CCBYNC Open access
Research

Exposure to bisphosphonates and risk of gastrointestinal cancers: series of nested case-control studies with QResearch and CPRD data

BMJ 2013; 346 doi: http://dx.doi.org/10.1136/bmj.f114 (Published 16 January 2013) Cite this as: BMJ 2013;346:f114

Re: Exposure to bisphosphonates and risk of gastrointestinal cancers: series of nested case-control studies with QResearch and CPRD data

Vinogradova et al revisited the issue of gastrointestinal cancers in bisphosphonate users. However, in their discussion they misinterpret the findings of our Danish national cohort study[1]. Moreover they did not address a potential problem in case-control studies of the type they performed, where there is a possible bias due to the longer survival of bisphosphonate users.

 

As regards our national cohort study, it is not correct that the risk was not reduced in long term users (defined by the authors as 6+ months). In fact, we reported a 31% reduced risk of colon cancer and a 38% reduced risk of colon cancer mortality at five years, both of which was statistically significant both before and after confounder adjustment (table 2 in our paper). The authors must have misunderstood our subsequent dose response analysis, which is not a comparison with the background population but a comparison of bisphosphonate users with different degrees of exposure.

 

The case-control design is problematic when survival itself is linked to exposure. This is perhaps not immediately obvious, but case-control studies only address runners in the race if they make it to the goal post, while cohort studies follow each runner from the start of the race.

 

It has been demonstrated in a number of studies that bisphosphonate users as a group have lower mortality[2–4], thus contributing more patient years to analyses. This is accurately captured in cohort studies, which capture event rates, but not in case-control studies as their currency is relative risk estimates.

Cohort studies certainly have lower resolving power for rare outcomes but they are at much lower risk of bias and they provide clinicians and researchers with meaningful absolute risks estimates.

 

[1] Pazianas M, Abrahamsen B, Eiken PA, Eastell R, Russell RGG. Reduced colon cancer incidence and mortality in postmenopausal women treated with an oral bisphosphonate-Danish National Register Based Cohort Study. Osteoporosis International  2012; 23(11):2693-701.

[2] Bolland MJ, Grey AB, Gamble GD, Reid IR. Effect of osteoporosis treatment on mortality: a meta-analysis. J Clin Endocrinol Metab 2010;95:1174–1181.

[3] Bondo L, Eiken P, Abrahamsen B. Analysis of the association between bisphosphonate treatment and survival in Danish hip fracture patients — a nationwide register-based open cohort study. Osteoporosis International 2013; 24(1):245-52

[4] Lyles KW, Colon-Emeric CS, Magaziner JS, Adachi JD, Pieper CF, Mautalen C, et al. Zoledronic acid and clinical fractures and mortality after hip fracture. N Engl J Med 2007;357:1799–1809.

Competing interests: B. Abrahamsen: Grant / Research support from Novartis, Nycomed, Amgen, Speakers Bureau with Merck, Eli Lilly. M. Pazianas: Grant / Research support from the Alliance for Better Bone Health and Warner Chilcott. R.G.G. Russell: Research support from Procter and Gamble, Sanofi-Aventis, and Warner Chilcott, Consultant/speaker and legal activities Amgen, GlaxoSmithKline, Roche, Procter and Gamble, Sanofi-Aventis, Novartis, Eli Lilly, Teva and Warner Chilcott.

01 February 2013
Bo Abrahamsen
Professor (SDU)
MIchael Pazianas (Oxford), R. Graham G. Russell (Oxford)
University of Southern Denmark
DK 5000, Odense C
Click to like:
32