What's The Evidence for the Link Between ARBs and Cancer? Ileana Piña Talks With Ilke Sipahi

Ileana L. Piña, MD


July 19, 2010

This feature requires the newest version of Flash. You can download it here.

Ileana L. Piña, MD: Hi, this is Ileana Piña, Professor of Medicine and Epidemiology and Biostatistics at Case Western Reserve University. I am here with one of my colleagues, Dr. Ilke Sipahi, who is part of our Heart Failure faculty at the Case Western Medical Center University Hospitals in Cleveland. He recently published an article[1] that has spawned a lot of discussion in medical circles and has even caused discussion in the news media concerning angiotensin II receptor blockers [ARBs] in cancer. Let me just preface by saying that ARBs are widely used due to their very good side effect profile in millions of hypertensive patients; [they are used] alone and in combination with other drugs (such as diuretics like hydrochlorothiazide) and, in heart failure, are indicated in the guidelines for patients who are angiotensin-converting enzyme inhibitor [ACE] intolerant. We have data, both in hypertension and certainly in heart failure, on the benefits and endpoints of these drugs.

Dr. Sipahi has been kind enough to join us today to participate in my blog and talk to us a little bit about the paper, so I'm going to be directing some questions to him. The paper has been published in Lancet Oncology just within the past 2 weeks, so I think we're really timely on this. [Dr. Sipahi], thank you very much for joining me.

First of all, talk to me a little bit about the biologic plausibility of this linkage. How did you come up with it?

Ilke Sipahi, MD: It has been long recognized that the renin-angiotensin-aldosterone system can play a role in the development of human malignancies, especially both the angiotensin II [type]-1 and [type]-2 receptors, which have been specifically studied.

Dr. Piña: Both of them to the same degree?

Dr. Sipahi: Both of them have been studied, but the data have been very conflicted, using basic science information -- animal data, so far. Some studies have suggested that they can prevent occurrence of cancers, whereas others have suggested that they increase angiogenesis very aggressively and --

Dr. Piña: The drugs or the receptors?

Dr. Sipahi: The drugs. There is an elegant study by Walther and colleagues[2], published in 2003. The study was done in Germany. It's an alginate model of tumor angiogenesis, and in [this study], [decreasing] the AT1R [angiotensin II type 1] receptor with losartan caused a 3-fold increase in angiogenesis. The reason I wanted to do this meta-analysis was not actually animal data, but I have been following the trials and...if you remember, in 2003, the CHARM [Candesartan in Heart Failure Assessment of Reduction in Mortality and Morbidity] trial reported a statistically significant increase in fatal cancers.

Dr. Piña: I'm going to go into that in a minute, so just to remind you, we did CHARM. We did it here and we did it at the VA [Cleveland Veterans Affairs Medical Center]. Let me go back a minute. Looking at the literature, I find a lot of literature on prevention or treatment, and yes, linking angiogenesis to the AT1 receptor. But the data are particularly robust with telmisartan and prostate cancer -- not only on prevention, but also treatment and augmentation of chemotherapy. I have also found data on breast cancer and even lung cancer, which you mentioned in the paper. There's a beautiful review by Rosenthal[3], which appeared in 2009 in the Journal of Human Hypertension, discussing the benefits of AT receptor blockers and even ACE inhibitors in the prevention and treatment of malignancy. In [your] paper, you have really given it a turnaround to a causation. Let me take you back to CHARM. CHARM was how many arms?

Dr. Sipahi: Two arms, placebo vs candesartan, but we had 3 different studies.

Dr. Piña: Actually, there were 4 studies. The total study is the fourth study. Which of those sub-studies or total study have the cancer incidence that you're talking about?

Dr. Sipahi: In the main publication of the CHARM trial, the cancer data --

Dr. Piña: Of the program or the trial?

Dr. Sipahi: Of the program. Cancer information was only presented for the overall program. When you go to the FDA [US Food and Drug Administration] Website, you can find additional cancer information.It appears that most of the excess cancers were coming from the CHARM added components --

Dr. Piña: So, in other words, for our viewers (to remind them), this is an addition of an ARB, namely candesartan, to an ACE inhibitor, and everybody in the trial has to be on an ACE inhibitor. The results were actually confident for reduction of events and no interaction with beta blockers. Do you think it's the ACE inhibitor or do you think it's the ARB?

Dr. Sipahi: Well, the ACE inhibitor was taken by both groups of patients and the excess cancer was in the candesartan arm, not the placebo arm.

Dr. Piña: Do you think it's the combination of the 2 drugs?

Dr. Sipahi:That's a great question. I have been trying to answer that. In our meta-analysis we gathered data, from not only the CHARM trial, but also from 5 other major trials. When you look specifically at the trials that did not allow concomitant ACE inhibition on the ARB arm, there is still a statistically significant excess of cancer development. However, when you look at the trial that mandates background ACE inhibition, the effect size seems to be larger. The hazard ratio, if there is background ACE inhibition, is around 1.14-1.15, and it's 1.08 when there is no background ACE inhibition.

Dr. Piña: But we're talking about a difference between an 8% and a 14%, which is still a small one. Now, in your analysis, how did you handle confounders such as smoking? Again, I want to go back to the viewers. In your analysis, you had hypertension, where patients may not have as many comorbidities. Then you have one major trial of heart failure, which is CHARM, that has 7000 and some patients, but they were older, they were sicker, and they had a lot of comorbidities. How did you handle smoking, for example, in confounders?

Dr. Sipahi: The nice thing about this meta-analysis is that it's a meta-analysis of only randomized controlled trials [RCTs], so there were no imbalances within the trial such as smoking, age, and other factors of concern. Because we do not have individual level data, we cannot control for these confounders. But as I've said, these are RCTs, so they should be taken care of.

Dr. Piña: Let me go to the FDA for a moment. The data that you say you have from the FDA, where did you acquire it from?

Dr. Sipahi:

Dr. Piña: There is a lot of data in the FDA [Website], but how did you get exactly to what you're talking about?

Dr. Sipahi: If you go to and [search for] telmisartan, you get all the documents about telmisartan. You can get telmisartan, cancer --

Dr. Piña: But do you get the raw data or do you get the transcripts?

Dr. Sipahi: You get the summaries. They're very detailed.

Dr. Piña: But there's a difference, right? A summary is not the actual data, and again...when we have panels, there are data that are presented in the panel, but that may not be the totality of the data that affect the FDA, because it's proprietary [information] of the sponsors. What you have is the transcripts, which is a lot of information, but you may not actually have the data set.

Dr. Sipahi: All the cancer data was presented rather extensively, of all the patients. Individual patient level data (eg, patient number one had cancer), obviously they did not disclose, but there is robust information there including all specific solid organ cancers. You can find lung cancer, prostate cancer information, breast cancer information for CHARM, for ONTARGET, for PROFESS, for TRANSCEND --

Dr. Piña: So CHARM is heart failure and ONTARGET is...?

Dr. Sipahi: Risk reduction.

Dr. Piña: Hypertension risk reduction. Now that you have looked at this, do you think that, with the millions of patients that are currently on these drugs that have been so successful, that this should have been noticed clinically? Because when the glitazones started coming out, we in the heart failure world noticed the problem immediately.

Dr. Sipahi: I don't think it can be noticed clinically. There are 2 reasons behind this. The first one is that cancer is a slowly developing disease. The second [reason] is that this is a modest increase in the risk. If these drugs had been increasing cancer risk 10-fold, we would recognize this, but because the relative risk increase is around 10%, you need to do a really careful analysis here with adequate follow-up.

Dr. Piña: Your meta-analysis is very interesting and...brings out an interesting hypothesis, which is really what it is. Yet, in the literature, the hypothesis is the opposite.

Dr. Sipahi: Not fully.

Dr. Piña: Well, I think there's more about prevention than there is about causation. What would you say would be the next step? We have a lot of meta-analyses in the literature that show one thing, and then when you go out and test it prospectively, it's totally the opposite or it comes out neutral. What would be your best advice for the scientific community to do next?

Dr. Sipahi: I think that right now the responsibility belongs to the regulatory bodies -- the FDA in the United States and EMA [European Medicines Agency] in Europe.

Dr. Piña: Why do you say that?

Dr. Sipahi: Because they have individual level data. As independent investigators, we have access to some of these clinical trials.

Dr. Piña: So you're putting the [responsibility] on the FDA?

Dr. Sipahi: Well, the FDA and EMA. Actually, we are really lucky on this matter because EMA (European Medicines Agency), last Friday, announced that it's starting a thorough investigation of the ARB cancer link. They posted it on their Website.

Dr. Piña: Lancet is a very European journal, so I can see that. Have you spoken to the FDA about this?

Dr. Sipahi: No, they did not consult me.

Dr. Piña: How about a randomized clinical trial, which is how we usually find the truth in --

Dr. Sipahi: That would be great. It has to be a big trial.

Dr. Piña: How would you do it? In what population?

Dr. Sipahi: It has to be a big trial. It should be perhaps over 40,000 patients. It needs to have a really long follow-up, preferably at least 5-6 years, because we are looking at cancer.

Dr. Piña: Probably more than that.

Dr. Sipahi: Right, the power calculations need to be done. I don't think it needs to be longer than 6-7 years because the trials that we used were, on average --

Dr. Piña: What population? Because you have a mix of the population --

Dr. Sipahi: The population that these drugs are indicated for. You can enroll heart failure patients, you can enroll hypertension patients --

Dr. Piña: Different populations though. So you would do it in all the populations?

Dr. Sipahi: Hypertension is, I think, the most important area, because that's the number one indication of ARB.

Dr. Piña: Shall we do it including heart failure patients?

Dr. Sipahi: I think you can include them. You could do ACE inhibitor vs ARB.

Dr. Piña: That's another trial. Give me the design. This is important if we're thinking scientifically.

Dr. Sipahi: The most important components here are: (1) size, (2) duration of follow-up, (3) compliance with the medication, and (4) the study drug dose.

Dr. Piña: How about patient population?

Dr. Sipahi: Patient population. Obviously, [the trial] should enroll patients that are older, because those are the patients who are at higher risk for the role of --

Dr. Piña: With hypertension?

Dr. Sipahi: Hypertension is the easiest one to enroll.

Dr. Piña: What's your control group?

Dr. Sipahi: ACE inhibitor vs ARBs.

Dr. Piña: But you just told me that in the trial that had the ACE and the ARB, the difference was --

Dr. Sipahi: That is correct, but ACE inhibitors by themselves have never been linked to cancer.

Dr. Piña: Well, they're already linked to treatments.

Dr. Sipahi: They are linked to treatments. Or you can do 3 arms like the ALLHAT trial. You can compare, for example, a diuretic or a calcium channel blocker with an ARB. That would have multiple aims because there are not very many antihypertensive trials with ARBs, comparing an ARB with a diuretic or with a beta-blocker. We just have the LIFE trial and the VALUE trial, and as you know, the VALUE trial [showed that] myocardial infarctions were more common with ARBs, so maybe you can set the record straight.

Dr. Piña: But that wasn't the case in CHARM; the myocardial infarctions were not higher.

Dr. Sipahi: That is correct, but that was placebo-controlled.

Dr. Piña: So they're all different. They're all very, very different. Who's going to fund it?

Dr. Sipahi: I'm not sure. The pharmaceutical industry is going to be --

Dr. Piña: Do you really think the pharmaceutical industry would --

Dr. Sipahi:The FDA can mandate that.

Dr. Piña: Actually, they can't. They can't mandate trials. Just for your information, they do not mandate clinical practice. They can't mandate clinical practice. Are you going to be the PI [principal investigator]?

Dr. Sipahi: If they make such an offer, I'll think about it.

Dr. Piña: Well, I want to thank you for your time today and I want to thank our audience for joining us. I hope this is of interest, certainly with the millions of patients on these drugs. These questions are interesting and they bring up a lot of concepts. I want to warn clinicians that it's not the time to remove the A2 receptor blockers from your patient's therapy, given the rate of stroke and hypertension and given the worsening of heart failure symptoms if the patient is truly ACE intolerant. The bottom line is: we will watch and wait, and see what happens. Thank you again for joining me. This is Ileana Piña and this is my blog. Good day.