September 19th, 2012

An Ad That Doesn’t Tell the Entire Story — Part 2: Niaspan

Although I do not include advertisements on my reading list, I can’t help but notice them as I leaf through journals (no, I’m still not quite completely digital). I wonder if anyone else was bothered by Abbott’s print advertisement for Niaspan, an extended-release formulation of niacin. The ad, shown below, focuses on Niaspan’s effect on HDL-cholesterol levels without any reference to patient outcomes.

Click to enlarge

This ad makes no mention of AIM-HIGH, the NIH trial showing that adding extended-release niacin (supplied by Abbott) to simvastatin provided no benefit to patients with established cardiovascular disease. The trial did show that extended-release niacin could increase HDL-c levels by 21%, consistent with the ad’s claims — but (and this is a big “but”) the drug did not reduce the risks for myocardial infarction, ischemic stroke, symptom-driven revascularization, or death from coronary heart disease.

I can’t blame Abbott for omitting information about AIM-HIGH. The FDA has retained an indication for Niaspan in lowering the risk for heart attack in people with high cholesterol who have previously had a heart attack. The Abbott ad does mention the lack of additional benefit in preventing heart disease from Niaspan in combination with simvastatin or lovastatin over that from simvastatin or lovastatin monotherapy. You can see the potential for confusion. Also confusing is the fact that the indication names specific statins, not statin therapy in general: Does the FDA think that Niaspan is likely to improve outcomes when added to, say, atorvastatin?

By the way, I was curious what the NIH says about extended-release niacin to consumers, given that it funded the multimillion-dollar AIM-HIGH trial. Interestingly, the NIH’s consumer webpages on niacin provide no context about AIM-HIGH. One page, revised in June 2012, doesn’t mention AIM-HIGH at all; another, last reviewed in August 2011, has AIM-HIGH in a reference list at the bottom but provides no findings from the study. Wouldn’t you want to educate the public about the science you funded? Maybe we should suggest that when the NIH publishes important research, it should update its information for patients as well.

A Merck-sponsored trial of extended-release niacin is due out next year. While we await any new information about this drug, I am concerned at seeing an ad that touts an effect on HDL-c levels without making clear that the only large contemporary trial of the drug in the statin era failed to show that it improved outcomes.

What do you think of this Niaspan ad?

10 Responses to “An Ad That Doesn’t Tell the Entire Story — Part 2: Niaspan”

  1. Furthermore, this suggests that clinicians and patients should focus on treating the vulnerable patient (i.e., to decrease overall risk of CV morbidity and mortality) with pharmacotherapy, rather than treating a vulnerable laboratory value such as a target goal for LDL (<70 or 40).

    As a tangible example, should a 70 yo patient with prior MI, prior PCI, poorly controlled diabetes, hypertension, current smoker, chronic renal failure and an LDL of 70 be treated with a statin? Conversely, should a 90 yo patient with no risk factors and no prior CV events with an LDL 150 be treated with a statin?

  2. Dan Hackam, MD PhD says:

    There is still a (small) role for nicotinic acid, largely in patients who cannot reach goals with statin therapy at maximal doses and/or those who are intolerant to statins (who are akin to patients enrolled in the second largest niacin trial conducted to date – the Coronary Drug Project). AIM-HIGH has been widely debated in many circles and remains controversial, largely because of the contaminating influence of statin escalation and ezetimibe addition in the placebo group – rates of fibrate and resin exposure were not reported, but these may also have been contaminating influences. I agree the ad is misleading but would like to wait until we have HPS2-THRIVE before entirely discounting niacin’s role in cardiovascular prevention.

  3. I agree with Dan regarding problems with Aim-high. The National Lipid Association as well as the European Society of Atherosclerosos have both released opinions that prescribing patterns should not change based on this study. Don’t forget that the placebo used in this study was niacin, the study was stopped early because of the assumption of no benefit, not the observation of no benefit, and remember that it takes niacin 9 months to reach steady state effects on lipids,

    A concern raised by Aim-High is related to a potential negative effect of high dose statin therapy on the benefit offered by niacin. Considering the fact that HATS and FATS used niacin with low dose statin or colestipol and demonstrated plaque regression and/or a dramatic reduction in heart attacks, perhaps the problem is not in the lack of effececy of niacin but rather a toxicity of high dose statin.

  4. Edgar Abovich, MD says:

    I disagree. In my opinion there is much more compelling data regarding high doses of statins than statin/niacin combos.

  5. H Robert Silverstein, MD says:

    This discussion is a tempest in a teapot. It is HK, MD, who is misleading. The patients in AIM-HIGH were already at “goal” with regard to non-HDL cholesterol & so no more “good” could be done. I published this point on a blog the day AIM-HIGH came out: I guess I need better PR. What is correct is that NIACIN does good when patients have not yet reached goal non-HDL (<90) by other means = less efective statin Rx or colestipol as in FATS & HATS as stated just above

  6. Robin Motz, M.D., Ph.D. says:

    This reminds me of the case of the FTC vs. Wise Potato Chips. They started running an ad stating that their potato chips contained no cholesterol at all, which of course is true for any vegetable product. The FTC made them pull the ad, claiming that consumers would interpret “no cholesterol” as meaning “no fat” (!)

  7. Dr. Silverstein: Thank you for your comment but what evidence do you have in a statin era that “niacin does good when patients have not yet reached goal non-HDL’? I do not see FATS and HATS as being helpful. For our readers…

    FATS, published in 2001, was a trial of 160 patients who were randomly assigned to receive one of four regimens: simvastatin plus niacin, antioxidants, simvastatin–niacin plus antioxidants, or placebos. The study could not isolate the effect of niacin. Moreover, the endpoints were arteriographic evidence of a change in coronary stenosis and the occurrence of a first cardiovascular event (death, myocardial infarction, stroke, or revascularization). The study had little power for clinical endpoints but there were fewer events in the simvastatin-niacin groups compared with the placebo group or the antioxidant vitamins group. So how does this help with understanding the value of Niaspan?

    HATS, published in 1990, was a trial of 142 men who were randomly assigned to lovastatin (20 mg twice a day) and colestipol (10 g three times a day); niacin (1 g four times a day) and colestipol (10 g three times a day); or conventional therapy with placebo (or colestipol if the low-density lipoprotein [LDL] cholesterol level was elevated). The primary end point was the average change between the two angiographic studies in the percent stenosis for the worst lesion in each of the nine proximal segments. Only 15 patients had clinical events. The evidence from this study, conducted more than 20 years ago, is not strong for niacin and cannot say anything about the value of niacin therapy above and beyond statins.

    Perhaps you could clarify why you support Niaspan.

  8. Jean-Pierre Usdin, MD says:

    In France we have not such advertisments.
    BUT we undergo some Mentors’ advice in the front page of public weekly papers warding off the whole population that “half of the prescribed drugs in France are useless!”
    I am not sure this is a better way to inform!

  9. Dear Harlan, regarding HATS and FATS:

    When we look at the effect of statins in reducing coronary endpoints, we are delighted to see a 40% reduction, indeed the average is a 26% reductions in the combined endpoint of MI and coronary death with the use of statin alone.

    In HATS and FATS, we saw a 70 to 90% reduction in events with a statistically significant P value. Although this does not fit into the model of a randomized trial of adding niacin to a statin (or as Eric Topol calls it, the “sanctimonious randomized controlled trial”), it certainly demonstrates endpoints that no statin study comes close to matching.

    Say what you want about the small nature of the trial, a statistically significant P value is a statistically significant P value. That is why we use statistics.

    Also to throw away a 20 year old trial because of its age is a bit insulting for us… uh shall we say… more mature physicians. 20 years ago, I had been practicing medicine for 12 years. This was new DATA at a time when I was an “experienced” physician. If we cannot learn the lessons from history, we are damned to repeat the mistakes.

    A few years ago, I did a simple DATA mining study looking at the association between the use of niacin and the stability or progression of calcified plaque by EBT imaging. Those subjects on niacin had a dramatically lower calcified plaque progression than did those subjects not on niacin. As stability of calcified plaque by serial EBT imaging has been shown to correlate dramatically reduction in coronary events at all levels of plaque, I believe that niacin is much more important than the current cardiovascular industry gives it credit for.

    I regret that AIM-HIGH has seemed to so inappropriately influence clinicians opinions on the value of niacin. The conclusions, limited as they are, should only apply to subjects on high dose statin with LDL cholesterol <70.

    To throw out this baby with the bathwater will cost lives.

  10. Dan Hackam, MD PhD says:

    Actually small trials are about as likely to lead to false positives as they are to false negatives. 95% confidence intervals are wider and the point where the point estimate hazard ratio falls is fairly random (and has nothing to do with the mean or average estimate of effect). AIM-HIGH had 28x as many subjects as HATS and was multicenter. Notwithstanding the issues mentioned regarding control group contamination with statin escalation and ezetimibe addition, AIM-HIGH is a much more robust study. FATS and HATS had primary endpoints consisting of atheroma progression/regression, which does not always correlate with plaque vulnerability and clinical events. AIM-HIGH had as primary endpoint a clinically meaningful patient-important endpoint. In addition, the trend towards increase in ischemic stroke in AIM-HIGH, which may be due to increased rates of atrial fibrillation on niacin (first noted in CDP in the 1970’s) is very concerning. While I have not entirely written off niaspan, it now remains a third or fourth line drug in my armamentarium, and certainly there is much more evidence for statin escalation first (TNT, PROVE-IT, ALLIANCE, AVERT, SPARCL, POST-CABG, etc).