20
Exposure to cyclo- oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova , Carol Coupland, Julia Hippisley-Cox

Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Embed Size (px)

Citation preview

Page 1: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer:nested case-control studies

IAE world Congress Epidemiology 2011 Edinburgh

Yana Vinogradova, Carol Coupland, Julia Hippisley-Cox

Page 2: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Background

• COX2i used for patients intolerant to NSAIDs• Lab investigations suggested mechanism for reducing

risk of cancer• Epidemiological studies

– Colorectal, RCT 36% decreased risk– Breast, lung - small studies – decreased risk– THIN (general population) no difference

Page 3: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Aim of the study

• To determine association between selective COX2i and risk of common cancers:

– Breast – Prostate– Colorectal – Lung– Blood – Melanoma– Gastric – Bladder– Pancreas – Oesophageal

Page 4: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Methods: Source of data

• QResearch• Currently largest database in the UK• 602 UK practices• > 6 practices in every Strategic Health Authority

(administrative area)• > 12 million patients including those who died,

left and still registered

Page 5: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Source of data: QResearch

• Patient level consolidated database• Anonymised data• Longitudinal data for 15+ years• Derived from GP clinical records• Validated against external and internal measures• Industry independent

Page 6: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Methods: Study design

• Nested case control study• 574 UK general practices• Study period Jan 1998-July 2008 • aged between 30 and 100 years• Up to 5 controls matched by

– Age– Sex– Practice– Calendar year

Page 7: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Methods: Assessment of Exposure

• at least 6 years of prescribing data• excluding 1 year before the index date• use: at least 1 script in prior 13 to 72 months• cumulative use: <3m, 3-12m, 13-24m, 25-60m• Short-term (<12m), long-term users• median dose: low/medium, high• different types of COX2i

Page 8: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Methods: Confounding factors

• Socio-economic status (Townsend score)• Body mass index• Smoking status • Morbidities (CVD, HBP, DM, RA, osteoarthritis)• Additional confounders (Benign breast disease,

family history; Colitis, Crohn’s disease)• Other medications (NSAIDs, Aspirin, statins,

contraceptive pills , HRT)

Page 9: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Methods: Statistical analysis

• Multiple imputations• Conditional logistic regression

– Odds ratios + 95% CI – 1% significance level

Page 10: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Results: Sample

118,780 incident cases

of cancer 1998/2008

113,879 cases

with primary cancer

88,125 cases

with 6 years of medical records

6,901

COX2i users

Page 11: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

0.8

0.8

1.8

4.4

0.8

0.8

1.7

4.1

0

2

4

6

8

Pro

por

tion

of

pat

ient

s (%

)

cases controls© QRESEARCH 2008 version 20

Proportion of cases and controls taking COX2i in prior 13-72 months

<33-1213-2425+

Months of use

Page 12: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Colorectal (11749) 0.76 (0.63 to 0.92)

Lung (10163) 0.79 (0.65 to 0.95)

Stomach (1992) 0.83 (0.53 to 1.31)

Oesophagus (3159) 0.98 (0.70 to 1.36)

Pancreas (2110) 1.04 (0.72 to 1.49)

Melanoma of skin (3249) 1.05 (0.75 to 1.47)

Prostate (14764) 1.10 (0.94 to 1.29)

Bladder (4227) 1.11 (0.84 to 1.46)

Breast (15666) 1.24 (1.08 to 1.42)

Haematological (7185) 1.38 (1.13 to 1.69)

Site of cancer (N of cases) OR (95%CI)

0 .5 1 1.5 2

Odds ratios and 95%CI are adjusted for deprivation, smoking, comorbidities, use of medicationReference group: No use of COX2 in 13 to 72 months prior the index date© QRESEARCH 2008 version 20

Adjusted odds ratios and 95% confidence intervals

Risk of cancer in patients using COX2 inhibitors for more than 365 daysin 13 to 72 months prior to the index date

Page 13: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

0.6

0.7

1.8

4.4

0.9

0.8

1.8

4.2

0

2

4

6

8

Pro

por

tion

of p

atie

nts

(%

)

cases controls

Proportion of COX2i users

25+m, 0.66 (0.51 to 0.86)

13-24m, 0.88 (0.69 to 1.12)

3-12m, 1.04 (0.88 to 1.21)

<3m, 1.07 (0.96 to 1.18)

Months of use, OR(95%CI)

.5 .75 1 1.25

Adjusted odds ratios

Odds ratios adjusted for deprivation, smoking, BMI, co-morbidities, use of drugs© QRESEARCH 2008 version 20

trend test for number of months, p=0.004Colorectal cancer and COX2i use

Page 14: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

0.9

0.9

2.1

4.4

0.8

0.8

1.8

4.6

0

2

4

6

8

Pro

por

tion

of p

atie

nts

(%

)

cases controls

Proportion of COX2i users

25+m, 1.19 (0.98 to 1.44)

13-24m, 1.30 (1.07 to 1.57)

3-12m, 1.21 (1.06 to 1.37)

<3m, 0.98 (0.90 to 1.07)

Months of use, OR(95%CI)

.75 1 1.25 1.5 1.75

Adjusted odds ratios

Odds ratios adjusted for deprivation, smoking, BMI, co-morbidities, use of drugs© QRESEARCH 2008 version 20

trend test for number of months, p=0.002Breast cancer and COX2i use

Page 15: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

1.0

1.0

2.4

4.5

0.7

0.7

1.8

4.0

0

2

4

6

8

Pro

por

tion

of p

atie

nts

(%

)

cases controls

Proportion of COX2i users

25+m, 1.47 (1.11 to 1.95)

13-24m, 1.31 (0.99 to 1.74)

3-12m, 1.25 (1.04 to 1.50)

<3m, 1.10 (0.96 to 1.25)

Months of use, OR(95%CI)

.75 1 1.25 1.5 1.75 2

Adjusted odds ratios

Odds ratios adjusted for deprivation, smoking, BMI, co-morbidities, use of drugs© QRESEARCH 2008 version 20

trend test for number of months, p<0.001Haematological malignancy and COX2i use

Page 16: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Summary of findings

Long-term use of COX2i : • 24% reduced risk of colorectal cancer • 24% increased risk of breast cancer • 38% increased risk of blood cancer

Page 17: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Strengths

• Large sample size and representative population• Data electronically collected – unlikely

misclassification bias • Data collected before the diagnosis – no recall

bias• Data in the last 12 months before the diagnosis

was excluded as might be misleading

Page 18: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Limitations

• Information on prescriptions only• Residual confounding• Missing data

Page 19: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Acknowledgements

• EMIS

Page 20: Exposure to cyclo-oxygenase-2 inhibitors and risk of cancer: nested case-control studies IAE world Congress Epidemiology 2011 Edinburgh Yana Vinogradova,

Thank you

Questions?

Y Vinogradova, C Coupland, J Hippisley-Cox “Exposure to cyclo-oxigenase-2

inhibitors and risk of cancer: nested case-control study” 2011, British Journal of

Cancer, 105(3), 452-459