All quite valid points but I have a bit of a different take and perhaps we are both correct.  Scientific truth should be replicated prior to acceptance - particularly in the field of Medicine where financial and personal interests abound.  The question is whether niacin in clinical practice is similar to ezetimibe -"Ezetimibe to any trial may be like adding a tiny drop of food coloring to a bucket of water." - perhaps niacin is similar when added to adequate dose statin.  The notion that CIMT is an accurate predictor of subsequent cardiac events and as such supports clinical trials using niacin is suspect due to the statin paradox - when there is a disease of "thickened carotid" I will treat those afflicted with niacin.  The HATS trial utilized low dose statin and did report unexpectedly (I think all of us were surprised) dramatic clinical benefit in a small trial not in any way powered to demonstrate same and the possibility of a type I error was minimized by the angiographic findings which were in concordance.  I agree it is difficult to explain away and would require a careful investigation of both the clinical endpoints and angiographic analysis to determine if that study was flawed in some way - dissecting trials should not be restricted to those with outcomes you disagree with.    Having said that I don't think such an investigation necessary or practical since AIM HIGH and the subsequent HPS2-Thrive Study will address the clinical question with sufficient rigor to provide what I would believe to be (hopefully) replication - if they disagree then confusion continues.  It will be interesting to dissect out compliance and lipid profile results as Dr. Bale has suggested but the entry criterion of a one month tolerance period prior to enrollment would make me wonder why compliance suddenly became an issue - I would have argued that this entry criteria defined a particularly unique patient population given your comments on niacin tolerability despite best cheerleading effort.  It would seem that the ezetimibe confounder argument is difficult to imagine as it would require both a significant randomization difference and a sufficient "danger" of the medication - I cannot rectify that with the much larger IMPROVE-IT trial which although not published has not been terminated due to harm unless Merck has some explaining to do.  It has been suggested in another blog on this site that the trial was terminated too early based on the late analysis of CDP (a dissimilar treatment trial) but that was not the case in HATS.  Compliance data is notoriously difficult to be certain of and would require estimated compliance of all the medicines involved not just niacin - although important what a slippery slope for an investigator with an agenda to explain away disappointing results.  Finally I would like to think that the NIH knows something about trial design and conduct and I suspect that the co-sponsor KOS/Abbot devoted a scientist or two to ensure that it was a fair test of their product so although I am by nature a skeptic without having the actual data I am inclined to accept the conclusion prior to publication of same.  So for all those reasons I have stopped niacin in every patient who fit this demographic (residual low HDL with adequate LDL on statin), I have never used ezetimibe and restrict fibrates to those with a history of diet refractory triglyceride elevations of greater than 500 mg/dl.  When medication recommendations are made the actual degree of benefit conferred by each medication should be able to be presented to the patient balanced by potential side effects and ultimately cost and then the patient can make an informed decision with regards to what treatment is acceptable.  I can do that with every anti-atherosclerotic measure I recommend based on replicated clinical trials (aspirin, ACE, beta blockers, statin, smoking cessation, diet, exercise) but not niacin in addition to adequate dose statin - not to mention "blind" maximal dose statin that probably is the most cost effective approach. 

Just want to clairfy for you Dr. K.  that I have the utmost respect for Drs. Brown and Boden and in fact consider them both heroes in the world of cardiovascular research. I am merely commenting on the strength of the study and per discussions around the world, there are questions as to the design of this trial AS THERE ARE WITH ANY TRIAL. Furthermore, in covering trials now for nearly a decade, I am a bit frustrated with the fact that we trot the results out for intense discussion with nary a subset analysis often times,  when in fact, all we treat in our patient populations are subsets so that information is as important and sometimes MORE important than the general message delivered by the initial analyses.  I anxiously await forthcoming information from this study and I do believe that all of the trial designers will welcome any discussion on it.

Melissa

Melissa, great post. I dropped off the AIM-HIGH conversation because it was getting so nasty, and some commenters have forgotten that we are just trying to figure out what is best for our patients.  

 

I complete agree that we need to see the final results of AIM-HIGH. It is tough to make much of a press release.  But even then, it is just one study. Niacin as used in other studies is an effective agent for reducing CVD events and there also are a lot of physiological and imaging data to support its benefits as well.

What was unique about AIM-HIGH is that none of the previous studies added niacin to such aggressive LDL-C lowering. In AIM-HIGH it averaged 71 mg/dL. The LDL-C in previous studies were notably higher. I really do wonder if you get LDL-C and total cholesterol low enough if HDL-C matters any more.   But as of today, all we can say is that as used in AIM-HIGH, niacin did not appear to help. Any more is an over-stretch. I agree that we need to see the final results and we HPS-II before making any bold proclamations before abandoning a tool in our arsenal.  

 

And we need to be civil and collegial in our communications!

Melissa,

I entirely agree that absent the actual data and the subsequent analysis of same it is hard to be an adamant commentator either way - I was just posing an alternative viewpoint.  I think it is quite likely that if this was entirely privately funded that the termination would not have been announced at all (and perhaps the results never published) so I do agree that the reputations of the principal authors and their subsequent comments dwarf anything posted on a blog.  Having said that I think RCTs have consequences and I know that you can recall numerous small studies consistently showing benefit of a treatment approach that were subsequently rebuked by large multicenter randomized studies (rotational and directional atherectomy, mucomist, vitamin E/C etc).  This trial is unique in its design in that it asks if adjunct niacin therapy to moderate dose statin is beneficial and the answer was no.  It doesn't mean niacin does nothing it just didn't add any measurable benefit to moderate dose statin therapy.  Again additional trials may clarify or confuse but I think I see some clarity here.

Additionally I appreciate the tenor of your response and I respect your opinion.  What I do not understand is the intense "cult of niacin" who seem to have some intense allegiance to the molecule - at least I do enjoy their banter and don't really mind the name calling.

Certainly, our speculations at this point are just that.  However, I feel some of the major studies done recently (ACCORD/ AIM-HIGH)/ ARBITER/ ENHANCE) are more a reflection on study design/ patient selection than on true treatment efficacy.  After all, in AIM-HIGH if the LDL-C was so low than most patients would be at or close to their Non-HDL-C goals (a rather good surrogate for apoB or LDL-P) despite having rather low HDL-C (Non-HDL-C= T.C-HDL-C).  Up to this point, we only use add-on therapy such as Niaspan only if one can't get to Non-HDL-C goals with moderate or high dose statin monotherapy (not max dose statin).  After all, the most important parameter to obtain is the athrogenic particle count.  Remember- there is zero level IA data that raising HDL-C is of any benefit esp. if apo B is at goal.....   I await the AIM-HIGH data on the subset of pt's who were not at or near Non-HDL-C (or apoB/ LDL-P) goal- I bet they would benefit from Niaspan add-on to the statin/ statin-zetia combo.-  Lets design trials with patients in whom we would actually use the medication in question.  In real life, we have thousands of patients on high dose statin (not max) and they are not near Non-HDL-C goals and in these, one of my best options is Niaspan to get to the Non-HDL-C goal (you still have to obtain your goals set by NCEP and Non-HDL-C is the next goal after LDL-C is at or near goal)-... can't ignore that-  Lets await the subset of data-

Dear Melissa,

     Long time no see.  In 2000 I pubished an analysis of eight published angiographic regresssion trials (J Cardiovasc Risk, 2000; 7:415-423).  As you know, I use a global risk factor graph approach using the cholesterol Retention Fraction (CRF, or[LDL-HDL]/LDL) on the ordinate and systolic blood pressure (SBP) on the abcissa.  The graph contains a threshold line with CRF-SBP loci of (0.74,100) and (0.49, 140).  The paper shows that any therapy that brings the patient's CRF-SBP plot below the threshold line results in angiographic stabilization/regression of coronary plaque in a minimum average of 75% of cases.  (Had POSCH been structured to controll blod pressure, many more CRF-SBP plots would have been vbelow the threshold line and the percentqage of angiograms with plaque stabilization/regeression would have approached 95% of cases.)

     Because POSCH was not structured to control hypertension and because when HDL is very low, the CRF will always be high, I also elected to look at LDL by itself.  When the LDL level was brought below 80 mg/dl (2.0 mmoles/L)--anywhere below 80 mg/dl--angiographic staqbilization/regression of coronary plaque occurred 93% of the time.  Since plaque stabilization/regeression is associated with a dramatic fall in subsequent clinical events, the results of AIM HIGH are not surprising in the least.

     Incidentally, if LDL is substituted for CRF in the graph, predictions are nowhere near as good as when CRF is used. (Feeman, presented at various international symposia, but not yet published)  Finally, if the direct measurtement of HDL is used, the result is 10 mg/dl (0.25 mmoles/L) higher than the HDL result when the older indirect measurement technique is used.  This results in a LDL level 10 mg/dl (0.25 mmoles/L) lower than would have been obtained had the indirect meausrement of HDL been used.  Hence the new goal would be 70 mg/dl (1.75 mmoles/L).

                                                                 Bill

I think this is going to simply come down to people weren't taking Niaspan.  They already know that more than 30% of the patients with ischemic strokes in the Niaspan group weren't taking it.  They also had to add Zetia to 515 people in the study.  94% of the patients were on a statin with an average LDL-cholesterol of 71mg/dl.  Adding Niaspan at 1,500mg-2,000mg should have made Zetia unnecessary in the treatment group.  I think the investigators underestimated the effort it takes to keep people taking Niaspan.  Anyone who routinely prescribes it knows how difficult it is to keep them on it.

Okay Dr.Brown and Boden did you count pills?  CJ thinks you cannot pronounce results unless you counted pills.  And if you didn't count pills well the results of the trial are negated. So did you count pills or not - I mean I don't count pills in my practice but apparently that is yet another thing I should be doing.  And I have reviewed the testimonial on the Bale-Doneen web site - I find this blog something considerably less than I had hoped.

 

High dose statin plus niacin is not the same as low dose statin plus niacin.

 As statins attenuate the production of cholesterol and niacin changes the quality of cholesterol, perhaps  the use of high dose statin negates the benefit of niacin.

A non-result from an “intent to treat” study with high dose statin does not trump prior studies and should not negatively affect our use of niacin in combination with stains, especially in combination with low dose statins.

We hold two truths to be self evident, RCTs are always correct, and the most recent RCT is better than an older RCT.  This basic tenant is wrong as RCTs are not infallible and older RCTs are not subservient to newer RCTs.

The AIM-High results are very interesting and demand further evaluation however, anyone who uses this study as a reason to stop using niacin is doing their patients a disservice. 

 

Dr. K,

 Don't blame you for being skeptical. So was I. Skepticism is healthy!!!

 

As for the blog, feel free NOT to read or participate.....however,  always glad to have a serious post about serious issues.

Melissa

 

Thanks James and I appreciate your insight. I based my support of niacin utilization on 10 years worth of data that looked like good data with what in my humble opinion seemed to be good trial design despite small numbers (I don't think that's always an oxymoron?).  I don't recommend things to my patients that I would not consider for myself or my family given the same situation. We should be careful to never accept anything as dogma, however and that's a lesson we can all learn from AIM-HI.  Everything in medicine deserves scrutiny!!

Melissa

I have abandoned the use of Niaspan on the basis of the preliminary AIM-HIGH data.

1) The study is 10X larger than the next largest niaspan study (CDP used IR niacin and was published in the pre-aspirin, pre-beta blocker, pre-ACE inhibitor, pre-clopidogrel, pre-PCI, pre-thrombolysis, pre-statin era; capiche?). The DSMB concluded that there was only a 1 in 10,000 chance that continuing the trial would show an efficacy difference. We call that a null trial.

2) Compliance was 92% as reported. So this is not an issue of poor adherence.

3) We do not look to subset analyses in trials with null primary endpoint - that violates good statistical practice for clinical trials.

4) The increase in stroke is worrisome and cannot be attributed to greater ezetimibe use in the control group (see SHARP).

5) Imaging data and small trials do not add up to much compared with a study of 3414 patients followed for 2.7 years with a primary event rate of 5.8% PER YEAR in the niaspan group. This is not an underpowered trial.

6) The data that are available support: a) statins; b) higher doses of statins are better than low dose statins; c) fibrates in patients with atherogenic dyslipidemia (low HDL and/or high TG); d) omega-3 fatty acids.  You can drop ezetimibe and niaspan from the therapeutic armentarium, given the lack of efficacy data. 

In summary, AIM-HIGH has rocked our beliefs which were based on smaller trials with imaging surrogates, alternate preparations of niacin (eg Slo-niacin, IR niacin), generally in poorly treated patients with higher starting LDLc levels.  Note that the baseline LDLc in AIM-HIGH was 72 (that was baseline). 

Dr. K,

Although you bring up some good points, it is generally hard to follow your train of thought due to all of the run on sentences.  I had to read, and re-read your posts many times just to try to figure out what you were saying.  In the future, please proofread your posts and add punctuation to provide some clarity as to what your point is.  I found myself getting lost in searching for a comma or a period, rather than focusing on the message you were attempting to convey.

Dr S.

I appreciate your comments.  It is indeed fortunate that I do not rely on my linguistic talent to make a living.  What I did not appreciate until now was the harm I was inflicting on others.  I will try to do better and I thank you for taking the time to consider my comments however tortuously constructed. 

DG

Well stated - particularly comment #3.  

The AIM-High results are very interesting and demand further evaluation however, anyone who uses this study as a reason to stop using niacin is doing their patients a disservice. 

Such an enormous disservice that it was not detectable in a sample size of 3400 patients.  Exactly what benefit is conferred by a medicine when a test of benefit proves absence of same?  I do agree that additional information may alter the discussion but to accuse practitioners of poor decision making based on a large RCT that looked specifically at the question at hand seems to demonstrate a bias that I must admit I do not understand.  It is also in my opinion an irresponsible statement directed at those of us treating patients based on the result of proven (and disproven) treatment strategies.  Are these discussions important - yes it turns out patients read comments on blogs such as this and defending the humble standard of care remains an important endeavor.  Likewise asserting something is superior to standard of care requires proof otherwise it is just advertising and in my opinion demeans our profession. 

 

What does this study prove?  It might prove that maximum dose of statin prevents the benefits of niacin from being expressed.  It might prove that there are un-accounted and uncontrolled for risk factors that have skewed the results.

 

There are older RCTs, HATS and FATS, that demonstrated dramatic benefits from niacin with therapy other than high dose statin.  I realize that high dose statin is the darling of the east coast thinkers, however there are potential serious issues with high dose statins, negation of niacin effect might be one. 

 

I am passionate about the flaws in this study because niacin is needed by many patients to control MI risk.  This study failed to prove otherwise; however impressionable, scientifically naive physicians will use this as an excuse to avoid the not insignificant headaches of initiating niacin.   

 

William,

 

"This study failed to prove otherwise; however impressionable, scientifically naive physicians will use this as an excuse to avoid the not insignificant headaches of initiating niacin."

Quite a remarkable proclamation.  The study showed no benefit of niacin  added to simvastatin in this patient subset.  Labeling colleagues who disagree with your opinion on the value of niacin as "scientifically naive" speaks volumes.  The notion that a large RCT would dispel previous suggestions of benefit should not be a foreign notion and numerous examples are given above. Which flaws did you encounter that leads you to the conclusion that the results are of no scientific value?  I think instead "believers" will dissect a study that refutes their preconceived notion of benefit conferred by niacin - if there is sufficient scientific merit so be it but point it out.  Finally what is this "passion" non-sense anyway?  Niacin as a therapy for atherosclerosis has data suggesting modest benefit that is dwarfed by statin trials (in terms of efficacy and volume of patients studied) and which is apparently not additional when added to statin therapy in this patient subset.  I think that comment is true on both coasts.

 

Ultimately, it comes down to why we do trials?  If we truly believe in a drug's efficacy a priori, i.e. if the evidence is that strong and convincing, no one would have performed, planned, executed AIM-HIGH and HPS2-THRIVE.  Then, when the results are in, why do we discard them if they do not dovetail with our prior beliefs ("confirmation bias")?  

I have seen nothing in the preliminary press release or the baseline characteristics and design paper or the Q & A posted on the NHLBI website to dispute or discard this trial.  

This was a well-conducted trial which achieved the requisite increases in HDL and decreases in TRGL, had high rates of compliance, high rates of events (including all-cause death), a large number of patients, and answered a scientifically important question.  Personally I am not waiting on HPS2-THRIVE in 2013-4, which tests a drug we don't even have access to (niacin+lapopirant), to act on these results.

In reply to Dr. K

What I do not understand is the intense "cult of niacin"

… because it requires the mindset of a Niacin basher to fabricate such an imaginative  cult.

And yes it’s a disservice to pull the plug on Niacin ..which Abbott did by the way in Europe discontinuing as of July Niaspan, its Niacin brand.

Well,beyond LDL there are  other lipoproteins with undisputed atherogenic effects to keep in mind  if you really are serious about lowering the total  risk of CVD.

One of them warranted  last year  the European Atherosclerosis Society to  recommend  future systematic screening.

What it has to do do with Niacin    …well Niacin is the only  drug able to significantly lower its level.

Do you see at last that elephant (not that small) sitting in the room ?

I hope you will excuse and allow a patient comment.

On my GP's advice, I've been taking 1000mg of instant release B3 twice daily with breakfast and dinner (and a baby aspirin) along with 10mg of Lipitor for the past 5 years.  I am quite happy with my 70/50 LDL/HDL ratio, the 50 TG, as well as with the measurable reduction in my carotid plaque.  So far, I've passed the age of both my younger brother's and father's first angioplasties and also hope to avoid a quadruple bypass such as my father had a few years ago.

If raising HDL (and especially the best kinds of HDL) and lowering TG mean nothing, then God only knows where we're to go.  I've no intention of giving up on B3, and I would miss that pleasurable "I know it's working" flush.

BTW. I am glad that I am not taking a megadose of statins in terms of possible side-effects.