Oral anticoagulation for prolonging
survival in patients with cancer
EA Akl, G Kamath, SY Kim, V Yosuico, M
Barba, I Terrenato, F Sperati, HJ Schünemann
DOI: 10.1002/14651858.CD006466
David
K. Cundiff, MD
Date received: June 28, 2007
The results, discussion, implications for
practice, and implications for research all focus on the possible benefit of
VKA in patients with small cell cancer of the lung. The SCCL subgroup analysis
is further subdivided into early and late stages. These subgroups were not
defined in the protocol, so post hoc statistical analyses of outcomes of these
subgroups should not be emphasized. In several places, the Cochrane Handbook
warns against problems with post hoc analyses:
(pg 19) “Changes in the protocol should not be
made on the basis of how they affect the results of the review. Post hoc
decisions (such as excluding selected studies) that are made when the impact on
the results of the review is known are highly susceptible to bias and should be
avoided”
(pg 63) “Post-hoc questions are more susceptible to
bias than those asked a priori, and data-driven questions can generate false
conclusions based on spurious results.”
(pg 143) “8.8.4.2 Specify characteristics in advance Authors should,
whenever possible, pre-specify characteristics in the protocol that later will
be subject to subgroup analyses or meta-regression. Pre-specifying
characteristics reduces the likelihood of spurious findings, first by limiting
the number of subgroups investigated and second by preventing knowledge of the
trials’ results influencing which subgroups are analysed.”
(pg 147) “Findings for the
most important comparisons and/or outcomes should be prominent in the text of the review, even when little relevant data were
available. Answers to post hoc analyses and less important questions for which
there happen to be plentiful data should not be oveemphasised. Post hoc analyses
should always be identified as such.”
In addition to the relatively insignificant financial conflict of interest acknowledged by Holger J. Schünemann, MD, he has received research funding from the anticoagulant producing companies AstraZeneca, Boehringer Ingelheim, Pfizer, and Amgen Inc. He has received honoraria and consultant fees from AstraZeneca, Boehringer Ingelheim, and Amgen that were deposited into research accounts at the University of Buffalo and McMaster University.1 He was co-author of six articles from the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy.2-7 AstraZeneca LP; Aventis Pharmaceuticals; Bristol-Myers Squibb/Sanofi-Syntholabo Partnership; GlaxoSmithKline; Organon Sanofi-Synthelabo LLC funded this anticoagulant-guideline producing conference.1
1. Financial Disclosures. Chest. September 1, 2004;126(3_suppl):167S-171. http://www.chestjournal.org/cgi/content/full/126/163_suppl/167S.
2. Stein PD, Schunemann HJ, Dalen JE, Gutterman D. Antithrombotic Therapy in Patients With Saphenous Vein and Internal Mammary Artery Bypass Grafts: The Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest. September 1, 2004;126(3_suppl):600S-608.
3. Schunemann HJ, Cook D, Grimshaw J, et al. Antithrombotic and Thrombolytic Therapy: From Evidence to Application: The Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest. September 1, 2004;126(3_suppl):688S-696.
4. Schunemann HJ, Munger H, Brower S, et al. Methodology for Guideline Development for the Seventh American College of Chest Physicians Conference on Antithrombotic and Thrombolytic Therapy: The Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest. September 1, 2004;126(3_suppl):174S-178.
5. Menon V, Harrington RA, Hochman JS, et al. Thrombolysis and Adjunctive Therapy in Acute Myocardial Infarction: The Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest. September 1, 2004 2004;126(3_suppl):549S-575.
6. Hirsh J, Guyatt G, Albers GW, Schunemann HJ. The Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy: Evidence-Based Guidelines. Chest. September 1, 2004;126(3_suppl):172S-173.
7. Guyatt G, Schunemann HJ, Cook D, Jaeschke R, Pauker S. Applying the Grades of Recommendation for Antithrombotic and Thrombolytic Therapy: The Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest. September 1, 2004;126(3_suppl):179S-187.
Author's reply
We thank Dr. Cundiff for the feedback.
We did define the subgroup analysis of SCLC a priori. In fact, we state in the protocol that "we will try to explain heterogeneity, if present, by conducting subgroup analyses. Our a priori factors to explain heterogeneity are the characteristics of participants…" and we list "type/location of cancer" as a patient characteristic.
The question is whether to consider SCLC, which was further divided into early and late stage, an a priori declared subgroup for analysis, because it is a cancer type.
Obviously the a priori definition could have been more explicit and given this was a subgroup analysis we attempted to be cautious in our conclusions. In spite of the statistically significant results for SCLC, we used in our "implications for practice" the verb "suggest": "It suggests however a survival benefit …". Similarly, we state in the "implications for research": "Future research should investigate the effects of oral anticoagulation in patients with different cancer subtypes (e.g. NSCLC) and different cancer stages (e.g. limited SCLC)". We are also willing to replace "benefit" by "potential benefit" in the "reviewers' conclusions" of the abstract as follows:
"The decision for a patient with extensive SCLC to start warfarin for a potential survival benefit should balance that benefit with the downsides of increased bleeding risk in light of patient values for these outcomes."
And in the Implications for practice as follows:
"This potential benefit is accompanied by a reduction in VTE but also by
an increased risk of minor and major bleeding and by the burden of warfarin
treatment (e.g. periodically checking INR levels). As a consequence, the
decision for a patient with extensive SCLC to start warfarin for a potential
survival benefit should balance the benefits and downsides and integrate that
patient's values and preferences (Haynes 2002)."
We are not sure what the commentator is trying to say. There has been no personal payments of these companies to Dr. Schünemann in relation to the ACCP conference. In the review, Dr. Schünemann has provided the following accurate statement of potential conflict of interest. The declarations are conform with the policy of Cochrane. In addition, as a policy of our research group, we do not take personal funding from for-profit organizations and we think this is clearly addressed in the conflict of interest statement:
"Schünemann: no personal funding, but received research grants and honoraria that were deposited into research accounts or received by a research group that he belongs to from Pfizer, Amgen, Roche, and AstraZeneca for development or consulting regarding quality of life instruments for chronic respiratory diseases."
The funding from Boehringer Ingelheim dates more than 5 years back. Usual declaration periods (e.g. WHO) date four years back. The funding for the ACCP conference has gone directly to the ACCP from the mentioned for profit funders. Unfortunately, it is possible that academic organizations that Dr. Schünemann is affiliated with receive other funds such as the ones Dr. Cundiff describes. For instance, major Universities such as the ones Dr. Schünemann works for, receive many grants from for-profit sponsors.
More importantly, the ACCP conference carries a very high standing for the process it uses to develop Evidence Based Guidelines. The methodology is among the most rigorous processes employed. Methods used to produce the guidelines are widely available and in fact described in one of the articles mentioned by Dr. Cundiff.
Finally, we want to once more make clear to the commentator that we are not emphasizing the use of oral anticoagulants by the way that we describe the implications for practice. This wording has been carefully composed for exactly the purpose: the use depends on values and preferences and this is what the data show.
We added a description saying:
"Institutions or organizations that Dr. Schünemann is affiliated with
likely receive funding from for-profit sponsors that are supporting
infrastructure and research that may serve to advance his work."
My Rebuttal
I thank Drs. Akl and Schünemann for their reply.
Regarding the interpretation of the subgroup analysis of SCLC, there would be two ways of treating the results statistically:
1. An a priori determined subgroup along with all 100+ other cell types of cancer – the p value, adjusted for multiple comparisons (e.g., Bonferroni Adjustment), would not be significant
2. A post hoc analysis that should not be emphasized in the implications of practice, abstract, or plain language summary.
In either case, oral anticoagulants are not evidence-based to benefit patients with SCLC.
Few patients with SCLC or their loved ones will be subscribing to the Cochrane Library for the full text of reviews such as this one. Instead, they will find the abstract and plain language summaries freely available on the Cochrane website. The plain language summary states, “Oral anticoagulation with vitamin K antagonists does improve mortality in the subgroup of patients with extensive small cell lung cancer.” Inclusion of this statement in the plain language summary is not justified by the data. Likewise, the authors’ conclusions in the abstract regarding SCLC would only be appropriate if another sufficiently powered RCT of SCLC patients confirmed the finding in the subgroup analysis. Unfortunately, the conclusion of this Cochrane review would make it virtually impossible to recruit patients with SCLC for a placebo controlled RCT with oral anticoagulants.
Thanks for the additional clarification of the potential competing financial interests.