Proprotein convertase subtilisin/kexin Type 9 inhibition in lipoprotein(a) lowering: more to be desired?

This editorial refers to ‘Impact of PCSK9 inhibitors on lipoprotein(a) levels: a multi-centre study’, by H.S. Bhatia et al. https://doi.org/10.1093/eurjpc/zwaf734 Lipoprotein(a) [Lp(a)] is among the most powerful genetic risk factors for atherosclerotic cardiovascular disease (ASCVD) and is elevated in ∼20% of the world’s population.1 Elevations in Lp(a), typically defined as >50 mg/dL (or 125 nmol/L) are associated with increased risk for a wide range of ASCVD outcomes, particularly myocardial infarction, revascularization, and aortic stenosis.1,2 While Lp(a) testing rates remain low, there is recent evidence that this is increasing3 and will likely continue to increase with recommendations for universal testing and growing interest in emerging pharmacotherapies. Proprotein convertase subtilisin/kexin type 9 inhibitors (PCSK9is) are the most widely used non-statin therapy that also lower Lp(a) levels, reported to be 23–27% from prior clinical trial data4,5; however, these trials did not specifically select for persons with elevated Lp(a) and levels were quite modest in most patients. While PCSK9is are FDA approved for the reduction of low density lipoprotein-cholesterol (LDL-C) as well as ASCVD events, they are not specifically approved for reducing Lp(a) levels. What has not been established, however, is the extent of Lp(a) lowering seen in real-world patient cohorts receiving PCSK9i and inclisiran. The recent report from Bhatia et al.6 provides valuable insight into this question as evaluated from data obtained from the University of California Health Data Warehouse (UCHDW) representing six major medical centres across California and the largest such study to date examining the effects of PCSK9i on Lp(a) in a real-world setting. They included 453 individuals who had at least two prescriptions of a PCSK9i or inclisiran (of which only eight patients were on inclisiran), at least one Lp(a) measurement before and after the prescription, and where there was a reduction in LDL-C (those with only a single prescription or no LDL-C change were excluded to minimize the impact of non-adherence). A control group not receiving these therapies who had Lp(a) measures at least 30 days apart was also included. Their cohort had median baseline Lp(a) levels of 91.3 mg/dL. They reported a median 17.3% and 11.3 mg/dL reduction in Lp(a) overall in the PCSK9i group, with those with baseline Lp(a) > 50 mg/dL showing median reductions of 17.7% corresponding to a 19.2 mg/dL reduction. Higher baseline Lp(a) levels predicted greater absolute Lp(a) reductions, whereas female sex was associated with smaller reductions. Those in the control group, in contrast, had no change in Lp(a) from serial measurements. As has been shown in past clinical trials, the authors also noted substantial variation in Lp(a) response with an interquartile range of 0–31.7 mg/dL (0–35%). Also of interest, among the subset of 274 patients with more than two Lp(a) measurements, there was sustained Lp(a) reduction beyond the second measurement with minimal further reductions. The data Bhatia et al. show somewhat lower percentage reductions in Lp(a) than that observed in past clinical trials, possibly because of varying adherence to PCSK9i that could be greater than that in past clinical trials, although their median LDL-C reduction of 51% is similar to that found in prior clinical trials,4,5 suggesting adherence was quite good. The timing between the repeat Lp(a) measures and the most recent PCSK9i prescription, which was not reported, could also have contributed to these differences, where there could have been an attenuation in response if the Lp(a) was reassessed after a significant amount of time elapsed since the last prescription. Of interest, the reductions in Lp(a) obtained by the second measurement were maintained among those patients who had up to 10 measures of Lp(a). However, despite the modest reduction in Lp(a), the authors point out that this may still provide a clinically significant reduction in ASCVD risk, based on data from ODYSSEY OUTCOMES showing that a 1 mg/dL reduction in Lp(a) was associated with a 0.6 reduction in risk,5 so this could translate to approximately an 11% reduction in risk on average for the 19 mg/dL median reduction in Lp(a) the authors observed. It is important to understand too that absolute risk reduction will be greater for those with higher baseline Lp(a) levels who on average have higher baseline ASCVD risk. Moreover, the FOURIER trial4 showed in fact that those above the median baseline Lp(a) level (which was 37 nmol/L, far lower than most patients in the current study) were the ones who had the most benefit in ASCVD risk reduction (hazarad ratio [HR] = 0.77, 95% confidence interval [CI] 0.67–0.88) compared with those at or below the median (HR = 0.93, 95% CI = 0.80–1.08), suggesting that most patients in the current study would likely derive ASCVD risk reduction benefit. Also of interest, the authors show in their control cohort who did not receive PCSK9i that many had ASCVD, elevated LDL-C, and elevated Lp(a), further demonstrating the underutilization of evidence-based therapies proven to reduce ASCVD risk, even in a large academic healthcare system. Beyond average population effects, patient-level risk stratification is likely to further refine the clinical impact of Lp(a) lowering. Individuals with imaging evidence of advanced or progressive atherosclerosis may be more likely to benefit from Lp(a)-lowering therapies. Coronary artery calcium scoring, coronary computed tomography angiography, and plaque characterization may help identify patients in whom elevated Lp(a) is contributing to atherosclerotic disease burden. In such patients, effects of PCSK9i therapy may be clinically meaningful despite modest average reductions in Lp(a). This approach aligns with contemporary preventive cardiology practice emphasizing personalized, risk-based treatment escalation.7 However, more meaningful reductions in ASCVD risk of 20% or more may require greater reductions in Lp(a) as have been estimated by several studies, including one involving Mendelian randomization estimating up to 100 mg/dL or greater reductions in Lp(a) may be needed to achieve the 20% risk reduction effects equivalent to an LDL-C reduction of 1 mmol/L.8 The truth will soon be known when results of ongoing cardiovascular outcomes trials of Lp(a) pharmacologic therapies in development become available. The first of these trials, Lp(a) HORIZON9 involving the antisense oligonucleotide (ASO) therapy pelacarsen, which lowers Lp(a) ∼80% from subcutaneous injections every 3 months, is due to report out this year 2026. The readout of the Ocean(a) Outcomes trial,10 involving olpasiran, which is a small interfering RNA (siRNA) therapy providing up to 100% reductions in Lp(a) from once-monthly injections, is expected by 2027. These first two trials specifically enrolled those with known ASCVD, however, and it will be important to demonstrate the benefit of these and future therapies in the much larger group of high-risk primary prevention patients, including those with diabetes. While newer pharmacotherapies specifically indicated for Lp(a) lowering are not yet available, it is crucial clinicians understand the importance of Lp(a) as a risk driver and the importance of testing Lp(a) in all their patients, as recommended by current guidelines.11 In the meantime, patients’ global risk should be addressed through more intensive lipid-lowering and other risk factor intervention while awaiting therapies specifically indicated for Lp(a) lowering and ASCVD risk reduction. Our recent survey of more than 2000 US clinicians12 showed that while 81% agreed that Lp(a) is a significant risk driver for ASCVD, only 41% believed in universal testing. Most physicians felt it to be important for newer Lp(a) therapies to have cardiovascular outcomes data (77%), demonstrate safety and efficacy (69%), and have real-world data (53%). Until Lp(a)-targeted therapies with cardiovascular outcomes data become available, PCSK9is may be viewed as part of a broader ‘bridging-era’ strategy for managing Lp(a)-associated risk. This strategy includes universal Lp(a) testing, aggressive LDL-C reduction, optimization of modifiable risk factors, and selective use of PCSK9is in patients with elevated Lp(a) and high clinical or imaging-defined risk. Moreover, the emergence of oral PCSK9is in the near future may offer an opportunity for those preferring oral therapies. This approach allows clinicians to mitigate Lp(a) associated risk before future Lp(a)-specific interventions are available, or even in combination with such therapies, especially when concomittant further LDL-C lowering is also needed. The current paper by Bhatia et al.6 nicely demonstrates in a real-world setting that meaningful reductions in Lp(a) can be achieved from PCSK9i use, which may potentially contribute to the ASCVD risk reduction from these therapies. Such information is invaluable to physicians, patients, and payers for the consideration of such therapies already demonstrated and indicated for ASCVD risk reduction, but remain severely underutilized among the many patients who could benefit. With the emergence of newer pharmacotherapies providing more substantial and durable reductions in Lp(a), it will be particularly important to examine the impact of these therapies in real world patient cohorts. Nathan D. Wong (Conceptualization [lead], Investigation [lead], Writing—original draft [lead], Writing—review & editing [lead]), Sina Kianoush (Writing—review & editing [supporting]), and Ailin Barseghian (Writing—review & editing [supporting]) No funding was received for the preparation of this paper.