I very much agree with your analysis and had done the manual subtraction to estimate the effects in the "secondary prevention" group myself. I'm saddened to see JAMA publish this without mentioning the counter presentation of the other 9764 participants and the possible increase in mortality (if you believe these results). If one is to believe the benefits are more significant in the primary prevention group, then shouldn't we also recognize the positive subgroup analyses that showed ONLY a benefit in males? In those aged 65-75? With LDL 130-160? And only diabetics?
The conventional wisdom (based on statins) is that a 1 mmol reduction in LDL should provide 25% RRR in CV mace, regardless of primary vs secondary prevention.
In order for us to accept the “primary prevention” subgroup benefits, it would require a biologically plausible explanation as to why this agent produces disparate clinical results relative to LDL lowering, based on “prevention” status. Also curious for something to work in lower risk and NOT work in higher risk. Something is weird here.
And of course, this is all predicated on “actual” statin intolerance, where other trials have shown a significant nocebo effect.
Has anyone considered that any benefits seen in mortality/morbidity with this agent, and with statins, has nothing to do with the lowering of LDL? Could it be that statins-and possibly bempedoic acid-provide a benefit because they lower triglycerides, the true bad actors in the opinion of some? And that some statins, but not all, provide some benefit because of their possible anti-inflammatory properties, particularly to the arterial endothelium? There seems to be something going on besides just lowering LDL/altering the patient's lipid profile. Pravastatin, in various trials, has usually generated pretty impressive outcomes data, without being a big number mover. Atorvastatin, clearly a more powerful statin than Pravastatin, and one that generates impressive "numbers," has had a far more difficult time producing clear, positive outcomes data. And has never, to my knowledge, produced positive outcomes data as clear as that produced by Pravastatin, Simvastatin or Rosuvastatin. So: could it be about something more than just the numbers?
I am reminded of a statement made by a psychiatrist of my acquaintance regarding SSRI's: "We know what they do...we just don't know how they work." By that he meant, of course, that we know what the pharmacological effects of SSRI's are--they increase serotonin levels in the synaptic spaces. We don't really know, though, how they improve depressive symptoms. Serotonin levels rise rather rapidly upon onset of treatment. Depression can take weeks or months to improve; if at all.
So could statins--and other agents which change someone's lipid profile--be working but not for the reasons we think? And therefore produce curious results as demonstrated by this study?
Please accept my apologies if this layman's questions seems silly or bizarre. And thank you, Dr. McCormick, for an interesting essay.
Totally agree and thanks - I find it much simpler to just care about "IF THEY WORK" rather than "HOW THEY WORK". We have been screwed by false mechanism of action way too many times to hang our hat on mechanisms. Personally I think statins reduce the risk of CVD by spreading fairy dust over intraarterial plaques - and I'm sticking to that belief until proven otherwise🤣🤣
I'm on record as the skeptical cardiologist (https://theskepticalcardiologist.com/2023/03/05/bempedoic-acid-an-option-for-those-with-true-statin-intolerance/) as writing "In my clinic we have been able to achieve target apo B levels in those with SAMS or other reasons for statin intolerance utilizing combinations of ezetimibe and PCSK9 inhibitor drugs. There are a few patients who don’t seem to tolerate any of these 3 agents although they work in entirely different ways. They tend to have very idiosyncratic side effects (like lower back pain or diarrhea) and also have side effects from multiple other medications.
I suspect that the patient intolerant of the 3 existing medications will experience the same side effects on bempedoic acid but I will offer it to them. I’ll be sure to mention the increase rate of gall stones, gout, and renal impairment in the bempedoic acid group. And the high price."
I also realize that subset analyses of RCTS are fraught-investigators like to tout them when the overall trial is positive and downplay them when it is negative.
Overall, these findings don't move me from my initial lukewarmness toward bempedoic acid
I agree - lukewarm is more than hot enough :) As an aside, and also as you are "The Skeptical Cardiologist" I was wondering if you could help me with the following issue. It is my understanding that the measurement variation in LDL levels is such that in an individual person one needs to see at least a 20-30% change before one can rule out the change being simply due to biologic/analytic variation https://pubmed.ncbi.nlm.nih.gov/32079593/. Given that, how do cardiologists deal with the fact that adding ezetimibe or increasing statin doses only lower LDL 10% - or maybe 15% at best - in other words the change is barely half what is needed to determine if a real change has occurred. Given that, wouldn't it be best to simply just give medications that have been shown to reduce CVD risk and just not bother measuring lipids. I posted a video about this a few months ago https://youtu.be/Q1MhrVB2tFE and I've never really received an adequate answer. Thanks in advance for your time.
I haven't looked at the literature on the topic of in detail and didn't have time to access your paywalled article but in 30 years of experience testing and treating high LDL I have found the changes induced by statin and ezetimibe therapy are dramatic obvious. There is a significant variation in LDL and apo B that I see in patients both on and off treatment. Typically on the order of 10% or so. Often, we can identify lifestyle factors that cause this, sometimes not. It is good to have a few measurements at different points prior to treatment and after to fully understand the effects of drug treatment. In the vast majority of patients, however, on average we see more than 10-15% incremental reductions when adding on ezetimibe onto a statin. I am aiming to goals in secondary prevention or high risk primary prevention that are quite aggressive and with a consistent reduction on repeated measurements below the goal apo B and LDL-C targets.
If you just give meds without verifying achieving target goals, you fail to ascertain problems with patient noncompliance with meds or lifestyle and you don't take into account individual factors (like lipo (a), varying intestinal cholesterol absorption, etc.)
These people unethically published the best of 38 subgroups to try to make their net harmful drug look better. I think that's disgusting and unethical. They gave it to people most likely to benefit from it, more died overall and of CV disease, the confidence intervals skewed above one for both so the drug may directly cause death, and when people saw through those terrible results they published this subgroup while ignoring that logic must dictate that if the results for primary are good the results for secondary must be terrible since they were bad overall. It was a mistake to grant this drug accelerated approval and it is a mistake to leave it on the market.
Thank you very much, James, you are correct. And you can actually see the results of the subgroup analysis in the online version of the ORGINAL study. Although the authors do not mention it, the results are presented in the supplementary materials. (Their entire subroup analysis Forest plot is shown). A benefit only in primary prevention, but not in secondary prevention. It is also (of course) very entertaining to read of the financial ties of many of the authors to Esperion, the company that is trying very hard to profit off this scam.
It is a sad comment on the state of medicine that the investigators' financial interests is the first thing that comes to mind whenever I read about a questionable analysis.
No way do I want to lower my cholesterol since the body makes 80-85% of its own (in at least a dozen different varieties) and for a million good reasons. Why mess with this natural process? And I still won't worry about it after I switch to a carnivore diet.
But go ahead and invent more drugs. After 50 years, the state of America's health is still going downhill despite all these life saving pharmaceuticals.
This is what I found, too. But Marilyn Mann, above, states there is a well-established causal relationship between LDL-cholesterol and atherosclerotic cardiovascular disease.
I’m missing something here. Again, not a medical anybody.
I have familial high cholesterol and refuse to take statins because from what I’ve witnessed they have more side effects vs doing any good, plus I just respond better to lifestyle changes vs meds.
After the past 3 years, I’ll be double dog damned to put any meds in my body that I don’t trust 100%.
There's an informative book by Duane Graveline: "The Statin Damage Crisis", worth reading. Statins work by cutting off a tree of biochemical reactions in our bodies. They do a good job of lowering cholesterol.. At a heavy price. By cutting off the "mevalonate pathway" at the base, they can cause loss of cholesterol in our brains, among other damage, leading to dementia and worse. And they cut off the body's supply of coenzyme Q10, which fights the inflammation that probably causes CVD.
Graveline was a NASA medical officer who was severely injured by statins, suffering transient global anemia twice, then ALS.
I fear that widespread statin use is quite likely the cause of the dementia that seems to be on the rise. I saw what happened to my father after he started on atorvastatin, a sad death from dementia of a man who was an active, vital airline pilot and Angus farmer.
Thanks for this information. I’m sad to hear about your dad; I know a few folks who had moderate to severe side effects from Lipitor, and I knew I didn’t want to go through that.
I don’t disagree with your analysis, but even if we didn’t look at the secondary prevention subgroup, analyzing subgroups of randomized controlled trials is not very reliable.
I love this very clever analysis! My takeaway is we should only be paying attention to the original study as designed, which showed no mortality benefit and a modest reduction in MACE which feels outweighed by the rates of kidney and liver toxicity. I think the most likely explanation of the subset analysis which confusingly has different effects on primary and secondary prevention is that this is noise due to a methodologically flawed approach to interpreting trials.
Caveat: I am not connected to any medical anything. I thought the whole “cholesterol causes heart problems “ theory was pretty much debunked. Or, is my lack of knowledge causing me to read this report incorrectly? Seems it’s just another cholesterol lowering drug.
There is a very well-established causal relationship between elevated LDL-cholesterol and atherosclerotic cardiovascular disease. I don’t know what sources you’re relying on, but that has not been debunked. To the contrary, it’s very well-established not just epidemiologically, but also in terms of randomized controlled trials and Mendelian randomization studies.
I don’t think that is correct. There is a reason people look at LDL particle number, small dense versus big fluffy, they look at ApoB and Lp(a), etc.
I think it has NOT been conclusively demonstrated that there is a direct causal relationship between LDL-cholesterol and ASCVD; there is certainly a correlation though.
First, I did not say LDL-C was the only biomarker that mattered. As for lipoprotein(a), it is important in the case of people with high levels, but there are as of yet no randomized controlled trial showing that lowering lp(a) reduces cardiovascular events. There are some drugs in development, so that could change.
First, I agree the failure to include the full analysis, without explanation, leads one to be suspicious. Hopefully there will be 2 additional studies, one for primary prevention only, and the other for secondary prevention only, so there will be no need for subgroup analysis and multiple comparisons. The study author referred to the primary prevention results as "striking," resulting in "really large reductions," based on RRR. But the absolute risk reduction in all-cause mortality was only 1.6%. Is that truly a "really large reduction," especially considering the fact that the concerning increase in impaired kidney function was 2.2%? I wish study authors would stop using relative risk reduction to draw any conclusions, because it is meaningless unless one knows the absolute risk to begin with, which is especially important in primary prevention. I also wish there were clearer standards on defining a "meaningful" ARR. Sometimes I've noticed that such low ARR's are interpreted as not being clinically meaningful, and sometimes they're called "really large." Perhaps it varies depending on whether one is talking about public health and the overall population vs. at the individual level. When I'm making the decision with my doctor whether to take a drug, I only care about the latter. When the risk of serious adverse events such as impaired kidney functioning is higher than the ARR, that is something that should be highly emphasized, but it usually is not.
Yes, and that is the first recommendation that AHA guidelines make. But making such changes often requires education and support, and insurance generally doesn't cover such prevention-oriented programs. Very backwards!
I very much agree with your analysis and had done the manual subtraction to estimate the effects in the "secondary prevention" group myself. I'm saddened to see JAMA publish this without mentioning the counter presentation of the other 9764 participants and the possible increase in mortality (if you believe these results). If one is to believe the benefits are more significant in the primary prevention group, then shouldn't we also recognize the positive subgroup analyses that showed ONLY a benefit in males? In those aged 65-75? With LDL 130-160? And only diabetics?
Very thought provoking.
The conventional wisdom (based on statins) is that a 1 mmol reduction in LDL should provide 25% RRR in CV mace, regardless of primary vs secondary prevention.
In order for us to accept the “primary prevention” subgroup benefits, it would require a biologically plausible explanation as to why this agent produces disparate clinical results relative to LDL lowering, based on “prevention” status. Also curious for something to work in lower risk and NOT work in higher risk. Something is weird here.
And of course, this is all predicated on “actual” statin intolerance, where other trials have shown a significant nocebo effect.
Great analysis, Prof. McCormack.
Has anyone considered that any benefits seen in mortality/morbidity with this agent, and with statins, has nothing to do with the lowering of LDL? Could it be that statins-and possibly bempedoic acid-provide a benefit because they lower triglycerides, the true bad actors in the opinion of some? And that some statins, but not all, provide some benefit because of their possible anti-inflammatory properties, particularly to the arterial endothelium? There seems to be something going on besides just lowering LDL/altering the patient's lipid profile. Pravastatin, in various trials, has usually generated pretty impressive outcomes data, without being a big number mover. Atorvastatin, clearly a more powerful statin than Pravastatin, and one that generates impressive "numbers," has had a far more difficult time producing clear, positive outcomes data. And has never, to my knowledge, produced positive outcomes data as clear as that produced by Pravastatin, Simvastatin or Rosuvastatin. So: could it be about something more than just the numbers?
I am reminded of a statement made by a psychiatrist of my acquaintance regarding SSRI's: "We know what they do...we just don't know how they work." By that he meant, of course, that we know what the pharmacological effects of SSRI's are--they increase serotonin levels in the synaptic spaces. We don't really know, though, how they improve depressive symptoms. Serotonin levels rise rather rapidly upon onset of treatment. Depression can take weeks or months to improve; if at all.
So could statins--and other agents which change someone's lipid profile--be working but not for the reasons we think? And therefore produce curious results as demonstrated by this study?
Please accept my apologies if this layman's questions seems silly or bizarre. And thank you, Dr. McCormick, for an interesting essay.
If I remember correctly Malcom Kendricks said the same - that Statins may help inflamation, LDL not the story
Totally agree and thanks - I find it much simpler to just care about "IF THEY WORK" rather than "HOW THEY WORK". We have been screwed by false mechanism of action way too many times to hang our hat on mechanisms. Personally I think statins reduce the risk of CVD by spreading fairy dust over intraarterial plaques - and I'm sticking to that belief until proven otherwise🤣🤣
False confidence in mechanism also can lead us to hang too much on putative markers of benefit rather than outcomes
Couldn't agree more - check out my video on why you don't need to measure cholesterol more than once in a lifetime. https://youtu.be/Q1MhrVB2tFE
Thanks, this is easy and logical to follow. I haven’t really been sold on this medicine as a primary care doc, so I’ll keep digesting this analysis. 👌
I'm on record as the skeptical cardiologist (https://theskepticalcardiologist.com/2023/03/05/bempedoic-acid-an-option-for-those-with-true-statin-intolerance/) as writing "In my clinic we have been able to achieve target apo B levels in those with SAMS or other reasons for statin intolerance utilizing combinations of ezetimibe and PCSK9 inhibitor drugs. There are a few patients who don’t seem to tolerate any of these 3 agents although they work in entirely different ways. They tend to have very idiosyncratic side effects (like lower back pain or diarrhea) and also have side effects from multiple other medications.
I suspect that the patient intolerant of the 3 existing medications will experience the same side effects on bempedoic acid but I will offer it to them. I’ll be sure to mention the increase rate of gall stones, gout, and renal impairment in the bempedoic acid group. And the high price."
I also realize that subset analyses of RCTS are fraught-investigators like to tout them when the overall trial is positive and downplay them when it is negative.
Overall, these findings don't move me from my initial lukewarmness toward bempedoic acid
I agree - lukewarm is more than hot enough :) As an aside, and also as you are "The Skeptical Cardiologist" I was wondering if you could help me with the following issue. It is my understanding that the measurement variation in LDL levels is such that in an individual person one needs to see at least a 20-30% change before one can rule out the change being simply due to biologic/analytic variation https://pubmed.ncbi.nlm.nih.gov/32079593/. Given that, how do cardiologists deal with the fact that adding ezetimibe or increasing statin doses only lower LDL 10% - or maybe 15% at best - in other words the change is barely half what is needed to determine if a real change has occurred. Given that, wouldn't it be best to simply just give medications that have been shown to reduce CVD risk and just not bother measuring lipids. I posted a video about this a few months ago https://youtu.be/Q1MhrVB2tFE and I've never really received an adequate answer. Thanks in advance for your time.
I haven't looked at the literature on the topic of in detail and didn't have time to access your paywalled article but in 30 years of experience testing and treating high LDL I have found the changes induced by statin and ezetimibe therapy are dramatic obvious. There is a significant variation in LDL and apo B that I see in patients both on and off treatment. Typically on the order of 10% or so. Often, we can identify lifestyle factors that cause this, sometimes not. It is good to have a few measurements at different points prior to treatment and after to fully understand the effects of drug treatment. In the vast majority of patients, however, on average we see more than 10-15% incremental reductions when adding on ezetimibe onto a statin. I am aiming to goals in secondary prevention or high risk primary prevention that are quite aggressive and with a consistent reduction on repeated measurements below the goal apo B and LDL-C targets.
If you just give meds without verifying achieving target goals, you fail to ascertain problems with patient noncompliance with meds or lifestyle and you don't take into account individual factors (like lipo (a), varying intestinal cholesterol absorption, etc.)
These people unethically published the best of 38 subgroups to try to make their net harmful drug look better. I think that's disgusting and unethical. They gave it to people most likely to benefit from it, more died overall and of CV disease, the confidence intervals skewed above one for both so the drug may directly cause death, and when people saw through those terrible results they published this subgroup while ignoring that logic must dictate that if the results for primary are good the results for secondary must be terrible since they were bad overall. It was a mistake to grant this drug accelerated approval and it is a mistake to leave it on the market.
Saddened to see Nissen with this look
Thank you very much, James, you are correct. And you can actually see the results of the subgroup analysis in the online version of the ORGINAL study. Although the authors do not mention it, the results are presented in the supplementary materials. (Their entire subroup analysis Forest plot is shown). A benefit only in primary prevention, but not in secondary prevention. It is also (of course) very entertaining to read of the financial ties of many of the authors to Esperion, the company that is trying very hard to profit off this scam.
It is a sad comment on the state of medicine that the investigators' financial interests is the first thing that comes to mind whenever I read about a questionable analysis.
Reminds me of other statistically questionable analyses put forward by Dr Nissen!
UBC, my alma mater! 🇨🇦
No way do I want to lower my cholesterol since the body makes 80-85% of its own (in at least a dozen different varieties) and for a million good reasons. Why mess with this natural process? And I still won't worry about it after I switch to a carnivore diet.
But go ahead and invent more drugs. After 50 years, the state of America's health is still going downhill despite all these life saving pharmaceuticals.
This is what I found, too. But Marilyn Mann, above, states there is a well-established causal relationship between LDL-cholesterol and atherosclerotic cardiovascular disease.
I’m missing something here. Again, not a medical anybody.
I have familial high cholesterol and refuse to take statins because from what I’ve witnessed they have more side effects vs doing any good, plus I just respond better to lifestyle changes vs meds.
After the past 3 years, I’ll be double dog damned to put any meds in my body that I don’t trust 100%.
There's an informative book by Duane Graveline: "The Statin Damage Crisis", worth reading. Statins work by cutting off a tree of biochemical reactions in our bodies. They do a good job of lowering cholesterol.. At a heavy price. By cutting off the "mevalonate pathway" at the base, they can cause loss of cholesterol in our brains, among other damage, leading to dementia and worse. And they cut off the body's supply of coenzyme Q10, which fights the inflammation that probably causes CVD.
Graveline was a NASA medical officer who was severely injured by statins, suffering transient global anemia twice, then ALS.
I fear that widespread statin use is quite likely the cause of the dementia that seems to be on the rise. I saw what happened to my father after he started on atorvastatin, a sad death from dementia of a man who was an active, vital airline pilot and Angus farmer.
No way I'd ever chance it with statins.
Thanks for this information. I’m sad to hear about your dad; I know a few folks who had moderate to severe side effects from Lipitor, and I knew I didn’t want to go through that.
I don’t disagree with your analysis, but even if we didn’t look at the secondary prevention subgroup, analyzing subgroups of randomized controlled trials is not very reliable.
I love this very clever analysis! My takeaway is we should only be paying attention to the original study as designed, which showed no mortality benefit and a modest reduction in MACE which feels outweighed by the rates of kidney and liver toxicity. I think the most likely explanation of the subset analysis which confusingly has different effects on primary and secondary prevention is that this is noise due to a methodologically flawed approach to interpreting trials.
Caveat: I am not connected to any medical anything. I thought the whole “cholesterol causes heart problems “ theory was pretty much debunked. Or, is my lack of knowledge causing me to read this report incorrectly? Seems it’s just another cholesterol lowering drug.
There is a very well-established causal relationship between elevated LDL-cholesterol and atherosclerotic cardiovascular disease. I don’t know what sources you’re relying on, but that has not been debunked. To the contrary, it’s very well-established not just epidemiologically, but also in terms of randomized controlled trials and Mendelian randomization studies.
I don’t think that is correct. There is a reason people look at LDL particle number, small dense versus big fluffy, they look at ApoB and Lp(a), etc.
I think it has NOT been conclusively demonstrated that there is a direct causal relationship between LDL-cholesterol and ASCVD; there is certainly a correlation though.
First, I did not say LDL-C was the only biomarker that mattered. As for lipoprotein(a), it is important in the case of people with high levels, but there are as of yet no randomized controlled trial showing that lowering lp(a) reduces cardiovascular events. There are some drugs in development, so that could change.
Well, ok. Guess I’ll have to do more checking into this. Thanks!
First, I agree the failure to include the full analysis, without explanation, leads one to be suspicious. Hopefully there will be 2 additional studies, one for primary prevention only, and the other for secondary prevention only, so there will be no need for subgroup analysis and multiple comparisons. The study author referred to the primary prevention results as "striking," resulting in "really large reductions," based on RRR. But the absolute risk reduction in all-cause mortality was only 1.6%. Is that truly a "really large reduction," especially considering the fact that the concerning increase in impaired kidney function was 2.2%? I wish study authors would stop using relative risk reduction to draw any conclusions, because it is meaningless unless one knows the absolute risk to begin with, which is especially important in primary prevention. I also wish there were clearer standards on defining a "meaningful" ARR. Sometimes I've noticed that such low ARR's are interpreted as not being clinically meaningful, and sometimes they're called "really large." Perhaps it varies depending on whether one is talking about public health and the overall population vs. at the individual level. When I'm making the decision with my doctor whether to take a drug, I only care about the latter. When the risk of serious adverse events such as impaired kidney functioning is higher than the ARR, that is something that should be highly emphasized, but it usually is not.
I'm no doctor, but often a small reduction in fairly low absolute risk can be had from improved diet and increased moderate excercise.
Yes, and that is the first recommendation that AHA guidelines make. But making such changes often requires education and support, and insurance generally doesn't cover such prevention-oriented programs. Very backwards!