Helical tomotherapy simultaneous integratedboost provides a dosimetric advantage in thetreatment of primary intracranial tumors

Joseph M Baisden1, Jason Sheehan2, Andrew G Reish3, Alyson F McIntosh1,Ke Sheng1, Paul W Read1, Stanley H Benedict1, James M Larner1

1Department of Radiation Oncology, University of Virginia, Charlottesville, VA, USA, 2Department of NeurologicalSurgery, University of Virginia Health System, Charlottesville, VA, USA, 3Medical School, University of Virginia,Charlottesville, VA, USA

Objective: The research quantitatively evaluates the dosimetric advantage of a helical tomotherapy (HT)intensity-modulated radiation therapy simultaneous integrated boost (SIB) compared to a conventional HTsequential (SEQ) boost for primary intracranial tumors.Methods: Hypothetical lesions (planning target volumes or PTVs) were contoured within computedtomography scans from normal controls. A dose of 50 Gy was prescribed to the larger PTV1, while theboost PTV2 received a total of 60 Gy. HT SEQ and HT SIB plans were generated and compared. Weevaluated the mean brain dose, the volume of normal brain receiving 45 Gy (V45), the volume of normalbrain receiving 5 Gy (V5), and the integral dose. In addition, patients who were treated with the HT SEQtechnique were replanned with the HT SIB technique and compared.Results: The average reduction in mean brain dose with the HT SIB plan compared to the composite HTSEQ plan was 11.0% [standard error (SE): 0.5]. The reductions in brains V45 and V5 were 43.7% (SE: 2.3)and 3.9% (SE: 0.6), respectively. The reduction in the integral dose was 11.0% (SE: 0.5). When comparingthe SIB plan to the first 50 Gy only of the SEQ plan, there was only a 2.5% increase in the mean brain doseand a 2.9% increase in brain V45. This increase was dependent on the relative volumes of PTV2 and PTV1.These results were confirmed for the patient plans compared.Conclusions: Treating primary brain tumors with the HT SIB technique provides significant sparing ofnormal brain parenchyma compared to a conventional HT SEQ boost.

Keywords: Brain sparing, Radiation-induced encephalopathy, Simultaneous integrated boost, Tomotherapy

IntroductionPrimary brain tumors have an incidence of 6.6 per

100 000, and this translates into 20 000 new cases per

year in the United States.1 A significant percentage of

these patients receive radiation therapy, which is

frequently delivered with an intensity-modulated

technique in order to decrease the dose to normal

brain and the associated likelihood of radiation-

induced neurocognitive decline. Although the mechan-

ism of radiation-induced central nervous system

toxicity is not well defined, the likelihood of adverse

central nervous system effects from brain radiation is

related to total dose, dose per fraction, and the volume

of normal brain irradiated.2–4

Conventional radiation treatment schedules for

primary brain tumors often involve the treatment of

an initial volume followed by a sequential (SEQ)

boost. These two treatments combined into one

constitute a simultaneous integrated boost (SIB).

Previous research has evaluated the feasibility of an

SIB with helical tomotherapy (HT SIB) in the

treatment of metastatic brain lesions.5,6 These studies

have shown that SIB by HT is able to deliver a

homogeneous brain dose and a fairly conformal dose

to the metastasis with surgical precision. In addition,

SIB offers other advantages compared to SEQ

treatment. First, SIB delivers a higher biologically

equivalent dose to the cancer cells which may result in

an advantage with regard to tumor control.7,8 In

addition, there is a theoretical dosimetric gain by

using the SIB compared to SEQ treatment. In the

ideal inverse optimization, a step function-shaped

dose is delivered to the boost volume without spillage

to the surrounding tissue. However, this is not

achievable in reality due to the lack of a ‘negative

Correspondence to: Jason Sheehan, Department of Neurological Surgery,University of Virginia Health System, Box 800212, HSC, Charlottesville,VA 22908, USA. Email: jps2f@virginia.edu

820� W. S. Maney & Son Ltd 2011DOI 10.1179/1743132811Y.0000000035 Neurological Research 2011 VOL. 33 NO. 8

dose’, which essentially undoes the dose delivered to

non-target tissue.9–12 However, the SIB technique can

create an effective ‘negative dose’ by reducing the dose

delivered to the initial target volume by an amount

that will match the dose delivered in the boost volume.

Therefore, we hypothesized that SIB would result in a

more conformal dose distribution than an SEQ

treatment plan. In this research, we investigated the

ability of HT SIB to spare normal brain parenchyma

compared to conventional HT SEQ techniques for

treatment of primary brain tumors.

Materials and MethodsComputed tomography scans of patients in the supine

position with an aquaplast immobilization mask were

simulated for treatment planning. Normal tissues were

contoured using AcQsim software (Philips Medical

Systems North America, Bothell, WA, USA). Hypo-

thetical planning target volumes (PTVs) consisting of

cylindrical structures of similar diameter and height

were placed at the central portion of the brain. For

every hypothetical lesion, two PTVs were created.

PTV1 corresponded to the initial larger treatment

volume of a traditional SEQ plan, based on the

peritumoral edema in the case of a typical high-grade

glioma. The boost volume, or PTV2 volume, corre-

sponded to the contrast-enhanced tumor plus a margin.

Multiple sizes of PTVs were evaluated (i.e. varied

combinations of PTV1 and PTV2 were employed).

Planning was performed using HI-Art Helical

Tomotherapy inverse planning software (Middleton,

WI, USA). SEQ plans were created by preparing

two separate doses: 50 Gy to PTV1 and 10 Gy to

PTV2. Composite dosimetry was performed using

proprietary HT software. For the evaluation of the

SIB in the treatment of primary brain tumors, a dose

of 50 Gy was prescribed to PTV1 and a dose of

60 Gy to PTV2. Both the SEQ and SIB plans were

optimized with similar constraints on normal tissue

to minimize any potential planning bias. A total of

42 comparisons between the two techniques were

analyzed.

Several dosimetric parameters were analyzed for

plan comparisons, including the mean and median

dose to the brain as well as the volume of brain

receiving 45 Gy (V45) and the volume of brain

receiving 5 Gy (V5). The mean dose was also

evaluated. Analysis of the dose to the normal brain

was performed using structures corresponding to the

entire brain minus the PTV1 and PTV2. The volume

for determining the integral dose was defined as the

external contour of the head with no internal

structures subtracted. Analysis and graphing of these

data were performed using the Microsoft Excel Data

Analysis Pak (Microsoft, Seattle, WA, USA) and

SPSS (SPSS, Inc., Chicago, IL, USA). MatLab

(Mathwork, Natick, MA, USA) was used in the

generation of color scale images demonstrating

differences in delivered dose.

To further confirm the results, three patients treated

for glioblastomas who had previously received HT

SEQ treatment for primary brain tumors were

replanned using the HT SIB technique for compar-

ison. In the HT SEQ plans that were delivered, all

patients received a combined total dose of 60 Gy in 30

fractions. HT SIB planning used similar dosimetric

constraints to the corresponding patient’s HT SEQ

plan in order to achieve a valid comparison.

Figure 1 Dosimetric comparison of the sequential (SEQ) and simultaneous integrated boost (SIB) plans. Panel A includes

dose volume histograms comparing helical tomotherapy (HT) SEQ and HT SIB plans for the treatment of a 5-cm planning target

volume (PTV) with 50 Gy (PTV1) and a 4-cm boost PTV with a total of 60 Gy (PTV2). Panel B is a representation of the

differences in absolute dose (Gy) between the HT SIB and HT SEQ plans, displayed in color. Brighter red indicates higher dose

delivered by the HT SEQ plan. This shows one axial image comparing the plans used to generate panel A.

Baisden et al. Brain sparing with tomotherapy integrated boost

Neurological Research 2011 VOL. 33 NO. 8 821

ResultsDifferences observed in simulated dose plansA representative comparison of the two types of

plans is included in Fig. 1. Panel A demonstrates that

the dose volume histogram line for the normal brain

is shifted to the left for the HT SIB plan, indicating

that the normal brain is spared at all doses compared

to the HT SEQ plan. The color wash (panel B)

demonstrates the absolute difference in dose delivered

between the two plans for one axial slice. Note that

the greatest sparing with the HT SIB plan compared

to the HT SEQ plan occurs just outside PTV2 and

PTV1 in the surrounding normal brain tissue at high

doses.

Figure 2 compares the SIB plans to the composite

SEQ plans using several standard dosimetric end-

points and illustrates significant sparing of normal

tissues in favor of the SIB technique. Compared to

SEQ, treatment via SIB resulted in decreases in the

mean and median brain doses of 11.0% [standard

error (SE): 0.5%] and 11.9% (SE: 0.5%), respectively.

The mean decrease in the V45 for the SIB technique

was 43.7% (SE: 2.3%), and in the V5, the reduction

was 3.9% (SE: 0.6%). These results indicate a greater

percentage of sparing at higher doses, which is also

shown in Fig. 1B. The mean decrease in the integral

dose was 11.0% (SE: 0.5%).

Given the significant sparing observed with HT

SIB (Figs. 1 and 2) treatment, we calculated the

magnitude of the dose increase to normal brain

during delivery of 60 Gy using an SIB technique

compared to 50 Gy with HT and no boost. Figure 3

demonstrates that delivering 60 Gy with the SIB

technique increases normal brain doses relatively

minimally compared to only the first 50 Gy of the

SEQ technique. The mean, V45, and integral brain

dose increases were 2.5% (SE: 0.5), 2.9% (SE: 0.5),

and 4.5% (SE: 0.8), respectively. Thus, a 10-Gy boost

can be delivered to a primary brain tumor using the

SIB technique with only a very minor increase in dose

to normal brain tissue.

In order to determine whether these increases were

dependent upon the relative volumes of PTV2 and

PTV1, we generated a scatter plot including the data

points of the 22 comparisons made on PTV2/PTV1

ratios ranging from 0.2 to 0.8 (Fig. 4). A linear

regression analysis of these data points was plotted as

a best-fit line; this showed an acceptable goodness of

fit (R250.831 for the mean dose and R250.81 for the

integral dose). Thus, for clinical situations wherein

the PTV2/PTV1 ratio is equal to 0.5 or less, the

increases in mean brain dose and integral dose

approach clinical insignificance provided that the

5% integral dose difference is used as a threshold.

Differences observed in three glioblastomapatientsTo determine whether the results from Figs. 1 and 2

which were based on hypothetical lesions accurately

reflect the results seen in patients with primary brain

tumors, we analyzed dosimetric data for three

glioblastoma patients who received HT SEQ treat-

ments and were later replanned using the HT SIB

technique for comparison (Table 1). In these three

patients, the SIB plans spared the normal brain in a

manner similar to the simulated results above, with

the highest decrease observed for brain V45. Mean

brain dose sparing ranged from 11 to 14%. Median

brain dose sparing ranged from 9 to 21%. Mean brain

V45 sparing ranged from 48 to 86%. Integral dose

sparing ranged from 4 to 12%. No sparing of the

brain V5 was observed. PTV coverage with both

planning techniques met the criteria that 95% of the

volume received the prescription doses of 50 and

60 Gy for PTV2 and PTV1, respectively. PTV1 doses

are 5% higher between the HT SEQ composite and

HT SIB plans. PTV2 is more inhomogeneous with the

SIB plans compared to SEQ plans. However, as the

Figure 2 Percentage reductions in the mean, V45, V5, and

integral doses to the normal brain using the helical

tomotherapy (HT) simultaneous integrated boost (SIB)

technique compared to the HT sequential (SEQ) technique.

Error bars show ¡standard error (SE).

Figure 3 Percentage increase in dose with helical tomother-

apy (HT) simultaneous integrated boost (SIB) plans com-

pared to the first 50 Gy only of the HT sequential (SEQ) plans.

Percentage increases in the mean dose to the brain, brain

V45, and the integral dose are shown. Error bars show

¡standard error (SE).

Baisden et al. Brain sparing with tomotherapy integrated boost

822 Neurological Research 2011 VOL. 33 NO. 8

measured integral dose decreases in the SIB treatment

compared to SEQ, the overall dosimetric gain in

normal brain outweighs the additional dose delivered

to the PTV.

DiscussionPrior reports have indicated that the use of an

intensity-modulated radiation therapy integrated

boost provides dosimetric advantages compared to

an SEQ boost, including shorter treatment time,

potential radiobiological gains, and normal tissue

sparing.13–18 Our HT-based data for primary brain

tumor patients are consistent with these reports. The

clinical significance of this dose sparing is unknown,

but the reduction in unwanted high dose spillage

could potentially reduce the risk of radiation-induced

edema and/or necrosis. The reduction in low dose

spillage to the entire brain, coupled with the

reduction in high dose spillage, may decrease the

chance of late cognitive deficits. In addition, the SIB

technique has the advantage that more of the total

dose is placed in the boost volume, since the PTV2

maximum dose is higher with the SIB than with the

SEQ technique. This increase in the maximum dose

reflects an increase in dose inhomogeneity in both the

PTV1 and PTV2 with the HT SIB technique, which

may increase the likelihood of tumor control while

only minimally increasing the risk for normal tissue.

As hypothesized, the conformality was improved

with SIB compared to SEQ treatment. The improve-

ment can be understood from the principle of inverse

treatment optimization. The more degrees of freedom

the optimization has, the more likely a clinician is to

reach the global optimum. In the case of SEQ, the

optimization is performed in two smaller subspaces

that do not communicate with each other. Therefore,

the solution is inferior to the SIB, which is able to

optimize in a larger, combined solution space. An

intuitive example of this is the aforementioned

negative dose. In order to deliver a perfectly

conformal boost dose, no dose spillage would occur

to the surrounding normal brain tissue. Theoretically,

a ‘negative dose’ could be delivered to the normal

tissue to offset the dose spillage from the boost

treatment. However, since the ‘negative dose’ cannot

actually be delivered, the boost plan in SEQ has to

incorporate a higher dose spillage compared to the

SIB plan, in which the boost can actually deliver a

‘negative dose’ by subtracting dose from the radiation

dose to the PTV1. To obtain such a theoretical

advantage, an inverse optimization routine has to be

able to explore the solution space thoroughly. HT,

with its ability to optimize radiation using 2pi degrees

and tens of thousands of beamlets, appears to be able

to convert the theoretical advantage into practical

gains. With the SIB plan, the software is able to

Figure 4 Percentage increases in mean brain dose and integral dose versus PTV2/PTV1 ratio with helical tomotherapy (HT)

simultaneous integrated boost (SIB) plans compared to the first 50 Gy only of the HT sequential (SEQ) plans. Panel A shows the

percentage increase in the mean brain dose. Data points show the percentage increase in the mean normal brain dose versus

the ratio of the volumes of PTV2/PTV1. A fit line is included (goodness of fit R250.831). Panel B shows the percentage increase

in the integral dose. A fit line is included (goodness of fit R250.81).

Table 1 Normal tissue doses seen using the simultaneousintegrated boost (SIB) technique compared to the compositeplan from the sequential (SEQ) technique

SIB versus SEQ composite

Patient no.Absolute

change (Gy) % change

Mean braindose

1 22.88 2112 22.97 2103 24.7 214

Brain V45 1 25.9 2582 25.2 2483 222.5 286

Brain V5 1 21 212 20.5 213 z0.9 z1

Integral dose 1 20.78 242 21.16 253 23.01 212

*Negative values indicate a decrease in dose with the SIBtechnique compared to the composite SEQ plan.

Baisden et al. Brain sparing with tomotherapy integrated boost

Neurological Research 2011 VOL. 33 NO. 8 823

account for the dose delivered to both PTV1 and

PTV2 and integrate this data into a single plan,

resulting in improved dosimetry. Conversely, the

SEQ technique does not integrate the information

from the two data sets into a composite plan;

therefore, the result is a significant decrease in

conformality, resulting in inferior dosimetry.

ConclusionsIn summary, using the HT SIB technique provides a

reduction in the radiation dose to the normal brain

compared to using an SEQ technique. Although the

clinical benefits of the use of the SIB technique are

unknown, it has the potential to decrease the likelihood

of adverse neurocognitive effects, particularly in patients

with good performance status and life expectancies.

References1 Fisher JL, Schwartzbaum JA, Wrensch M, Wiemels JL.

Epidemiology of brain tumors. Neurol Clin 2007;25:867–90.2 Crossen JR, Garwood D, Glatstein E, Neuwelt EA.

Neurobehavioral sequelae of cranial irradiation in adults: areview of radiation-induced encephalopathy. J Clin Oncol1994;12:627–42.

3 Klein M, Heimans JJ, Aaronson NK, van der Ploeg HM, Grit J,Muller M, et al. Effect of radiotherapy and other treatment-related factors on mid-term to long-term cognitive sequelae inlow-grade gliomas: a comparative study. Lancet 2002;360:1361–8.

4 Surma-aho O, Niemela M, Vilkki J, Kouri M, Brander A,Salonen O, et al. Adverse long-term effects of brain radio-therapy in adult low-grade glioma patients. Neurology2001;56:1285–90.

5 Bauman G, Yartsev S, Coad T, Fisher B, Kron T. Helicaltomotherapy for craniospinal radiation. Br J Radiol2005;78:548–52.

6 Gutierrez AN, Westerly DC, Tome WA, Jaradat HA, MackieTR, Bentzen SM, et al. Whole brain radiotherapy withhippocampal avoidance and simultaneously integrated brainmetastases boost: a planning study. Int J Radiat Oncol BiolPhys 2007;69:589–97.

7 Lauve A, Morris M, Schmidt-Ullrich R, Wu Q, Mohan R,Abayomi O, et al. Simultaneous integrated boost intensity-modulated radiotherapy for locally advanced head-and-necksquamous cell carcinomas: II–clinical results. Int J RadiatOncol Biol Phys 2004;60:374–87.

8 Wu Q, Mohan R, Morris M, Lauve A, Schmidt-Ullrich R.Simultaneous integrated boost intensity-modulated radiother-apy for locally advanced head-and-neck squamous cell carci-nomas. I: dosimetric results. Int J Radiat Oncol Biol Phys2003;56:573–85.

9 Brahme A. Optimization of stationary and moving beamradiation therapy techniques. Radiother Oncol 1988;12:129–40.

10 Holmes TW, Mackie TR, Reckwerdt P. An iterative filteredbackprojection inverse treatment planning algorithm fortomotherapy. Int J Radiat Oncol Biol Phys 1995;32:1215–25.

11 Mackie TR, Holmes T, Swerdloff S, Reckwerdt P, Deasy JO,Yang J, et al. Tomotherapy: a new concept for the delivery ofdynamic conformal radiotherapy. Med Phys 1993;20:1709–19.

12 Holmes MA, Mackie TR, Sohn W, Reckwerdt PJ, Kinsella TJ,Bielajew AF, et al. The application of correlated sampling tothe computation of electron beam dose distributions inheterogeneous phantoms using the Monte Carlo method.Phys Med Biol 1993;38:675–88.

13 Li XA, Wang JZ, Jursinic PA, Lawton CA, Wang D.Dosimetric advantages of IMRT simultaneous integrated boostfor high-risk prostate cancer. Int J Radiat Oncol Biol Phys2005;61:1251–7.

14 Guerrero M, Li XA, Ma L, Linder J, Deyoung C, Erickson B.Simultaneous integrated intensity-modulated radiotherapyboost for locally advanced gynecological cancer: radiobiologi-cal and dosimetric considerations. Int J Radiat Oncol Biol Phys2005;62:933–9.

15 Guerrero M, Li XA, Earl MA, Sarfaraz M, Kiggundu E.Simultaneous integrated boost for breast cancer using IMRT: aradiobiological and treatment planning study. Int J RadiatOncol Biol Phys 2004;59:1513–22.

16 Mott JH, Livsey JE, Logue JP. Development of a simultaneousboost IMRT class solution for a hypofractionated prostatecancer protocol. Br J Radiol 2004;77:377–86.

17 Dogan N, King S, Emami B, Mohideen N, Mirkovic N,Leybovich LB, et al. Assessment of different IMRT boostdelivery methods on target coverage and normal-tissue sparing.Int J Radiat Oncol Biol Phys 2003;57:1480–91.

18 Fogliata A, Bolsi A, Cozzi L, Bernier J. Comparativedosimetric evaluation of the simultaneous integrated boostwith photon intensity modulation in head and neck cancerpatients. Radiother Oncol 2003;69:267–75.

Baisden et al. Brain sparing with tomotherapy integrated boost

824 Neurological Research 2011 VOL. 33 NO. 8