PCSK9 inhibitors (vs statins) and diabetes

 There is concern about an increased risk of diabetes in patients taking statins. A recent study of patients taking PCSK9 inhibitors, however, found that there was not a clear relationship for them with developing diabetes (see lipid PCSK9 and diabetes CurrentAtheroRep2022, or doi.org/10.1007/s11883-022-01074-y).

 

Details: 

-- this study involved an evaluation of the cardiovascular outcomes of the FOURIER (using evolocumab) and ODYSSEY (using alirocumab) trials, meta-analyses, Mendelian randomization trials, and post-marketing safety reports

-- Main outcome: new-onset diabetes (NOD) rates in patients on PCSK9 inhibitors

 

Results: 

-- FOURIER trial with evolocumab:

    -- 25,982 patients on evolocumab found a monotonic relationship between achieved LDL and major cardiovascular outcomes, down to an LDL <20mg/dL (cad high risk LDL less than 20 fourier studyCirc2023 in dropbox, or DOI: 10.1161/CIRCULATIONAHA.122.063399)

    -- no differences vs placebo in the fasting plasma glucose, A1c, or rate of NOD after two and three years of follow-up

    -- at the 7.1 year follow-up, the overall annualized incidence rate of NOD was 1.2% compared to 2.3% in the placebo arm [ie, it was actually lower with the PCSK9 med.  BUT it turns out that buried in the supplementary materials, more of the patients decreased or stopped their statins in the PCSK9 wing, though no details on any of the specifics….]

    -- this lack of relationship with NOD was also found in the BANTING study

 

-- ODYSSEY trial with alirocumab:

    -- pooled analysis of 14 trials where the LDL reached concentrations of <15 mg/dL: no increased risk of diabetes or diabetic complications, and this was confirmed in secondary analysis of the subsequent ODYSSEY OUTCOME trial

 

-- Meta-analysis of all of these PCSK9 trials:

    -- 68,123 participants, median follow-up 78 weeks, did find a significantly increased fasting blood glucose (1.88 mg/dL) and A1c ( 0.032%) with PCSK9’s, with these changes increasing stepwise with duration of treatment; the dysglycemia started to be evident at a mean follow-up of 1.5 years

    -- but these were pretty miniscule changes in glucose control, and neither the incidence of diabetes nor the worsening of diabetes appeared to be higher

 

-- Metanalysis of statins and statins/PCSK9 inhibitors (in the ever increasing numbers of patients who need to have some lipid-lowering therapy):

        -- 163,688 patients without diabetes assigned to more intensive versus less intensive lipid-lowering therapy: neither LDL lowering nor the use of PCSK9 inhibitors was associated with the incidence of diabetes

        -- the results of statin studies suggest that those developing NOD were basically confined to those already at a high risk of developing diabetes, and that statins were more likely to accelerate the process than lead to de novo development of diabetes: Statin Safety and Associated Adverse Events: A Scientific Statement From the American Heart Association | Arteriosclerosis, Thrombosis, and Vascular Biology (ahajournals.org). And they note that high risk of developing diabetes itself is a cardiovascular risk factor as per below (and, therefore, should have a higher likelihood of needing statins....)

 

-- Post-Marketing safety reports:

    -- data from the FDA Adverse Event Reporting System (FAERS): 87,724 reports were identified from patients taking a PCSK9 inhibitor, finding an increased reporting of hyperglycemia of 2.1% vs what they found in the full database, leading to an adjusted-reporting odds ratio of 1.14 (1.07-1.22); they found a disproportionate increase in NOD in those on evolocumab, where there was an adjusted OR of 1.24 (1.15-1.32), but no increase in those on alirocumab. The increase in those on evolocumab was mostly for hyperglycemia rather than diabetes

    -- most of the hyperglycemic events occurred during the first month of therapy with PCSK9 meds, as opposed to statins where hyperglycemia occurred mostly around 3 months: both time intervals were likely related to differences in the rapidity of LDL lowering with these 2 meds (PCSK9 being much faster)

    -- patients with diabetes who were on PCSK9 inhibitors had more hyperglycemia than those without diabetes: 11.3% vs 2.1%

  

Commentary: 

-- atherosclerotic cardiovascular disease, as we all know, is associated with a huge global burden of disease (more than 30% of the global adult population), and is the leading cause of death and disability worldwide, as well as in the United States. This is despite pretty dramatic advances in the prevention and treatment of heart disease both pharmacologically as well as with interventional therapies

-- the PCSK9 inhibitors present a dramatically effective way of lowering LDL cholesterol levels, though their use (ie, from high cost) is limited to people with documented significant heart disease or extremely high LDL levels, when there is documented insufficient response to other medications. Of note, the PCSK9 ability to lower cardiovascular outcomes is actually pretty much the same as with other lipid-lowering medications (including bile acids sequestrants, ezetimibe, and statins) on a per mg/dL reduction of LDL (ie, PCSK9 inhibitors are more clinically effective just because they get more LDL lowering)

-- human studies have shown that LDL lowering to <20 mg/dL, which is largely achieved through PCSK9 inhibitors, seems to provide benefit without significant risks: https://gmodestmedblogs.blogspot.com/2018/08/very-low-ldl-levels-benefit-without-harm.html and https://gmodestmedblogs.blogspot.com/2021/12/ldl-less-than-40-in-high-risk-patients.html . though some patients do get very low LDL levels just with statins, and sometimes even without taking the most powerful ones...

-- it is notable that the incidence of diabetes in those on statins seems to be largely confined to those with "prediabetes"

-- and, this brings up the concern about using “diabetes” as the clinical endpoint, since the cutpoint of A1c of 6.5% was established because it reflects the beginning of an increase in microvascular complications (specifically retinopathy). BUT the major cause of mortality is definitively cardiovascular causes, around 80% of all mortality in patients with diabetes

    -- “prediabetes” is associated with a significant increase in cardiovascular disease, more significantly so in men, with increased cardiovasc risk even below an A1c of 5.5%: http://gmodestmedblogs.blogspot.com/2022/11/prediabetes-increases-risk-heart.html

    -- and, unfortunately, many patients as well as clinicians do not pay enough attention to aggressively documenting and treating prediabetes as an important cardiovascular risk factor (the name “prediabetes” undercuts its clinical significance)

    -- so, using the marker of new-onset diabetes (NOD) may be misleading:

        -- NOD is a specific issue: eg A1c reaching the 6.5% mark, and prediabetes is a range of 5.7% up to 6.5%: 

            -- the rate of progression to diabetes in those with "prediabetes" is, not surprisingly, much higher in those closer to that 6.5% threshold

            -- therefore, assessing the development of NOD may vary significantly depending on the clusters of patients involved (were their average A1c's 6.4% vs 5.7%??)

            -- it is therefore important in studies to actually assess the specifics of “prediabetes” and stratify the results accordingly (which is not done very often)

            -- in two of the statin trials, patients with prediabetes developed diabetes while on the statins, but at a five-fold rate of those without prediabetes, over a five-year follow-up: https://www.sciencedirect.com/science/article/pii/S0735109714070375?via%3Dihub

        -- but the bottom line: to the extent that the rate of NOD is increased with statins (and studies do vary some in this association), the consensus opinion is that the long-term risk of the potential major complications of diabetes is far outweighed by the benefits of statins

 

-- genetic studies have found that there is an association between 3-Hydroxy-3-Methylglutaryl-CoA reductase (HMGCR) variants that have decreased functionality leading to low levels of LDL clinically, but these people still have increased risk of new onset diabetes

    -- four Mendelian randomization studies have assessed the genes encoding PCSK9, 3 finding an increased NOD risk at a rate of each 38.7 mg/dL decrease of LDL being associated with 1.69 mg/dL higher fasting glucose, and a 29% increase in NOD, pretty much the same effect as those with dysfunctional HMGCR.

       -- there have therefore been suggestions that the culprit in diabetes risk may not be the statins but simply the lower LDL levels: https://gmodestmedblogs.blogspot.com/2015/09/low-ldl-and-diabetes-risk.html

       -- the current PCSK9 studies suggest strongly that the issue is NOT the low LDL levels, but the low HMGCR function (which can either be by statins or genetic polymorphisms)

       -- hence the importance of human clinical outcomes studies: Mendelian randomization is a useful and important technique, but does not always elucidate what happens with patients….

    -- of note, in PCSK9-deficient genetically-altered mice, there is impaired glucose-stimulated insulin secretion, though not peripheral insulin resistance. different from the human studies

       -- hence, the importance (again) of human studies….

 

-- one other important perspective is diabetes prevention: both the Finnish Diabetes Prevention Study and the US Diabetes Prevention Program (DPP) found that people with impaired glucose tolerance in a rigorous program of weight loss, diet, and exercise had dramatic decreases in the development of diabetes, with a few comments:

    -- the DPP had 3 groups (lifestyle modification, metformin 850mg bid, and placebo), finding that lifestyle modification was about twice as effective as metformin in decreasing fasting plasma glucose (!!), and the benefit lasted 10 years after the study started, with some people benefiting even 15 years later

        -- the lifestyle intervention was even more effective in people >60yo

    -- 7-year followup of the Finnish study also found that the diabetes reduction continued

    -- so, even in those who might develop NOD on statins (and, per above, the group developing NOD being largely in those with some glucose intolerance), it is a reasonable assumption that non-drug, lifestyle changes targeting those with glucose intolerance or high probability of developing diabetes might have a profound, long-term effect, and may decrease progression to diabetes for 10 years or so

         -- which all means that we should be aggressively treating prediabetes, basically with motivational interviewing about appropriate lifestyle changes.

 

Limitations:

-- the group of patients on PCSK9 inhibitors is a highly select one: those at very high cardiovascular risk (mostly secondary prevention, though some with primary severe hypercholesterolemia, etc) who are unable to take other meds or have insufficient response. This restriction limits generalizability to the general population

-- post-marketing reporting is only as good as the reporting is: likely many patients who had small increases in their A1c were never reported (ie, may never have made it into their full database), so these results are helpful but not determinant

-- and the usual concern about meta-analyses: they are combining results of disparate patients of disparate demographics of different doses of different statins with disparate inclusion and exclusion criteria into one (reductionist) amalgam.  which really does raise questions about their generalizability. that being said, there is a benefit in the numbers, especially when these disparate groups in the different studies have pretty homogeneous results

 

so, a few points:

-- PCSK9 inhibitors do not seem to increase the risk of new onset diabetes

-- they are not so available anyway for the vast majority of primary care patients (and, though i am quite aggressive in lipid management, both pharmacologically and through nonpharmacologic lifestyle counseling, i have never prescribed them). some combination of powerful statins (eg rosuvastatin 40mg, which is more potent than any of the others on the "high intensity" palette) perhaps augmented with ezetimibe seems to work well enough for many needing aggressive medical lipid management along with lifestyle changes

-- this article, yet again, shows that some of our assumptions are just wrong and need to be tested clinically in people: many (including me) considered the real possibility that the (small) increase in diabetes with statins might be from the lower LDL levels themselves (and dismiss the statin/diabetes link); this current study found that it was likely actually the effect on HMG-CoA reductase blocking by statins.  we need to be careful/cognizant/always questioning about our assumptions.....

-- And we need to be careful about relying on the results of animal studies to predict human effects:

    -- part of the issue is likely that there are actually real differences between "mice and men"

    -- and part is that with animal studies, genetic researchers tend to isolate a single entity (eg changes in a single polymorphism) and assess the results. but in humans, there is pretty much always an interplay with the array of genetics (with potentially many different genetic polymorphisms with complex interactions between them vs targeting a single one) and the much more complex social environments in which we live (which can lead to profound epigenetic changes). Many of these criticisms also apply to Mendelian randomization studies: humans are much more complex than finding people with a few genetic polymorphisms and assuming that these changes are determinant, given the many interactions between the social environment and the genes, some mediated through epigenetics where the complex environment turns on or off genes, some by other means (eg changes in the microbiome).

 

-- and, as an important side note: we should be careful about using "diabetes" as “the” risk factor for atherosclerotic events: there is a continuum of increasing cardiovascular risk as the level of A1c increases from about 5.5% until extremely high levels. The term "prediabetes" may lead to complacency by patients and by us (eg see https://gmodestmedblogs.blogspot.com/2023/10/update-ascvd-risk-factor-critique.html ). And this confusion could potentially lead to profound conclusions (eg that statins “cause” diabetes) incorrectly, instead of concluding that perhaps there are small changes in glucose tolerance that push some people from the threshold of “prediabetes” to “diabetes”, maybe without really significantly increasing their cardiovascular or other diabetes-related risk…  this needs to be studied more specifically, but it would be a disservice for people reading this study, or patients hearing this news, to decrease their statin usage: statins are a really important intervention in preventing cardiovascular disease….

 

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

  

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

  -- click on 3 parallel lines top left, if you want to see blogs by category, then click on "labels" and choose a category​

  -- or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them

 

if you would like access to the dropbox for articles, go to https://www.dropbox.com/scl/fo/vj803z91w1trd471h9fj8/h?rlkey=klpxdjpdhcdt3sahnpirzz730&dl=0

 

please feel free to circulate this to others. also, if you send me their emails (gmodest@bidmc.harvard.edu), i can add them to the list