Similar LDL-C Percent Reductions But Fewer Reaching LDL-C Targets in a Real-world PCSK9 Inhibitor Cohort


Real-world LDL-C Reduction in a Nested Cohort of PCSK9 Inhibitor Patients

Feolo, DA, Biskupiak, JE

Objectives: Proprotein Convertase Subtilisin Kexin Type 9 inhibitors (PCSK9i) are human monoclonal antibodies approved for primary and secondary prevention of ASCVD in the following populations: very high ASCVD risk, established ASCVD, HeFH and HoFH patients. Long-term real-world data and high cost has limited the use of PCSK9i. This study determined real world data on LDL-C reduction in a nested cohort, and compared that data with the FDA clinical studies for evolocumab* and alirocumab±. Real-world data from studies like this are essential for clinicians and payers to determine the future role of these agents in prevention of ASCVD.

Study Design: Data was pulled from a regional health insurance database on patients approved for PCSK9i (N=24). Then using an uncontrolled longitudinal study design (before and after treatment) LDL-C data was collected.

Methods: Percent LDL-C change from baseline and 95% confidence intervals were calculated. The last LDL-C value pretreatment was used to calculate baseline, and the first LDL-C representing a significant change was used to calculate mean LDL-C after treatment and percent change under baseline.

Simple descriptive statistics were used to characterize the baseline and demographic characteristics of the cohorts. The LDL-C percentage reduction endpoint was chosen due to the small sample size and limited follow-up time to measure hard endpoints such as ASCVD events. These percent reduction values were then analyzed using a simple t-test to determine the p-value and 95% confidence interval. Percent achieving an LDL-C <70mg/dL was also calculated. These results were then compared to clinical trials which help established the efficacy of Repatha and Praluent.

Results: Cohort mean LDL-C change was -58.9% (95%CI: -69%, -49%). LAPLACE-2 study (Evolocumab*) LDL-C change was -45% (95%CI: -52%, -39%). Study 2 (Alirocumab±) LDL-C change was -46.0% (95% CI: -53%, -39%). Percent achieving LDL-C goal of our cohort was 67%, while the Repatha study reported 93% and the Praluent study 82%. The baseline median LDL-C of our cohort was 162 mg/dL, while baselines for Repatha and Praluent cohorts were 102 mg/dL.

Conclusions: Despite the similar LDL-C reductions between the real-world cohort and the clinical trials data, the percentage reaching recommended LDL-C goals was significantly less in the real-world. This may indicate that factors such cost, intolerance, non-adherence, lifestyle, and not maximizing conventional lipid modifying background therapies is contributing to the smaller percentage of patients meeting goal in the real-world. Future research could look to describe the reasons why patients fail to meet these goals despite PCSK9 inhibitor therapy.

Published in College of Pharmacy, Virtual Poster Session Spring 2020


  1. Did you expect the real-world patients to have more trouble reaching their goals before getting your results? Any surprises? It seems like it would be useful to have more patients with higher LDL-C levels in the trials. Good work!

    1. I did not expect this at first, but after I really started looking at the data I realized this was a major difference. It became more clear to me upon reading recently published real-world studies similar to mine that they were also seeing this same thing occurring and to an even greater extent. I think this underlines the importance of looking at the inclusion and exclusion criteria and the baseline characteristics in a study when determining the strength of evidence. It may be less important in this case since the dramatic reductions are present in the real-world and clinical settings, but if you were expecting a patient to almost certainly meet the recommended LDL-C for their risk category then you might end up a little disappointed in the real world.

  2. Given the price of these drugs, they must be pretty disappointed in the results. Will they continue testing?

    1. As the prices come down (which they have modestly) I suspect you will see more patients at very high risk despite statins and ezetimibe utilizing PCSK9i therapy. The percent LDL-C reductions in the real-world are as good as the clinical trials. What is not completely clear is this large discrepancy between percent reaching goal. Since the percent reductions are similar and the baseline LDL-C discrepant between the real world and clinical participants I am theorizing the major difference is simply the optimization of conventional therapy in the real-world patients. In other words if you are wanting to meet recommended goals then make sure you are optimizing underlying therapy first and this will increase your chances and further improve the patient outcomes. It would be an interesting study to pull the adherence data and conduct chart reviews to describe the optimization in the real-world cohort, because we know the clinical trial cohort had a 4 week lipid-stabilizing statin regimen prior to randomization.

  3. It is not clear to me what the “study cohort” group is? Both drugs? Prior treatment? Please clarify. Also, the proportion of each treated group of patients reaching LDL < 70 was fairly high. Do you expect to see 100% less than 70? Why/why not?

    1. Our real world cohort consisted of patients approved for PCSK9 inhibitor therapy for that time period. Since prior authorization generally required patients to have elevated LDL-C and either clinical ASCVD or familial hypercholesterolemia (HoFH, HeFH) as well as a trial of maximally tolerated statins and ezetimibe, we made an assumption that our cohort had optimized standard of care or attempted to and was still elevated despite this regimen. Since our cohort baseline LDL-C was significantly higher this may in retrospect have been a bad assumption to make. For example, how in depth does the prior authorization process go in determining who has actually been adherent to conventional therapy? Do they take the patient and providers word for it, or do they check pharmacy claims data. Why are patients unable to tolerate their statins or tolerate them at a higher doses? Were re-challenges attempted or alternative statins tried? Did the provider attempt to deferentially diagnose the side effects (such as myalgias)? The clinical trials also had an open-label specific statin regimen for a 4-week lipid stabilization period, which may have selected for adherence. Since the clinical trials had a significantly higher number meeting goal compared to our study as well as other real world studies (who had even less meeting goal by the way), it seems clear the percent of patients in the real world meeting goal is not near 90% as the clinical trials suggest, but more likely in the 40 to 70% range. I would expect if you could really hone in on the reasons why the clinical participants have a lower baseline LDL-C, then you may be able to determine why less people in the real world actually meet goal. Regardless the LDL-C percent reductions are significant and similar in the real world and clinical trials.

  4. Hi David,

    Very interesting study. Do you think something a simple as adherence is the issue?

    1. I do think optimizing management prior to PCSK9i therapy is essential to meeting goals at the levels met in the clinical trials. This could come down to adherence ultimately because the prior authorization criteria usually require maximum tolerated statins and ezetimibe. Real-world studies on PSCK9i’s like ours show significantly higher baseline LDL-C when compared to the clinical trials as well. This may be because the clinical trials include an open-label specific statin regimen for a 4-week lipid stabilization period prior to randomization. This may indicate that although LDL-C percent reductions are attainable, if underlying therapy is not optimized then many patients may never meet the goal of <70 mg/dL after PCSK9i.

  5. David, nice work. I found this so interesting. While it looks like fewer meet goal, given their higher starting levels, the change in your cohort is quite remarkable, even if they don’t reach goal. I wonder if the difference isn’t so much a lack of real-world efficacy of the drug, but rather, as you suggest, that the real world patients in your cohort are not being as optimally managed (either because they’re non-adherent or their management is lacking) as the patients in the original clinical trials. Do you have thoughts on what might be the main issue?

    1. Although our sample size was small I would not have expected to see as large of difference at baseline between the real world and the clinical trial participants (other real world studies found largely the same discrepancy). One specific reason why I suspect this is the case is management/adherence, as I suggested. For example in the Laplace-2 trial for Repatha, all participants underwent an open-label specific statin regimen for a 4-week lipid stabilization period (as did the Praluent Study-2 participants). It would have been interesting to learn more about my study cohort participants actual lipid modifying regimen and compare that with pharmacy claims data to determine if there is any correlation between adherence and meeting recommended goals of therapy. Regardless the LDL-C reduction appears to be “as advertised” and regardless of meeting goals the benefits for prevention of stroke, MI, revascularization and hospitalization appears to be clear.

  6. David – I love this topic. Good drugs used poorly result in not amazing outcomes! Wonder if a clinical pharmacist was involved in the care of these pts…..
    Dr G

Comments are closed.