 Gy in five daily fractions. The outcomes for these two patient groups who

received radiotherapy will be reported in this systematic review. A total of 358 patients were

randomised to immediate radiotherapy; 177 to standard radiotherapy regimen and 181 to

short course radiotherapy regimen. Median follow-up was 16.9 years. The primary outcome

of the study was 5 year locoregional relapse rate, with survival and locoregional control

reported as secondary outcomes.19


UK FAST trial 2011

The FAST trialists group reported results of a randomised trial comparing standard

radiotherapy to two hypofractionated radiotherapy schedules following BCS (mastectomy

patients were excluded).20 Nine-hundred and fifteen women with node negative early breast

cancer were randomised to receive either standard radiotherapy regimen of 50 Gy in 25

fractions over 5 weeks (n=302) or a hypofractionated radiotherapy regimen of either 30 Gy in

5 once weekly fractions of 6.0 Gy over 5 weeks (n=308) or 28.5 Gy in 5 once weekly fractions

of 5.7 Gy over 5 weeks (n=305). Median follow-up was 37.3 months. The first published results

of the trial reported change in photographic breast appearance, radiation induced breast

changes, and local tumour control.20

Conference abstracts

An additional three RCTs published as conference abstracts were also identified however

these included limited results. The characteristics of these studies are presented in table 3

Table 3 Study characteristics of conference abstracts

Author

Location

Patients

(n)

Median

follow-

up

Intervention Comparator Outcomes

Barsoum,

2010, Egypt

308 Not

reported

40 Gy in 15

fractions over 3

weeks

50 Gy in 25

fractions over 5

weeks

Locoregional DFS,

distant-free survival,

overall survival

Patni, 2012,

India

40 7mths 40Gy in 15

fractions over 3

weeks

50Gy in 25

fractions over 5

weeks

Locoregional control,

disease-free survival

Fragandrea,

2012,

Greece

61 Not

reported

43.2 Gy in 16

fractions over

22 days with

boost 10 Gy in 5

fractions over 1

week

50 Gy in 25

fractions over 5

weeks with boost


over 1 week

NR

Quality

The quality of each of the four included trials with full publications was considered to be fair

(moderate risk of bias). All trials were randomised, with the methods of randomisation

described, usually of a high quality. The majority of trials reported survival outcomes by

intention-to-treat analysis and limited numbers of patients were lost to follow-up. Trials were

not blinded. All trials had standardised assessment of outcomes and had well matched

population characteristics between treatment arms at baseline. Most of the phase III trials

were powered to detect a significant difference in primary outcomes.


Outcomes

3.3.1 Overall survival

Three studies reported on overall survival; START A, START B and Spooner 2012. OS was not a

primary outcome in any of the RCTs included in the evidence base.

START A and START B trials both reported all-cause mortality rates at 5 and 10 years as well as

causes of death, including breast cancer and other causes.17

START A reported a total 392 deaths in the study population with no significant difference in

all-cause 10 year mortality rates between the hypofractionated radiotherapy schedules and

standard radiotherapy.17 See table 4. At the time of analysis, 69.6% (273 deaths) of deaths in

START A were from breast cancer, 6.6% (26 deaths) were related to cardiac disease only,

8.7% (34 deaths) were from other cancers, 11.2% (44 deaths) were from other non-cancer

causes and 3.8% (15 deaths) were from unknown cause. Table 5 details causes of death for

each of the radiotherapy schedules.17

In the START B trial there were a total of 351 deaths.17 The 10 year all-cause mortality rate was

significantly lower in the hypofractionated radiotherapy arm than the standard radiotherapy

arm; HR=0.80 (95% CI 0.65-0.99), p= 0.042. See table 4. Of the 351 deaths in START B, 67.2%

(236 deaths) were from breast cancer, 4.8% (17 deaths) were related to cardiac disease

only, 13.7% (48 deaths) were from other causes, 11.4% (40 deaths) were from other non-

cancer causes, and 2.8% (10 deaths) were from unknown cause. Table 5 details causes of

death for each of the radiotherapy schedules.17

Table 4 All-cause mortality reported in START A and START B trials (2013)17

Events

(n/patients; %)

Estimated

proportion of

patients with

event by 5 years

(%; 95% CI)

Estimated

proportion of

patients with

event by 10

years (%; 95% CI)

Crude hazard

ratio (95% CI)

P value

START A

50 Gy 130/749 (17.4%) 10.5% (8.5-13.0) 19.8% (16.8-23.2) 1.00

41.6 Gy 128/750 (17.1%) 10.7% (8.7-13.2) 18.4% (15.7-21.6) 0.96 (0.75-1.17

)

0.74

39 Gy 134/737 (18.2%) 9.9% (8.0-12.4) 20.3% (17.3-23.7) 1.05 (0.82-1.34) 0.69

START B

50 Gy 192/1105 (17.4%) 10.9% (9.1-12.9) 19.2% (16.8-21.9) 1.00

40 Gy 159/1110 (14.3%) 7.9% (6.4-9.6) 15.9% (13.7-18.4) 0.80 (0.65-0.99) 0.042

Table 5 Causes of death in START A and START B trials (2013)17

Breast cancer Cardiac disease

only

Other cancers Other non-

cancer causes

Unknown

cause

START A

50 Gy 92/130 (71%) 7/130 (5%) 9/130 (7%) 16/130 (12%) 6/130 (5%)


41.6 Gy 86/128 (67%) 13/128 (10%) 10/128 (8%) 16/128 (13%) 3/128 (2%)

39 Gy 95/134 (71%) 6/134 (4%) 15/134 (11%) 12/134 (9%) 6/134 (4%)

START B

50 Gy 130/192 (68%) 12/192 (6%) 25/192 (13%) 21/192 (11%) 4/192 (2%)

40 Gy 106/159 (67%) 5/159 (3%) 23/159 (14%) 19/159 (12%) 6/159 (4%)

The study by Spooner et al (2012) reported 2, 5, 10, and 15 year estimates for overall

survival.19 The 2, 5, 10 and 15 year overall survival estimates for short course radiotherapy

were 94%, 85%, 70%, 53% respectively. For long course radiotherapy the 2, 5, 10 and 15 year

overall survival estimates were 92%, 81%, 67%, 52% respectively. There was no significant

difference between long- and short-course radiotherapy schedules for 2, 5, 10 and 15 year

overall survival estimates (HR 1.02, 95% CI 0.76-1.35).19

One additional RCT, published as a conference abstract only, also reported overall survival.

Barsoum et al (2013) reported no significant difference between hypofractionated

radiotherapy and standard radiotherapy (84.8% vs. 79.2% respectively, p=0.408).21

3.3.2 Disease-free survival

Haviland et al (2013) reported disease-free survival (DFS) for both START A and START B trials.

For START A there was no significant difference in DFS between the hypofractionated

radiotherapy schedules and standard radiotherapy (41.6 Gy vs. 50 Gy HR=0.94, 95% CI 0.75-

1.17, p=0.57; 39 Gy vs. 50 Gy HR=1.08, 95% CI 0.87-1.35, p=0.48).17 Whereas START B reported

significantly higher rates of DFS in patients receiving hypofractionated radiotherapy

compared to standard radiotherapy (40 Gy vs. 50 Gy HR=0.79, 95% CI 0.65-0.97, p=0.022).17

See table 6.

Table 6 Any breast cancer-related events reported in START A and START B trials (2013)17

Events

(n/patients; %)

Estimated

proportion of

patients with

event by 5 years

(%; 95% CI)

Estimated

proportion of

patients with

event by 10

years (%; 95% CI)

Crude hazard

ratio (95% CI)

P value

START A

50 Gy 154/749 (20.6%) 14.0% (11.6-16.7) 22.6% (19.5-26.1) 1.00

41.6 Gy 149/750 (19.9%) 11.7% (9.5-14.2) 22.7% (19.5-26.3) 0.94 (0.75-1.17) 0.57

39 Gy 163/737 (22.1%) 15.5% (13.0-18.3) 24.3% (21.1-28.0) 1.08 (0.87-1.35) 0.48

START B

50 Gy 222/1105 (20.1%) 14.3% (12.3-16.5) 22.2% (19.7-25.0) 1.00

40 Gy 182/1110 (16.4%) 10.4% (8.7-12.4) 18.3% (16.0-20.9) 0.79 (0.65-0.97) 0.022

3.3.3 Relapse-free survival

Spooner et al (2012) reported relapse-free survival estimates at 2, 5, 10 and 15 years. For

short-course radiotherapy the 2, 5, 10 and 15 year relapse-free estimates were 89%, 81%, 61%


and 46% respectively.19 The 2, 5, 10 and 15 year relapse-free estimates for long-course

radiotherapy were 86%, 73%, 59% and 44% respectively. There was no significant difference

between short- and long-course radiotherapy (HR 0.98, 95% CI 0.75-1.29).19

3.3.4 Event-free rates

Spooner et al (2012) reported 2, 5, 10 and 15 year cumulative incidence-free estimates

including: locoregional relapse event-free rates, distant relapse event-free rates, death

event-free rates, and overall competing event-free rates.19 No significant differences were

reported by short- or long-course radiotherapy treatment group. Table 7 reports the

cumulative incidence-free estimates.19

Table 7 Cumulative incidence-free estimates reported by Spooner et al (2012)19

Factor n (events) Cumulative incidence-free estimates (%)*

2 year 5 year 10 year 15 year

Locoregional relapse event-free rates

Short course 181 (25) 97 94 87 86

Long course 177 (21) 97 91 88 87

Distant relapse event-free rates

Short course 181 (29) 96 92 87 83

Long course 177 (24) 95 89 86 86

Death event-free rates

Short course 181 (98) 95 86 72 54

Long course 177 (92) 93 83 67 50

Overall competing event-free rates

Short course 181 (110)^ 88 75 54 38

Long course 177 (105)^ 86 67 51 38

* based on 1-cumulative incidence function for competing risks

^ events are the number of women with at least one competing event. Total numbers of competing

events: short course 152, long course 137.

3.3.5 Recurrence/relapse

Haviland et al (2013) reported that in the START A trial at a median follow-up 9.3 years 76% of

patients were alive and without relapse, 2.5% were alive with local-regional relapse (without

distant relapse), 3.5% were alive with distant relapse (including patients with local-regional

relapse), 17.5% had died (including 66 with local-regional relapse), and 0.4% had had no

follow-up.17 In START B, at a median follow-up of 9.9 years, 78.2% of patients were alive and

without relapse, 2.3% were alive with local-regional relapse (without distant relapse), 2.8%

were alive with distant relapse (including 10 patients with local-regional relapse), 15.8% had

died (including 35 patients with local-regional relapse) and 0.9% had no follow-up.17

Spooner et al (2012) reported that 15% of short-course radiotherapy patients and 21% of

long-course radiotherapy patients had relapse within 5 years.19 Overall relapse, at analysis

(median follow-up 16.9 years) occurred in 32% of hypofractionated radiotherapy patients

and in 29% of long-course radiotherapy patients.19


Local relapse

Four of the RCTs reported on local relapse; START A, START B, Spooner and UK FAST trial.

Both START A and START B trials reported 10 year local relapse rates, with both trials reporting

no significant difference between hypofractionated radiotherapy and conventional

radiotherapy regimens.17 See table 8.

Spooner et al (2012) reported on 5 year local relapse and overall local relapse.19 Overall

local relapse rates were similar across long- and short-course radiotherapy regimens. In the

short course radiotherapy arm 6.6% of patients relapsed within 5 years compared with 9.6%

of long-course radiotherapy patients (p value was not reported). Overall local relapse

occurred in 13.8% of short course radiotherapy patients compared with 11.9% of long course

radiotherapy patients.19

The UK FAST trial reported two local tumour relapses at median follow-up of 37.3 months; both

relapses were in the standard radiotherapy arm.20

Table 8 START A and START B relapses17

Events

(n/patients; %)

Estimated

proportion of

patients with

event by 5

years (%; 95%

CI)

Estimated

proportion of

patients with

event by 10

years (%; 95% CI)

Crude hazard

ratio (95% CI)

P value

START A

Local relapse

50 Gy 40/749 (5.3%) 3.4% (2.3-5.1) 6.7% (4.9-9.2) 1.00

41.6 Gy 37/750 (4.9%) 3.1% (2.0-4.7) 5.6% (4.1-7.8) 0.90 (0.57-1.40) 0.63

39 Gy 47/737 (6.4%) 4.4% (3.1-6.2) 8.1% (6.1-10.7) 1.20 (0.79-1.83) 0.39

Local-regional relapse

50 Gy 45/749 (6.0%) 4.0% (2.8-5.7) 7.4% (5.5-10.0) 1.00

41.6 Gy 42/750 (5.6%) 3.8% (2.6-5.5) 6.3% (4.7-8.5) 0.91 (0.59-1.38) 0.65

39 Gy 52/737 (7.1%) 5.1% (3.7-7.1) 8.8% (6.7-11.4) 1.18 (0.79-1.76) 0.41

Distant relapse

50 Gy 100/749 (13.3%) 9.8% (7.9-12.3) 14.7% (12.2-17.7) 1.00

41.6 Gy 110/750 (14.7%) 9.5% (7.6-11.9) 16.8% (14.0-20.0) 1.08 (0.82-1.41) 0.58

39 Gy 121/737 (16.4%) 11.8% (9.7-14.4) 18.0% (15.1-21.2) 1.24 (0.95-1.61) 0.11

START B

Local relapse

50 Gy 50/1105 (4.5%) 3.3% (2.4-4.6) 5.2% (3.9-6.9) 1.00

40 Gy 36/1110 (3.2%) 1.9% (1.2-3.0) 3.8% (2.7-5.2) 0.70 (0.46-1.07) 0.10


50 Gy 53/1105 (4.8%) 3.5% (2.5-4.8) 5.5% (4.2-7.2) 1.00

40 Gy 42/1110 (3.8%) 2.3% (1.5-3.4) 4.3% (3.2-5.9) 0.77 (0.51-1.16) 0.21

Distant relapse

50 Gy 158/1105 (14.3%) 10.5% (8.8-12.5) 16.0% (13.8-18.5) 1.00

40 Gy 121/1110 (10.9%) 7.5% (6.0-9.2) 12.3% (10.3-14.6) 0.74 (0.59-0.94) 0.014



Three RCTs reported on local-regional relapse; START A, START B, and UK FAST trial.

Haviland et al (2013) reported no significant difference in 10 year local-regional relapse rates

between hypofractionated radiotherapy regimens and standard radiotherapy regimen in

both START A and START B trials.17 See table 8. The authors reported that at the time of

analysis, 6.2% of patients in START A had local-regional tumour relapse compared with 4.3% of

patients in START B. It was noted that the lower proportion of local-regional tumour relapses in

START B was probably a result of the slightly better prognosis of patients recruited into START B

compared with START A.17

In START A Haviland et al (2013) reported that the estimated absolute differences in the

proportion of patients with local-regional relapses at 10 years compared with 50 Gy were

-0.6% (95% CI -3.0 to 2.7) for 41.6 Gy and 1.3% (95% CI -1.5 to 5.2) for 39 Gy.17 The upper limits

of the one-sided 95% CI for the absolute difference in 10 year local-regional relapse rates

indicated an estimated maximum 2.0% excess risk with 41.6 Gy and 4.5% with 39 Gy

compared with 50 Gy.17 See figure 1.

The estimated α/β value for local-regional relapse in START A was 4 Gy (95% CI 0.0-8.9),

adjusting for age, tumour size, type of primary surgery, use of adjuvant chemotherapy, use of

tamoxifen, lymphatic radiotherapy, and tumour bed boost radiotherapy. 17 Meta-analysis of

START A and the START pilot trial (349 events, 3646 women), provided an adjusted α/β value

for local regional relapse of 3.5 Gy (95% CI 1.2-5.7). 17

In the START B trial the estimated absolute difference in the proportion of patients with 10

year local-regional relapse for 40 Gy compared with 50 Gy was -1.2% (95% CI -2.6 to 1.0%).17

The upper limit of the one-sided 95% CI for the absolute difference in 10 year local-regional

relapse rates suggested an estimated 0.4% excess risk associated with the 15 fraction

schedule. See figure 1.


Figure 1: Cumulative risk of local-regional tumour relapse In START-A (A) and START-B (B).

Haviland et al (2013) reported that in a post-hoc subgroup meta-analysis, which compared

the combined hypofractionated radiotherapy regimens of START A and B and the pilot trial

(n=5861) versus the control groups for local-regional relapse, the treatment effect was not

http://www.sciencedirect.com/science/article/pii/S1470204513703863#gr1


significantly different irrespective of age, type of primary surgery, axillary node status, tumour

grade, adjuvant chemotherapy use, or use of tumour bed boost radiotherapy, see figure 2.17

Figure 2 Meta-analysis of local-regional relapse comparing hypofractionated regimens versus 50 Gy in

25 fractions Includes 5861 patients from the START pilot trial, START-A, and START-B.

The FAST trialists group reported regional relapse at median follow-up of 37.3 months.20 A total

of three regional relapses were reported in the UK FAST trial; two in the 28.5 Gy schedule and

one in the standard radiotherapy arm.20

Distant relapse

Four of the identified RCTs reported distant relapses; START A, START B, Spooner 2012 and UK

FAST trial.



START A trial reported no significant difference in 10 year distant relapse rates between the

hypofractionated radiotherapy schedules and standard radiotherapy.17 START B reported

significantly less distant relapses at 10 years in the hypofractionated radiotherapy group (40

Gy in 15 fractions over 3 weeks) compared with the standard radiotherapy group; 12.3% vs.

16% respectively (HR 0.74, 95% CI 0.59-0.94, p=0.014).17 See table 8.

The RCT by Spooner et al (2012) reported overall distant relapses and distant relapses within 5

years.19 Overall, distant relapses occurred in 16% of short-course radiotherapy patients and

13.6% of long-course radiotherapy patients. Within 5 years 8.8% of short course radiotherapy

patients and 11.3% of long-course radiotherapy patients experienced distant relapse. P

values were not reported.19

The UK FAST trial reported at total of 17 distant relapses at median follow-up; 10 in the 28.5 Gy

arm, two in the 30 Gy arm and five in the standard radiotherapy arm.20

3.3.6 Adverse events

Three of the identified RCTs reported on adverse events; START A, START B, and UK FAST trial

(2011). Additional RCTs reported as conference abstracts only also reported on adverse

events.

Late normal tissue effects

Both START A and START B reported late normal tissue effects as a primary endpoint. Normal

tissue effects in the breast, arm, and shoulder were assessed by physician, photographic

comparison with baseline, and patient self-reports.17 Table 9 and table 10 present physician-

assessed normal tissue effects for START A and START B respectively, including breast

shrinkage, breast induration, telangiectasia, breast oedema, shoulder stiffness, arm oedema,

and other.17

In START A the most common normal tissue effects at 10 years were breast shrinkage and

induration. In comparison to standard radiotherapy, patients in the 39 Gy regimen were

significantly less likely to have moderate or marked breast induration, telangiectasia, and

breast oedema, see table 9.17 Moderate or marked normal tissue effects did not differ

significantly between 41.6 Gy and 50 Gy groups.

Similar to START A, breast shrinkage and induration were the most common late normal tissue

effects at 10 years in START B.17 In comparison to standard radiotherapy patients, those

receiving hypofractionated radiotherapy (40 Gy) were significantly less likely to experience

moderate or marked breast shrinkage, telangiectasia, and breast oedema, see table 10.17

Haviland et al (2013) reported that in a post-hoc analysis, the incidence of any moderate or

marked physician-assessed normal tissue effects in the breast (shrinkage, induration,

oedema, or telangiectasia) for the 4,660 women with data available from START A, START B,

and the pilot study showed that the treatment effect was similar irrespective of age, breast

size, use of tumour bed boost radiotherapy, adjuvant chemotherapy, or tamoxifen, see figure

3.17


Figure 3: Meta-analysis of any moderate or marked physician-assessed normal tissue effects in the

breast comparing hypofractionated regimens versus 50 Gy in 25 fractions. Includes 4672 patients from

START pilot trial, START-A, and START-B. *Assessed from baseline photographs.

Table 9 Physician-assessed normal tissue effects by fraction schedule in START A (2013)17

Moderate or

marked events;

n/patients (%)

Estimated

proportion of

patients with

event by 5 years

(%; 95% CI)

Estimated

proportion of

patients with

event by 10

years (%; 95% CI)

Crude hazard

ratio (95% CI)

P value

Breast shrinkage*

50 Gy 165/616 (26.8%) 14.1% (11.5-17.2) 34.2% (29.8-39.2) 1.00

41.6 Gy 168/627 (26.8%) 17.8% (14.9-21.1) 31.4% (27.2-36.0) 0.98 (0.79-1.21) 0.83

39 Gy 140/617 (22.7%) 14.7% (12.0-18.0) 30.0% (25.7-34.8) 0.86 (0.69-1.08) 0.19

Breast induration (tumour bed)*

50 Gy 142/616 (23.0%) 18.5% (15.6-21.9) 27.1% (23.3-31.3) 1.00

41.6 Gy 150/627 (23.9%) 18.9% (16.0-22.3) 28.2% (24.2-32.7) 1.01 (0.80-1.27) 0.95

39 Gy 110/617 (17.8%) 15.0% (12.3-18.3) 21.6% (18.1-25.7) 0.76 (0.59-0.98) 0.034



Telangiectasia

50 Gy 42/730 (5.7%) 4.2% (3.0-6.1) 7.2% (5.2-9.8) 1.00

41.6 Gy 43/733 (5.9%) 4.9% (3.5-6.8) 7.1% (5.2-9.5) 1.00 (0.65-1.53) 0.99

39 Gy 18/723 (2.5%) 1.3% (0.6-2.5) 3.0% (1.8-5.0) 0.43 (0.25-0.75) 0.003

Breast oedema*

50 Gy 78/616 (12.7%) 12.1% (9.7-15.0) 13.5% (10.9-16.6) 1.00

41.6 Gy 67/627 (10.7%) 9.2% (7.1-11.7) 11.8% (9.2%-14.8) 0.82 (0.59-1.14) 0.24

39 Gy 43/617 (7.0%) 7.3% (5.5-9.7) 7.3% (5.5-9.7) 0.54 (0.37-0.78) 0.001

Shoulder stiffness^

50 Gy 14/117 (12.0%) 8.8% (4.7-16.4) 17.5% (10.2-29.1) 1.00

41.6 Gy 10/95 (10.5%) 7.1% (3.3-15.2) 14.8% (8.0-26.6) 0.85 (0.38-1.90) 0.69

39 Gy 8/92 (8.7%) 7.5% (3.4-16.0) 11.0% (5.6-21.0) 0.74 (0.31-1.76) 0.49

Arm oedema^

50 Gy 15/117 (12.8%) 12.8% (7.6-21.2) 16.3% (9.9-26.2) 1.00

41.6 Gy 16/95 (16.8%) 11.9% (6.6-21.0) 22.5% (14.1-34.7) 1.31 (0.65-2.66) 0.45

39 Gy 6/92 (6.5%) 6.4% (2.7-14.7) 8.2% (3.7-17.6) 0.50 (0.20-1.30) 0.16

Other

50 Gy 18/729 (2.5%) 1.3% (0.7-2.6) 3.4% (2.1-5.4) 1.00

41.6 Gy 20/733 (2.7%) 2.0% (1.2-3.4) 3.7% (2.3-6.1) 1.09 (0.58-2.06) 0.79

39 Gy 24/724 (3.3%) 2.3% (1.4-3.8) 3.9% (2.6-5.9) 1.37 (0.74-2.52 0.31

*Only assessed in women who had BCS. ^ Restricted to women who received lymphatic radiotherapy

(to axilla or supraclavicular).

Table 10 Physician-assessed normal tissue effects by fraction schedule in START B (2013)17

Moderate or

marked events;

n/patients (%)

Estimated

proportion of

patients with

event by 5 years

(%; 95% CI)

Estimated

proportion of

patients with

event by 10

years (%; 95% CI)

Crude hazard

ratio (95% CI)

P value

Breast shrinkage*

50 Gy 256/1003 (25.5%) 15.8% (13.6-18.3) 31.2% (27.9-34.9) 1.00

40 Gy 221/1006 (22.0%) 11.4% (9.5-13.6) 26.2% (23.1-29.6) 0.80 (0.67-0.96) 0.015

Breast induration (tumour bed)*

50 Gy 153/1003 (15.3%) 12.1% (10.2-14.4) 17.4% (14.9-20.3) 1.00

40 Gy 129/1006 (12.8%) 9.6% (7.9-11.6) 14.3% (12.1-16.9) 0.81 (0.64-1.03) 0.084

Telangiectasia

50 Gy 52/1081 (4.8%) 3.8% (2.8-5.2) 5.8% (4.4-7.7) 1.00

40 Gy 34/1094 (3.1%) 1.8% (1.1-2.8) 4.2% (2.9-5.9) 0.62 (0.40-0.96) 0.032

Breast oedema*9.0% (7.3-11.0)

50 Gy 86/1003 (8.6%) 8.1% (6.6-10.1) 9.0% (7.3-11.0) 1.00

40 Gy 49/1006 (4.9%) 4.7% (3.5-6.2) 5.1% (3.9-6.7) 0.55 (0.39-0.79) 0.001

Shoulder stiffness^

50 Gy 4/73 (5.5%) 2.9% (0.7-11.0) 8.2% (2.9-21.8) 1.00

40 Gy 3/81 (3.7%) 3.1% (0.8-11.9) 3.1% (0.8-11.9) 0.76 (0.17-3.39) 0.71

Arm oedema^

50 Gy 7/73 (9.6%) 6.0% (2.3-15.3) 13.5% (6.4-27.0) 1.00

40 Gy 3/81 (3.7%) 2.8% (0.7-10.7) 4.7% (1.5-14.0) 0.42 (0.11-1.63) 0.21


Other

50 Gy 77/1082 (7.1%) 5.6% (4.3-7.2) 8.1% (6.5-10.2) 1.00

40 Gy 53/1095 (4.8%) 3.3% (2.4-4.6) 6.4% (4.8-8.4) 0.65 (0.46-0.93) 0.018

*Only assessed in women who had BCS. ^ Restricted to women who received lymphatic radiotherapy

(to axilla or supraclavicular.

Late adverse effects

Haviland et al (2013) reported on late adverse effects for both START A and START B including

symptomatic rib fracture, symptomatic lung fibrosis, ischaemic heart disease and brachial

plexopathy.17 For both START A and START B, ischaemic heart disease, symptomatic rib

fracture and symptomatic lung fibrosis were rare at 10 years and incidence was similar

between radiotherapy schedules, see table 11.17

Table 11 Incidence of other late adverse effects according to fractionation schedule in START

(2013)17

START A START B

50 Gy

(n=749)

41.6 Gy

(n=750)

39 Gy

(n=737)

Total

(n=2236)

50 Gy

(n=1105)

40 Gy

(n=1110)

Total

(n=2215)

Symptomatic rib fracture*

Reported 5 (0.7%) 8 (1.1%) 9 (1.2%) 22 (1.0%) 17 (1.5%) 24 (2.2%) 41 (1.9%)

Confirmed^ 0 0 1 (0.1%) 1 (<0.1%) 3 (0.3%) 3 (0.3%) 6 (0.3%)

Symptomatic lung fibrosis

Reported 6 (0.8%) 9 (1.2%) 8 (1.1%) 23 (1.0%) 19 (1.7%) 19 (1.7%) 38 (1.7%)

Confirmed^ 0 2 (0.3%) 1 (0.1%) 3 (0.1%) 2 (0.2%) 8 (0.7%) 10 (0.5%)

Ischaemic heart disease#

Reported 14 (1.9%) 11 (1.5%) 8 (1.1%) 33 (1.5%) 23 (2.1%) 17 (1.5%) 40 (1.8%)

Confirmed^

Total

Left-sided

7 (0.9%)

4 (0.5%)

5 (0.7%)

1 (0.1%)

6 (0.8%)

4 (0.5%)

18 (0.8%)

9 (0.4%)

16 (1.4%)

5 (0.5%)

8 (0.7%)

4 (0.4%)

24 (1.1%)

9 (0.4%)

Brachial

plexopathy

0 1 (0.1%) 0 1 (<0.1%) 0 0 0

*reported cases include seven after trauma (5 START A, 2 START B), and 10 after metastases (5 in START A

and 5 in START B). # 26 patients in START A and 22 in START B had pre-existing heart disease at enrolment

and were excluded. ^ after imaging and further investigations.

Change in breast appearance

The UK FAST trial’s primary endpoint was change in photographic breast appearance

measured by photographic assessments at baseline and at 2 years and 5 years.20

Assessments of 2-year change in photographic breast appearance were available for 81% of

patients still alive and disease free (729/901). The trial reported the risk ratio for mild or marked

change in 2 year photographic breast appearance for 30 Gy vs. 50 Gy was 1.70 (95% CI 1.26-

2.29, p=<0.001) and for 28.5 Gy vs. 50 Gy the risk ratio was 1.15 (95% CI 0.82-1.60, p=0.489).

The trial demonstrated a clear and statistically significant dose response between 28.5 Gy

and 30 Gy with worse results for change in photographic breast appearance at 2 years in

the 30 Gy patients. Outcomes were comparable between the 28.5 Gy schedule and 50 Gy


schedule, see table 12.20 Physician-assessed adverse effects in the breast confirmed the

findings of photographic assessment of breast appearance. Moderate or marked adverse

effects in the breast were reported in 155 patients overall. Three-year rates of physician-

assessed moderate/marked adverse effects in the breast were 17.3% (13.3-22.3%) for 30 Gy

and 11.1% (7.9-15.6%) for 28.5 Gy compared with 9.5% (6.5-13.7%) after 50 Gy; the rate in the

30 Gy group was significantly higher than in 50 Gy (p=<0.001) and in 28.5 Gy (p=<0.006). The

rates were similar between the 28.5 Gy and 50 Gy groups (p=0.18).20

The UK FAST trial reported that the most common moderate or marked adverse effects in the

breast were shrinkage (n=106) and induration (n=40), with some reports of oedema (n=27)

and telangiectasia (n=15).20 Rates of induration and shrinkage were similar in the 50 Gy and

28.5 Gy schedules but higher in the 30 Gy group.20 Results for breast shrinkage were also

significantly higher among patients in the 30 Gy group; 30 Gy vs. 50 Gy p=0.002 and 30 Gy vs.

28.5 Gy p=0.016 and similar between the 28.5 Gy and 50 Gy groups; p=0.455. 20

Change in photographic breast appearance gave an estimate of α/β of 2.6 Gy (95% CI 1.4-

3.7) in the UK FAST trial.20 Using this estimate, the isoeffect doses expressed in 2.0 Gy

equivalents for 30 and 28.5 Gy in 5 fractions are 56.3 and 51.6 Gy, respectively. Adjustment

for breast size and surgical deficit at baseline made little difference to the α/β estimate viz.

2.5 Gy (95% CI 1.2-3.7). Estimates of α/β for the physician-assessed adverse effects are similar

to the result for the photographic endpoint.20

Table 12 Change in photographic breast appearance at 2 years by fractionation schedule20

Fractionation schedule Total

n=729 (%)

RR for 30 Gy

vs. 50 Gy

(95% CI), p-

value

RR for 28.5

Gy vs. 50

Gy (95%

CI), p-

value

RR for 30

Gy vs.

28.5 Gy

(95% CI),

p-value

50 Gy,

n=239 (%)

30 Gy,

n=248 (%)

28.5 Gy

n=242 (%)

No

change

189 (79.1) 160 (64.5) 184 (76.0) 533 (73.1) 1, p=<0.001 1, p=0.26 1, p=0.002

Mild

change

46 (19.2) 65 (26.2) 49 (20.2) 160 (22.0) 1.48

(1.06-2.05)

1.07

(0.75-15.4)

1.37

(1.00-1.90)

Marked

change

4 (1.7) 23 (9.3) 9 (3.7) 36 (4.9) 6.06

(2.14-17.20)

2.25

(0.70-7.18)

2.70

(1.28-5.67)

Abbreviations: RR=risk ratio

In the conference abstract by Patni et al (2012), acute breast pain was observed more

frequently in the hypofractionated radiotherapy arm (52.63% vs. 80%, p=0.141) at the

completion of radiotherapy.22 This became statistically significant at 7-10 days of completion

of radiotherapy (57.89% vs. 95%, p=0.018). Breast pain was similar in both groups after 6

months of treatment. Breast oedema and heat sensation were similar at both points of

assessment in both study groups.22

Skin toxicity

Data from 327 consecutive patients on acute skin reactions in the UK FAST trial showed milder

acute reactions in both hypofractionated radiotherapy schedules, see table 13.20


Table 13 Acute skin reactions during treatment by fractionation schedule in UK FAST trial (2011)20

RTOG Grade Fractionation schedule Total (%)

50 Gy (%) 30 Gy (%) 28.5 Gy (%)

0=no visible change 8 (7.3) 28 (25.2) 42 (39.6) 78 (23.9)

1=faint/dull erythema 51 (46.4) 67 (60.4) 53 (50.0) 171 (52.3)

2=tender/bright erythema ± dry

desquamation

39 (35.5) 13 (11.7) 9 (8.5) 61 (18.7)

3=patchy moist desquamation, moderate

oedema

12 (10.9) 3 (2.7) 2 (1.9) 17 (5.2)

4=confluent moist desquamation, pitting

oedema

0 0 0 0

Total with known RTOG grade for acute

skin reaction

110 (100) 111 (100) 106 (100) 327 (100)

Not recorded 187 192 196 575

Not known 5 5 3 13

Total randomised 302 308 305 915

The conference abstract by Patni et al (2012) reported that radiation dermatitis was slightly

higher in the hypofractionated radiotherapy arm at completion of radiotherapy (90% vs.

78.95%, p=0.608) and at 7-10 days of radiotherapy (90% vs. 68.42%, p=0.204).22

Fragandrea et al (2012) also reported results of an RCT in a conference abstract.23 No

significantly difference was observed between the hypofractionated radiotherapy arm and

standard radiotherapy arm for radiation-induced skin toxicity (18% vs. 16% respectively).23

3.3.7 Quality of life

Quality of life outcomes were not reported in the included studies.


3.4 Ongoing trials

Trial Name Study Design Participants Intervention Control Status

NCT0000156130

Randomised,

prospective,

phase IV trial

Patients diagnosed with early

(invasive) breast cancer followed

by breast conversing surgery or

Mastectomy

42.5 Gy in 16 fractions over 22 days

Boost not reported

50 Gy in 25 fractions over 35

days

Boost not reported

Active, not

recruiting

NCT01349322 Randomised,

phase III trial

Patients diagnosed with early

stage breast cancer removed by

surgery

Patients undergo accelerated

hypofractionated radiotherapy with

a concurrent boost 5 days a week

for 3 weeks

Patients undergo standard

whole-breast radiotherapy

(WBI) for 5 days a week for

3-5 weeks followed by

sequential radiotherapy

boost

Recruiting

ISRCTN19906132

‘FAST-forward’

Randomised,

phase III trial

Patients diagnosed with invasive

carcinoma of the breast removed

by breast conservation surgery

n=4000

Patients receive 27Gy in 5 fractions

over 5 days

Patients receive 26Gy in 5 fractions

over 5 days

40Gy in 15 fractions over 15

days

Ongoing

SHARE trial Randomised,

phase III trial

Patients diagnosed with invasive

carcinoma of the breast

n=2800

Hypofractionated: Patients receive

42.5 Gy in 16 fractions over 3 weeks

Patients receive 40 Gy in 15

fractions over 3 weeks

Accelerated partial breast

irradiation:

40 Gy in 10 fractions over 3 weeks

50 Gy in 25 fractions over 35

days followed by a 10 to 16

Gy boost in 5 to 8 fractions

Recruiting


Investigators of ongoing trials were contacted in May 2014 to seek additional information on the trial and anticipated availability of results. The

information obtained from the investigators is listed below:.

Trial When will there be additional

information/data available

on trial?

Does the trial have an

expected date for

publication of results?

Will the trial include cost

effectiveness analysis?

Will patient preferences be

examined?

ISRCTN19906132 ‘FAST-

forward’

The trial has recently closed

(N=4000)

Data available probably in

2016 (secondary endpoints

focusing on clinical

assessment and patient self-

assessments of adverse

effects)

No, primary endpoint analysis

at a median 5 year (likely

2017)

Yes No

SHARE/NCT01247233 trial 2020

No Yes in a subset of patients

Yes, in Quality of Life

questionnaire


4 Discussion


radiotherapy for early (operable) breast cancer in November 2011. The guidelines were


January 2001 to March 2010 from randomised controlled trials. The RCT evidence in the first

systematic review came from five trials: RMH/GOC, Canadian trial, two Standardisation of

Breast Radiotherapy Trials (A and B) and Spooner (conference abstract only).

This current systematic review was undertaken to identify new and updated evidence on the

use of hypofractionated radiotherapy for the treatment of early breast cancer. The literature

search identified one new RCT (UK FAST 2011) and updated full publications from trials

included in the first systematic review: the START A, START B trials (Haviland 2013) and the

Spooner trial (Spooner 2012). Three further RCTs were identified that have only been

published as conference abstracts to date. The updated search identified three international

guidelines , but no systematic reviews. One meta-analysis was identified in the search

(Haviland 2010).

The body of evidence on hypofractionated radiotherapy for early (operable) breast cancer

from these two systematic reviews includes:

six primary RCTs: START A trial, START B trial, a trial by Spooner et al, the UK FAST trial,

the Canadian trial and the United Kingdom Royal Marsden Hospital/Gloucestershire

Oncology Centre (RMH/GOC) trial

three RCTs published as conference abstracts only.

The RMH/GOC, START A and the UK FAST trials tested two hypofractionated radiotherapy

regimens. The Canadian trial, the START B and Spooner trial each tested one

hypofractionated radiotherapy regimen. In all trials, the conventional radiotherapy regimen

used as a comparator was 50 Gy in 25 fractions, delivered over 5 weeks.

A range of hypofractionated radiotherapy regimens were examined, including

39 Gy in 13 fractions over 35 days(RMH/GOC trial and START A)

40 Gy in 15 fractions over 21 days (START B and Spooner trial)

41.6 Gy in 13 fractions over 35 days (START A)

42.5 Gy in 16 fractions over 22 days(Canadian trial)

42.9 Gy in 13 fractions over 35 days(RMH/GOC trial)

30 Gy in 5 once weekly fractions of 6Gy over 5 weeks (UK FAST)

28.5 Gy in 5 once weekly fractions of 5.7Gy over 5 weeks (UK FAST)

Four trials included women who had undergone breast conserving surgery only (Spooner

trial, UK FAST trial, Canadian trial, RMH/GOC).9,15,19,20,24,25 Two trials included women who had

undergone breast conserving surgery or mastectomy (START A and START B).16-18

Median follow up ranged from 37.3 months in the UK FAST trial to 16.9 years in the Spooner

trial.


From this updated systematic review and the 2011 systematic review, three studies examined

hypofractionated radiotherapy in patients with high grade tumours; the Canadian trial and

START trials as well as an additional retrospective population based cohort study included in

the original guideline, by Herbert et al (2012).26 The 2010 publication of the Canadian trial by

Whelan et al included an unplanned subgroup analysis including tumour grade.25 An

unplanned subgroup analysis of the Canadian trial showed that for patients with high grade

tumours, the hypofractionated radiotherapy regimen of 42.5 Gy in 16 fractions over 22 days

was associated with a higher local recurrence rate compared with conventionally

fractionated radiotherapy at 12 years follow-up (p=0.01).25 However, an updated analysis of

the Canadian trial by Bane et al (2014), reported no statistically significant difference for

local recurrence between grade 1-2 and grade 3 breast cancers (p=0.11).27 In the 2013

meta-analysis of the START A, START B trial and their pilot study no significant difference in

locoregional relapse between grade 1 and 2 tumours and grade 3 tumours (p=0.12) was

reported.17 A retrospective population based cohort study by Herbert et al (2012) of patients

with grade 3 breast cancer reported the 10-year cumulative incidence of local relapse was

6.9% in the hypofractionated group and 6.2% in the conventionally fractionated

radiotherapy group (p=0.99).26

In this updated systematic review, overall survival outcomes were consistent with the overall

survival outcomes reported in the 2011 systematic review. In both systematic reviews the

START B trial was the only study to report significantly higher overall survival rates in the

hypofractionated radiotherapy arm (40 Gy in 15 fractions over 3 weeks) compared with the

standard radiotherapy arm (updated systematic review: HR 0.80, 95% CI 0.65-0.99, p=0.042).

The START A, the Spooner 2012 trial and the Canadian trial (Whelan 2010) reported no

statistically significant differences in overall survival between hypofractionated radiotherapy

and standard radiotherapy.

START B reported significantly longer disease-free survival at 10 years in patients receiving

hypofractionated radiotherapy compared with patients receiving standard radiotherapy (HR

0.79, 95% CI 0.65-0.97, p=0.022). START A reported no significant difference for disease-free

survival between the hypofractionated radiotherapy schedules and standard radiotherapy.

Similar 2, 5, 10 and 15 year relapse-free survival estimates and cumulative incidence-free

estimates between the hypofractionated radiotherapy and standard radiotherapy regimens

were reported in the Spooner 2012 trial.

Five trials reported on local recurrence (START A, START B, Spooner 2012, RMH/GOC trial,

Canadian trial). All trials reported similar rates of local relapse for women treated with

hypofractionated radiotherapy and standard radiotherapy.9,17,19,25 RMH/GOC noted a

statistically significant difference in recurrence rates between the two hypofractionated

regimens (42.9 Gy vs. 39 Gy: 9.6% vs. 14.8%, p=0.027) but not when either of the

hypofractionated regimens was compared to 50 Gy in 25 fractions.9

START A and START B also reported no significant difference in 10 year local-regional relapse

rates between hypofractionated radiotherapy schedules and standard radiotherapy. Post-

hoc meta-analysis also demonstrated no significant difference in local-regional relapse rates

when stratified by age, type of primary surgery, axillary node status, tumour grade, adjuvant

chemotherapy use, or use of tumour bed boost radiotherapy.


START B reported significantly less distant relapses in patients receiving hypofractionated

radiotherapy (40 Gy in 15 fractions over 3 weeks) compared with patients receiving standard

radiotherapy (distant relapse rate: HR 0.74, 95% CI 0.59-0.94, p=0.014). Similar rates of distant

relapse were reported between patients receiving hypofractionated radiotherapy and

patients receiving conventionally fractionated radiotherapy in the START A, Spooner 2012

and UK FAST trials.17,19,20

Moderate or marked breast shrinkage, telangiectasia, and breast oedema were

experienced significantly less in patients who received 39 Gy in 13 fractions over 5 weeks in

the START A trial and 40 Gy in 15 fractions over 3 weeks in the START B trial compared with

patients receiving standard radiotherapy. There was no significant difference between 41.6

Gy in 13 fractions over 5 weeks and standard radiotherapy in the START A trial.

The UK FAST trial reported risk ratios for mild or marked change in 2 year photographic breast

appearance. The trial reported the risk ratio for mild or marked change in 2 year

photographic breast appearance for 30 Gy vs. 50 Gy was 1.70 (95% CI 1.26-2.29, p=<0.001)

and for 28.5 Gy vs. 50 Gy the risk ratio was 1.15 (95% CI 0.82-1.60, p=0.489). The trial

demonstrated a clear and statistically significant dose response between 28.5 Gy and 30 Gy

with worse results for change in photographic breast appearance at 2 years in the 30 Gy

patients. Outcomes were comparable between the 28.5 Gy schedule and 50 Gy schedule..

Three-year rates of physician-assessed moderate/marked adverse effects in the breast were

also significantly higher in the hypofractionated radiotherapy group of 30 Gy in 5 once

weekly fractions over 5 weeks compared with the standard radiotherapy group (p=<0.001)

and the hypofractionated radiotherapy group of 28.5 Gy in 5 once weekly fractions over 5

weeks (p=<0.006). The rates were similar between the 28.5 Gy and 50 Gy groups (p=0.18).20

Results for breast shrinkage were also significantly higher among patients in the 30 Gy group;

30 Gy vs. 50 Gy p=0.002 and 30 Gy vs. 28.5 Gy p=0.016 and similar between the 28.5 Gy and

50 Gy groups; p=0.455.

The Canadian trial reported that following assessments at baseline, three, five and ten years

after randomisation, the global cosmetic outcome worsened over time however there were

no significant differences observed between the 42.5 Gy group and the 50 Gy group at any

time.25 At ten years follow-up, 71.3% of women in the 50 Gy group compared to 69.8% of

women in the hypofractionated radiotherapy treatment group had an excellent or good

cosmetic outcome.25 Cosmetic outcome was shown to be affected by time from

randomisation, patient’s age and tumour size but there was no interaction with the

treatment.25

For photographically assessed changes in breast appearance, the RMH/GOC trial found a

higher risk of developing any radiation effect for patients allocated to 42.9 Gy in 13 fractions,

compared to those allocated to 39 Gy in 13 fractions or 50 Gy in 25 fractions (p=<0.001 for

comparison of three fractionation schedules).15 Clinical assessment of patients also indicated

significant differences between the three fractionation schedules, with the 42.9 Gy group

experiencing the highest incidence of events for overall breast cosmesis (p=<0.001), breast

shrinkage (p=0.026), breast distortion (p=0.005), breast oedema (p=0.004), induration

(p=0.001) and shoulder stiffness (p=0.001).15


5 Conclusion

This updated systematic review considered the evidence on the effectiveness of

hypofractionated radiotherapy for the treatment of early breast cancer compared with

standard radiotherapy following surgery.

The updated 10 year follow-up results from START A and START B and the published results

from the Spooner trial and UK FAST trial provide further evidence that hypofractionated

radiotherapy is equivalent to standard radiotherapy for women with early (operable) breast

cancer. Primary outcomes of overall survival, disease-free survival and recurrence were

similar for hypofractionated radiotherapy compared to standard radiotherapy and in some

instances were improved for those receiving hypofractionated radiotherapy.


Appendix A Contributors

Working group members

The hypofractionated radiotherapy for the treatment of early breast cancer systematic

review was developed with input from an expert multidisciplinary working group with the

following members:

Dr Marie-Frances Burke (Chair) Radiation Oncologist

Ms Jan Rice Breast Care Nurse

Dr Patsy Soon Breast Surgeon

Dr Kirsty Stuart Radiation Oncologist

Ms Bronwyn Wells Consumer Representative

Cancer Australia staff

The following Cancer Australia staff were involved in the development of the

hypofractionated radiotherapy for the treatment of early breast cancer systematic review:

Ms Katrina Anderson Senior Project Officer, Evidence Review

Ms Ornella Care Manager, Breast Cancer

Ms Medora Lee Project Officer, Evidence Review

Dr Anne Nelson Manager, Evidence Review


Appendix B Literature databases searched

Source Results/Retrievals

Medline (OVID) 172

Embase 322

Pubmed 86


Appendix C Search strategy

Breast cancer 1. "Breast Neoplasms"[MeSH] OR “breast cancer” OR

“breast carcinoma” OR “breast neoplasm” OR

“breast tumour” OR “breast tumor”

Hypofractionated radiotherapy 2. “radiation dose fractionation”[MeSH] OR

“radiation dose fractionation” OR “dose

fractionation”[MeSH] OR “dose fractionation” OR

“hypofractionated radiotherapy” OR

“fractionated radiotherapy”

3. fractionated OR hypofractionated OR fraction OR

fractio*)

4. “radiotherapy”[MeSH] OR “radiotherapy” OR

“irradiation therapy”[MeSH] OR “irradiation

therapy” OR “irradiation treatment”[MeSH] OR

“irradiation treatment” OR “radiation

therapy”[MeSH] OR “radiation therapy” OR

“radiation treatment”[MeSH] OR “radiation

treatment” OR “therapeutic radiology”[MeSH] OR

“therapeutic radiology” OR “radiation, therapy”

OR “treatment, irradiation”[MeSH] OR “treatment,

irradiation”

Search combination for hypofractionated radiotherapy

term: (#3 AND #4) OR #2

RCTS and meta-analyses 5. ("Randomized Controlled Trial"[MeSH] OR

"randomized controlled trial" OR "randomized

controlled trials" OR "randomised controlled trial*"

OR "random*" OR "random allocation" OR

"controlled clinical trial" OR "controlled trial" OR

"double blind method" OR "single blind method"

OR ("meta-analysis"[MeSH] OR "meta-analysis" OR

"meta analysis") OR "systematic review" OR "pooled

analysis")

FINAL COMBINATION #1 AND #hypofractionated radiotherapy terms combination AND #5


Appendix D Guideline and clinical trial sites searched

Acronym Organisation Website

Australia

ANZHSN Australia and New

Zealand Horizon

Scanning Network

http://www.horizonscanning.gov.au/

MSAC Medical Services

Advisory Committee

http://www.msac.gov.au/

NHMRC National Health and

Medical Research

Council

http://www.nhmrc.gov.au/

Canada

CCO Cancer Care Ontario http://www.cancercare.on.ca/

CADTH Canadian Agency for

Drugs and Technologies

in Health

http://www.cadth.ca/

International

HTAi Health Technology

Assessment International

http://www.htai.org/

Scotland

SIGN Scottish Intercollegiate

Guidelines Network

http://www.sign.ac.uk/

UK

CCT Current Controlled Trials http://www.controlled-trials.com/

NICE National Institute for

Health and Clinical

Excellence

http://www.nice.org.uk/

NRR National Research

Register

http://www.nrr.nhs.uk/

US

CT ClinicalTrials.gov http://www.clinicaltrials.gov/

NCI National Cancer Institute

Clinical Trials

http://www.cancer.gov/clinicaltrials

NGC National Guideline

Clearinghouse

http://www.guideline.gov/

AHRQ Agency for Healthcare

Research and Quality

http://www.ahrq.gov/

http://www.horizonscanning.gov.au/

http://www.msac.gov.au/

http://www.nhmrc.gov.au/

http://www.cancercare.on.ca/

http://www.cadth.ca/

http://www.htai.org/

http://www.sign.ac.uk/

http://www.controlled-trials.com/

http://www.nice.org.uk/

http://www.nrr.nhs.uk/

http://www.clinicaltrials.gov/

http://www.cancer.gov/clinicaltrials

http://www.guideline.gov/

http://www.ahrq.gov/


Appendix E Flowchart of inclusion/exclusion


Abbreviations

ASCO American Society of Clinical Oncology

ASTRO American Society for Radiation Oncology

BCS Breast conserving surgery

DEGRO German Society of Radiation Oncology

DFS Disease-Free Survival

EBCTG Early Breast Cancer Trialists Group

ESMO European Society of Medical Oncology

GIN Guidelines International Network

Gy Gray

HR Hazards ratio

IMRT Intensity Modulated Radiotherapy

NICE National Institute of Clinical Excellence

NZGG New Zealand Guidelines Group

OS Overall survival

RCTs Randomised controlled trials

RR Risk ratio

SABCS San Antonio Breast Cancer Symposium

SIGN Scottish Intercollegiate Guidelines Network

SHARE trial Standard or Hypofractionated Radiotherapy versus

Accelerated Breast Irradiation for Breast Cancer

START trials Standardisation of Breast Radiotherapy trials


References

1. James ML, Lehman M, Hider PN, et al. Fraction size in radiation treatment for breast

conservation in early breast cancer. Cochrane Database Syst Rev. 2010;11):CD003860.

2. Freedman GM. Hypofractionated radiation therapy in the treatment of early-stage breast

cancer. Curr Oncol Rep. 2012;14(1):12-9.

3. Australian Institute of Health and Welfare & Australasian Association of Cancer Registries.

Cancer in Australia: an overview, 2012. Cancer series no. 74. Cat. no. CAN 70. AIHW,

Canberra, 2012.

4. Australian Institute of Health and Welfare. Cancer incidence and projections: Australia,

2011 to 2020. Cancer series no. 66. Cat no. CAN 62. AIHW, Canberra, 2012.

5. Australian Institute of Health and Welfare & Cancer Australia 2012. Breast cancer in

Australia: an overview. Cancer series no. 71. Cat no. CAN 67. AIHW, Canberra, 2012.

6. National Breast Cancer Centre (NBCC). Clinical practice guidelines for the management

of early breast cancer (2nd edition). Commonwealth of Australia, Canberra, 2001.

7. Early Breast Cancer Trialists' Collaborative G, Darby S, McGale P, et al. Effect of

radiotherapy after breast-conserving surgery on 10-year recurrence and 15-year breast

cancer death: meta-analysis of individual patient data for 10,801 women in 17 randomised

trials. Lancet. 2011;378(9804):1707-16.

8. Smith BD, Bentzen SM, Correa CR, et al. Fractionation for Whole Breast Irradiation: An

American Society for Radiation Oncology (ASTRO) Evidence-Based Guideline. Int J Radiat

Oncol Biol Phys. 2011;81(1):59-68.

9. Owen JR, Ashton A, Bliss JM, et al. Effect of radiotherapy fraction size on tumour control in

patients with early-stage breast cancer after local tumour excision: long-term results of a

randomised trial. Lancet Oncol. 2006;7(6):467-71.

10. Cuncins-Hearn AV, Boult M, Babidge W, et al. National breast cancer audit: overview of

invasive breast cancer management. ANZ J Surg. 2006;76(8):745-50.

11. Cancer Australia. Recommendations for use of Hypofractionated radiotherapy for early

(operable) breast cancer. Cancer Australia, Surry Hills, NSW, 2011.

12. Senkus E, Kyriakides S, Penault-Llorca F, et al. Primary breast cancer: ESMO Clinical

Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2013;24 Suppl 6(vi7-

23.


13. Sedlmayer F, Sautter-Bihl ML, Budach W, et al. DEGRO practical guidelines: radiotherapy

of breast cancer I: radiotherapy following breast conserving therapy for invasive breast

cancer. Strahlenther Onkol. 2013;189(10):825-33.

14. Bourgier C, Aimard L, Bodez V, et al. Adjuvant radiotherapy in the management of

axillary node negative invasive breast cancer: a qualitative systematic review. Crit Rev Oncol

Hematol. 2013;86(1):33-41.

15. Yarnold J, Ashton A, Bliss J, et al. Fractionation sensitivity and dose response of late

adverse effects in the breast after radiotherapy for early breast cancer: long-term results of a

randomised trial. Radiother Oncol. 2005;75(1):9-17.

16. Bentzen SM, Agrawal RK, Aird EG, et al. The UK Standardisation of Breast Radiotherapy

(START) Trial A of radiotherapy hypofractionation for treatment of early breast cancer: a

randomised trial. Lancet Oncol. 2008;9(4):331-41.

17. Haviland JS, Owen JR, Dewar JA, et al. The UK Standardisation of Breast Radiotherapy

(START) trials of radiotherapy hypofractionation for treatment of early breast cancer: 10-year

follow-up results of two randomised controlled trials. Lancet Oncol. 2013;14(11):1086-94.

18. Bentzen SM, Agrawal RK, Aird EG, et al. The UK Standardisation of Breast Radiotherapy

(START) Trial B of radiotherapy hypofractionation for treatment of early breast cancer: a

randomised trial. Lancet. 2008;371(9618):1098-107.

19. Spooner D, Stocken DD, Jordan S, et al. A Randomised Controlled Trial to Evaluate both

the Role and the Optimal Fractionation of Radiotherapy in the Conservative Management of

Early Breast Cancer. Clinical Oncology. 2012;24(10):697-706.

20. FAST Trialists group, Agrawal RK, Alhasso A, et al. First results of the randomised UK FAST Trial

of radiotherapy hypofractionation for treatment of early breast cancer (CRUKE/04/015).

Radiother Oncol. 2011;100(1):93-100.

21. Barsoum MS, El Mongi MM, Khalil EM, et al. Prospective randomized trial comparing

postoperative adjuvant concurrent versus sequential hormonal and different radiation

fractionation schedule in breast cancer patients. Journal of Clinical Oncology.

2010;28(15s):abstr 544.

22. Patni N, Jain M, Patni S and Bapna A. A comparison of acute and chronic toxicity profile

between conventional and hypofractionated whole breast irradiation in patients undergoing

breast conserving surgery. International Journal of Radiation Oncology Biology Physics.

2012;84(Issue 3 Supplement):S232.

23. Fragandrea I, Kouloulias V, Sotiropoulou A, et al. Radiation induced skin toxicity following

hypofractionated radiotherapy treatment in early breast cancer: Single institution

experience. European Journal of Cancer. 2012;48(Supplement 1):S158.


24. Whelan T, MacKenzie R, Julian J, et al. Randomized trial of breast irradiation schedules

after lumpectomy for women with lymph node-negative breast cancer. J Natl Cancer Inst.

2002;94(15):1143-50.

25. Whelan TJ, Pignol JP, Levine MN, et al. Long-term results of hypofractionated radiation

therapy for breast cancer. N Engl J Med. 2010;362(6):513-20.

26. Herbert C, Nichol A, Olivotto I, et al. The Impact of Hypofractionated Whole Breast

Radiotherapy on Local Relapse in Patients with Grade 3 Early Breast Cancer: A Population-

based Cohort Study. Int J Radiat Oncol Biol Phys. 2012;82(5):2086-92.

27. Bane AL, Whelan TJ, Pond GR, et al. Tumor Factors Predictive of Response to

Hypofractionated Radiotherapy in a Randomized Trial Following Breast Conserving Therapy.

Ann Oncol. 2014.

1

Hypofractionated radiotherapy for early breast cancer

Left- vs. right-sided breast cancers

October 2014

Background

Cancer Australia is currently updating the 2011 topic-specific guideline Recommendations

for use of hypofractionated radiotherapy for the treatment of early (operable) breast cancer

to incorporate new and updated evidence and revise the statements of evidence and

recommendations. At working group meetings it was agreed that consideration needed to

be given to whether or not the recommendations be updated to include right-sided tumours

as a patient inclusion criterion for offering hypofractionated radiotherapy and left-sided

tumours be included as a criterion with insufficient evidence to recommend for or against

hypofractionated radiotherapy.

The systematic review to support the update of the guideline did not identify/report

evidence for left-vs. right-sided tumours. This section of the Technical Document provides

evidence on cardiotoxicity left- vs. right-sided breast cancers from the randomised

controlled trials (RCTs) included in the guideline. Evidence from additional non-randomised

trials has also been included.

Method

Evidence on left- vs. right-sided tumours was extracted from RCTs identified in either the 2011

systematic review and/or the updated 2014 systematic review. Additional evidence from

non-randomised studies was identified in a supplementary literature search. The search was

undertaken in August 2014 in the electronic database PubMed and included key words

“breast cancer”, “hypofractionated radiotherapy” and “left”. The search identified seven

citations of which four are included in this evidence summary.

1. Evidence from randomised controlled trials (RCTs)

The topic-specific guideline Recommendations for use of hypofractionated radiotherapy for

the treatment of early (operable) breast cancer is based on evidence from six RCTs

comparing hypofractionated radiotherapy to conventionally fractionated radiotherapy;

START A, START B, Spooner 2012, UK FAST trial, Canadian trial, and RMH/GOC trial. None of the

trials included left- and right-sided tumours in their inclusion or exclusion criteria. Only three of

the six trials reported cardiac outcomes by left- vs right-sided tumours: START A, START B and

UK FAST. Table 1 shows the total number of deaths reported in each of these three trials, with

breakdowns by cardiac disease related, breast cancer related, other cancer related or

other non-cancer related. Cardiac disease related deaths are shown by tumour side, where

reported. The median duration of follow-up most recently reported for START A, START B and

UK FAST is 9.3, 9.9 and 3.1 years, respectively.

2

Table 1: Mortality in hypofractionated radiotherapy RCTs showing cardiac disease-related deaths by tumour side

Start A (Haviland 2013)1

(median follow-up 9.3yrs)

Start B (Haviland 2013)1

(median follow-up 9.9yrs

UK-FAST 20113

(median follow-up 3.1yrs)

50Gy 41.6Gy 39Gy Total 50Gy 40Gy Total 30Gy 28.5Gy 50Gy Total

N=749 N=750 N=737 N=2236 N=1105 N=1110 N=2215 N=308 N=305 N=302 N=915

All cause 130 128 134 392 192 159 351 5 12 6 23

Breast cancer related 92 86 95 273 130 106 236 2 6 2 10

Other cancer related 9 10 15 34 25 23 48 NR NR NR NR

Cardiac disease related only 7 13 6 26 12 5 17 NR NR NR 4

Left sided 4 10 1 15 8 3 11 NR NR NR 2

Right sided 3 3 5 11 4 2 6 NR NR NR 2

Othera 16 16 12 44 21 19 40 3 6 4 13

Unknown 6 3 6 15 4 6 10 NR NR NR NR

Abbreviations NR: not reported; a defined as non-cancer, non-cardiac disease in START A and START B, and as non-breast cancer related in UK FAST

3

As shown in table 1, the most common cause of death in patients across these three RCTs

was breast or other cancer. In START A after 9.3 years median follow-up, 26/392 (6.6%) deaths

were related to cardiac disease (seven with 50 Gy, 13 with 41.6 Gy, and six with 39 Gy).

Fifteen (57.7%) of the 26 deaths from cardiac disease were in women with left-sided primary

tumours (four of seven with 50 Gy, ten of 13 with 41.6 Gy, and one of six with 39 Gy). In START

B, after 9.9 years median follow-up, 17/351 (4.8%) deaths were related to cardiac disease (12

with 50 Gy and five with 40 Gy). Eleven (64.7%) of the 17 deaths from cardiac disease were in

women with left-sided primary tumours (eight of 12 with 50 Gy and three of five with 40 Gy).

In UK FAST after 3.1 years median follow-up, 4/23 (17.4%) deaths were attributed to cardiac

disease, with two deaths in women with left sided tumours, and two deaths in women with

right-sided tumours. However, the UK FAST publication does not report the treatment group

assignment for any of these cardiac disease related deaths.In addition, while the Canadian

trial did not report results for left- and right-sided breast cancers, the authors did note that at

a median follow-up of 12 years few cardiac-related deaths were observed and no increase

occurred in patients who received the hypofractionated regimen.4

When interpreting the mortality rates from START A, B and UK FAST a number of factors should

be kept in mind. The number of events in each study is low and the number of deaths due to

unknown causes is relatively high. For example, in START A only four of the fifteen deaths from

unknown causes need to be cardiac-disease related in right-sided tumours for the observed

event rates to be equivalent. Furthermore, although women with pre-existing heart disease

were excluded from START A and START B, none of the three studies stratified patients at

baseline by cardiac risk factors. And finally, interpretation of the available evidence is

difficult given that subsequent chemotherapy regimens will likely have differed among

women and not been matched across radiotherapy treatment groups. Differences in the

cardiotoxicity of subsequent chemotherapy regimens may confound the attribution of

cardiotoxicity to a particular radiotherapy regimen.

Haviland et al (2013) concluded that the START A and B trial results showed that although

follow-up was still shorter than would be desired for cardiac events, there was no major

difference between the fractionation schedules for the number of cases of heart disease in

women with left-sided primary tumours, see table 2.1 Haviland et al (2013) also note that the

heart is sensitive to radiation whatever fractionation is used with no lower dose threshold for

adverse effects. A commentary on the 2013 START trial results agreed with the START trial

authors that techniques to protect the heart are important for both radiotherapy schedules

and the choice of fractionation should not be affected by whether the tumour is in the right

or left breast.6

Table 2: Incidence of ischaemic heart disease according to fractionation schedule (START

trials) 1

Ischaemic

heart

disease*

START A START B

50 Gy

n=749

41.6 Gy

n=750

39 Gy

n=737

Total

n=2236

50 Gy

n=1105

40 Gy

n=1110

Total

n=2215

Reported 14 (1.9%) 11 (1.5%) 8 (1.1%) 33 (1.5%) 23 (2.1%) 17 (1.5%) 40 (1.8%)

Confirmed

Total 7 (0.9%) 5 (0.7%) 6 (0.8%) 18 (0.8%) 16 (1.4%) 8 (0.7%) 24 (1.1%)

Left sided 4 (0.5%) 1 (0.1%) 4 (0.5%) 9 (0.4%) 5 (0.5%) 4 (0.4%) 9 (0.4%)

4

*26 patients in START A and 22 in START B had pre-existing heart disease at enrolment and were

excluded

2. Evidence from additional non-randomised studies

Four additional non-randomised studies were identified which include outcomes for left- and

right-sided breast cancers.

A population-based retrospective study by Chan et al was reported in two 2014 publications.

The first paper aimed (median follow-up 13.2 years; Ontario) to determine if there is an

increase in hospital-related morbidity from cardiac causes with either hypofractionated

radiotherapy (40-44 Gy in 16 fractions) or conventional radiotherapy (45-50 Gy in 25 fractions

or 50.4 Gy in 28 fractions).7 For left-sided cases, 15-year cumulative hospital-related morbidity

from cardiac causes was not different between the two radiotherapy regimens (both 21%,

p=0.93). The difference was also not significant for right-sided cases (hypofractionated 18%,

conventional 19%; p=0.76). The 15-year cumulative mortality before first cardiac

hospitalisation between hypofractionated radiotherapy and conventional radiotherapy was

not statistically different; 20.7% vs. 23.8% respectively (p= NR). Right-sided cases were also not

significantly different.

The authors concluded that for women with left-sided early-stage breast cancer who

received postoperative radiation therapy to the whole breast or chest wall, there was no

difference in the 15-year cumulative morbidity due to cardiac causes, between

conventionally fractionated and hypofractionated treatment schedules. Even after

adjustment for baseline patient, tumour, and treatment factors between the two radiation

therapy fractionation groups, no difference was seen in the incidence of cardiac morbidity.7

The second publication from the population-based retrospective study by Chan et al (2014)

(median follow-up 14 years; Ontario) reported on if there is an increase in cardiac mortality

with hypofractionated radiotherapy relative to conventional radiotherapy.10 For left-sided

cases, at 15-years follow-up, the rate of cardiac mortality for hypofractionated radiotherapy

was 4.8% and for conventional radiotherapy it was 4.2%, this difference was not statistically

significant (p = 0.74). The difference was also not significant for right-sided cases

(hypofractionated 4.9%, conventional 3.5%, p=0.21). The 15-year cumulative mortality due to

breast cancer, cardiovascular and other causes was calculated separately, comparing

hypofractionated and conventional radiotherapy for the left and right-sided cases; none of

the results were statistically significant.

The authors concluded that at 15-years follow-up, cardiac mortality is not statistically

different among left-sided breast cancer patients treated with hypofractionated

radiotherapy and conventional radiotherapy.

The authors noted that additional unpublished analyses were undertaken in which cardiac

cause of death was prioritised a priori. In other words, for cases where breast cancer and a

cardiac event were listed as cause of death, the main analysis indexed breast cancer as the

cause of death. However, in the additional analyses the cardiac event was indexed as the

cause of death. These analyses are therefore highly conservative with respect to cardiac

mortality (i.e., likely to over-estimate cardiac attribution).

From personal correspondence with the study authors two additional analysis were provided,

with:

1. cardiac cause as a priority

2. cardiac cause not including cardiac arrest as a priority

5

For both analyses there was no significant difference between hypofractionated

radiotherapy and conventional radiotherapy for death from cardiac causes at 10 or 15 years

(p=0.206 and p=0.406, respectively).

The additional analyses are consistent with the overall conclusion of the study that at 15-

years follow-up, cardiac mortality is not statistically different among left-sided breast cancer

patients treated with hypofractionated radiotherapy or conventionally fractionated

radiotherapy.

Appelt et al (2013) analysed dose plans for 60 left-sided breast cancer patients to compare

fraction size-corrected dose distributions to the heart for four hypofractionated schedules

with the normofractionated schedule of 50 Gy in 25 fractions, for a range of α/β values.8 All

patients were planned with tangential fields for whole breast irradiation. Dose distributions

were corrected to the equivalent dose in 2 Gy fractions (EQD2) using the linear quadratic

model for five different fractionation schedules (50 Gy/25 fractions and four

hypofractionated regimens) and for a range of α/β values (0-5 Gy). The mean EQD2 to the

heart (DEQD2 mean) and the volume receiving 40 Gy (VEQD2 40 Gy), both as calculated from

the EQD2 dose distributions, were compared between schedules.

The authors concluded that for standard tangential field whole breast irradiation, most of the

examined hypofractionated schedules are estimated to spare the heart when compared

with normofractionation. The dose to the heart, adjusted for fraction size using the linear

quadratic model, will generally be lower after hypofractionated compared with

normofractionated schedules, even for very low values of α/β values.8

Holm Tjessem (2013) reported results of a retrospective case-control study that analysed the

20-year risk of death from ischemic heart disease (IHD) in breast cancer patients who

received hypofractionated locoregional radiation therapy at the Norwegian Radium

Hospital between 1975 and 1991 (median follow-up: all patients 4.5yrs, surviving patients

20yrs). Two hypofractionated radiotherapy regimens were used: 4.3 Gy x 10 given as 2

weekly fractions (n=1107) and 2.5 Gy x 20 given as 4 weekly fractions (n=459). Controls were

cancer-free (n=10). In the 4.3 Gy group, 86% of patients died, compared with 84% in the 2.5

Gy group. In the 4.3 Gy group, 4% died of IHD compared with 2% in the 2.5 Gy group

(HR=2.37, 95% CI 1.06-5.32, p=0.036), with the difference emerging first after 10 to 15 years. In

multivariate analysis the 4.3 Gy still had an associated increased risk of IHD but with

borderline significance only (HR=2.90, 95% CI 0.97-8.76, p=0.057). Patients in the 4.3 Gy group

had an increased risk of dying of IHD compared with controls (HR 1.59, 95% CI 1.13-2.23,

p=0.008). No elevated risk of death resulting from IHD was found in the 2.5 Gy group

compared with controls.

Treatment before 1984 was associated with an increased risk of dying of IHD (HR 2.87, 95% CI

1.34-6.17, p=0.006) compared with treatment from 1984 onwards. Patients treated for left-

sided cancer did not have increased risk of dying of IHD compared with right-sided breast

cancer.

The authors concluded that the degree of fractionation and photon beams in the

parasternal contributed to increased IHD mortality in the patient cohort. The differences in

IHD mortality emerged 12 to 15 years after treatment and the increased risk among breast

cancer patients treated with hypofractionated radiation therapy thus indicates that a follow-

up time of at least two decades is needed to evaluate safety of such irradiation.

6

Marhin et al (2007) in an earlier retrospective study of the Ontario cases(median follow-up 7.9

years), assessed whether an adjuvant hypofractionated radiotherapy schedule with fraction

sizes >2 Gy would increase the risk of cardiac mortality in women with localised left-sided

breast cancer compared with schedules with fraction sizes ≤2 Gy.9 The study reported there

was no significant difference in cardiac mortality for women ≤60 or >60 years of age who

received adjuvant radiotherapy for left-sided vs. right-sided cancer. There was no difference

in cardiac mortality for women who received adjuvant radiotherapy with fraction sizes ≤2 vs.

>2 Gy for left- or right-sided cancer. The study reported that the relative risk of cardiac death

for left- relative to right-sided radiotherapy with >2 Gy fractions was 1.07 (95% CI 0.68-1.69).

The authors concluded that the absence of a difference in cardiac mortality between

women treated for left-sided breast cancer with hypofractionated vs. conventional

fractionated radiotherapy adds further support to the efficacy of hypofractionated regimens

in this clinical setting.9

An additional randomised study of 60 patients was identified (Ibrahim 2014), however the

study population was small and had limited results. The authors concluded hypofractionated

radiotherapy decreased cardiac toxicity though not statistically significant, however it is

more cost effective and time consuming.

References

1. Haviland JS, Owen JR, Dewar JA, et al. The UK Standardisation of Breast Radiotherapy

(START) trials of radiotherapy hypofractionation for treatment of early breast cancer:

10-year follow-up results of two randomised controlled trials. Lancet Oncol.

2013;14(11):1086-94.

2. Spooner D, Stocken DD, Jordan S, et al. A Randomised Controlled Trial to Evaluate

both the Role and the Optimal Fractionation of Radiotherapy in the Conservative

Management of Early Breast Cancer. Clinical Oncology. 2012;24(10):697-706.

3. FAST Trialists group, Agrawal RK, Alhasso A, et al. First results of the randomised UK FAST

Trial of radiotherapy hypofractionation for treatment of early breast cancer

(CRUKE/04/015). Radiother Oncol. 2011;100(1):93-100

4. Whelan TJ, Pignol JP, Levine MN, et al. Long-term results of hypofractionated radiation

therapy for breast cancer. N Engl J Med. 2010;362(6):513-20.

5. Owen JR, Ashton A, Bliss JM, et al. Effect of radiotherapy fraction size on tumour

control in patients with early-stage breast cancer after local tumour excision: long-

term results of a randomised trial. Lancet Oncol. 2006;7(6):467-71.

6. Haffty BG, Buchholz TA. Hypofractionated breast radiation: preferred standard of

care? Lancet Oncol. 2013;14(11):1032-4.

7. Chan EK, Woods R, McBride ML, et al. Adjuvant hypofractionated versus conventional

whole breast radiation therapy for early-stage breast cancer: long-term hospital-

related morbidity from cardiac causes. Int J Radiat Oncol Biol Phys. 2014;88(4):786-92.

8. Appelt AL, Vogelius IR, Bentzen SM. Modern hypofractionation schedules for

tangential whole breast irradiation decrease the fraction size-corrected dose to the

heart. Clin Oncol (R Coll Radiol). 2013;25(3):147-52.

9. Marhin W, Wai E, Tyldesley S. Impact of fraction size on cardiac mortality in women

treated with tangential radiotherapy for localized breast cancer. Int J Radiat Oncol

Biol Phys. 2007;69(2):483-9.

7

10. Chan EK, Woods R, Virani S, et al. Long-term mortality from cardiac causes after

adjuvant hypofractionated vs. conventional radiotherapy for localized left-sided

breast cancer. Radiother Oncol. 2014 early online 13th September 2014.

Hypofractionated radiotherapy for the treatment of early breast cancer: a systematic

review

November 2011

Hypofractionated radiotherapy for the treatment of early breast cancer: a systematic review was

developed by:

Cancer Australia

Locked Bag 3 Strawberry Hills NSW 2012 Australia

Tel: +61 2 9357 9400 Fax: +61 2 9357 9477

Website: www.canceraustralia.gov.au

© Cancer Australia 2011

ISBN Online: 978-1-74127-180-5

Recommended citation

Cancer Australia. Hypofractionated radiotherapy for the treatment of early breast cancer: a systematic review.

National Breast and Ovarian Cancer Centre, Surry Hills, NSW, 2011

Copyright statements:

Paper-based publications

This work is copyright. You may reproduce the whole or part of this work in unaltered form for your own personal use

or, if you are part of an organisation, for internal use within your organisation, but only if you or your organisation do

not use the reproduction for any commercial purpose and retain this copyright notice and all disclaimer notices as

part of that reproduction. Apart from rights to use as permitted by the Copyright Act 1968 or allowed by this

copyright notice, all other rights are reserved and you are not allowed to reproduce the whole or any part of this

work in any way (electronic or otherwise) without first being given the specific written permission from Cancer

Australia to do so. Requests and inquiries concerning reproduction and rights are to be sent to the Publications and

Copyright contact officer, Cancer Australia, Locked Bag 3, Strawberry Hills, NSW 2012.

Internet sites

This work is copyright. You may download, display, print and reproduce the whole or part of this work in unaltered

form for your own personal use or, if you are part of an organisation, for internal use within your organisation, but only

if you or your organisation do not use the reproduction for any commercial purpose and retain this copyright notice

and all disclaimer notices as part of that reproduction. Apart from rights to use as permitted by the Copyright Act

1968 or allowed by this copyright notice, all other rights are reserved and you are not allowed to reproduce the

whole or any part of this work in any way (electronic or otherwise) without first being given the specific written

permission from Cancer Australia to do so. Requests and inquiries concerning reproduction and rights are to be sent

to the Publications and Copyright contact officer, Cancer Australia, Locked Bag 3, Strawberry Hills, NSW 2012

Copies of Hypofractionated radiotherapy for the treatment of early breast cancer: a systematic review can be

downloaded from the Cancer Australia website: www.canceraustralia.gov.au.

Acknowledgments

This report was prepared on behalf of National Breast and Ovarian Cancer Centre (NBOCC)* by

Lisa Elliott, Gregory Merlo and Adele Watson of Health Technology Analysts.

NBOCC gratefully acknowledges the contribution of the Hypofractionated Radiotherapy Working

Group, chaired by Associate Professor Boon Chua (see Appendix A).

Funding

Funding for the development of this systematic review was provided by the Australian

Government Department of Health and Ageing.

*On 1 July 2011, National Breast and Ovarian Cancer Centre (NBOCC) amalgamated with Cancer Australia to form a single national agency, Cancer Australia, to provide leadership in cancer control and improve outcomes for Australians affected by cancer.


Hypofractionated radiotherapy for the treatment of early breast cancer:

a systematic review

v

Contents

Executive summary ...................................................................................................................... ix

1 Introduction ........................................................................................................................ 1

2 Method ................................................................................................................................ 3

2.1 Criteria for determining study eligibility ................................................................... 3

2.2 Literature search methodology ............................................................................... 3

2.3 Assessment of study eligibility ................................................................................ 6

2.4 Included trials .......................................................................................................... 7

2.5 Appraisal of included trials ...................................................................................... 8

2.6 Limitations of the review ....................................................................................... 10

3 Description of included studies ..................................................................................... 12

3.1 Systematic reviews ............................................................................................... 12

3.2 Primary studies ..................................................................................................... 13

4 Results of included trials ................................................................................................ 23

4.1 Local recurrence ................................................................................................... 23

4.2 Local-regional recurrence ..................................................................................... 28

4.3 Distant relapse ...................................................................................................... 30

4.4 Overall survival ...................................................................................................... 32

4.5 Adverse events and toxicity .................................................................................. 35

4.6 Cosmetic outcome ................................................................................................ 41

4.7 Quality of life ......................................................................................................... 54

5 Guidelines ........................................................................................................................ 58

5.1 Guidelines search ................................................................................................. 58

5.2 Results .................................................................................................................. 58

6 Conclusions ..................................................................................................................... 63

7 References ....................................................................................................................... 65

Appendix A Contributors ........................................................................................................ 67

vi Hypofractionated radiotherapy for the treatment of early breast cancer:

a systematic review

Tables

Table 1 Summary of key results for local recurrence ................................................................ x

Table 2 Summary of key results for regional recurrence .......................................................... xi

Table 3 Summary of key results for distant relapse .................................................................. xi

Table 4 Summary of key results for overall survival ................................................................. xii

Table 5 Summary of key results for adverse events and toxicity ............................................ xiii

Table 6 Summary of key results for adverse cosmetic outcomes ........................................... xiv

Table 7 Criteria for determining study eligibility ......................................................................... 3

Table 8 Search strategy ............................................................................................................. 5

Table 9 Exclusion criteria ........................................................................................................... 6

Table 10 Included and excluded citations .................................................................................... 7

Table 11 Included systematic reviews ......................................................................................... 8

Table 12 Included RCTs ................................................................................................................. 8

Table 13 NHMRC Dimensions of evidence21

................................................................................. 9

Table 14 NHMRC Interim Levels of Evidence (NHMRC 2009) for evaluating interventions and diagnostic accuracy studies

22 ........................................................... 9

Table 15 Quality criteria for different levels of evidence21

........................................................ 10

Table 16 Reporting biases in systematic reviews23

...................................................................... 11

Table 17 Key characteristics of included studies ....................................................................... 14

Table 18 RMH/GOC trial: Demographic and clinical characteristics of 1410 patients randomised

8 ................................................................................................................ 16

Table 19 Canadian trial: Patient characteristics7 ....................................................................... 17

Table 20 START A: Patient characteristics4 .............................................................................. 19

Table 21 START B: Patient characteristics5 .............................................................................. 21

Table 22 RMH/GOC trial: Survival analysis of local relapse according to fractionation schedule

1 ..................................................................................................................... 23

Table 23 START A: Survival analyses of relapse and mortality according to fractionation schedule (Local relapse)

4 ....................................................................... 27

Table 24 START B: Survival analyses of relapse and mortality according to fractionation schedule (local relapse)

5 ........................................................................ 27

Table 25 Summary of key results for local recurrence .............................................................. 28

Table 26 START A: Survival analyses of relapse and mortality according to fractionation schedule (Local-regional relapse)

4 ......................................................... 29

Table 27 START B: Survival analyses of relapse and mortality according to fractionation schedule (Local-regional Relapse)

5 ....................................................... 29

Table 28 Summary of key results for local-regional recurrence4-5

............................................. 30

Table 29 START A: Survival analyses of relapse and mortality according to fractionation schedule (Distant relapse)

4 ................................................................... 30

Table 30 START B: Survival analyses of relapse and mortality according to fractionation schedule (Distant relapse)

5 .................................................................... 31


a systematic review

vii

Table 31 Summary of key results for distant relapse4-5

............................................................. 31

Table 32 Canadian trial: Cause of deaths2 ................................................................................ 33

Table 33 START A: Survival analyses of relapse and mortality according to fractionation schedule (All-cause mortality)

4 ............................................................... 33

Table 34 START B: Survival analyses of relapse and mortality according to fractionation schedule (All-cause mortality)

5 ............................................................... 34

Table 35 Summary of key results for overall survival ................................................................ 34

Table 36 Canadian trial: Late toxic effects of radiation, assessed according to the RTOG-EORTC late radiation morbidity scoring scheme

a2 ......................................... 35

Table 37 START A: Incidence of ischemic heart disease, symptomatic rib fracture, and symptomatic lung fibrosis according to fractionation schedule

4 ................................. 36

Table 38 START A: Contralateral and other secondary cancers4 ............................................. 36

Table 39 START B: Incidence of ischemic heart disease, symptomatic rib fracture, and symptomatic lung fibrosis according to fractionation schedule

5 ................................. 37

Table 40 START B: Contralateral and other secondary cancers5 ............................................. 37

Table 41 Summary of key results for adverse events and toxicity ............................................ 40

Table 42 RMH/GOC trial: Survival analyses of change in breast appearance and clinical assessments of late radiation effects according to fractionation schedule

8 .................................................................................................................... 44

Table 43 START A AND B: Survival analyses of moderate or marked grade normal tissue effects from patients’ self-assessments, according to fractionation schedule, type of primary surgery

6.............................................................................. 46

Table 44 Canadian trial: Global cosmetic outcome assessed according to the EORTC scale

a2 .......................................................................................................................... 47

Table 45 START A AND B: Survival analyses of moderate or marked grade normal tissue effects from patients’ self-assessments according to fractionation schedule, type of primary surgery

6.............................................................................. 48

Table 46 START A: Mild or marked change in breast appearance4 .......................................... 49

Table 47 START B: Mild or marked change in breast appearance5 .......................................... 50

Table 48 Summary of key results for cosmetic outcomes ......................................................... 53

Table 49 START A AND B: Breast, arm, or shoulder symptoms and body image scale scores at 5 years

a according to radiotherapy regimen, type of primary

surgery6 ....................................................................................................................... 55

Table 50 START A AND B: Breast, arm, or shoulder symptoms and body image scale scores at 5 years

a according to radiotherapy regimen, type of primary

surgery6 ....................................................................................................................... 56

Table 51 START A AND B: Breast, arm, or shoulder symptoms and body image scale scores, according to radiotherapy regimen, over time from randomisation

6 .............. 57

Table 52 Search terms for guidelines websites ......................................................................... 58

viii Hypofractionated radiotherapy for the treatment of early breast cancer:

a systematic review

Figures

Figure 1 RMH/GOC trial: Local ipsilateral relapse in the breast according to fractionation scheudule

1 .............................................................................................. 24

Figure 2 Canadian trial: Kaplan-Meier estimates for local recurrencea2

................................... 25

Figure 3 Canadian trial: Hazard ratios for Ipsilateral recurrence of breast cancer in subgroups of patients

2 ................................................................................................. 26

Figure 4 Canadian trial: Kaplan-Meier estimate for overall survival2 ........................................ 32

Figure 5 START A AND B: Forest plots of normal tissue effects assessed as moderate or marked by patients, according to radiotherapy regimens

6 ..................................... 39

Figure 6 RMH/GOC trial: Probability of any change in breast appearance late radiation effect ten years after radiotherapy by fractionation schedule

8 .................................... 42

Figure 7 RMH/GOC trial: Probability of marked change in breast appearance late radiation effect ten years after radiotherapy by fractionation schedule

8..................... 42

Figure 8 RMH/GOC trial: Probability of palpable breast induration ten years after radiotherapy by fractionation schedule

8 ...................................................................... 43

Figure 9 START A: Kaplan-Meier plot of mild/marked change in breast appearance (photographic) in 1055 patients with breast conserving surgery

4 ............................... 49

Figure 10 START B: Kaplan-Meier plot of mild/marked change in breast appearance (photographic) in 923 patients with breast conserving surgery

5 ................................. 50

Figure 11 START A AND B: Forest plots of normal tissue effects assessed as moderate or marked by patients, according to radiotherapy regimen

6 ....................................... 52


a systematic review

ix

Executive summary

Introduction and methods

This review was commissioned by the National Breast and Ovarian Cancer Centre (NBOCC).

The use of hypofractionated radiotherapy in early breast cancer has been identified as a topic for

evidence review and guideline recommendation development. The following clinical question was

selected as the focus of the systematic literature review:

What are the key outcomes associated with different dose fractionation

(dosage/scheduling) for radiotherapy treatment of early (invasive) breast cancer?

A systematic method of literature searching and selection was employed in the preparation of

this review. Searches were conducted in EMBASE and Medline (via EMBASE.com) and the

Cochrane Database of Systematic Reviews to identify citations published between January 2001

and March 2010. A search of conference websites was also conducted. These were the

American Society of Clinical Oncology, American Society of Radiation Oncology and San

Antonio Breast Cancer Symposium. A total of 682 non-duplicate citations were identified. The

exclusion criteria was applied to all citations, with a total of 10 publications meeting the inclusion

criteria.

Key findings

Local recurrence

All five included trials reported local recurrence (RMH/GOC, Canadian, Spooner, START A and

START B) (see Table 1). There was no evidence that any hypofractionated radiotherapy regimen

was associated with a statistically significant difference in local recurrence rate when compared

with a control arm. The Royal Marsden Hospital/Gloucester Oncology Centre (RMH/GOC) trial

noted a statistically significant difference in recurrence rates when the two hypofractionated

radiotherapy regimens were compared (42.9 Gy vs 39 Gy: 9.6% vs 14.8%, p=0.027), but not

when each regimen was compared to the control arm (50 Gy in 25 fractions).1

Subgroup analyses were performed in the Canadian trial.2 There were no significant differences

in any subgroup, with the exception of tumour grade. The impact of the 42.5 Gy regimen on local

recurrence was less in patients with high-grade tumours compared to patients with low-grade

tumours (p=0.01).2

On 1 July 2011, National Breast and Ovarian Cancer Centre (NBOCC) amalgamated with Cancer Australia to form a single national agency, Cancer Australia, to provide leadership in cancer control and improve outcomes for Australians affected by cancer.

x Hypofractionated radiotherapy for the treatment of early breast cancer:

a systematic review

Table 1 Summary of key results for local recurrence

Study ID Study arms Results

Post breast conserving surgery

RMH/GOC1 39 Gy in 13 fractions over 5 weeks


50 Gy in 25 fractions over 5 weeks*

5 year local recurrence

39 Gy vs 50 Gy: 9.1% vs 7.9%, p=NR

42.9 Gy vs 50 Gy: 7.1% vs 7.9%, p=NR

42.9 Gy vs 39 Gy: 7.1% vs 9.1%, p=NR


39 Gy vs 50 Gy: 14.8% vs 12.1%, p=NS

42.9 Gy vs 50 Gy: 9.6% vs 12.1%, p=NS

42.9 Gy vs 39 Gy: 9.6% vs 14.8%, p=0.027

39 Gy: HR 1.33 (95% CI 0.92, 1.92), p=NS

42.9 Gy: HR 0.86 (95% CI 0.57, 1.30), p=NS

Canadian2 42.5 Gy in 16 fractions over 22 days

50 Gy in 25 fractions over 35 days*

10 year cumulative incidence of local recurrence

42.5 Gy vs 50 Gy: 6.2% vs. 6.7%, p=NS

10 year cumulative incidence of invasive or non-invasive local recurrence 42.5 Gy vs 50 Gy: 7.4% vs. 7.5%, p=NS

Subgroup analyses

Patient age, tumour size, oestrogen-receptor status, tumour grade, systemic therapy, p=NS

High-grade vs low grade tumours, p=0.01

Any surgery

Spooner3

40 Gy in 15 fractions once a day


Delayed salvage treatment

17 year relapse frequency

No difference, data not reported

START A4 39 Gy in 13 fractions over 5 weeks



5 year local relapse rate

50 Gy vs 41.6 Gy vs 39 Gy: 3.2% vs 3.2% vs 4.6%, p=NR

5 year local relapse

39 Gy: HR 1.25 (95% CI 0.74, 2.12), p=0.40

41.6 Gy: HR 1.09 (95% CI 0.64, 1.88), p=0.74

START B5 40 Gy in 15 fractions over 3 weeks



50 Gy vs 40 Gy: 3.3% vs 2.0%, p=NR


40 Gy: HR 0.72 (95% CI 0.43, 1.21), p=0.21

Abbreviations: CI=confidence interval, HR=hazard ratio, NR=not reported, NS=not significant * control arm

Local-regional recurrence

The Standardisation of Breast Radiotherapy Trials A and B (START A and START B) reported

regional recurrence (see Table 2).4-5

There was no evidence that any hypofractionated

radiotherapy regimen was associated with a statistically significant difference in local recurrence

rate when compared with 50 Gy in 25 fractions over 5 weeks (control).


a systematic review

xi

Table 2 Summary of key results for regional recurrence


Any surgery




5 year local-regional relapse rate


5 year local-regional relapse

39 Gy: HR 1.26 (95% CI 0.77, 2.08), p=0.35

41.6 Gy: HR 1.05 (95% CI 0.63, 1.75), p=0.86




50 Gy vs 40 Gy: 3.3% vs 2.2%, p=NR


40 Gy: HR 0.79 (95% CI 0.48, 1.29), p=0.35

Abbreviations: CI=confidence interval; HR=hazard ratio; NR=not reported * control arm

Distant relapse

Two trials reported distant relapse (START A and START B) (see

Table 3).4-5

In START A, there was no statistical difference between either of the

hypofractionated regimen compared with the control arm.4 START B reported that the 40 Gy

study arm had a statistically significantly lower rate of distant relapse when compared with the

control arm (HR 0.69 95% CI 0.53, 0.91, p=0.01).5

Table 3 Summary of key results for distant relapse


Any surgery




5 year distant relapse rate



39 Gy: HR 1.29 (95% CI 0.95, 1.76), p=0.10

41.6 Gy: HR 0.92 (95% CI 0.66, 1.28), p=0.64




50 Gy vs 40 Gy: 10.2% vs 7.6%, p=NR


40 Gy: HR 0.69 (95% CI 0.53, 0.91), p=0.01

Abbreviations: CI=confidence interval; HR=hazard ratio; NR=not reported * control arm

Overall survival

A total of four trials reported overall survival (RMH/GOC, Spooner, START A and START B) (see

Table 4).2-5

Most studies reported that there was no evidence that hypofractionated radiotherapy

was associated with a statistically significantly difference in overall survival. START B found that

40 Gy in 15 fractions over three weeks was associated with a statistically significantly lower all-

cause mortality rate when compared with 50 Gy in 25 fractions over five weeks (HR 0.76 95% CI

xii Hypofractionated radiotherapy for the treatment of early breast cancer:

a systematic review

0.59, 0.98, p=0.03).5 Therefore, there was no evidence that any hypofractionated radiotherapy

regimen was associated with a worse overall survival rate (i.e. the only study that reported a

significant difference showed lower mortality for patients treated with hypofractionated

radiotherapy).

Table 4 Summary of key results for overall survival





10 year survival

42.5 Gy vs 50 Gy: 84.6% vs 84.4%, p=0.79

Any surgery

Spooner3 40 Gy in 15 fractions once a day


17 year survival





5 year all cause mortality

39 Gy: HR 1.00 95% CI 0.74, 1.36, p=0.99

41.6 Gy: HR 1.04 95% CI 0.77, 1.40, p=0.81




40 Gy: HR 0.76 95% CI 0.59, 0.98, p=0.03

Abbreviations: CI=confidence interval, HR=hazard ratio * control arm


A total of three trials reported adverse events and toxicity outcomes (Canadian, START A and

START B) (see Table 5).2, 4-5

Most studies reported that there was no difference in adverse

events and toxicity. Combined results from the START A and START B trials found that a change

in skin appearance occurred significantly less often in the 39 Gy and 40 Gy arms when

compared with the control arm (39 Gy HR 0.63 95% CI 0.47, 0.84, p=0.0019 and 40 Gy HR 0.76

95% CI 0.60, 0.97, p=0.0262).6


a systematic review

xiii

Table 5 Summary of key results for adverse events and toxicity





Late toxic radiation effects, : NS

Any surgery




Ischemic heart disease, symptomatic rib fracture, symptomatic lung fibrosis, contralateral breast cancer, other secondary primary cancers: NS




Combined QoL data from START A and B6

As for START A and START B Tissue effects, arm and shoulder symptoms: NS

Skin appearance: 39 Gy HR 0.63 (95% CI 0.47, 0.84), p=0.0019

40 Gy HR 0.76 (95% CI 0.60, 0.97), p=0.0262

Abbreviations: CI=confidence interval, HR=hazard ratio, NS=not significant * control arm a Assessed 3, 5, and 10 years after randomisation

Cosmetic outcome

A total of four trials reported cosmetic outcome (RMH/GOC, Canadian, START A and START B)

(see Table 6).2, 4-5, 7-8

There was no statistically significant difference in the majority of cosmetic

outcomes assessed by the included publications. RMH/GOC reported that the risk of developing

any late radiation effect was statistically significantly lower for patients in the 39 Gy arm

compared to the 50 Gy arm (p=0.01). For most clinically assessed breast and arm outcomes

estimated at 10 years, compared to the 50 Gy arm, there were fewer events for patients in the 39

Gy arm and more in the 42.9 Gy arm.

The START A trial reported that the 39 Gy arm was associated with significantly less mild or

marked change in photographic breast appearance by photographic assessment (HR 0.69 95%

CI 0.52, 0.91, p=0.01),4 and change in skin appearance by patient self-assessment (HR 0.63

95% CI 0.47, 0.84, p=0.0019).6 The 40 Gy arm of the START B trial was associated with

significantly less change in skin appearance by patient self assessment (40 Gy: HR 0.76 95% CI

0.60, 0.97, p=0.0262).6

In subgroup analyses for the START A and START B trials, the relative effects of the randomised

radiation schedules on patient reported symptoms did not vary significantly according to type of

primary surgery (breast conserving surgery or mastectomy).6

xiv Hypofractionated radiotherapy for the treatment of early breast cancer:

a systematic review

Table 6 Summary of key results for adverse cosmetic outcomes






39 Gy: adverse cosmetic outcomes were reported less frequently when compared to the 50 Gy arm (p=0.01)

42.9 Gy: Cosmetic outcomes were reported more frequently when compared to the 50 Gy arm (p=0.05)

Canadian2, 7 42.5 Gy in 16 fractions over 22 days


No statistically significant differences in any cosmetic outcome

Any surgery




41.6 Gy: No statistically significant differences in any cosmetic outcome

39 Gy: No statistically significant differences in cosmetic outcome, with the exception of mild or marked change in breast appearance (HR 0.69 95% CI 0.52, 0.91, p=0.01)



0.77 (95% CI 0.61-0.98) p=0.02

Combined data from START A and B6

As for START A and START B Change in skin appearance

39 Gy: HR 0.63 (95% CI 0.47, 0.84), p=0.0019

40 Gy: HR 0.76 (95% CI 0.60, 0.97), p=0.0262

Subgroup analysis by breast conserving surgery and mastectomy: NS

Abbreviations: CI=confidence interval, HR=hazard ratio, NS=not significant * control arm

Quality of life

A total of two studies reported quality of life outcomes (START A and START B). 6 There was no

evidence that any hypofractionated radiotherapy regimen was associated with a statistically

significant difference in quality of life score as measured by the BR23 breast symptom subscale.

Subgroup analysis was performed, with results analysed by surgery type. There were no

statistically significant differences in outcomes, nor were any interaction tests significant overall.

Guidelines

In order to identify current recommendations in existing radiotherapy guidelines, a systematic

search of guidelines was undertaken.

The American Society for Radiation Oncology (ASTRO)9 reported that evudebce supports the

equivalence of hypofractionated whole breast irradiation with conventionally fractionated whole

breast irradiation for patients who satisfy all these criteria:

Patient is 50 years or older at diagnosis.

Pathologic stage is T1-2 N0 and patient has been treated with breast-conserving surgery.

Patient has not been treated with systemic chemotherapy.


a systematic review

xv

Within the breast along the central axis, the minimum does is no less than 93% and

maximum dose is no greater than 107% of the prescription does ±7% (as calculated with

2-dimensional treatment planning without heterogeneity corrections)

For patients who do not satisfy all of these criteria, the task force could not reach consensus and

therefore chose not to render a recommendation.9

The New Zealand Ministry of Health guidelines10

made the following recommendation

regarding hypofractionated radiotherapy:

Recommendation

• Radiotherapy treatment for early invasive breast cancer should use an accepted regimen such

as: 50 Gy in 25 fractions over 5 weeks (Grade A†), 45 Gy in 20 fractions over 5 weeks (Grade

B‡), 42.5 Gy in 16 fractions over 3.5 weeks for those with small or medium breasts, not requiring

boost or nodal radiation (Grade B†), 40 Gy in 15 fractions over 3 weeks (Grade B

2)

Good practice points

• If boost radiotherapy is used after a hypofractionated regimen it should be at the standard 2 Gy

per fraction

• Women with large breasts and those with significant postoperative induration, oedema,

erythema, haematoma or infection should be considered for extended fractionation, with smaller

daily doses over 5–6 weeks

The NICE 2009 guidelines11

made the following recommendation regarding dose fractionation:

Recommendation

• Use external beam radiotherapy giving 40 Gy in 15 fractions as standard practice for patients

with early invasive breast cancer after breast conserving surgery or mastectomy.

Qualifying statement: This recommendation is based on RCT evidence of clinical effectiveness

and the guideline development group agreeing that a regimen using fewer fractions would

probably be cost effective.

The Scottish Intercollegiate Guidelines Network (SIGN)12

management of breast cancer in

women guidance paper was developed in 2005, prior to the publication of a number of key RCTs

(such as the START trials). No formal recommendations were made.

† ie, body of evidence can be trusted to guide practice.

‡ ie, body of evidence can be trusted to guide practice in most situations


A systematic literature review of hypofractionated radiotherapy for the treatment of early breast cancer Page 1

1 Introduction

Aim

This review was commissioned by the National Breast and Ovarian Cancer Centre

(NBOCC), Australia’s national authority and source of evidence-based information on

breast and ovarian cancer. In 2001, NBOCC published the second edition of the “Clinical

practice guidelines for the management of early breast cancer”13

which replaced the first

edition released in 1995. NBOCC’s approach to maintaining the currency of these

guidelines is to produce timely topic-specific guideline recommendations in key areas of

changing evidence.

Based on input from a multidisciplinary Steering Committee and additional consultation by

NBOCC, the use of hypofractionated radiotherapy in early breast cancer was identified as a

topic for evidence review and guideline recommendation development. The

Hypofractionated Radiotherapy Working Group selected the following clinical question as

the focus of the systematic literature review:

What are the key outcomes associated with different dose fractionation

(dosage/scheduling) for radiotherapy treatment of early (invasive) breast cancer?

In order to answer this clinical question, a systematic literature search was conducted. The

methods and results are described in detail in the following sections.

Radiotherapy for treatment of early breast cancer

Early breast cancer has been defined as tumours of not more than five centimetres

diameter, with either impalpable or palpable but not fixed lymph nodes and with no

evidence of distant metastases.13

This corresponds to tumours that are T1-2, N0-1, M0 as

currently defined by the International Union Against Cancer (UICC).7

Early breast cancer can be treated with a range of therapies including surgery, radiotherapy

and systematic adjuvant therapy. A recent Cochrane Review noted that over the last three

decades standard management practices have changed.14

Previously, most women with

early breast cancer underwent removal of the whole breast (mastectomy). However,

following a number of clinical studies, breast conserving surgery followed by radiotherapy

has become the recommended option for women with early breast cancer.13-15

Different tissue types (including malignant tissue types) have different sensitivities to

radiation and therefore respond to different radiotherapy fraction sizes. In clinical oncology

a model is used in which the sensitivity to fraction size (measured by the degree of tissue

damage for normal tissues, and tumour recurrence rates for malignant tumours) is

represented by the constants α and β.4 The lower the ratio of these constants (expressed in

Gy), the greater the effect of fraction size on the tissue. The most appropriate radiotherapy

regimen may therefore differ between tissue types. The choice of dose in radiotherapy

must balance the risk of local cancer recurrence against the harmful effects on healthy

tissues.

2 Hypofractionated radiotherapy for the treatment of early breast cancer:

a systematic review

In a “standard” whole breast radiotherapy regime, radiation is delivered over a period of 5

to 6 weeks using a standard 2 Gy radiation dose per fraction, in 25 to 30 treatment

episodes, to a total dose of 50 to 60 Gy.14

Some regimens include an additional boost of

radiation. In hypofractionated radiotherapy, patients receive fewer fractions, however each

fraction contains a larger dose of radiation. Although the dose of each individual fraction is

higher than the conventional regimen, the total dose of radiotherapy is lower. Concerns

have been raised, however, as to whether shorter fractionation schedules have equivalent

outcomes in terms of local tumour control, breast appearance (cosmesis), overall survival,

and patient satisfaction. The concern with larger fraction sizes has been raised as

radiobiological principles state that the fraction size is the dominant factor in determining

late side effects.14

Higher fraction size could lead to increased scarring and retraction of

breast tissue as well as skin atrophy (thinning) and telangiectasia (dilated blood vessels).

Economic considerations

One consequence of the increase in breast conserving surgery and radiotherapy is the

extra demand placed on health services.14, 16-17

Shorter fractionation schedules have the

advantages of using machine and staff time more efficiently and reducing patient

inconvenience. 18

Hypofractionated radiotherapy significantly reduce the amount of time

that patients require radiotherapy equipment. A reduction of time per patient on the

machine of about 50 minutes and 100 minutes for the patients exposed to 35 Gy over two

weeks and those exposed to 27 Gy over one week respectively, compared with those

exposed to 40 Gy in three weeks has been reported.18

A cohort study (n=313) conducted in

Brisbane’s Princess Alexandra Hospital found that, compared to conventional radiotherapy,

hypofractionated radiotherapy was associated with a 26% reduction per patient in the cost

to Medicare *.19

* Dwyer et al 2009 is a conference abstract and little detail is provided of the resource utilization. The estimation for the reduction in Medicare cost appears to be based on the inference from the cohort study results that had the entire cohort received hypofractionated radiotherapy, 288 fractions per month would be ‘saved’ and available for treatment in other cancer patients.


2 Method

2.1 Criteria for determining study eligibility

The criteria for determining study eligibility are shown in Table 7. Publications were eligible

for inclusion in the systematic literature review if they described a randomised controlled

trial (RCT) which recruited women with early (invasive) breast cancer and patients treated

by breast conserving surgery or total mastectomy.

Hypofractionated radiotherapy was defined by the Hypofractionated Radiotherapy Working

Group as ‘giving larger doses of radiotherapy per fraction, but giving fewer fractions

compared with standard radiotherapy’. RCTs must have compared hypofractionated

radiotherapy with either standard radiotherapy, or an alternative regimen of

hypofractionated radiotherapy.

The following outcomes were extracted from the publications: 1) local recurrence, 2) overall

survival, 3) adverse events, 4) toxicity, 5) cosmetic outcome and 6) quality of life (assessed

using a quality of life instrument). Results for subgroups were extracted where available.

The pre-defined subgroups were age, breast size/width, tumour grade, nodal status,

surgical margins, use in conjunction with nodal irradiation, post breast conserving surgery

and total mastectomy.

Table 7 Criteria for determining study eligibility

Study design Randomised, controlled trials

Population Women with early (invasive) breast cancer treated with surgery

Intervention Hypofractionated radiotherapy

Comparator 1. Standard radiotherapy

2. Other regimens of hypofractionated radiotherapy

Outcomes a 1. Local recurrence

2. Overall survival

3. Adverse events

4. Toxicity (including burns and blisters)

5. Cosmetic outcome

6. Quality of life (assessed using a quality of life instrument)

a The predefined subgroups were: age, breast size/width, tumour grade, nodal status, surgical margins, use in conjunction with nodal irradiation, post breast conserving surgery and total mastectomy

2.2 Literature search methodology

A systematic method of literature searching and selection was employed in the preparation

of this review. Searches for full-length publications and abstracts were conducted in

EMBASE and Medline (via EMBASE.com) and the Cochrane Database of Systematic

Reviews. At the request of the NBOCC and the Hypofractionated Radiotherapy Working

Group, searches were restricted to English language studies published from 2001 onwards.

Search terms were approved by the NBOCC prior to searches being conducted. The

reference lists of included papers were reviewed to identify any peer-reviewed evidence


a systematic review

that may have been missed in the literature search. Contacting of authors for unpublished

research was not undertaken. The searches were conducted prior to the 31st of March,

2010. Therefore, studies published after this time were not eligible for inclusion in the

systematic review.

The search strategy for the online bibliographic databases is shown in Table 8. A search of

conference websites was also conducted. Three conferences were selected by the NBOCC

and Hypofractionated Radiotherapy Working Group: American Society of Clinical Oncology,

American Society of Radiation Oncology and San Antonio Breast Cancer Symposium. The

same search terms were used for all conference abstracts. A total of 682 non-duplicate

citations were identified.


Table 8 Search strategy

Database Date searched

# Search terms Citations

EMBASE + Medline

<1950 – 26 Mar 2010

1

'breast cancer'/exp OR 'breast cancer' OR 'breast gland cancer'/exp OR 'breast gland cancer' OR 'breast gland neoplasm'/exp OR 'breast gland neoplasm' OR 'mammary cancer'/exp OR 'mammary cancer' OR 'mammary gland cancer'/exp OR 'mammary gland cancer' OR 'breast neoplasm'

227,592

2 ('breast' OR 'breast'/exp OR breast) AND ('cancer' OR 'cancer'/exp OR cancer OR 'carcinoma' OR 'carcinoma'/exp OR carcinoma OR 'tumour' OR 'tumour'/exp OR tumour OR 'tumour' OR 'tumour'/exp OR tumour OR 'neoplasm'/exp OR neoplasm)

288,233

3 #1 OR #2 288,591

4

'radiotherapy'/exp OR 'radiotherapy' OR 'irradiation therapy'/exp OR 'irradiation therapy' OR 'irradiation treatment'/exp OR 'irradiation treatment' OR 'radiation therapy'/exp OR 'radiation therapy' OR 'radiation treatment'/exp OR 'radiation treatment' OR 'therapeutic radiology'/exp OR 'therapeutic radiology' OR 'radiation; therapy' OR 'treatment, irradiation'/exp OR 'treatment, irradiation'

385,244

5 fractionated OR hypofractionated OR fraction OR fractio* 418,274

6 #4 AND #5 26,442

7 'radiation dose fractionation'/exp OR 'radiation dose fractionation' OR 'dose fractionation'/exp OR 'dose fractionation' OR 'hypofractionated radiotherapy' OR 'fractionated radiotherapy'

10,331

8 #6 OR #7 26,554

9 #3 AND #8 2,275

10 #9 AND [1-1-2001]/sd NOT [9-4-2010]/sd 1,336

11

'clinical trial'/exp OR 'clinical trial' OR 'randomized controlled trial'/exp OR 'randomized controlled trial' OR 'randomised controlled trial'/exp OR 'randomised controlled trial' OR 'randomization' OR 'randomization'/exp OR randomization OR 'randomisation' OR 'randomisation'/exp OR randomisation OR 'meta-analysis'/exp OR 'meta-analysis' OR 'systematic review'/exp OR 'systematic review' OR 'guideline' OR 'single blind procedure'/exp OR 'single blind procedure' OR 'double blind procedure'/exp OR 'double blind procedure' OR 'triple blind procedure'/exp OR 'triple blind procedure' OR 'crossover procedure'/exp OR 'crossover procedure' OR 'placebo' OR 'placebo'/exp OR placebo OR randomi?ed:ab,ti OR rct:ab,ti OR 'random allocation':ab,ti OR 'randomly allocated':ab,ti OR 'allocated randomly':ab,ti OR (allocated NEAR/2 random*):ab,ti OR 'single blind':ab,ti OR 'single blinded':ab,ti OR 'double blind':ab,ti OR 'double blinded':ab,ti OR 'treble blind':ab,ti OR 'treble blinded':ab,ti OR 'triple blind':ab,ti OR 'triple blinded':ab,ti OR placebo*:ab,ti OR 'prospective study'/exp OR 'prospective study' NOT ('case study'/exp OR 'case study' OR 'case report':ab,ti OR 'abstract report'/exp OR 'abstract report' OR 'letter' OR 'letter'/exp OR letter)

1,344,424

12 #11 AND #10 535

Cochrane Library

1800 –30 Mar 2010

1 "breast cancer" or "breast gland cancer" OR "breast gland neoplasm" OR mammary cancer" OR "mammary gland cancer" OR "breast neoplasm"

11,726

2 breast AND (cancer OR carcinoma OR tumour OR tumour OR neoplasm) 13,831

3 radiotherapy OR "irradiation therapy" OR "irradiation treatment" OR "radiation therapy" OR radiology OR "therapeutic radiology"

17,142

4 fractionated OR hypofractionated OR fraction OR fractio* 11,864

5 ( #1 OR #2 ) AND #3 AND #4 211

ASCO 31 Mar 2010

1 “hypofractionated radiotherapy” OR “fractionated radiotherapy” OR “irradiation therapy” OR “irradiation treatment” OR “hypofractionated radiation treatment” OR “fractionated radiation treatment” OR “therapeutic radiology”

1

ASTRO 0

SABCS 3

Manual search 4

Total number of citations 756

Total number of non-duplicate citations 682

Abbreviations: ASCO=American Society of Clinical Oncology, ASTRO=American Society of Radiation Oncology, SABCS=San Antonio Breast Cancer Symposium

http://www3.interscience.wiley.com/cochrane/searchHistory?mode=runquery&qnum=1








a systematic review

2.3 Assessment of study eligibility

Publications identified in the literature search were reviewed and the exclusion criteria

shown in Table 9 applied hierarchically. As the Cochrane search was not restricted by date,

the first exclusion criteria were all publications published prior to 2001 (the cut-off date for

inclusion in the systematic review). Publications were excluded if they were the wrong

study type (not an RCT), if they were in the wrong population (not women with early breast

cancer treated with surgery), evaluated the wrong intervention (not hypofractionated

radiotherapy) or the wrong comparator (not standard radiotherapy or another regimen of

hypofractionated radiotherapy). Publications were excluded if they reported the wrong

outcomes (as described in Table 9). Only English language publications were eligible for

inclusion.

Table 9 Exclusion criteria

Wrong year Study published prior to 2001

Wrong study type Not randomised controlled trials

Wrong population Not in women with early (invasive) breast cancer or patients treated with surgery

Wrong intervention Not hypofractionated radiotherapy

Wrong comparator Not standard radiotherapy or another regimens of hypofractionated radiotherapy

Wrong outcome Study did not report local recurrence, overall survival, adverse events, toxicity, cosmetic outcome, quality of life or a subgroup analysis of age, breast size/width, tumour grade, nodal status, surgical margins, use in conjunction with nodal irradiation, post breast conserving surgery or total mastectomy

Insufficient follow-up Follow-up of less than 5 years

Not in English Not in English

The exclusion criteria was applied to all citations by reviewing the abstract and title, with

665 publications excluded (shown in Table 10). A total of 17 publications remained, and the

full text version of each publication was retrieved and reviewed. The same exclusion criteria

were then applied to the full text articles. A total of 10 publications met the inclusion criteria.

In Taher et al. 2004,20

women with early breast cancer (n=30) were randomised to either i)

50 Gy in 25 fractions, followed by a boost to the tumour bed of 10 Gy in five fractions over

five days or ii) 42.5 Gy in 16 fractions over 22 days with no boost. The RCT was excluded

as the median follow-up was only 1.7 years. No statistical difference in acute skin reactions

or cosmetic outcomes was observed between patients in the treatment arms. The study did

not report local or regional recurrence, survival, or quality of life as outcomes.

Shahid et al. 200918

was excluded because although it included women with early breast

cancer, it was not limited to this population. For inclusion into this RCT, the primary lesion

must have been T2, T3, or T4, with a nodal status of N1, N2, N3, Nx or N0. The study did

not conduct subgroup analyses of women with early breast cancer.

In Shahid et al. 200918

, women were randomised to either i) 27 Gy in five fractions over one

week or ii) 35 Gy in 10 fractions over two weeks or iii) 40 Gy in 15 fractions over three

weeks (the control arm). The percentage of patients with local recurrence was 11%, 12%,

and 10% for those patients in the 27 Gy, 35 Gy, and 40 Gy arms respectively (p=0.91). The

five year overall survival was 87%, 83%, and 82% (p=0.89). Grade 1 skin reactions

occurred significantly more often in the 40 Gy arm compared with the 27 Gy arm and the 35


Gy arm (62% vs 33% and 35%, p<0.05). However, the 40 Gy arm had a significantly lower

incidence of Grade 3 and 4 skin reactions compared with the 27 Gy arm, but not the 35 Gy

arm (14% vs 37% and 28%, p<0.05). No other adverse outcomes had a statistically

significant difference in incidence rates.

Table 10 Included and excluded citations

Exclusion criteria Number

Total citations 682

Citations excluded after review of abstract/title

Wrong year

Wrong study type

Wrong population

Wrong intervention

Wrong comparator

Wrong outcome

Not in English

Total excluded citations

88

186

57

334

0

0

0

665

Full papers reviewed 17

Citations excluded after review of full publication

Wrong study type

Wrong population

Wrong intervention

Wrong comparator

Wrong outcome

Insufficient follow-up

Not in English

Total excluded citations

2

2

1

1

0

1

0

7

Total included citations 10

2.4 Included trials

The literature search identified two systematic literature reviews (shown in Table 11) and

eight publications describing five RCTs (shown in Table 12). Hopwood et al. 20106

described combined results from both START A and START B and is therefore shown

twice in Table 12 (i.e. there are 11 citations in the table but only 10 included publications).

For all included trials (with the exception of Spooner 20083) there were at least two

publications describing the results of the same clinical trial. In almost all instances, the first

paper describes interim results (e.g. 5 years of follow-up) whereas the second paper

describes the full study results (e.g. 10 years of follow-up). Throughout the report data has

been taken from the most recent publication. In rare instances where this was not the case

(e.g. the earlier paper reporting different outcome to the more recent paper) this has clearly

been stated. Hopwood et al 2010 reported combined results from both the START A and

START B trials and therefore contained data not reported in the individual START A and

START B publications.6 This has been clearly referenced throughout the report.


a systematic review

Table 11 Included systematic reviews

Study ID Citation

James 2008 James ML, Lehman M, Hider PN, Jeffery M, Francis DP, Hickey BE. Fraction size in radiation treatment for breast conservation in early breast cancer. Coch Data Syst Rev 2008;(3).

Kalogeridi 2009 Kalogeridi MA, Kelekis N, Kouvaris J, Platoni K, Kyrias G, Pectasides D et al. Accelerated hypofractionated radiotherapy schedules in breast cancer: A review of the current literature. Rev Recent Clin Trials 2009; 4(3):147-151.

Table 12 Included RCTs

Study ID Citations

Royal Marsden Hospital/Gloucester Oncology Centre (RMH/GOC)

Yarnold J, Ashton A, Bliss J, Homewood J, Harper C, Hanson J et al. Fractionation sensitivity and dose response of late adverse effects in the breast after radiotherapy for early breast cancer: Long-term results of a randomised trial. Radiother Oncol 2005; 75(1):9-17.

Owen JR, Ashton A, Bliss JM, Homewood J, Harper C, Hanson J et al. Effect of radiotherapy fraction size on tumour control in patients with early-stage breast cancer after local tumour excision: long-term results of a randomised trial. Lancet Oncology 2006; 7(6):467-471.

Spooner (abstract only)

Spooner D, Stocken DD, Jordan S, Bathers S, Dunn JA, Jevons C, Morrison M, Oates G, Grieve R. A randomised controlled trial to evaluate both the role and optimal fractionation of radiotherapy in the conservative management of early breast cancer. San Antonio Breast Cancer Symposium, 2008.

Standardisation of Breast Radiotherapy Trial A (START A)

-START-Trialists'-Group, Bentzen SM, Agrawal RK, Aird EG, Barrett JM, Barrett-Lee PJ et al. The UK Standardisation of Breast Radiotherapy (START) Trial A of radiotherapy hypofractionation for treatment of early breast cancer: a randomised trial. Lancet Oncol 2008; 9:331-341.

Hopwood P, Haviland JS, Sumo G, Mills J, Bliss JM, Yarnold JR. Comparison of patient-reported breast, arm, and shoulder symptoms and body image after radiotherapy for early breast cancer: 5-year follow-up in the randomised Standardisation of Breast Radiotherapy (START) trials. Lancet Oncol 2010; 11(3):231-240.

Standardisation of Breast Radiotherapy Trial B (START B)

-START-Trialists'-Group, Bentzen SM, Agrawal RK, Aird EG, Barrett JM, Barrett-Lee PJ et al. The UK Standardisation of Breast Radiotherapy (START) Trial B of radiotherapy hypofractionation for treatment of early breast cancer: a randomised trial. Lancet 2008; 371:1098-1107.

Hopwood P, Haviland JS, Sumo G, Mills J, Bliss JM, Yarnold JR. Comparison of patient-reported breast, arm, and shoulder symptoms and body image after radiotherapy for early breast cancer: 5-year follow-up in the randomised Standardisation of Breast Radiotherapy (START) trials. Lancet Oncol 2010; 11(3):231-240.

Canadian Whelan T, MacKenzie R, Julian J, Levine M, Shelley W, Grimard L et al. Randomized trial of breast irradiation schedules after lumpectomy for women with lymph node-negative breast cancer. J Natl Inst 2002; 94(15):1143-1150.

Whelan TJ, Pignol JP, Levine MN, Julian JA, MacKenzie R, Parpia S et al. Long-term results of hypofractionated radiation therapy for breast cancer. New Engl J Med 2010; 362(6):513-520.

2.5 Appraisal of included trials

Dimensions of evidence

The aim of this review was to find the highest quality evidence to answer the clinical

questions being asked. In accordance with National Health and Medical Research Council

(NHMRC) guidance, the following dimensions of evidence were reviewed for each of the


included studies (shown in Table 13). It is important to recognise that the value of a piece

of evidence is determined by all of these dimensions, not just the level of evidence.

Table 13 NHMRC Dimensions of evidence21

Dimension Reviewers definition

Strength of the evidence

Level

Quality

Statistical precision

The study design used, as an indication of the degree to which bias has been eliminated by the design alone. The levels reflect the effectiveness of the study design to answer the research question.

The methods used to minimise bias within an individual study (i.e., other than design per se)

An indication of the precision of the estimate of effect reflecting the degree of certainty about the existence of a true effect, as opposed to an effect due to chance

Size of effect Determines the magnitude of effect and whether this is of clinical importance

Relevance of evidence Considers the relevance of the study to the specific research question and the context in which the information is likely to be applied, with regard to a) the nature of the intervention, b) the nature of the population and c) the definition of the outcomes.

Each study was also assigned a level of evidence in accordance with the NHMRC (2009)

interim levels of evidence (see Table 14).22

The highest level of evidence available is a

systematic review of RCTs, which is considered the study type least subject to bias.

Individual RCTs also represent high-level evidence. Only systematic reviews and RCTs

were eligible for inclusion in this systematic review.

Table 14 NHMRC Interim Levels of Evidence (NHMRC 2009) for evaluating interventions and diagnostic accuracy studies22

Level Intervention

I * A systematic review of level II studies

II A randomised controlled trial

III-1 A pseudorandomised controlled trial

(i.e., alternate allocation or some other method)

III-2 A comparative study with concurrent controls:

• Non-randomised, experimental trial a

• Cohort study

• Case-control study

• Interrupted time series with a control group

III-3 A comparative study without concurrent controls:

• Historical control study

• Two or more single arm studies b

• Interrupted time series without a parallel control group

IV Case series with either post-test or pre-test/post-test outcomes

Source: National Health and Medical Research Council (2009)22 Note: When a level of evidence is attributed in the text of a document, it should also be framed according to its corresponding research question e.g., level II intervention evidence; level IV diagnostic evidence; level III-2 prognostic evidence. * A systematic review will only be assigned a level of evidence as high as the studies it contains, excepting where those studies are of level II evidence. a This also includes controlled before-and-after (pre-test/post-test) studies, as well as indirect comparisons (i.e., utilise A vs B and B vs C, to determine A vs C). b Comparing single arm studies i.e., case series from two studies.


a systematic review

Even within the levels of evidence stated above, there is considerable variability in the

quality of evidence. In accordance with NHMRC guidelines, it was necessary to consider

the quality of each of the included studies. Quality assessment was based on criteria

reported by the NHMRC (2000), as shown in Table 15, with studies rated as good, fair or

poor quality.21

Table 15 Quality criteria for different levels of evidence21

Study type Quality criteria

Systematic review

Was an adequate search strategy used?

Were the inclusion criteria appropriate and applied in an unbiased way?

Was a quality assessment of included studies undertaken?

Were the characteristics and results of the individual studies appropriately summarised?

Were the methods for pooling the data appropriate?

Were sources of heterogeneity explored?

Randomised controlled trials

Was allocation to treatment groups concealed from those responsible for recruiting patients?

Was the study double-blinded?

Where outcomes assessors blinded to treatment allocation?

Were all randomised participants included in the analysis?

Were treatment groups well matched at baseline?

Was the study powered to detect a difference in primary outcome?

Source: Adapted from NHMRC (2000)21

Data synthesis

In addition to the level and quality of evidence of individual studies, the review considered

the body of evidence in total. This involved consideration of the volume of evidence and its

consistency.

The review presented the statistical precision of the estimated effect size, together with a

discussion of its clinical significance. Finally, the review considered the relevance of the

evidence, both with regard to the generalisablity of the patient population and the

intervention, as well as the applicability to the Australian health care setting.

2.6 Limitations of the review

This review used a structured approach to review the literature. However, there are some

inherent limitations with this approach. All types of study are subject to bias, with

systematic reviews, such as the one conducted here, being subject to the same biases

seen in the original studies they include, as well as biases specifically related to the

systematic review process. Reporting biases are a particular problem related to systematic

reviews and include publication bias, time-lag bias, multiple publication bias, language bias

and outcome reporting bias. A brief summary of the different types of reporting bias is

shown in Table 16. Other biases can result if the methodology to be used in a review is not

defined a priori (i.e., before the review commences). Detailed knowledge of studies

performed in the area of interest may influence the eligibility criteria for inclusion of studies

in the review and may therefore result in biased results. For example, studies with more

positive results may be preferentially included in a review, thus biasing the results and

overestimating treatment effect.


Table 16 Reporting biases in systematic reviews23

Type of bias Definition and effect on results of review

Publication bias The publication or non-publication of research findings.

Small, negative trials tend not to be published and this may lead to an overestimate of results of a review if only published studies are included.

Time-lag bias The rapid or delayed publication of research findings.

Studies with positive results tend to be published sooner than studies with negative findings and hence results may be overestimated until the negative studies ‘catch up’.

Multiple publication bias The multiple or singular publication of research findings.

Studies with significant results tend to be published multiple times which increases the chance of duplication of the same data and may bias the results of a review.

Citation bias The citation or non-citation of research.

Citing of trials in publications is not objective so retrieving studies using this method alone may result in biased results. Unsupported studies tend to be cited often which may also bias results.

Language bias The publication of research findings in a particular language.

Significant results are more likely to be published in English so a search limited to English-language journals may result in an overestimation of effect.

Outcome reporting bias The selective reporting of some outcomes but not others.

Outcomes with favourable findings may be reported more. For example, adverse events have been found to be reported more often in unpublished studies. This may result in more favourable results for published studies.

Source: Adapted from Egger et al. (2001).23

Some of these biases are potentially present in this review. The search was limited to

English-language publications only, so language bias is a potential problem. Outcome

reporting bias and inclusion criteria bias are unlikely as the reviewers had no detailed

knowledge of the topic literature, and the methodology used in the review and the scope of

the review were defined a priori.

The majority of studies included in this review were conducted outside Australia, and

therefore, their generalisablity to the Australian population and context needs to be

considered. This review was confined to an examination of the efficacy and safety of the

interventions and did not consider ethical or legal considerations associated with those

interventions.

The studies were initially selected by examining the abstracts of these articles. Therefore, it

is possible that some studies were inappropriately excluded prior to examination of the full

text article. However, where detail was lacking, ambiguous papers were retrieved as full

text to minimise this possibility. Data extraction, critical appraisal and report preparation

was performed by one reviewer and double-checked by another. The review was

conducted over a limited timeframe (March 2010 – June 2010).


a systematic review

3 Description of included studies

3.1 Systematic reviews

The literature search identified two systematic literature reviews.

3.1.1 James 2008

This Cochrane Review aimed to assess the effects of altered fraction size on women with

early breast cancer who have undergone breast conserving surgery. The Cochrane Breast

Cancer Group Specialised Register, MEDLINE, EMBASE, reference lists for articles and

relevant conference proceeding were searched. The inclusion criteria were randomised

controlled trials of unconventional versus conventional fractionation in women with early

breast cancer who had undergone breast conserving surgery. Two trials were included:

RMH/GOC1, 8

and Canadian.2, 7

Both trials were identified in the literature search conducted

for this systematic literature review, and have been discussed in detail in the following

sections.

There were no significant differences between the fractionation regimens in regard to

cosmesis, late skin toxicity, or late radiation toxicity. For overall survival, there was no

significant difference between the regimens. No data were available for costs, quality of life,

or women’s preference. The publication acknowledged the limitations related to

assessment of subjective outcomes, such as cosmesis and breast induration.

Although both trials independently showed no difference in local control with altered

fractionation, the reporting did not allow combination of data. The findings of this review

provided reassurance that the practice of offering shortened radiation fractionation regimes

to carefully selected groups of patients is unlikely to be detrimental in terms of breast

appearance, late radiation breast toxicity, or survival.

3.1.2 Kalogeridi 2009

The authors conducted a literature search using MEDLINE, although the search strategy

was not reported in the publication. Eight non-randomised trial and four randomised trials

(RMH/GOC1, 8

, Canadian2, 7

, START A4 and START B

5) were identified. All randomised

studies were identified in the literature search conducted for this systematic literature

review, and have been discussed in detail in the following sections.

The publication concluded that there are some significant concerns when short

fractionation schedules are used for breast radiotherapy. Particularly, the authors

concluded that a large dose per fraction could increase late normal tissue toxicity.

However, case series, cohort studies and recently published randomised trials support the

idea of hypofractionation for breast cancer, giving comparable control rates with

conventional fractionation and acceptable toxicity. The authors cautioned that at least 15-

20 years of follow-up was needed to assess the long-term sequelae and confirm the safety

of the hypofractionated regimens.


3.2 Primary studies

The literature search identified five RCTs which met the inclusion criteria. The key

characteristics are shown in Table 12. Two trials were in patients who had undergone

breast conserving surgery (RMH/GOC and Canadian trials)1-2, 7-8

and three were in patients

who had undergone any form of surgery (START A, START B and Spooner trials)3-5

. All

studies were considered fair quality, with the exception of Spooner which was considered

poor*. Spooner 2008 is a conference abstract; no full publication of this abstract was

identified in this systematic review conducted for this technical report.3 The studies

evaluated a range of hypofractionated radiotherapy regimens.

The following section includes further discussion of each study design and patient

population.

* The conference abstract provided insufficient study detail to rate it as either fair or good. It is unclear from the abstract whether allocation was concealed from those responsible for recruiting subjects, whether outcome assessement was blinded, and whether there was loss to follow up,


a systematic review

Table 17 Key characteristics of included studies

Study ID Study type

Quality Population, median follow-up

Country Intervention Comparator Outcomes


RMH/GOC1, 8 RCT

Fair

T1-3, N0-1, M0, <75years N=1,410

9.7 years (range 7.8-11.8 years)

UK

39 Gy in 13 fractions over 5 weeks (N=474)

42.9 Gy in 13 fractions over 5 weeks (N=466)


Local recurrence

Cosmetic outcomes

Canadian2, 7 RCT

Fair

Invasive carcinoma with negative axillary nodes,

N=1,234

12 years (range not reported)

Canada

42.5 Gy in 16 fractions over 22 days (N=622)

50 Gy in 25 fractions over 35 days (N=612)

Local recurrence (including subgroup analysis)

Overall survival


Cosmetic outcome

Post breast conserving surgery or mastectomy

START A4, 6 RCT

Fair

T1-3a, N0-1, M0

N=2,236

5.1 years (range 4.4-6.0)

UK


41.6 Gy in 13 fractions over 5 weeks (N=750)


Local recurrence

Overall survival


Cosmetic outcome (including subgroup analysis)

Quality of life (including subgroup analysis)

START B5, 6 RCT

Fair

T1-3a, N0-1, M0

N=2,215

6 years (range 5.0-6.2)

UK

40 Gy in 15 fractions over 3 weeks (N=1,110)

50 Gy in 25 fractions over 5 weeks (N=1,105)

Local recurrence

Overall survival


Cosmetic outcome (including subgroup analysis)

Quality of life (including subgroup analysis)

Spooner3 RCT (conference abstract)

Poora

Stage 1 or 2, median tumour size 2cm N=707

16.9 years (range 15.4-18.8 years)

UK

40 Gy in 15 daily fractions (N=NR)

50 Gy in 25 daily fractions (N=NR)


Time to first relapse

Abbreviations: NR=not reported, RCT=randomised controlled trial a The conference abstract provided insufficient study detail to rate it as either fair or good. It is unclear from the abstract whether allocation was concealed from those responsible for recruiting subjects, whether outcome assessment was blinded, and whether there was loss to follow up,

Hypofractionated radiotherapy for the treatment of early breast cancer: a systematic review Page 15

3.2.1 Royal Marsden Hospital/Gloucester Oncology Centre trial (RMH/GOC)1, 8

Patients with operable invasive breast cancer (T1–3, N0–1, M0) who needed radiotherapy were

eligible for the trial. Inclusion criteria were <75 years of age, completed breast conserving

surgery and complete macroscopic resection of invasive carcinoma. The study was conducted in

the UK.

Patients (N=1,410) were randomised to either i) 50 Gy in 25 fractions (control group), or ii) 39 Gy

in 13 fractions or iii) 42.9 Gy in 13 fractions (experimental schedules). All regimens were

administered over five weeks. Patients with a complete macroscopic resection who were judged

eligible by the clinician were further randomly allocated to receive a tumour bed boost or no

boost. This process ended in July 1997, and all patients were offered an elective boost

thereafter. The proportion of patients who received a boost was similar in all three treatment

groups: 348 (74%) patients for 50 Gy, 348 (75%) for 42.9 Gy, and 351 (74%) for 39 Gy.

The trial was powered to detect a difference in tumour recurrence between each study arm, with

the aim to enrol 2,250 subjects. However, recruitment was stopped prior to this number being

reached as this trial was superseded by the START trials. Recruitment occurred between 1986

and 1998. The median follow-up period was 9.7 years, with a maximum of 18.4 years of follow-

up. The publications did not report baseline characteristics in each study arm. Patient

characteristics are shown in Table 18.


a systematic review

Table 18 RMH/GOC trial: Demographic and clinical characteristics of 1410 patients randomised8

Patient characteristic Number (%)

Age at randomisation

20-29 9 (0.9)

30-39 98 (7.0)

40-49 316 (22.4)

50-59 503 (35.7)

60-69 425 (30.1)

70-79 59 (4.2)

Breast size (from photographs)

Small 186 (13.2)

Medium 952 (67.5)

Large 203 (14.4)

Not known 69 (4.9)

Surgical deficit (from photographs)

Small 845 (59.9)

Medium 415 (29.4)

Large 76 (5.4)

Not known 74 (5.2)

cT stage

T0 59 (4.2)

T1 749 (53.1)

T2 575 (40.8)

T3 22 (1.6)

T4 2 (0.1)

TX 3 (0.2)

cN stage

N0 1,187 (84.7)

N1 219 (15.5)

N2 3 (0.2)

NX 1 (0.1)

Number of nodes pathologically involved

0 564 (67.3)

1-3 202 (24.1)

4+ 72 (8.6)

No axillary surgery 572 (40.6)

Adjuvant treatment

None 289 (20.5)

Tamoxifen only 918 (65.1)

Chemotherapy only 40 (2.8)

Tamoxifen and chemotherapy 156 (11.1)

Other 7 (0.5)

Axillary supraclavicular fossa treatment

None 337 (23.9)

Axillary supraclavicular fossa radiotherapy, no axillary surgery 231 (16.4)

Surgery, no radiotherapy 782 (55.5)

Surgery and supraclavicular fossa radiotherapy 59 (4.2)

Not known 1 (0.1)

Breast boost

Randomised to no boost 359 (25.5)

Randomised to boost 364 (25.8)

Non-randomised boost given 687 (48.7)

Source: Yarnold 20058, Table 1 page 12


3.2.2 Canadian trial2, 7

Patients with invasive carcinoma of the breast treated by breast conserving surgery and axillary

dissection with pathologically negative axillary lymph nodes were eligible for inclusion in the trial.

Key exclusion criteria were invasive disease or ductal carcinoma in situ involving the margins of

excision, tumours that were larger than 5 cm in diameter, and a breast width of more than 25 cm

at the posterior border of the medial and lateral tangential beams. The study was conducted in

Canada.

Patients (N=1,234) were randomised to either i) 50 Gy in 25 fractions over 35 days or ii) 42.5 Gy

in 16 fractions over 22 days. The 50 Gy arm was considered the control arm.

The study was designed to assess the non-inferiority of the hypofractionated regimen relative to

the standard schedule for radiation therapy in terms of local recurrence. Recruitment occurred

between 1993 and 1996. The study had a median follow-up of 12 years. The authors noted that

the patients in each treatment arm were reasonably comparable with respect to their baseline

characteristics. This is shown in Table 19.

Table 19 Canadian trial: Patient characteristics7

Characteristic Number (%)

42.5 Gy N=622

Number (%)

50 Gy N=612

Age

<50 y 157 (25) 148 (24)

50–59 y 186 (30) 155 (25)

60–69 y 181 (29) 200 (33)

≥70 y 98 (16) 109 (18)

Tumour size

≤1 cm 183 (29) 192 (31)

>1–2 cm 317 (51) 302 (49)

>2 cm 122 (20) 118 (19)

Tumour grade

I 215 (35) 209 (34)

II 244 (39) 236 (39)

III 117 (19) 116 (19)

Unknown 46 (7) 51 (8)

Oestrogen receptor status

Positive 440 (71) 434 (71)

Negative 165 (27) 157 (26)

Unknown 17 (3) 21 (3)

Systemic therapy

None 298 (48) 295 (48)

Tamoxifen 254 (41) 251 (41)

Chemotherapy 70 (11) 66 (11)

Source: Whelan 20027 Table 1 page 1146


a systematic review

3.2.3 Standardisation of Breast Radiotherapy Trial A (START A)4

The START A and START B trials were conducted in parallel.4-5

Centres in the UK could elect to

enter either Trial A (17 centres) or Trial B (23 centres). Due to earlier completion of recruitment in

Trial B, those centres were invited to join Trial A after accrual to Trial B was complete. The study

was conducted in the UK.

Women with operable invasive breast cancer (T1-3a, N0-1 M0) requiring radiotherapy after

surgery (breast-conserving surgery or mastectomy, with clear tumour margins ≥1 mm) were

eligible for the trial if they were aged over 18 years and did not have an immediate surgical

reconstruction.

Patients (N=2,236) were randomised to either i) 50 Gy in 25 fractions (control group) or ii) 41.6

Gy in 13 fractions or iii) 39 Gy in 13 fractions (experimental schedules). All regimens were

administered over five weeks to eliminate treatment time as a variable.

The trial was powered to detect a difference in local-regional tumour relapse between each 13

fraction schedule and the control group. Recruitment occurred between 1998 and 2002. The

median follow-up period was 5.1 years, with a maximum of 8 years of follow-up. Demographic

and clinical characteristics at randomisation were well balanced between treatment groups. This

is shown in Table 20.


Table 20 START A: Patient characteristics4

Characteristic

Fractionation schedule

50 Gy n=749 (%)

41.6 Gy n=750 (%)

39 Gy n=737 (%)

Total N=2,236 (%)

Age years

20–29 5 (0.7) 4 (0.5) 3 (0.4) 12 (0.5)

30–39 38 (5.1) 40 (5.3) 38 (5.2) 116 (5.2)

40–49 116 (15.5) 136 (18.1) 129 (17.5) 381 (17.0)

50–59 280 (37.4) 283 (37.7) 286 (38.8) 849 (38.0)

60–69 215 (28.7) 192 (25.6) 194 (26.3) 601 (26.9)

70–79 87 (11.6) 85 (11.3) 78 (10.6) 250 (11.2)

80– 8 (1.1) 10 (1.3) 9 (1.2) 27 (1.2)

Mean (SD) 57.6 (10.5) 57.0 (10.7) 57.1 (10.5) 57.2 (10.6)

Time from surgery to randomisation (weeks)

Median time (IQR) 8.8 (5.3–20.8) 9.4 (5.9–20.2) 9.3 (5.4–21.1) 9.1 (5.4–20.7)

Range 0.4–71.3 1.0–50.3 1.1–53.6 0.4–71.3

Primary surgery

Breast conserving surgery 631 (84.2) 641 (85.5) 628 (85.2) 1900 (85.0)

Mastectomy 118 (15.8) 109 (14.5) 109 (14.8) 336 (15.0)

Histological type

Invasive ductal 581 (77.6) 585 (78.0) 584 (79.2) 1750 (78.3)

Invasive lobular 88 (11.7) 95 (12.7) 83 (11.3) 266 (11.9)

Mixed ductal/lobular 21 (2.8) 17 (2.3) 17 (2.3) 55 (2.5)

Other 57 (7.6) 51 (6.8) 52 (7.1) 160 (7.2)

Not known 2 (0.3) 2 (0.3) 1 (0.1) 5 (0.2)

Pathological node status

Positive 222 (29.6) 197 (26.3) 224 (30.4) 643 (28.8)

Negative 514 (68.6) 536 (71.5) 497 (67.4) 1547 (69.2)

Not known (no axillary surgery) 12 (1.6) 17 (2.3) 15 (2.0) 44 (2.0)

Not known (missing data) 1 (0.1) 0 (0.0) 1 (0.2) 2 (0.1)

Tumour size (cm)

<1 24 (3.2) 26 (3.5) 24 (3.3) 74 (3.3)

1– 362 (48.3) 347 (46.3) 355 (48.2) 1064 (47.6)

2– 202 (27.0) 203 (27.1) 198 (26.9) 603 (27.0)

3– 156 (20.8) 169 (22.5) 157 (21.3) 482 (21.6)

Not known 5 (0.7) 5 (0.7) 3 (0.3) 13 (0.6)

Tumour grade

1 157 (21.0) 150 (20.0) 149 (20.2) 456 (20.4)

2 369 (49.3) 379 (50.5) 368 (49.9) 1116 (49.9)

3 212 (28.3) 207 (27.6) 210 (28.5) 629 (28.1)

Not known (not applicable) a 11 (1.5) 10 (1.3) 6 (0.8) 27 (1.2)

Not known 0 (0.0) 4 (0.6) 4 (0.5) 8 (0.4)

Adjuvant therapy

None 52 (6.9) 53 (7.1) 67 (9.1) 172 (7.7)

Tamoxifen/no chemotherapy 416 (55.5) 418 (55.7) 376 (51.0) 1210 (54.1)

Chemotherapy/no tamoxifen 86 (11.5) 77 (10.3) 82 (11.1) 245 (11.0)

Tamoxifen + chemotherapy 173 (23.1) 187 (25.0) 188 (25.5) 548 (24.5)


a systematic review

Characteristic


50 Gy n=749 (%)

41.6 Gy n=750 (%)

39 Gy n=737 (%)

Total N=2,236 (%)

Other endocrine therapy b 17 (2.3) 13 (1.7) 17 (2.3) 47 (2.1)

Not known 5 ( 0.7) 2 ( 0.2) 7 ( 0.9) 14 ( 0.6)

Lymphatic treatment

None 8 (1.1) 14 (1.9) 13 (1.8) 35 (1.6)

Surgery/no radiotherapy 610 (81.4) 636 (84.8) 620 (84.1) 1866 (83.5)

Radiotherapy/no surgery 3 (0.4) 4 (0.5) 2 (0.3) 9 (0.4)

Surgery + radiotherapy 119 (15.9) 95 (12.7) 95 (12.9) 309 (13.8)

Not known 9 (1.2) 1 (0.1) 7 (0.9) 17 (0.8)

Boost (breast conserving surgery patients only)

Number n=631 n=641 n=628 n=1,900

Yes 381 (60.4) 391 (61.0) 380 (60.5) 1152 (60.6)

No 242 (38.3) 249 (38.8) 241 (38.4) 732 (38.5)

Not known 8 (1.3) 1 (0.2) 7 (1.1) 16 (0.8)

From Baseline photographs

Number n=413 n=421 n=416 n=1250

Breast size

Small 43 (10.4) 47 (11.2) 41 (9.9) 131 (10.5)

Medium 294 (71.2) 324 (77.0) 322 (77.4) 940 (75.2)

Large 76 (18.4) 50 (11.9) 53 (12.7) 179 (14.3)

Surgical deficit

Small 232 (56.2) 235 (55.8) 249 (59.9) 716 (57.3)

Medium 142 (34.4) 146 (34.7) 132 (31.7) 420 (33.6)

Large 39 (9.4) 40 (9.5) 35 (8.4) 114 (9.1)

Source: START A4 Table 1 page 332 Abbreviations: IQR=Interquartile range, SD=Standard deviation a Lobular and other histological types. b Other endocrine therapies include combinations of tamoxifen/anastrozole/letrozole/exemestane/goserelin, mostly within randomised trials.

3.2.4 Standardisation of Breast Radiotherapy Trial B (START B)5

As noted above, the START A and START B trials were conducted in parallel. Women with

operable invasive breast cancer (T1-3a, N0-1, M0) requiring radiotherapy after surgery (breast-

conserving surgery or mastectomy, with clear tumour margins ≥1 mm) were eligible for the trial if

they were aged over 18 years and did not have an immediate surgical reconstruction.

Patients (N=2,215) were randomised to either i) 50 Gy in 25 fractions over five weeks or ii) 40 Gy

in 15 fractions over three weeks. The 50 Gy arm was considered the control arm. The trial was

powered to detect a difference in local-regional tumour relapse between each study arm.

Recruitment occurred between 1998 and 2001. The median follow-up period was 6.0 years, with

a maximum of 8 years of follow-up. Demographic and clinical characteristics at randomisation

were well balanced between treatment groups. This is shown in Table 21.


Table 21 START B: Patient characteristics5

Characteristic


50 Gy n=1,105 (%)

40 Gy n=1,110 (%)

Total N=2,215 (%)

Age years

20–29 7 (0.6) 0 (0.0) 7 (0.3)

30–39 62 (5.6) 39 (3.5) 101 (4.6)

40–49 179 (16.2) 170 (15.3) 349 (15.8)

50–59 427 (38.6) 447 (40.3) 874 (39.5)

60–69 304 (27.5) 327 (29.5) 631 (28.5)

70–79 117 (10.6) 119 (10.7) 236 (10.7)

80– 9 (0.8) 8 (0.7) 17 (0.8)

Mean (SD) 57.0 (10.4) 57.8 (9.5) 57.4 (10.0)

Time from surgery to randomisation (weeks)

Median time (IQR) 7.3 (4.9–12.3) 7.1 (4.9–11.9) 7.3 (4.9–12.0)

Range 0.9–45.3 0.6–49.3 0.6–49.3

Primary surgery

Breast conserving surgery 1020 (92.3) 1018 (91.7) 2038 (92.0)

Mastectomy 85 (7.7) 92 (8.3) 177 (8.0)

Histological type

Invasive ductal 865 (78.3) 843 (75.9) 1708 (77.1)

Invasive lobular 122 (11.0) 132 (11.9) 254 (11.5)

Mixed ductal/lobular 20 (1.8) 25 (2.3) 45 (2.0)

Other 95 (8.6) 103 (9.3) 198 (8.9)

Not known 3 (0.3) 7 (0.6) 10 (0.5)

Pathological node status

Positive 238 (21.5) 266 (24.0) 504 (22.8)

Negative 831 (75.2) 804 (72.4) 1635 (73.8)

Not known (no axillary surgery) 36 (3.3) 39 (3.5) 75 (3.4)

Not known (missing data) 0 (0.0) 1 (0.1) 1 (0.04)

Tumour size (cm)

<1 151 (13.7) 167 (15.0) 318 (14.4)

1– 552 (50.0) 542 (48.8) 1094 (49.4)

2– 287 (26.0) 288 (25.9) 575 (26.0)

3– 113 (10.2) 107 (9.6) 220 (9.9)

Not known 2 (0.2) 6 (0.5) 8 (0.4)

Tumour grade

1 306 (27.7) 311 (28.0) 617 (27.9)

2 518 (46.9) 532 (47.9) 1050 (47.4)

3 261 (23.6) 248 (22.3) 509 (23.0)

Not known (not applicable) a 15 (1.4) 15 (1.3) 30 (1.3)

Not known 5 (0.4) 4 (0.4) 9 (0.4)

Adjuvant therapy

None 37 (3.3) 47 (4.2) 84 (3.8)

Tamoxifen/no chemotherapy 782 (70.8) 810 (73.0) 1592 (71.9)

Chemotherapy/no tamoxifen 77 (7.0) 78 (7.0) 155 (7.0)

Tamoxifen + chemotherapy 181 (16.4) 155 (14.0) 336 (15.2)


a systematic review

Characteristic


50 Gy n=1,105 (%)

40 Gy n=1,110 (%)

Total N=2,215 (%)

Other endocrine therapy b 16 (1.4) 11 (1.0) 27 (1.2)

Not known 12 (1.1) 9 (0.8) 21 (0.9)

Lymphatic treatment

None 32 (2.9) 36 ( 3.2) 68 (3.1)

Surgery/no radiotherapy 980 (88.7) 984 (88.6) 1964 (88.7)

Radiotherapy/no surgery 5 (0.4) 3 (0.3) 8 (0.4)

Surgery + radiotherapy 74 (6.7) 79 (7.1) 153 (6.9)

Not known 14 (1.3) 8 (0.7) 22 (1.0)

Boost (breast conserving surgery patients only)

Number n=1020 n=1018 n=2038

Yes 422 (41.4) 446 (43.8) 868 (42.6)

No 584 (57.3) 565 (55.5) 1149 (56.4)

Not known 14 (1.4) 7 (0.7) 21 (1.0)

From Baseline photographs

Number n=522 n=514 n=1036

Breast size

Small 49 (9.4) 42 (8.2) 91 (8.8)

Medium 377 (72.2) 390 (75.9) 767 (74.0)

Large 96 (18.4) 82 (16.0) 178 (17.2)

Surgical deficit

Small 307 (58.8) 286 (55.6) 593 (57.2)

Medium 164 (31.4) 177 (34.4) 341 (32.9)

Large 51 (9.8) 51 (9.9) 102 (9.8)

Source: START B5 Table 1 page 1100 Abbreviations: IQR=Interquartile range, SD=Standard deviation a Lobular and other histological types. b Other endocrine therapies include combinations of tamoxifen/anastrozole/letrozole/exemestane/goserelin, mostly within randomised trials.

3.2.5 Spooner3

This citation was a conference abstract, therefore limited information was available. Patients with

clinical stage 1 and 2 disease (N=707) were randomised to receive immediate postoperative

radiotherapy or delayed salvage treatment (no radiotherapy). The study was conducted in the

UK.

Patients receiving radiotherapy (N=NR) were further randomised to short (40 Gy in 15 daily

fractions) or long (50 Gy in 25 daily fractions). The 50 Gy arm was considered the control arm.

Recruitment occurred between 1985 and 1992. Patients were followed for a mean of 16.9 years.

No information about patient characteristics was available.


4 Results of included trials

The results of the identified systematic reviews are not shown here as all trials that were

identified in their literature searches have been described in detail below.

4.1 Local recurrence

All five included trials reported local recurrence. Two trials described patients who had

undergone breast conserving surgery (RMH/GOC and Canadian)1-2

and three trials did not

include type of surgery as an inclusion criteria (Spooner, START A and START B).3-5

4.1.1 Post breast conserving surgery

RMH/GOC trial1

This trial compared i) 39 Gy in 13 fractions and ii) 42.9 Gy in 13 fractions with iii) 50 Gy in 25

fractions (the control arm). All fractions were administered over a five week period.

The results are shown in Table 22. Compared with the 50 Gy arm, the hazard ratio for the 42.9

Gy arm was 0.86 (95% CI 0.57, 1.30), and for the 39 Gy arm was 1.33 (95% CI 0.92, 1.92). The

Kaplan-Meier estimates of local recurrence at 10 years were 12.1% (95% CI 8.8, 15.5) for the 50

Gy arm, 9.6% (95% CI 6.7, 12.6) for the 42.9 Gy arm and 14.8% (95% CI 11.2, 18.3) for the 39

Gy arm. There was a significant difference in the probability of local recurrence after 10 years

between the 42.9 Gy and 39 Gy arms of the study (3・7%, 95% CI 0・3–8・3; χ2 test, degrees

of freedom [df]=1,p=0.027), with a higher rate of recurrence in the 39 Gy arm.

Table 22 RMH/GOC trial: Survival analysis of local relapse according to fractionation schedule1

Arm Number of local

recurrence / person-years

Crude hazard ratio

(95% CI) a

Kaplan-Meier estimates of local recurrence

(95% CI)

Smoothed estimate of absolute difference (95% CI) a

5 years follow-up

10 years follow-up

5 years follow-up

10 years follow-up

50 Gy 50/3965 1 7.9% (5.4, 10.4)

12.1% (8.8, 15.5)

NR NR

42.9 Gy 42/3840 0.86 (0.57, 1.30)

7.1% (4.6, 9.5)

9.6% (6.7, 12.6)

-1.1% (-3.3, 2.3)

-1.6% (-5.0, 3.3)

39 Gy 66/3890 1.33 (0.92, 1.92)

9.1% (6.4, 11.7)

14.8% (11.2, 18.3)

2.5% (-0.6, 6.7)

3.7% (-0.9, 9.8)

Source: Owen 20061, Table page 469 Abbreviations: CI=confidence interval a Compared with 50 Gy

Figure 1 shows the recurrence-free Kaplan-Meier curves. The curves for each fractionation

schedule begin to diverge after 5 years of follow-up. Hazard ratios were calculated by comparing

the 42.9Gy arm and the 39 Gy arm with the 50 Gy arm. The hazard ratios for the first five years


a systematic review

of follow-up were 0.90 (95% CI 0.55, 1.46) for 42.9 Gy arm and 1.14 (95% CI 0.72, 1.79) for the

39 Gy arm. The hazard ratios for the period from 5 years until the end of follow-up were 0.77

(95% CI 0.36, 1.69) for the 42.9 Gy arm and 1.81 (95% CI 0.96, 3.41) for 39 Gy arm. This

difference was not significant (p=0.1).

Figure 1 RMH/GOC trial: Local ipsilateral relapse in the breast according to fractionation scheudule1

Source: Owen 20061, Figure 2 page 470

Canadian trial2

This trial compared i) 42.5 Gy in 16 fractions over 22 days with ii) 50 Gy in 25 fractions over 35

days (the control arm). The cumulative incidence of local invasive recurrence was similar in the

two groups (as shown in Figure 2). At 10 years, the cumulative incidence of local invasive

recurrence was 6.2% in the 42.5 Gy arm compared with 6.7% in the 50 Gy arm (absolute

difference,0.5 percentage points; 95% CI, −2.5, 3.5). The pre-defined criteria for non-inferiority

was met (with a p value for non-inferiority of <0.001), indicating that the 42.5 Gy arm was not

inferior to the 50 Gy arm.

Non-invasive recurrences occurred in six patients in the 42.5 Gy arm and seven patients in the

50 Gy arm. The 10 year cumulative incidence of invasive or non-invasive local recurrence was

7.4% in the 42.5 Gy arm as compared with 7.5% in the 50 Gy arm (absolute difference, 0.1

percentage points; 95% CI, −3.1, 3.3). A hazard ratio for the entire population was not reported.


Figure 2 Canadian trial: Kaplan-Meier estimates for local recurrencea2

Source: Whelan 20102 Figure 1a page 516 a p<0.001 for non-inferiority Standard regimen = 50 Gy arm Hypofractionated regimen = 42.5 Gy arm

A subgroup analysis of local recurrence rate showed that the treatment effect of the

hypofractionated protocol was similar regardless of patient age, tumour size, oestrogen-receptor

status, or use of systemic therapy (shown in Figure 3). The hypofractionated regimen appeared

to be less effective at preventing local recurrence in patients with high-grade tumours (p=0.01).

For these patients, the 10 year cumulative incidence of local recurrence was 4.7% in the control

group as compared with 15.6% in the hypofractionated-radiation group (absolute difference, -

10.9; 95% CI: -19.1, -2.8). In the high grade (grade 3) patient group, the hazard ratio was 3.08

(95% CI 1.22, 7.76), compared with 0.70 (95% CI 0.31, 1.58) in the low grade patient group.


a systematic review

Figure 3 Canadian trial: Hazard ratios for Ipsilateral recurrence of breast cancer in subgroups of patients2

Source: Whelan 20102 Figure 2 page 517

4.1.2 Any surgery

Spooner3

This study compared i) 40 Gy in 15 fractions once a day or ii) 50 Gy in 25 fractions once a day

with iii) delayed salvage treatment. The abstract stated that there was no difference in relapse

frequency, or site between the two radiotherapy study arms, however no data was reported in the

abstract.

START A4

This trial compared i) 39 Gy in 13 fractions over five weeks or ii) 41.6 Gy in 13 fractions over five

weeks with iii) 50 Gy in 25 fractions over five weeks (the control arm). The relapse rates are

shown in

Table 23. The crude hazard ratios for local relapse, relative to the 50 Gy arm, were 1.09 (95% CI

0.64, 1.88) for the 41.6 Gy arm and 1.25 (95% CI 0.74, 2.12) for the 39 Gy arm. These

differences were not statistically significant.


Table 23 START A: Survival analyses of relapse and mortality according to fractionation schedule (Local relapse)4

Arm Events/total (%) Estimated % with event by 5 years

(95% CI) Crude hazard ratio (95%

CI) P value

50 Gy 25/749 (3.3) 3.2 (1.9, 4.6) 1 -

41.6 Gy 28/750 (3.7) 3.2 (1.9, 4.5) 1.09 (0.64, 1.88) 0.74

39 Gy 31/737 (4.2) 4.6 (3.0, 6.2) 1.25 (0.74, 2.12) 0.40

Source: Bentzen 20084 Table 2 page 335 Abbreviations: CI=confidence interval

START B5

This trial compared i) 40 Gy in 15 fractions over three weeks with ii) 50 Gy in 25 fractions over

five weeks (the control arm). The relapse rates are shown in Table 24. The crude hazard ratio for

local relapse, relative to the 50 Gy arm, was 0.72 (95% CI 0.43, 1.21) for the 40 Gy arm. This

was not statistically significant.

Table 24 START B: Survival analyses of relapse and mortality according to fractionation schedule (local relapse)5



CI) P value

50 Gy 34/1105 (3.1) 3.3 (2.2, 4.4) 1 -

40 Gy 25/1110 (2.2) 2.0 (1.1, 2.8) 0.72 (0.43, 1.21) 0.21

Source: Bentzen 20085 Table 2 page 1102 Abbreviations: CI=confidence interval

4.1.3 Conclusions: Local recurrence

The results are summarised in Table 25. There was no evidence that any hypofractionated

radiotherapy regimen was associated with a statistically significant difference in local recurrence

rate when compared with a control arm. The RMH/GOC trial noted a statistically significant

difference in recurrence rates when the two hypofractionated radiotherapy regimens were

compared (42.9 Gy vs 39 Gy: 9.6% vs 14.8%, p=0.027), but not when each regimen was

compared to the control arm.1

Subgroup analyses were performed in one publication.2 The Canadian trial analysed local

recurrence by patient age, tumour size, oestrogen-receptor status, tumour grade or use of

systemic therapy. There were no significant differences in any subgroup, with the exception of

tumour grade. The 42.5 Gy regimen was least effective in patients with high-grade tumours

compared to patients with low grade tumours (p=0.01).2

These results must be considered in the context of the range of hypofractionated radiotherapy

regimes evaluated and different study designs used in these publications. Any imbalance

between study arms in the treatments received may influence long term outcomes such as local

recurrence.


a systematic review

Table 25 Summary of key results for local recurrence







39 Gy vs 50 Gy: 9.1% vs 7.9%, p=NR

42.9 Gy vs 50 Gy: 7.1% vs 7.9%, p=NR

42.9 Gy vs 39 Gy: 7.1% vs 9.1%, p=NR


39 Gy vs 50 Gy: 14.8% vs 12.1%, p=NS

42.9 Gy vs 50 Gy: 9.6% vs 12.1%, p=NS

42.9 Gy vs 39 Gy: 9.6% vs 14.8%, p=0.027

39 Gy: HR 1.33 (95% CI 0.92, 1.92), p=NS

42.9 Gy: HR 0.86 (95% CI 0.57, 1.30), p=NS



10 year cumulative incidence of local recurrence

42.5 Gy vs 50 Gy: 6.2% vs. 6.7%, p=NS

10 year cumulative incidence of invasive or non-invasive local recurrence 42.5 Gy vs 50 Gy: 7.4% vs. 7.5%, p=NS

Subgroup analyses

Patient age, tumour size, oestrogen-receptor status, tumour grade, systemic therapy, p=NS

High-grade vs low grade tumours, p=0.01

Any surgery

Spooner3




17 year relapse frequency








39 Gy: HR 1.25 (95% CI 0.74, 2.12), p=0.40

41.6 Gy: HR 1.09 (95% CI 0.64, 1.88), p=0.74




50 Gy vs 40 Gy: 3.3% vs 2.0%, p=NR


40 Gy: HR 0.72 (95% CI 0.43, 1.21), p=0.21

Abbreviations: CI=confidence interval, HR=hazard ratio, , NR= not reported, NS=not significant * control arm

4.2 Local-regional recurrence

Two trials reported local-regional recurrence, both in studies that did not include surgery type as

an inclusion criteria (START A and START B).4-5


4.2.1 Any surgery

START A4

This trial compared i) 39 Gy in 13 fractions over five weeks and ii) 41.6 Gy in 13 fractions over

five weeks with iii) 50 Gy in 25 fractions over five weeks (the control arm). The relapse rates are

shown in Table 26. The hazard ratios for local-regional tumour relapse were 1.05 (95% CI 0.63,

1.75) for the 41.6 Gy arm and 1.26 (95% CI 0.77, 2.08) for the 39 Gy arm. These differences

were not statistically significant.

Table 26 START A: Survival analyses of relapse and mortality according to fractionation schedule (Local-regional relapse)4



CI) P value


50 Gy 28/749 (3.7) 3.6 (2.2, 5.1) 1 –

41.6 Gy 30/750 (4.0)) 3.5 (2.1, 4.3) 1.05 (0.63, 1.75) 0.86

39 Gy 35/737 (4.7) 5.2 (3.5, 6.9) 1.26 (0.77, 2.08) 0.35

Source: Bentzen et al 20084 Table 2 page 335 Abbreviations: CI=confidence interval

START B5


five weeks (the control arm). The relapse rates are shown in Table 27. The hazard ratios for

local-regional tumour relapse was 0.79 (95% CI 0.48, 1.29) for the 40 Gy arm, which was not

statistically significant.

Table 27 START B: Survival analyses of relapse and mortality according to fractionation schedule (Local-regional Relapse)5



CI) P value


50 Gy 36/1105 (3.2) 3.3 (2.2, 4.5) 1 –

40 Gy 29/1110 (2.6) 2.2 (1.3, 3.1) 0.79 (0.48, 1.29) 0.35


4.2.2 Conclusions: Local-regional recurrence

The results are summarised in Table 28. There was no evidence that any hypofractionated

radiotherapy regimen was associated with a statistically significant difference in regional

recurrence rate when compared with a control arm.




a systematic review

between study arms in the treatments received may influence long term outcomes such as local-

regional recurrence.

Table 28 Summary of key results for local-regional recurrence4-5


Any surgery





50 Gy vs 41.6 Gy vs 39 Gy: 3.6% vs 3.5%vs 5.2%, p=NR


39 Gy: HR 1.26 (95% CI 0.77, 2.08), p=0.35

41.6 Gy: HR 1.05 (95% CI 0.63, 1.75), p=0.86




50 Gy vs 40 Gy: 3.3% vs 2.2%, p=NR


40 Gy: HR 0.79 (95% CI 0.48, 1.29), p=0.35

Abbreviations: CI=confidence interval, HR=hazard ratio, NR=not reported * control arm

4.3 Distant relapse

Two trials reported distant relapse, both in studies that did not include surgery type as an

inclusion criteria (START A and START B).4-5

4.3.1 Any surgery

START A4


five weeks with iii) 50 Gy in 25 fractions over five weeks (the control arm). The distant relapse

rates are shown in Table 29. The hazard ratios for distant relapse, compared to the 50 Gy arm,

were 0.92 (95% CI 0.66, 1.28) for the 41.6 Gy arm and 1.29 (95% CI 0.95, 1.76) for the 39 Gy

arm. These ratios were not statistically significant.

Table 29 START A: Survival analyses of relapse and mortality according to fractionation schedule (Distant relapse) 4



CI) P value

Distant relapse

50 Gy 73/749 (9.7) 9.8 (7.5, 12.0) 1 –

41.6 Gy 69/750 (9.2) 9.5 (7.3, 11.7) 0.92 (0.66, 1.28) 0.64

39 Gy 93/737 (12.6) 11.9 (9.5, 14.4) 1.29 (0.95, 1.76) 0.10



START B5


five weeks (the control arm). The distant relapse rates are shown in Table 30. The hazard ratios

for distant relapse was 0.69 (95% CI 0.53, 0.91) for the 40 Gy arm. This result was statistically

significant (p=0.01).

Table 30 START B: Survival analyses of relapse and mortality according to fractionation schedule (Distant relapse)5



CI) P value

Distant relapse

50 Gy 122/1105 (11.0) 10.2 (8.4, 12.1) 1 –

40 Gy 87/1110 (7.8) 7.6 (6.0, 9.2) 0.69 (0.53, 0.91) 0.01


4.3.2 Conclusions: Distant relapse

The results are summarised in Table 31. START A reported no statistical difference between

either of the hypofractionated regimens compared with control. START B reported that the 40 Gy

study arm had a significantly lower rate of distant relapse when compared with the control arm

(HR 0.69 95% CI 0.53, 0.91, p=0.01).



between study arms in the treatments received may influence long term outcomes such as local

recurrence.

Table 31 Summary of key results for distant relapse4-5


Any surgery




5 year distant relapse rate



39 Gy: HR 1.29 (95% CI 0.95, 1.76), p=0.10

41.6 Gy: HR 0.92 (95% CI 0.66, 1.28), p=0.64




50 Gy vs 40 Gy: 10.2% vs 7.6%, p=NR


40 Gy: HR 0.69 (95% CI 0.53, 0.91), p=0.01

Abbreviations: CI=confidence interval, HR=hazard ratio, NR=not reported * control arm


a systematic review

4.4 Overall survival

A total of four trials reported overall survival (Canadian, Spooner, START A and START B). 2, 4-5,

7 The Canadian trial

2 described patients who had undergone breast conserving surgery and

three publications did not include surgery type as an inclusion criteria (Spooner, START A and

START B).3-5


Canadian trial2


days (the control arm). During the study there were 126 deaths in the 50 Gy arm (20.6%) and

122 deaths in the in the 42.5 Gy arm (19.6%). At 10 years, the probability of survival was 84.4%

in the 50 Gy arm compared with 84.6% in the 42.5 Gy arm (shown in Figure 4). The absolute

difference was -0.2 percentage points (95% CI -4.3, 4.0). This difference was not statistically

significant (p=0.79).

Figure 4 Canadian trial: Kaplan-Meier estimate for overall survival2

Source: Whelan 20102 Figure 1b page 516 p=0.79

The cause of death is shown in Table 32. In the 50 Gy arm 13.4% of deaths were related to

cancer, 1.5% were related to cardiac disease and 5.7% were due to other causes. In the 42.5 Gy

arm 13.2% were related to cancer, 1.9% were related to cardiac disease and 4.5% were due to

other causes. None of these differences were statistically significant.


Table 32 Canadian trial: Cause of deaths2

Arm 50 Gy n (%)

42.5 Gy n (%)

P value

Deaths related to cancer 82 (13.4) 82 (13.2) NS

Deaths related to cardiac disease 9 (1.5) 12 (1.9) NS

Deaths related to other causes 35 (5.7) 28 (4.5) NS

Source: Whelan 20102 page 517 Abbreviations: NS=not significant

4.4.2 Any surgery

Spooner3

This study compared i) 40 Gy in 15 fractions once a day or ii) 50 Gy in 25 fractions once a day

with delayed salavage treatment. The abstract noted that there was no difference in overall

survival between the two study arms, however no data was reported in the abstract.

START A4


five weeks with iii) 50 Gy in 25 fractions over five weeks (the control arm). The START A trial

reported all-cause mortality, rather than survival. As shown in Table 33, the hazard ratios for all-

cause mortality, compared with the 50 Gy arm, were 1.04 (95% CI 0.77, 1.40) for the 41.6 Gy

arm and 1.00 (95% CI 0.74, 1.36) for the 39 Gy arm. These ratios were not statistically

significant.

Table 33 START A: Survival analyses of relapse and mortality according to fractionation schedule (All-cause mortality)4



CI) P value

50 Gy 84/749 (11.2) 11.1 (8.7, 13.4) 1 –

41.6 Gy 89/750 (11.9) 11.3 (8.9, 13.7) 1.04 (0.77, 1.40) 0.81

39 Gy 83/737 (11.3) 10.7 (8.3, 13.1) 1.00 (0.74, 1.36) 0.99


START B5


five weeks (the control arm). The START B trial reported all-cause mortality, rather than survival.

As shown in Table 34, the hazard ratio for all-cause mortality for the 40 Gy arm was 0.76 (95%

CI 0.59, 0.98) compared with the 50 Gy arm. This was statistically significant (p=0.03).


a systematic review

Table 34 START B: Survival analyses of relapse and mortality according to fractionation schedule (All-cause mortality)5



CI) P value

50 Gy 138/1105 (12.5) 11.0 (9.1, 12.9) 1 –

40 Gy 107/1110 (9.6) 8.0 (6.4, 9.7) 0.76 (0.59, 0.98) 0.03


4.4.3 Conclusions

The results are summarised in Table 35. Most studies reported that there was no evidence that

hypofractionated radiotherapy was associated with a statistically significantly difference in overall

survival. START B found that 40 Gy in 15 fractions over three weeks was associated with a

statistically significantly lower all-cause mortality rate when compared with 50 Gy in 25 fractions

over five weeks (HR 0.76 95% CI 0.59, 0.98, p=0.03). Therefore, there was no evidence that any

hypofractionated radiotherapy regimen was associated with a worse overall survival rate (i.e. the

only study that reported a significant difference showed lower mortality for patients treated with

hypofractionated radiotherapy).


regimes evaluated and different study designs used in these publications. None of the

publications described the treatments that patients in each study arm received after radiotherapy

(e.g. chemotherapy). Any imbalance between study arms in the treatments received may

influence long term outcomes such as overall survival.

Table 35 Summary of key results for overall survival





10 year survival

42.5 Gy vs 50 Gy: 84.6% vs 84.4%, p=0.79

Any surgery

Spooner3 40 Gy in 15 fractions once a day


17 year survival






39 Gy: HR 1.00 (95% CI 0.74, 1.36), p=0.99

41.6 Gy: HR 1.04 (95% CI 0.77, 1.40), p=0.81




40 Gy: HR 0.76 (95% CI 0.59, 0.98), p=0.03

Abbreviations: CI=confidence interval, HR=hazard ratio * control arm


4.5 Adverse events and toxicity

A total of three trials reported adverse events and toxicity outcomes (Canadian, START A and

START B).2, 4-5, 7

The Canadian trial2, 7

described patients who had undergone breast conserving

surgery and two trials did not include surgery type as an inclusion criteria (START A4 and START

B5, as well as combined data from both studies

6).


Canadian trial2


days (the control arm). The late toxic effects of radiation are shown in Table 36. The results were

generally consistent between study arms. No grade 4 skin ulceration or soft-tissue necrosis was

observed in any subjects.

Table 36 Canadian trial: Late toxic effects of radiation, assessed according to the RTOG-EORTC late radiation morbidity scoring schemea2

Site and Grade

5 year follow-up 10 year follow-up

50 Gy n=424 %

42.5 Gy n=449 %

50 Gy n=220 %

42.5 Gy n=235 %

Skin

0b 82.3 86.1 70.5 66.8

1 14.4 10.7 21.8 24.3

2 2.6 2.5 5.0 6.4

3 0.7 0.7 2.7 2.5

Subcutaneous tissue

0c 61.4 66.8 45.3 48.1

1 32.5 29.5 44.3 40.0

2 5.2 3.8 6.8 9.4

3 0.9 0.9 3.6 2.5

Source: Whelan 20102 Table 1 page 518 Abbreviations: CI=confidence interval, RTOG-EORTC=Radiation Therapy Oncology Group and the European Organization for Research and Treatment of Cancer. a Effects of radiation therapy on skin and subcutaneous tissue were graded on a scale of 0 to 4 (with 0 indicating no toxic effects and grade 4 indicating skin ulceration or soft-tissue necrosis). RTOG–EORTC denotes the Radiation Therapy Oncology Group and the European Organization for Research and Treatment of Cancer. b The absolute difference at 5 years was -3.8 percentage points (95% CI: -8.7, 1.0), and at 10 years the absolute difference was 3.7 percentage points (95% CI: -4.9, 12.1). c The absolute difference at 5 years was -5.4 percentage points (95% CI: -11.9, 0.9), and at 10 years the absolute difference was -2.8 percentage points (95% CI: -11.7, 6.5).

4.5.2 Any surgery

START A4

This trial compared i) 39 Gy in 13 fractions over five weeks or ii) 41.6 Gy in 13 fractions over five

weeks with iii) 50 Gy in 25 fractions over five weeks (the control arm). The incidence of ischemic


a systematic review

heart disease, symptomatic rib fracture and symptomatic lung fibrosis was low, with no

differences between the study arms (shown in Table 37).

Table 37 START A: Incidence of ischemic heart disease, symptomatic rib fracture, and symptomatic lung fibrosis according to fractionation schedule4

Outcome Arm Reported

n, (%) Confirmed

n, (%) a

Ischemic heart disease b

50 Gy 12 (1.6) 3 (0.4) [1]c

41.6 Gy 7 (0.9) 2 (0.3) [0]c

39 Gy 8 (1.1) 5 (0.7) [4]c

Total 27(1.2) 10 (0.4) [5]c

Symptomatic rib fractures d

50 Gy 8 (1.1) 1 (0.1)

41.6 Gy 9 (1.2) 2 (0.3)

39 Gy 10 (1.4) 1 (0.1)

Total 27 (1.2) 4 (0.2)


50 Gy 5 (0.7) 0 (0)

41.6 Gy 6 (0.8) 2 (0.3)

39 Gy 7 (0.9) 1 (0.1)

Total 18 (0.8) 3 (0.1)

Source: Bentzen et al 20084 Table 3 page 337 a Cases confirmed after imaging and further investigations. b 18 patients had pre-existing heart disease at randomisation and were excluded. c Confirmed cases of ischemic heart disease in patients with left-sided primary tumours. d Reported cases include three with rib fracture after bone metastases and nine after trauma.

Other adverse events were discussed in the publication. In the 41.6 Gy arm, there was one case

of pneumonitis which occurred nine months after treatment and one patient who developed mild

symptoms and signs of brachial plexopathy two years after treatment. Two patients in the 50 Gy

arm experienced an unusually marked acute skin reaction during their radiotherapy treatment,

culminating in extensive moist desquamation. Neither patient had received adjuvant

chemotherapy.4 It was assumed that no cases were reported in the other study arms, although

this was not explicitly stated in the publication.

As shown in Table 38, a small number of patients had contralateral breast cancer (26 patients

(1.2%)) or a secondary primary cancer (44 patients (2%))..4

Table 38 START A: Contralateral and other secondary cancers4

Outcome Arm n, (%)

Contralateral breast cancer

50 Gy 13 (1.7)

41.6 Gy 5 (0.7)

39 Gy 8 (1.1)

Other secondary primary cancers

50 Gy 15 (0.7)

41.6 Gy 10 (0.4)

39 Gy 19 (0.8)

Source: Bentzen et al 20084 page 338


Patient self-assessments of late normal tissue effects are discussed in the combined START A

and B publication results.6

START B5

This study compared i) 40 Gy in 15 fractions over three weeks with ii) 50 Gy in 25 fractions over

five weeks (the control arm). The incidence of ischemic heart disease, symptomatic rib fracture

and symptomatic lung fibrosis was low, with no differences between the study arms (shown in

Table 39).

Table 39 START B: Incidence of ischemic heart disease, symptomatic rib fracture, and symptomatic lung fibrosis according to fractionation schedule5

Outcome Arm Reported (%) Confirmed (%) a

Ischemic heart disease b

50 Gy 19 (1.7) 12 (1.1) [4]c

40 Gy 15 (1.3) 7 (0.6) [3]c

Total 34 (1.5) 19 (0.9) [7] c

Symptomatic rib fractures d

50 Gy 17 (1.5) 2 (0.2)

40 Gy 16 (1.4) 2 (0.2)

Total 33 (1.5) 4 (0.2)


50 Gy 15 (1.4) 1 (0.1)

40 Gy 16 (1.4) 3 (0.3)

Total 31 (1.4) 4 (0.2)

Source: Bentzen et al 20085 Table 3 page 1104 a Cases confirmed after imaging and further investigations. b 11 patients had pre-existing heart disease at randomisation and were excluded. c Confirmed cases of ischemic heart disease in patients with left-sided primary tumours. d Reported cases include four with rib fracture after bone metastases and three after trauma.

Other adverse events were discussed in the publication. There were no cases of brachial

plexopathy in either study arm. Thirteen patients in the 50 Gy arm (1.2%) and three patients in

the 40 Gy arm (0.3%) reported a marked acute reaction during radiotherapy.5

As shown in Table 40, a small number of patients had contralateral breast cancer (36 patients

(1.6%)) or a secondary primary cancer (58 patients (2.6%)).5

Table 40 START B: Contralateral and other secondary cancers5

Outcome Arm n, (%)

Contralateral breast cancer 50 Gy 19 (1.7)

40 Gy 17 (1.5)

Other secondary primary cancers 50 Gy 32 (2.9)

40 Gy 26 (2.3)

Source: Bentzen et al 20085 page 1103

Patient self-assessments of late normal tissue effects are discussed in the combined START A

and B publication results.6


a systematic review

START A and B6

The START A trial compared i) 39 Gy in 13 fractions over five weeks and ii) 41.6 Gy in 13

fractions over five weeks with iii) 50 Gy in 25 fractions over five weeks (the control arm). The

START B trial compared i) 40 Gy in 15 fractions over three weeks with ii) 50 Gy in 25 fractions

over five weeks (the control arm). Combined patient-reported quality-of-life results from Start A

and B are reported in Hopwood et al 2010.6

Of the 4,451 patients enrolled in either the START A or START B trials, 2,208 patients were

accrued into the quality-of-life study (1,129 from START A and 1,079 from START B).6

Figure 5 shows the forest plot for patient reported normal tissue effects in the START A and

START B trials. Two sets of outcomes were reported: breast symptoms and arm or shoulder

symptoms. Many of the breast symptoms can also be considered adverse events as well as

cosmetic outcomes, and have therefore been discussed here as well as in the following section

on cosmetic outcomes. Breast symptom outcomes are discussed first, followed by arm or

shoulder symptoms.

Breast Symptoms

In the START A trial, there were no statistically significant differences between the 41.6 Gy and

50 Gy arms. Change in breast appearance, breast hardness, breast shrinkage, pain in area of

affected breast and oversensitivity in area of affected breast favoured the 50 Gy arm, while

change in skin appearance and swelling in area of affected breast favoured the 41.6 Gy arm.

Skin problems on or in area of affected breast had a hazard ratio of 1.01.6

When the 39 Gy and 50 Gy arms in the START A trial were compared, all outcomes favoured the

39 Gy arm (with the exception of breast shrinkage which had a hazard ratio of 1.00). The only

outcome which statistically significantly favoured the 39 Gy arm was change in skin appearance

(hazard ratio 0.63 95% CI 0.47, 0.84, p=0.0019).6

In the START B trial, all outcomes favoured the 40 Gy arm, with the exception of oversensitivity

in area of affected breast. The only outcome which statistically significantly favoured the 40 Gy

arm was change in skin appearance (hazard ratio 0.76 95% CI 0.60, 0.97, p=0.0262).6

Arm or shoulder symptoms

In the START A trial, shoulder stiffness, difficult in raising or moving arm sideways and swelling

in arm or hand favoured the 41.6 Gy arm compared to the 50 Gy arm, although this was not

statistically significant. Pain in arm or shoulder had a hazard ratio of 1.00. When the 39 Gy and

50 Gy arms were compared, shoulder stiffness and swelling in arm or hand favoured the 39 Gy

arm, while pain in arm or shoulder favoured the 50 Gy arm, but these differences were not

statically significant..6

In the START B trial there were no differences in arm or shoulder symptoms. The hazard ratios

ranged from 0.95 to 1.05.6


Figure 5 START A AND B: Forest plots of normal tissue effects assessed as moderate or marked by

patients, according to radiotherapy regimens6

Source: Hopwood et al 20106 Figure 2 page 6 Positions of squares in the forest plot show the estimate of the hazard ratio describing relative effect of the test schedule compared with control, with the 95% CI represented by horizontal lines. Squares to the left of the vertical line indicate when rates of adverse effects are lower in the test schedule compared with control; estimates to the right of the line indicate whether rates are higher in the test schedule. Size of squares is proportional to the precision of the estimate, with larger squares indicating greater precision. *In patients who had completed breast-conserving surgery only.


a systematic review

4.5.3 Conclusions

The results are summarised in Table 41. Most trials reported that there was no difference in

adverse events and toxicity. Combined results from the START A and START B trials found that

a change in skin appearance occurred significantly less often in the 39 Gy and 40 Gy arms when

compared with the control arm (39 Gy HR 0.63 95% CI 0.47, 0.84, p=0.0019 and 40 Gy HR 0.76

95% CI 0.60, 0.97, p=0.0262).6

These results should be considered in the context of the difference in adverse event and toxicity

outcomes assessed in each publication, and the difference in hypofractionated radiotherapy

regimens evaluated.

Table 41 Summary of key results for adverse events and toxicity





Late toxic radiation effects: NS

Any surgery








Combined QoL data from START A and B6^

As for START A and START B Tissue effects, arm and shoulder symptoms: NS

Skin appearance: 39 Gy HR 0.63 95% CI 0.47, 0.84, p=0.0019

40 Gy HR 0.76 (95% CI 0.60, 0.97), p=0.0262

Abbreviations: CI=confidence interval, HR=hazard ratio, NS=not significant, QoL=Quality-of-life * control arm ^ Of the 4,451 patients enrolled in either the START A or START B trials, 2,208 patients were accrued into the quality-of-life study (1,129 from START A and 1,079 from START B)


4.6 Cosmetic outcome

A total of four trials reported cosmetic outcomes. Four of the trials described patients who had

undergone breast conserving surgery (RMH/GOC, START A, START B and Canadian trials). 2, 8

6 Two trials described patients who had undergone mastectomy (START A and START B).

6. The

same two trials reported results for surgery types combined.4-5


RMH/GOC trial8

This study compared i) 39 Gy in 13 fractions and ii) 42.9 Gy in 13 fractions with iii) 50 Gy in 25

fractions (the control arm). All fractions were administered over a five week period. Note that the

cosmetic outcomes described here were all reported in Yarnold 20058, but not the follow-up

publication.1 Therefore, the source for all data in this section is Yarnold 2005.

8 A subset of 1,202

patients underwent breast appearance assessment. Change in breast appearance was assessed

using photographs which were scored by blinded assessors. Clinician assessments were also

performed.

The change in breast appearance, based on photographic and clinical assessment, is shown in

Table 42. There was a significant difference between the three study arms in terms of any

change in breast appearance, and a marked change in breast appearance (p<0.001). Patients in

the 39 Gy arm were the least likely to have a marked change in breast appearance at 10 years

(93.4% with no event) compared with the 50 Gy arm and 42.9 Gy arm (90.2% and 84.4% with

no event respectively).8

The publication noted for change in breast appearance based on photographic assessment, that

there was some evidence of variation in the difference between the fractionation schedules as

follow-up increased, as the 50 and 39 Gy arms appear to be converging, but analysis shows this

not to be statistically significant (p=0.08). The probability of any change in breast appearance ten

years after radiotherapy is shown in Figure 6. The risk of developing any radiation effect was

much lower for patients allocated to the 39 Gy arm compared with those allocated to the 42.9 Gy

arm. There was a statistically significant difference between the 50 and 39 Gy arms of the trial

over this time period (p=0.01), but weaker evidence for the difference between 50 and 42.9 Gy

(p=0.05).8 Evidence was observed of a lower risk of marked changed in the patients treated with

39 Gy compared to 42.9 Gy. The probability of no change was highest in the 39 Gy arm, followed

by the 50 Gy arm and the 42.9 Gy arm. A similar pattern was seen for a marked change in breast

appearance, as shown in Figure 7.


a systematic review

Figure 6 RMH/GOC trial: Probability of any change in breast appearance late radiation effect ten years after

radiotherapy by fractionation schedule8

Source: Yarnold 20058, Figure 2 page 13

Figure 7 RMH/GOC trial: Probability of marked change in breast appearance late radiation effect ten years

after radiotherapy by fractionation schedule8


Clinical assessments of overall breast cosmesis (involving an aesthetic judgement), breast

shrinkage, breast distortion, breast oedema, induration and shoulder stiffness showed a

significant difference between the treatment arms. For all outcomes (except shoulder stiffness),

the estimated percent of subjects with no event at 10 years was highest in the 39 Gy arm and

lowest in the 42.9 Gy arm.

The probability of palpable breast induration ten years after radiotherapy is shown in Figure 8.

The 39 Gy arm had the highest probability of no change, followed by the 50 Gy arm and the 42.9

Gy arm.


Figure 8 RMH/GOC trial: Probability of palpable breast induration ten years after radiotherapy by

fractionation schedule8


For telangiectasia, the estimated percent of subjects with no event at 10 years was 88.0% in the

39 Gy arm, 81.9% in the 50 Gy arm, and 82.0% in the 42.9 Gy arm. The only outcome which

failed to show a dose response was arm oedema (92.3% 50 Gy vs 89.5% 42.9 Gy vs 93.0% 39

Gy). The authors noted that this may be due to the small proportion (20.6%) of patients who

experienced any form of lymphatic radiotherapy.


a systematic review

Table 42 RMH/GOC trial: Survival analyses of change in breast appearance and clinical assessments of late radiation effects according to fractionation schedule 8

Endpoint Arm Events/total (%) Estimated % with no event at 5 years (95%

CI)

Estimated % with no event at 10 years (95%

CI) P value

Photographic assessment

Any change in breast appearance

50 Gy 140/396 (35.4) 60.4 (54.9, 65.8) 46.6 (37.2, 55.9)

<0.001 42.9 Gy 168/397 (42.3) 54.3 (48.9, 59.7) 42.0 (33.0, 51.0)

39 Gy 112/409 (27.4) 69.7 (64.6, 74.8) 43.9 (30.8, 57.0)

Marked change in breast appearance

50 Gy 22/396 (5.6) 93.6 (90.8, 96.4) 90.2 (85.0, 95.5)

<0.001 42.9 Gy 40/397 (10.1) 88.8 (85.3, 92.2) 84.4 (77.7, 91.1)

39 Gy 14/409 (3.4) 96.1 (93.9, 98.2) 93.4 (87.8, 99.0)

Clinical assessment

Cosmesis (fair/poor)

50 Gy 165/271 (60.9) 44.1 (37.7, 50.4) 28.8 (22.3, 35.4)

<0.001 42.9 Gy 175/266 (65.8) 37.9 (31.7, 44.1) 25.6 (19.3, 31.8)

39 Gy 136/269 (50.6) 54.6 (48.3, 60.9) 42.0 (34.9, 49.1)

Breast shrinkage (moderate/marked)

50 Gy 147/271 (54.6) 49.9 (43.5, 56.3) 36.2 (29.3, 43.1)

0.026 42.9 Gy 148/266 (55.8) 47.2 (40.8, 53.7) 34.2 (27.0, 41.5)

39 Gy 124/269 (46.1) 56.9 (50.6, 63.2) 44.4 (37.0, 51.7)

Breast distortion (moderate/marked)

50 Gy 132/271 (48.9) 54.6 (48.2, 61.0) 41.5 (34.4, 48.6)

0.005 42.9 Gy 148/266 (55.8) 45.7 (39.9, 52.1) 38.0 (31.4, 44.6)

39 Gy 115/269 (42.8) 59.3 (53.1, 65.4) 51.4 (44.4, 58.4)

Breast oedema (moderate/marked)

50 Gy 34/271 (12.6) 87.6 (83.6, 91.7) 86.2 (81.8, 90.7)

0.004 42.9 Gy 54/266 (20.3) 80.2 (75.3, 85.2) 78.5 (73.1, 83.9)

39 Gy 29/269 (10.8) 89.4 (85.6, 93.2) 88.5 (84.4, 92.7)

Induration (moderate/marked)

50 Gy 77/271 (28.6) 76.9 (71.5, 82.3) 63.7 (56.6, 70.7)

<0.001 42.9 Gy 108/266 (40.8) 64.4 (58.1, 70.6) 48.9 (41.5, 56.4)

39 Gy 55/269 (20.4) 84.0 (79.2, 88.8) 72.3 (65.5, 79.2)

Telangiectasia (moderate/marked)

50 Gy 37/271 (13.8) 88.0 (83.8, 92.3) 81.9 (76.5, 87.3)

0.065 42.9 Gy 38/266 (14.3) 87.0 (82.7, 91.4) 82.0 (76.5, 87.5)

39 Gy 23/269 (8.6) 94.4 (91.4, 97.4) 88.0 (83.0, 92.9)

Arm oedema (moderate/marked)

50 Gy 17/271 (6.3) 93.8 (90.7, 97.0) 92.3 (88.6, 96.1)

0.494 42.9 Gy 22/266 (8.3) 91.7 (88.1, 95.3) 89.5 (85.1, 93.8)

39 Gy 16/269 (5.9) 95.4 (92.7, 98.1) 93.0 (89.2, 96.8)

Shoulder stiffness (moderate/marked)

50 Gy 21/271 (7.8) 94.1 (91.2, 97.0) 90.0 (85.6, 94.3)

<0.001 42.9 Gy 48/266 (18.1) 84.0 (79.3, 88.6) 78.2 (72.3, 84.0)

39 Gy 19/269 (7.1) 94.2 (91.2, 97.2) 89.9 (85.3, 94.6)

Source: Yarnold 20058, Table 2 page 13


START A and B6




over five weeks (the control arm).6


accrued into the quality-of life study (1,129 from START A and 1,079 from START B). Of these

2,208 patients, 1831 had undergone breast-conserving surgery before radiotherapy (885 from

START A and 946 from START B).6

Table 43 reports the self-reported cosmetic outcomes for the women who underwent breast

conserving surgery before radiotherapy. These outcomes were similar between treatment arms

across the two RCTs. The low numbers of patients and events in some subgroups limited the

statistical power of these analyses.6


a systematic review

Table 43 START A AND B: Survival analyses of moderate or marked grade normal tissue effects from patients’ self-assessments, according to fractionation schedule, type of primary surgery6

Study Arm Hazard ratio (95% CI)

Change in skin appearance since radiotherapy

START A

50 Gy 1

41.6 Gy 0.92 (0.68–1.25)

39 Gy 0.63 (0.45–0.88)

START B 50 Gy 1

40 Gy 0.80 (0.63–1.03)

Skin problems on or in area of affected breast a

START A

50 Gy 1

41.6 Gy 1.02 (0.70–1.50)

39 Gy 0.87 (0.58–1.30)

START B 50 Gy 1

40 Gy 0.86 (0.65–1.15)

Pain in area of affected breast a

START A

50 Gy 1

41.6 Gy 1.29 (0.92–1.82)

39 Gy 1.01 (0.70–1.45)

START B 50 Gy 1

40 Gy 0.97 (0.74–1.26)

Oversensitivity in area of affected breast a

START A

50 Gy 1

41.6 Gy 1.14 (0.81–1.58)

39 Gy 0.79 (0.55–1.12)

START B 50 Gy 1

40 Gy 1.18 (0.91–1.53)

Swelling in area of affected breast a

START A

50 Gy 1

41.6 Gy 0.87 (0.58–1.32)

39 Gy 0.68 (0.44–1.05)

START B 50 Gy 1

40 Gy 0.89 (0.62–1.29)

Arm or shoulder pain a

START A

50 Gy 1

41.6 Gy 1.05 (0.79–1.39)

39 Gy 1.11 (0.83–1.48)

START B 50 Gy 1

40 Gy 1.03 (0.83–1.29)

Shoulder stiffness a

START A

50 Gy 1

41.6 Gy 0.94 (0.65–1.37)

39 Gy 0.98 (0.67–1.41)

START B 50 Gy 1

40 Gy 0.94 (0.70–1.25)

Difficulty in raising or moving arm sideways a

START A

50 Gy 1

41.6 Gy 1.16 (0.76–1.77)

39 Gy 1.15 (0.75–1.76)

START B 50 Gy 1

40 Gy 0.99 (0.73–1.35)

Arm or hand swelling a

START A

50 Gy 1

41.6 Gy 0.77 (0.49–1.21)

39 Gy 0.92 (0.60–1.42)

START B 50 Gy 1

40 Gy 1.12 (0.78–1.60)

Source: Hopwood et al 20106 Table 3 page 7 and Table 4 page 8 Abbreviations: CI=confidence interval a Results adjusted for baseline scores.


Canadian trial2


days (the control arm). Table 44 reports the cosmetic outcome at baseline, 5 years, and 10

years. Although the global cosmetic outcome worsened over time, no significant differences were

observed between the groups at any time. The repeated-measures logistic-regression analysis

suggested that the cosmetic outcome was affected by the time from randomisation as well as by

the patient’s age and tumour size, but there was no interaction with treatment.

Table 44 Canadian trial: Global cosmetic outcome assessed according to the EORTC scalea2

Rating

Baseline 5-year follow-up 10-year follow-up

50 Gy %

42.5 Gy %

50 Gy %

42.5 Gy %

50 Gy %

42.5 Gy %

Excellent 46.3 46.8 34.3 36.4 27.8 30.6

Good 36.3 37.0 44.9 41.5 43.5 39.2

Fair 15.1 14.6 17.3 19.0 25.5 25.4

Poor 2.3 1.6 3.5 3.1 3.2 4.8

Excellent or good

82.6 83.8 b 79.2 77.9 c 71.3 69.8 d

Source: Whelan 20102 Table 2 page 519 Abbreviations: CI=confidence interval a Absolute differences were calculated as the value in the group that received the standard regimen minus the value in the group that received the hypofractionated regimen. b Absolute difference at baseline, -1.2 percentage points (95% CI: -5.4, 3.1). c Absolute difference at 5-year follow-up, 1.3 percentage points (95% CI: -4.2, 6.7). d Absolute difference at 10-year follow-up, 1.5 percentage points (95% CI: -6.9, 9.8).

4.6.2 Post mastectomy

START A and B6




over five weeks (the control arm).



2,208 patients, 377 had undergone a mastectomy before radiotherapy (244 from START A and

133 from START B).

Table 45 reports the self-reported cosmetic outcomes for the women who underwent

mastectomy before radiotherapy. These outcomes were similar between treatment arms across

the two RCTs. The low numbers of patients and events in some subgroups limited the statistical

power of these analyses.


a systematic review

Table 45 START A AND B: Survival analyses of moderate or marked grade normal tissue effects from patients’ self-assessments according to fractionation schedule, type of primary surgery6

Study Arm Hazard ratio (95% CI)

Change in skin appearance since radiotherapy

START A

50 Gy 1

41.6 Gy 0.53 (0.28–0.99)

39 Gy 0.64 (0.34–1.17)

START B 50 Gy 1

40 Gy 0.48 (0.20–1.16)

Skin problems on or in area of affected breast a

START A

50 Gy 1

41.6 Gy 0.90 (0.39–2.10)

39 Gy 1.07 (0.48–2.38)

START B 50 Gy 1

40 Gy 2.26 (0.43–11.80)

Pain in area of affected breast a

START A

50 Gy 1

41.6 Gy 0.82 (0.42–1.61)

39 Gy 0.87 (0.45–1.69)

START B 50 Gy 1

40 Gy 0.63 (0.22–1.79)

Oversensitivity in area of affected breast a

START A

50 Gy 1

41.6 Gy 0.87 (0.42–1.81)

39 Gy 0.81 (0.39–1.68)

START B 50 Gy 1

40 Gy 0.55 (0.19–1.54)

Swelling in area of affected breast a

START A

50 Gy 1

41.6 Gy 0.40 (0.11–1.51)

39 Gy 0.98 (0.36–2.61)

START B 50 Gy 1

40 Gy 4.18 (0.61–28.37)

Arm or shoulder pain a

START A

50 Gy 1

41.6 Gy 0.83 (0.47–1.47)

39 Gy 0.97 (0.56–1.69)

START B 50 Gy 1

40 Gy 0.92 (0.47–1.80)

Shoulder stiffness a

START A

50 Gy 1

41.6 Gy 0.74 (0.37–1.46)

39 Gy 0.45 (0.20–0.99)

START B 50 Gy 1

40 Gy 1.10 (0.47–2.58)

Difficulty in raising or moving arm sideways a

START A

50 Gy 1

41.6 Gy 0.61 (0.30–1.24)

39 Gy 0.61 (0.31–1.23)

START B 50 Gy 1

40 Gy 0.79 (0.35–1.77)

Arm or hand swelling a

START A

50 Gy 1

41.6 Gy 1.09 (0.50–2.35)

39 Gy 0.88 (0.40–1.95)

START B 50 Gy 1

40 Gy 0.65 (0.20–2.17)

Source: Hopwood et al 20106 Table 3 page 7 and Table 4 page 8 Abbreviations: CI=confidence interval a Results adjusted for baseline scores.


4.6.3 Any surgery

START A4


five weeks with iii) 50 Gy in 25 fractions over five weeks (the control arm). The change in breast

appearance was assessed by photograph in 1,055 patients who had undergone breast

conserving surgery.

The hazard ratios for a mild or marked change in breast appearance, compared with the 50 Gy

arm, were 1.09 (95% CI 0.85, 1.40) for the 41.6 Gy arm and 0.69 (95% CI 0.52, 0.91) for the 39

Gy arm. This difference was not significant for the 41.6 Gy arm (p=0.62), but was significant for

the 39 Gy arm (p=0.01). This is shown in Table 46. A Kaplan-Meier plot is shown in Figure 9.

Table 46 START A: Mild or marked change in breast appearance4

Study arm Crude hazard ratio (95% CI) P value

50 Gy 1 –

41.6 Gy 1.09 (95% CI 0.85, 1.40) 0.62

39 Gy 0.69 (95% CI 0.52, 0.91) 0.01

Source: Bentzen et al 20084 page 337 Abbreviations: CI=confidence interval

Figure 9 START A: Kaplan-Meier plot of mild/marked change in breast appearance (photographic) in 1055

patients with breast conserving surgery4

Source: Bentzen et al 20084 Figure 3 page 337

Other cosmetic outcomes are discussed on page 50 as part of the combined START A and B

study data.


a systematic review

START B5


five weeks (the control arm). The change in breast appearance was assessed by photograph in

923 patients who had undergone breast conserving surgery.

The hazard ratio for a mild or marked change in breast appearance for the 40 Gy arm (compared

with the 50 Gy arm) was 0.83 (95% CI 0.66, 1.04; p=0.06). This is shown in Table 47. A Kaplan-

Meier plot is shown in Figure 10.

Table 47 START B: Mild or marked change in breast appearance5

Study arm Crude hazard ratio (95% CI) P value

50 Gy 1 –

40 Gy 0.83 (0.66, 1.04) 0.06

Source: Bentzen et al 20085 page 1103 Abbreviations: CI=confidence interval

Figure 10 START B: Kaplan-Meier plot of mild/marked change in breast appearance (photographic) in 923

patients with breast conserving surgery5

Source: Bentzen et al 20085 Figure 4 page 1103

Other cosmetic outcomes are discussed below as part of the of the combined START A and B

trials data.

START A and B6







accrued into the quality-of life study (1,129 from START A and 1,079 from START B).

Figure 11 shows the forest plot for patient reported normal tissue effects in the START A and

START B trials. Two sets of outcomes were reported: breast symptoms and arm or shoulder

symptoms. Arm or shoulder symptoms have been discussed in the previous section on adverse

events and toxicity.

In the START A trial, there were no statistically significant differences between the 41.6 Gy and

50 Gy arms. Change in breast appearance, breast hardness, breast shrinkage, pain in area of

affected breast and oversensitivity in area of affected breast favoured the 50 Gy arm, while

change in skin appearance and swelling in area of affected breast favoured the 41.6 Gy arm.

Skin problems on or in area of affected breast had a hazard ratio of 1.01.

When the 39 Gy and 50 Gy arms in the START A trial were compared, all outcomes favoured the

39 Gy arm (with the exception of breast shrinkage which had a hazard ratio of 1.00 ). The only

outcome which statistically significantly favoured the 39Gy arm was change in skin appearance

(hazard ratio 0.63 95% CI 0.47, 0.84, p=0.0019).

In the START B trial, all outcomes favoured the 40 Gy arm, with the exception of oversensitivity

in area of affected breast which favoured the 50 Gy arm. The only statistically significant

difference was change in skin appearance, which favoured the 40 Gy arm (hazard ratio 0.76 95%

CI 0.60, 0.97; p=0.0262).


a systematic review

Figure 11 START A AND B: Forest plots of normal tissue effects assessed as moderate or marked by

patients, according to radiotherapy regimen6

Source: Hopwood et al 20106 Figure 2 page 6 Positions of squares in the forest plot show the estimate of the hazard ratio describing relative effect of the test schedule compared with control, with the 95% CI represented by horizontal lines. Squares to the left of the vertical line indicate when rates of adverse effects are lower in the test schedule compared with control; estimates to the right of the line indicate whether rates are higher in the test schedule. Size of squares is proportional to the precision of the estimate, with larger squares indicating greater precision. *In patients who had completed breast-conserving surgery only. (A) START trial A, 41.6 Gy vs 50 Gy. (B) START trial A, 39 Gy vs 50 Gy. (C) START Trial B, 40 Gy vs 50 Gy.


4.6.4 Conclusions

The results are summarised in Table 48. There was no statistically significant difference in the

majority of cosmetic outcomes assessed by the included publications.

RMH/GOC reported the risk of developing any late radiation effect was statistically significantly

lower for patients in the 39 Gy arm compared to the 50 Gy arm (p=0.01). For most clinically

assessed breast and arm outcomes estimated at 10 years, compared to the 50 Gy arm, there

were fewer events for patients in the 39 Gy arm and more in the 42.9 Gy arm.8

The START A trial reported that the 39 Gy arm was associated with significantly less mild or

marked change in photographic breast appearance (HR 0.69 95% CI 0.52, 0.91, p=0.01)4 and

change in skin appearance (HR 0.63 95% CI 0.47, 0.84, p=0.0019).6 The 40 Gy arm of the

START B trial was associated with significantly less change in skin appearance (40 Gy: HR 0.76

95% CI 0.60, 0.97, p=0.0262).6

In subgroup analyses for the START A and START B trials, the relative effects of the

randomised radiation schedules on patients reported symptoms did not vary significantly

according to type of primary surgery (breast conserving or mastectomy). . 8

These results should be considered in the context of the difference in cosmetic outcomes

assessed in each publication, the methods of assessing cosmetic outcome and the difference in

hypofractionated radiotherapy regimens.

Table 48 Summary of key results for cosmetic outcomes






39 Gy: adverse cosmetic outcomes were reported less frequently when compared to the 50 Gy arm (p=0.01)



No statistically significant differences in any cosmetic outcome

Any surgery




41.6 Gy: No statistically significant differences in any cosmetic outcome

39 Gy: No statistically significant differences in cosmetic outcome, with the exception of mild or marked change in breast appearance (HR 0.69 95% CI 0.52, 0.91, p=0.01)



0.77 (95% CI 0.61-0.98), p=0.02

Combined data from START A and B6^

As for START A and START B Change in skin appearance

39 Gy: HR 0.63 95% CI 0.47, 0.84, p=0.0019

40 Gy: HR 0.76 95% CI 0.60, 0.97, p=0.0262

Subgroup analysis by breast conserving surgery and mastectomy: NS

Abbreviations: CI=confidence interval, HR=hazard ratio, NS=not significant * control arm ^ Of the 4,451 patients enrolled in either the START A or START B trials, 2,208 patients were accrued into the quality-of-life study (1,129 from START A and 1,079 from START B)


a systematic review

4.7 Quality of life

The START A and START B trials reported quality of life outcomes .6


START A and B6







2,208 patients, 1831 had undergone breast-conserving surgery before radiotherapy (885 from

START A and 946 from START B).

Table 49 shows the results of three self-reported scales assessed using the EORTC (European

Organisation for Research and Treatment of Cancer) modules for subjects who underwent

breast-conserving surgery: the BR23 breast symptoms subscale, the BR23 arm or shoulder

symptoms subscale and the body image scale.6 There was no significant difference in outcomes

based on study arm. The low numbers of patients and events in some subgroups limited the

statistical power of these analyses.


Table 49 START A AND B: Breast, arm, or shoulder symptoms and body image scale scores at 5 yearsa according to radiotherapy regimen, type of primary surgery6

Study Arm Median (IQR)

BR23 breast symptoms subscale (0–100)

START A

50 Gy 8.3 (0–25.0)

41.6 Gy 8.3 (0–16.7)

39 Gy 8.3 (0–16.7)

START B 50 Gy 8.3 (0–16.7)

40 Gy 8.3 (0–16.7)

BR23 arm or shoulder symptoms subscale (0–100)

START A

50 Gy 11.1 (0–22.2)

41.6 Gy 11.1 (0–22.2)

39 Gy 11.1 (0–22.2)

START B 50 Gy 11.1 (0–22.2)

40 Gy 11.1 (0–22.2)

Body image scale (0–30)

START A

50 Gy 1.0 (0–5.0)

41.6 Gy 2.0 (0–7.0)

39 Gy 1.0 (0–5.0)

START B 50 Gy 1.0 (0–5.0)

40 Gy 1.0 (0–5.0)

Source: Hopwood et al 20106 Table 5 page 9 Higher scores indicate more symptoms or concerns. a Subgroup analyses undertaken with all follow-up data in generalised estimating equation models, but only 5-year data are shown for simplicity of presentation

4.7.2 Post mastectomy

START A and B6







2,208 patients, 377 had undergone a mastectomy before radiotherapy (244 from START A and

133 from START B).

Table 50 shows the results of three self-reported scales for subjects who underwent mastectomy:

the BR23 breast symptoms subscale, the BR23 arm or shoulder symptoms subscale and the

body image scale. There was no significant difference in outcomes based on study arm. The low

numbers of patients and events in some subgroups limited the statistical power of these

analyses.


a systematic review

Table 50 START A AND B: Breast, arm, or shoulder symptoms and body image scale scores at 5 yearsa according to radiotherapy regimen, type of primary surgery6

Study Arm Mastectomy Median (IQR)


START A

50 Gy 8.3 (0–20.8)

41.6 Gy 8.3 (0–25.0)

39 Gy 8.3 (0–22.9)

START B 50 Gy 8.3 (0–16.7)

40 Gy 8.3 (0–16.7)


START A

50 Gy 11.1 (5.6–33.3)

41.6 Gy 11.1 (0–33.3)

39 Gy 11.1 (0–22.2)

START B 50 Gy 11.1 (0–22.2)

40 Gy 11.1 (0–22.2)


START A

50 Gy 8.0 (2.2–15.5)

41.6 Gy 3.0 (0–8.0)

39 Gy 6.0 (1.0–11.0)

START B 50 Gy 4.0 (1.0–9.5)

40 Gy 4.0 (0–7.0)

Source: Hopwood et al 20106 Table 5 page 9 Higher scores indicate more symptoms or concerns. a Subgroup analyses undertaken with all follow-up data in generalised estimating equation models, but only 5-year data are shown for simplicity of presentation

4.7.3 Any surgery

START A and B6





Of the 4,451 patients enrolled in either the START A or START B trials, 2,208 patients entered

the quality-of life study (1,129 from START A and 1,079 from START B). Quality of life was

assessed using the European Organisation for Research and Treatment of Cancer (EORTC)

breast-cancer module BR23. This consists of six subscales, of which three were used in the

analysis: breast symptoms subscale (pain, swelling, oversensitivity, and skin problems in the

breast), arm subscale (swelling in arm or hand, arm or shoulder pain, and difficulty moving the

arm), and body image (containing four items which were not reported in the publication).

Table 51 shows the results of three self-reported scales: the BR23 breast symptom subscale, the

BR23 arm or shoulder symptoms subscale and the BR23 body image scale. There was no

significant differences between regimens for the breast symptom subscale (START A p=0.5558,

START B p=0.8757), the arm or shoulder symptoms subscale (START A p=0.2071, START B

p=0.3101) or the body image scale (START A p=0.9990, START B p=0.3405). For all


radiotherapy regimens across each of the three scales, scores declined significantly indicating

improved body image, between baseline and 60 months (p=<0.0001).

Table 51 START A AND B: Breast, arm, or shoulder symptoms and body image scale scores, according to radiotherapy regimen, over time from randomisation6

0 months

median (IQR) 6 months




median (IQR)


Trial A

50 Gy 16.7 (8.3–25.0) 16.7 (8.3–33.3) 16.7 (8.3–25.0) 8.3 (0–25.0) 8.3 (0–25.0)

41.6 Gy 16.7 (8.3–25.0) 16.7 (8.3–33.3) 16.7 (0–25.0) 8.3 (0–25.0) 8.3 (0–16.7)

39 Gy 16.7 (8.3–25.0) 16.7 (8.3–25.0) 16.7 (0–25.0) 8.3 (0–16.7) 8.3 (0–16.7)

Trial B

50 Gy 16.7 (8.3–25.0) 16.7 (8.3–33.3) 16.7 (8.3–25.0) 8.3 (0–16.7) 8.3 (0–16.7)

40 Gy 16.7 (8.3–25.0) 16.7 (8.3–33.3) 16.7 (0–25.0) 8.3 (0–16.7) 8.3 (0–16.7)


Trial A

50 Gy 22.2 (0–33.3) 11.1 (0–22.2) 11.1 (0–22.2) 11.1 (0–22.2) 11.1 (0–22.2)

41.6 Gy 11.1 (0–33.3) 11.1 (0–22.2) 11.1 (0–22.2) 11.1 (0–22.2) 11.1 (0–22.2)

39 Gy 22.2 (11.1–33.3) 11.1 (0–33.3) 11.1 (0–22.2) 11.1 (0–22.2) 11.1 (0–22.2)

Trial B

50 Gy 11.1 (11.1–33.3) 11.1 (0–22.2) 11.1 (0–22.2) 11.1 (0–22.2) 11.1 (0–22.2)

40 Gy 22.2 (11.1–33.3) 11.1 (0–33.3) 11.1 (0–22.2) 11.1 (0–22.2) 11.1 (0–22.2)


Trial A

50 Gy 3.0 (1.0–8.0) 3.0 (0–7.0) 3.0 (0–7.0) 2.0 (0–7.0) 2.0 (0–7.0)

41.6 Gy 4.0 (1.0–8.7) 2.0 (0–6.0) 2.0 (0–6.0) 2.0 (0–6.0) 2.0 (0–7.0)

39 Gy 4.0 (1.0–9.0) 2.0 (0–7.0) 2.0 (0–7.0) 2.0 (0–7.0) 2.0 (0–6.0)

Trial B

50 Gy 3.0 (0–8.0) 2.0 (0–6.0) 1.0 (0–5.0) 1.0 (0–5.7) 1.5 (0–6.0)

40 Gy 3.0 (0–7.0) 2.0 (0–6.0) 1.0 (0–5.0) 1.0 (0–5.0) 1.0 (0–5.0)

Source: Hopwood et al 20106 Table 2 page 5 Abbreviations: IQR=interquartile range Higher scores indicate more symptoms or concerns.

4.7.4 Conclusions

The START A and START B trials evaluated quality of life outcomes.6 There was no evidence

that any hypofractionated radiotherapy regimen was associated with a statistically significant

difference in quality of life score as measured by the BR23 breast symptom subscale.24

Subgroup analysis was performed, with results analysed by surgery type. There were no

statistically significant differences in outcomes, nor were any interaction tests significant overall.


regimes and single quality of life instrument used to assess this outcome.


a systematic review

5 Guidelines

5.1 Guidelines search

In order to identify current recommendations in existing radiotherapy guidelines, a systematic

search of guidelines was undertaken in March 2010. The nine guideline websites which were

searched are shown in Table 52. The same search terms were used for all websites. Manual

searching of reference lists was also performed. A total of three citations were identified. One set

of guidelines9, that was published after the literature search date, was included post hoc.

Table 52 Search terms for guidelines websites

Guideline websites Search terms Citations

Agency for Healthcare Research and Quality

“hypofractionated radiotherapy” OR “fractionated radiotherapy” OR “irradiation therapy” OR “irradiation treatment” OR “hypofractionated radiation treatment” OR “fractionated radiation treatment” OR “therapeutic radiology” OR (“breast cancer” AND radiotherapy)

0

EuroScan 0

Australia and New Zealand Horizon Scanning Network 0

Centre for Reviews and Dissemination 0

MSAC 0

National Guideline Clearinghouse 0

NHMRC 0

NHS Evidence 2

CADTH 0

Manual 1

Total 3

5.2 Results

The American Society for Radiation Oncology (ASTRO) guidelines on fractionation for

whole breast irradiation 20109

Based on the same body of evidence evaluated in this systematic review, ASTRO found that

evidence supports the equivalence of hypofractionated whole breast irradiation with

conventionally fractionated whole breast irradiation for patients who satisfy all these criteria:

Patient is 50 years or older at diagnosis.

Pathologic stage is T1-2 N0 and patient has been treated with breast-conserving surgery.

Patient has not been treated with systemic chemotherapy.

Within the breast along the central axis, the minimum dose is no less than 93% and

maximum dose is no greater than 107% of the prescription dose (±7%) (as calculated

with 2-dimensional treatment planning without heterogeneity corrections).

For patients who do not satisfy all of these criteria, the task force could not reach consensus and

therefore chose not to render a recommendation.


There were few data to define the indications for and toxicity of a tumour bed boost in patients

treated with hypofractionated radiotherapy. The task force agreed that the use of

hypofractionated radiotherapy alone (without a boost) is not appropriate when a tumour bed

boost is thought to be indicated. When a boost is indicated, there was lack of consensus

regarding the appropriateness of hypofractionated radiotherapy. Although the majority of the task

force members thought that there were sufficient data showing safety of hypofractionated

radiotherapy followed by a tumour bed boost to recommend its use in otherwise suitable patients,

a minority believed that conventional radiotherapy should be used instead when a tumour bed

boost is indicated.

For patients not receiving a tumour-bed boost, the task force favoured a dose of 42.5 Gy in 16

fractions over approximately 22 days when hypofractionated radiotherapy was planned. The

optimal hypofractionated radiotherapy regimen to use when a boost is given and the optimal

tumour-bed boost dose-fractionation to use in conjunction with hypofractionated radiotherapy

have not been determined.

Two-dimensional treatment planning with optimisation of dose homogeneity in the central axis is

the minimum acceptable standard for hypofractionated radiotherapy treatment planning.

However, CT-guided treatment planning using three-dimensional dose compensation is strongly

recommended to optimise dose homogeneity throughout the entire breast. As a conservative

measure, the task force recommended exclusion of the heart from the primary treatment fields

provided that coverage of the primary tumour site is not compromised.

The New Zealand Ministry of Health Guidelines for Management of Early Breast Cancer

200910

The New Zealand Ministry of Health guidelines made the following recommendation regarding

hypofractionated radiotherapy:

Recommendation

• Radiotherapy treatment for early invasive breast cancer should use an accepted regimen such

as:

50 Gy in 25 fractions over 5 weeks (Grade A)

45 Gy in 20 fractions over 5 weeks (Grade B)

42.5 Gy in 16 fractions over 3.5 weeks for those with small or medium breasts, not requiring

boost or nodal radiation (Grade B)

40 Gy in 15 fractions over 3 weeks* (Grade B)

Good practice points

• If boost radiotherapy is used after a hypofractionated regimen it should be at the standard 2 Gy

per fraction

• Women with large breasts and those with significant postoperative induration, oedema,

erythema, haematoma or infection should be considered for extended fractionation, with smaller

daily doses over 5–6 weeks

* It should be noted that the data for long-term follow-up in the latter three schedules of this recommendation is still awaited Notes Grades indicate the strength of the supporting evidence, rather than the importance of the recommendations A good practice point represents the opinion of the Guideline Development Team, or feedback from consultation within New Zealand where no evidence is available


a systematic review

A summary of the findings of the systematic review used to inform the guidelines is presented

below.

Survival

The SIGN guideline12

reported data from the Canadian trial that found no significant difference in

overall survival rate at the five-year follow-up with a hypofractionated compared with standard

regimen.7 The START Trialists Group reported that in the START B trial there were significant

differences in disease-free and overall mortality rates in favour of the group who received the

hypofractionated regimen (40 Gy in 15 fractions over 3 weeks).5 The authors anticipate that this

effect will diminish over time, and the long-term follow-up of the trial continues.

Loco-regional recurrence

In the Canadian trial study reported in the SIGN guideline, no significant difference in local

recurrence free rate at the five-year follow-up was seen (96.8% with 25 fractions vs 97.2% with

16 fractions; 95% CI 1.5–2.4).7 At the five-year follow-up, RMH/GOC reported hazard ratios

comparing 50 Gy to 42.9 Gy of 0.90 (95% CI 0.55–1.46) and 1.14 (95% CI 0.72–1.79) for 39 Gy

compared to 50 Gy.1 After 10 years, the probability of recurrence was significantly greater in the

39 Gy than in the 42.9 Gy group (difference 3.7%, 95% CI 0.3–8.3, p=0.027). RMH/GOC

concluded that the results were consistent with the hypothesis that fewer, larger fractions are at

least as safe and as effective as ‘standard’ regimens but that the shorter schedule should be

restricted to clinical trials.1 At the five-year follow-up, the START Trialists Group reported of the

START B trial that the absolute difference in loco-regional recurrence could be up to 1.7% better

and at most 1% worse with the hypofractionated regimen. The trial authors concluded that the

delivery of 40 Gy in 15 fractions appeared to result in a loco-regional recurrence rate that was at

least as favourable as the ‘standard’ 50 Gy in 25 fractions.5

Other outcomes

Cosmetic results at five years were similar between fractionation schedules. However, in a 12-

year update of the Canadian trial data, the incidence of moderate to severe late radiation

morbidity (subcutaneous fibrosis) at 10 years doubled (8% vs 4%)††

in the shorter fractionation

schedule.7 The START A trial reported in a quality of life assessment that changes in breast

appearance and breast hardness were the most commonly reported side effects. These side

effects were less marked in the 39 Gy group and similar in the 41.6 Gy and 50 Gy groups, in

contrast to those found at 10 years by RMH/GOC. (11.2% for 42.9 Gy in 13 fractions of 3.3 Gy vs

6.4% for 50 Gy in 25 fractions of 2 Gy). Both the START A and START B trials reported that the

follow-up period of five years was too short to assess potential late normal tissue effects, such as

cardiac damage.4-5

Follow-up continues for these trials. The long-term safety of the short fractionation schedule for

the nodal areas has not been established.

Conclusions

Based on a review of the published evidence it was noted that there is currently insufficient

evidence to identify one optimum fractionation schedule. The results of ongoing clinical trials will

††

The 8% vs 4% values come directly from the text of the The New Zealand Ministry of Health Guidelines. It has been noted that these values are not consistent with the results reported in Whelan 2010, however the The New Zealand Ministry of Health Guidelines do not cite a page or table reference, or note if this data was calculated post hoc or obtained from the authors.


inform guidelines in the future. To minimise late tissue damage whilst maximising tumour control,

the Guideline Development Team supported the administration of boost dose radiotherapy at 2

Gy per fraction where indicated following a hypofractionated regimen. The Guideline

Development Team also noted that extended fractionation with smaller doses over five to six

weeks should be considered in women with large breasts and postoperative side effects.

NICE 2009 guidelines for early and locally advanced breast cancer11

The NICE 2009 guidelines made the following recommendation regarding dose fractionation:

Recommendation

• Use external beam radiotherapy giving 40 Gy in 15 fractions as standard practice for patients

with early invasive breast cancer after breast conserving surgery or mastectomy.

Qualifying statement: This recommendation is based on RCT evidence of clinical effectiveness

and the guideline development group agreeing that a regimen using fewer fractions would

probably be cost effective.

The summary of the clinical evidence noted that rates of local recurrence were not significantly

different between conventional 50 Gy fractions and hypofractionated schedules. One study

(START B) found that distant relapse was lower in the hypofractionated arm which improved the

rates of disease-free survival and overall survival.5 Assessments of cosmetic outcomes were less

consistent, and depended on the comparisons made. One RCT7 reported no significant

difference between the 50 Gy and 42.5 Gy arms, whilst another 8 reported a significantly poorer

cosmetic outcome in the 42.9 Gy arm when compared to the 39 Gy arm. The hazard ratio for no

change in breast appearance was significantly improved in the 39 Gy arm of the START A trial4

compared to 50 Gy; whilst there was no difference between the 50 Gy and 41.6 Gy arms in

START A4 or between 50 Gy and 40 Gy in START B

5.

Global cosmetic outcomes were also less consistent since effects were reported at different

times and between different fractionation doses. Breast oedema, fibrosis, lymphoedema and

telangiectasia were reported in few studies.

The guidelines advised that careful treatment planning is required for all patients to avoid

potential hotspots in the breast but this may be particularly important with hypofractionated

schedules. Patients with breast reconstruction/augmentation or large breast size may have a

better cosmetic result using conventional dose radiotherapy of 50 Gy in 25 fractions (lower dose

per fraction), although 3D radiotherapy planning may make hypofractionated regimens

equivalent.

The use of hypofractionated regimes should result in considerable saving of resources, both

human and financial.

SIGN 200512

The Scottish Intercollegiate Guidelines Network (SIGN) management of breast cancer in women

guidance paper was developed in 2005, prior to the publication of a number of key RCTs (such

as the START trials). No formal recommendations were made.


a systematic review

The guidelines noted that “current evidence is not able to identify an optimal dose/fractionation

for post-operative radiotherapy. It is therefore reasonable to treat patients with currently accepted

regimens such as 50 Gy in 25 daily fractions over five weeks, 45 Gy in 20 fractions, or 40 Gy in

15 or 16 fractions. Results of ongoing trials investigating fractionated are awaited”.


6 Conclusions

A systematic review of the literature identified five clinical trials of hypofractionated radiotherapy

for the treatment of early breast cancer. Two trials were in patients who had undergone breast

conserving surgery (RMH/GOC and Canadian trials)1-2, 7-8

and three were in patients who had

undergone any form of surgery (START A, START B and Spooner)3-5

. The studies evaluated a

number of different hypofractionated radiotherapy regimens.

All five trials reported local recurrence rates (RMH/GOC, Canadian, Spooner, START A and

START B trials). There was no evidence that any hypofractionated radiotherapy regimen was

associated with a statistically significantly difference in local recurrence rate when compared to a

control arm. The Canadian trial reported no difference when subgroups were analysed, with the

exception of tumour grade. The impact of the 42.5 Gy regimen on local recurrence was less in

patients with high-grade tumours compared to patients with low-grade tumours (p=0.01).2

Two studies (START A and START B) reported local-regional and distant relapse. The studies

found no significant difference in local-regional relapse.4-5

START B reported that the 40 Gy

study arm had a statistically significantly lower rate of distant relapse when compared with the

control arm (p=0.01).5

Four studies reported overall survival (Canadian, START A, START B and Spooner). Most

studies reported that there was no evidence that hypofractionated radiotherapy was associated

with a statistically significantly difference in overall survival. The START B study found that 40 Gy

in 15 fractions over three weeks was associated with a statistically significantly lower all-cause

mortality rate when compared with 50 Gy in 25 fractions over five weeks (p=0.03).5 Therefore,

there was no evidence that any hypofractionated radiotherapy regimen was associated with a

worse overall survival rate (i.e. the only study that reported a significant difference showed lower

mortality for patients treated with hypofractionated radiotherapy).

Three studies reported adverse events and toxicity outcomes (Canadian, START A and START

B). There were generally no differences in adverse events and toxicity outcomes between the

study arms. START A and START B noted a statistically significant difference in skin appearance

(favouring the 39 Gy arm, p=0.0019 and 40 Gy arm, p=0.0262 when compared with the control

arm).6

Four studies reported cosmetic outcomes (RMH/GOC, Canadian, START A and START B).

Some studies reported differences in cosmetic outcomes between study arms. The RMH/GOC

trial reported that the risk of developing any late radiation effect was statistically significantly

lower for patients in the 39 Gy arm compared to the 50 Gy arm (p=0.01), but more often in the

42.9 Gy arm compared to the control (p=0.05).8 START A and START B reported that changes

in skin appearance were less frequent in patients receiving 39 Gy in 13 fractions (p=0.0019) and

patients receiving 40 Gy in 13 fractions (p=0.0262), when compared to 50 Gy in 25 fractions.6

START A and START B reported quality of life outcomes. There was no evidence that any

hypofractionated radiotherapy regimen was associated with a statistically significant difference in

quality of life score as measured by the BR23 breast symptom subscale. There was no

difference in cosmetic outcomes or quality of life when outcomes were evaluated by surgery

type.6


a systematic review

Overall, there was no evidence that the evaluated hypofractionated radiotherapy regimens were

associated with an increased rate of local recurrence or reduced survival. There were some

differences in adverse events and toxicity, as well as cosmetic outcomes.

All of the reported outcomes need to be considered in the context of range of hypofractionated

radiotherapy regimens that were evaluated. Although 50 Gy in 25 fractions was used as a control

arm in all trials, no two trials compared the same two radiotherapy regimens, making

comparisons difficult. The study design should also be considered. Not all studies were powered

to detect a difference in all outcomes. RMH/GOC1, 8

was powered to detect late change in breast

appearance and the Canadian2, 7

trials were powered to detect a difference in local recurrence.

START A4 and START B

5 were powered to detect a difference in local-regional tumour relapse

rate. The included studies had a follow-up period of 6 – 10 years. This length of follow-up should

be sufficient to detect a difference in outcome.

The generalisablity the findings must also be considered. For example, the Canadian trial

excluded patients who had node-positive invasive carcinoma, or patients with a breast width of

more than 25 cm.7 It may be that the results of this study may not apply to women with ductal

carcinoma in situ only, women with node-positive cancer or women with large breasts.

The literature review did not identify any studies which specifically assessed the use of

hypofractionated radiotherapy in conjunction with chemotherapy or other biological therapies.

The Canadian trial assessed local recurrence by systematic therapy use, and found no difference

recurrence rates.2

Overall, there was no evidence that any hypofractionated radiotherapy regimen was associated

with a statistically significantly difference in local recurrence rate or a significantly worse overall

survival rate. There were some differences in adverse event, toxicity and cosmetic outcomes,

although this was not consistent across all hypofractionated radiotherapy protocols. These

results should be considered in the context of the included patient populations, statistical power

of the studies, different hypofractionated radiotherapy regimens used and length of follow-up.


7 References

1. Owen JR, Ashton A, Bliss JM, et al. Effect of radiotherapy fraction size on tumour control in patients

with early-stage breast cancer after local tumour excision: long-term results of a randomised trial.

Lancet Oncol 2006;7(6):467-71.

2. Whelan TJ, Pignol JP, Levine MN, et al. Long-term results of hypofractionated radiation therapy for

breast cancer. N Engl J Med 2010;362(6):513-20.

3. Spooner D, Stocken DD, Jordan S, et al. A randomised controlled trial to evaluate both the role and

optimal fractionation of radiotherapy in the conservative management of early breast cancer. San

Antonio Breast Cancer Symposium, 2008.

4. Bentzen SM, Agrawal RK, Aird EG, et al. The UK Standardisation of Breast Radiotherapy (START)

Trial A of radiotherapy hypofractionation for treatment of early breast cancer: a randomised trial.

Lancet Oncol 2008;9(4):331-41.

5. Bentzen SM, Agrawal RK, Aird EG, et al. The UK Standardisation of Breast Radiotherapy (START)

Trial B of radiotherapy hypofractionation for treatment of early breast cancer: a randomised trial.

Lancet 2008;371(9618):1098-107.

6. Hopwood P, Haviland JS, Sumo G, et al. Comparison of patient-reported breast, arm, and shoulder

symptoms and body image after radiotherapy for early breast cancer: 5-year follow-up in the

randomised Standardisation of Breast Radiotherapy (START) trials. Lancet Oncol 2010;11(3):231-40.

7. Whelan T, MacKenzie R, Julian J, et al. Randomized trial of breast irradiation schedules after

lumpectomy for women with lymph node-negative breast cancer. J Natl Cancer Inst

2002;94(15):1143-50.

8. Yarnold J, Ashton A, Bliss J, et al. Fractionation sensitivity and dose response of late adverse

effects in the breast after radiotherapy for early breast cancer: long-term results of a randomised trial.

Radiother Oncol 2005;75(1):9-17.

9. Smith BD, Bentzen SM, Correa CR, et al. Fractionation for Whole Breast Irradiation: An American

Society for Radiation Oncology (ASTRO) Evidence-Based Guideline. Int J Radiat Oncol Biol Phys

2010;July 15(Epub ahead of print):1-10.

10. New Zealand Guidelines Group (NZGG). Management of early breast cancer. NZGG: Wellington

New Zealand, 2009.

11. National Institute for Health and Clinical Excellence (NICE). Early and locally advanced breast

cancer: diagnosis and treatment. NICE: UK, 2009.

12. Scottish Intercollegiate Guidelines Network (SIGN) Management of breast cancer in women: A

national clinical guideline. SIGN: Edinburgh Scotland, 2005.

13. National Breast Cancer Centre. Clinical practice guidelines for the management of early breast

cancer (2nd edition). Commonwealth of Australia: Canberra, 2001.

14. James ML, Lehman M, Hider PN, et al. Fraction size in radiation treatment for breast conservation

in early breast cancer. Cochrane Database Syst Rev 2008;3):CD003860.

15. NIH consensus conference. The treatment of early breast cancer. JAMA 1991;265(3):391-5.

16. Ash D. Waiting times for cancer treatment. Clin Oncol (R Coll Radiol) 2000;12(3):140.

17. Mackillop WJ, Fu H, Quirt CF, et al. Waiting for radiotherapy in Ontario. Int J Rad Oncol, Biol,

Phys 1994;30(1):221-8.

18. Shahid A, Athar MA, Asghar S, et al. Post mastectomy adjuvant radiotherapy in breast cancer: A

comparision of three hypofractionated protocols. J Pak Med Assoc 2009;59(5):282-87.

19. Dwyer P, Hickey B and Burmeister B. Hypofractionated breast radiotherapy: impact on

departmental waiting times and cost. J Med Imag Radiat Oncol 2009;54(3):229-34.


a systematic review

20. Taher AN, El-Baradie MM, Essa H, et al. Hypofractionation versus conventional fractionation

radiotherapy after conservative treatment of breast cancer: early skin reactions and cosmetic results.

J Egypt Natl Canc Inst 2004;16(3):178-187.

21. National Health and Medical Research Council. How to use the evidence: assessment and

application of scientific evidence. Canberra: Commonwealth of Australia, 2000.

22. National Health and Medical Research Council. NHMRC additional levels of evidence and grades

for recommendations for developers of guidelines. Commonwealth of Australia: Canberra, 2009.

23. Egger M, Smith GD and Sterne JA. Uses and abuses of meta-analysis. Clin Med 2001;1(6):478-

84.

24. Sprangers MA, Groenvold M, Arraras JI, et al. The European Organization for Research and

Treatment of Cancer breast cancer-specific quality-of-life questionnaire module: first results from a

three-country field study. J Clin Oncol 1996;14(10):2756-2768.


Appendix A Contributors

Working group members

The following people were members of the working group:

Associate Professor Boon Chua (Chair) Radiation oncologist

Dr Marie-Frances Burke Radiation oncologist

Professor Geoff Delaney Radiation oncologist

Dr Jane O’Brien Surgeon

Ms Jan Rice Breast care nurse

Ms Geraldine Robertson Consumer representative

Dr Kirsty Stuart Radiation oncologist

Cancer Australia staff

The following Cancer Australia staff were involved in the project:

Ms Katrina Anderson Project Officer – Research

Ms Phillipa Hastings Project Officer

Dr Anne Nelson Evidence Review & Research Leader

Ms Sue Sinclair General Manager

Ms Heidi Wilcoxon Program Manager

Documents

Hypofractionated radiotherapy for early (operable) breast ... · Hypofractionated radiotherapy for early (operable) breast cancer: technical document Part 1 - Hypofractionated Radiotherapy