Heartfelt with Dr Melissa Walton-ShirleyView all posts »
"Given the limited data," I won't be stopping my dad's ARBJul 5, 2010 21:18 EDT
"Given the limited data, it is not possible to draw conclusions about the exact risk of cancer associated with each particular drug."
This direct quote from the Lancet June 14, 2010 issue should serve as both the beginning and the end of this blog. However, the fact remains that angiotensin-receptor blockers (ARBs) (particularly telmisartan) stand informally accused of carcinogenicity, and all the noncommittal interviews and "innocent-until-proven-guilty" stances in the world may not be able to convince some physicians and patients of telmisartan's innocence. Once accused of such an atrocity as carcinogenicity, the fate of a compound is often left to spin and hype. Apathy will be its greatest ally, because no one wants to disrupt a well-controlled blood-pressure level by stopping a medication that is so well tolerated. What we as practitioners should do in this instance is to become detectives, examine the data, and arm ourselves with enough information to explain the issues to our patients so they can make an informed decision.
The first hint that something isn't quite right about concluding that telmisartan increases the risk of lung cancer is the type of cancer that occurred: non–small cell. Now ask yourself the question, "What exactly is the most common risk factor for non–small-cell lung cancer?" That's right . . . the first, second, third, fourth-through-infinity greatest risk for non–small-cell lung cancer is SMOKING. According to the Mayo Clinic Proceedings 2009: 83:584–594, 85% of all lung cancers are caused by tobacco use. Looking at the ONTARGET and TRANSCEND patient population contributed to this meta-analysis, a total of 3580 patients were active smokers in just those two studies. Many of the requirements for eligibility into these two trials were surrogates for smoking, which included limb/foot amputation, claudication, greater than 50% diameter stenosis of a limb vessel, multivessel CABG, angina, and multivessel PCI. The smoking issue is a tricky one. There was a distinction between former smokers and active smokers, but I didn't see how those determinations were made. Did the patient just quit smoking last week or five years ago? The most salient point, however, is that the treatment arms were NOT randomized between smokers and nonsmokers. Primary smoking, along with the incidence of passive smoking, as well as the duration of smoking cessation are all severely confounding patient characteristics when looking at any increased non–small-cell cancer risk.
The second hint that the increased incidence of lung cancer isn't just the ARB utilization is the fact that in the PROFESS trial, with 21% "current" smokers (4479) and 36% (7319) former smokers, there was NO increased incidence of lung cancer, despite the fact that telmisartan was also the ARB utilized. One could argue, however, that the 2.5-year follow-up wasn't long enough for PROFESS cohorts compared with the 56-month follow-up for the TRANSCEND and ONTARGET studies. There again, the treatment arms were not randomized between smokers and nonsmokers, at least in my analysis.
A good trialist will probably shred my premise to bits, but the fact remains that this was NOT a randomized controlled trial powered to look at cancer incidence. Because of this fact, I will NOT be contacting my patients to stop their ARB, and I will NOT stop my dad's valsartan, because his BP is 120/70 mm Hg and it took a while to get him onto meds he could tolerate. Having said that, if a well-designed and adequately powered trial proves a link to cancer, I'll be the first to stop all the ARBS I can get my hands on.
In the words of the authors, I'll conclude my blog just as I started it: "Given the limited data, it is not possible to draw conclusions about the exact risk of cancer associated with each particular drug," which leads me to ask the most obvious question of all: "Why exactly did you ask the question with such 'limited data' to begin with?"