Efficacy and safety of proprotein convertase subtilisin/kexin type 9 inhibitors for adults with familial hypercholesterolemia: A network meta-analysis

Abstract

Purpose

The comparative efficacy and safety profiles of PCSK9 inhibitors in familial hypercholesterolemia (FH), including genotype-dependent treatment responses, remain unclear.

Methods

This systematic review was conducted in accordance with the Preferred Reporting Items for Meta-Analyses guidelines. A network meta-analysis of randomized clinical trials evaluating the use of PCSK9 inhibitors for the treatment of FH patients, including subgroup analyses of efficacy, was performed.

Results

Fifteen randomized clinical trials (n = 2954 patients) were included. All PCSK9 inhibitors significantly improved lipid parameters compared to control. In heterozygous FH (HeFH) populations, ongericimab showed the greatest reductions in LDL-C (mean difference [MD]: −74.98 %), ApoB (MD: −64.64 %), and Lp(a) (MD: −59.66 %), with SUCRA rankings of 68.7 %, 63.6 %, and 95.0 %, respectively. However, these results are based on a single trial and require further validation. No significant lipid-lowering effects were observed in HoFH patients. In terms of safety, lerodalcibep showed the most favorable profile for injection-site reactions and ALT >3 × ULN, with SUCRA values of 98.5 % and 96.7 %, respectively. Inclisiran was associated with a significantly higher risk of injection-site reactions.

Conclusion

PCSK9 inhibitors generally show favorable efficacy and safety in FH patients. However, comparative rankings and point estimates should be interpreted with caution due to funnel plot asymmetry for LDL-C and imbalances in trial data. Ongericimab demonstrated promising results in HeFH, but further validation is required. Inclisiran's efficacy may be underestimated due to short-term follow-up. Monotherapy with PCSK9 inhibitors has limited efficacy in HoFH patients, highlighting the need for combination therapies.

Highlights

• •PCSK9 inhibitors significantly reduced LDL-C, ApoB, and Lp(a) levels in HeFH.
• •Monotherapy with PCSK9 inhibitors showed limited efficacy in homozygous FH patients.
• •Inclisiran demonstrated a higher risk of injection-site reactions compared to others.
• •Personalized FH treatment based on genotype is critical for optimizing therapy.

1. Introduction

Cardiovascular diseases (CVDs) remain the leading cause of global mortality and disability, accounting for an estimated 19.2 million deaths and 437 million disability-adjusted life years (DALYs) in 2023. Atherosclerotic cardiovascular disease (ASCVD) is a significant contributor to this burden, with ischemic heart disease and ischemic stroke being primary causes. Among the modifiable risk factors for CVD, elevated low-density lipoprotein cholesterol (LDL-C) stands out as one of the foremost drivers, contributing to 90.7 million CVD DALYs in 2023. Despite the well-established causal role of LDL-C in ASCVD, global efforts to reduce high LDL-C have faced challenges, with limited success outside a few high-income nations [1].

Familial hypercholesterolemia (FH), an autosomal dominant disorder, represents a significant contributor to the development of early-onset ASCVD. The condition is primarily preventable through appropriate intervention, making it a major target for public health strategies aimed at reducing global cardiovascular risk [2]. The global prevalence of heterozygous FH (HeFH) is estimated at 1/200–1/250, whereas homozygous FH (HoFH) affects 1/16,000–320,000 individuals [3,4]. Patients experience lifelong exposure to significantly elevated LDL-C levels (typically >190 mg/dL) [5], resulting in a 16–22-fold increased risk of ASCVD, coronary heart disease, and other cardiovascular events compared with the general population [[6], [7], [8]]. Although combination therapy with statins and ezetimibe reduces LDL-C by approximately 50 % [9], more than 70 % of HeFH patients and nearly all HoFH patients fail to reach the LDL-C target (<70 mg/dL) [10,11].

Proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors have transformed the management of FH. As a key regulator of cholesterol metabolism, PCSK9 binds to the epidermal growth factor-like repeat A (EGF-A) domain of the hepatic low-density lipoprotein receptor (LDLR). This interaction promotes LDLR internalization and lysosomal degradation, preventing its recycling to the cell surface and thereby impairing LDLR-mediated LDL-C clearance [12]. Targeting the PCSK9/LDLR interaction represents a groundbreaking lipid-lowering strategy. Approved PCSK9 inhibitors include two classes: monoclonal antibodies (alirocumab, evolocumab) and small interfering RNA (siRNA) therapies (inclisiran), which differ fundamentally in their molecular targets and mechanisms of action. Monoclonal antibodies bind to the catalytic domain of circulating PCSK9, preventing its interaction with the LDL-R EGF-A region, thereby protecting LDL-R from lysosomal degradation. This process restores LDL-R recycling, resulting in a 50–60 % reduction in LDL-C [[13], [14], [15]]. These monoclonal antibodies are administered subcutaneously every 2–4 weeks, given their ∼11–20 day half-life [14,16]. In contrast, siRNA therapy such as inclisiran is a GalNAc-conjugated double-stranded siRNA that is selectively taken up by hepatocytes via the asialoglycoprotein receptor. Once inside the cell, the antisense strand is incorporated into the RNA-induced silencing complex (RISC), which cleaves PCSK9 mRNA, thereby inhibiting de novo PCSK9 synthesis [[17], [18], [19]]. The reduction of both intracellular and extracellular PCSK9 enhances LDL-R expression and recycling, achieving a comparable reduction in LDL-C with an initial dose, a 3-month booster, and subsequent administration every six months [20,21]. Large-scale trials have confirmed that adding PCSK9 inhibitors to statins further reduces major adverse cardiovascular events by 15–20 % [22], suggesting a crucial therapeutic option for high-risk ASCVD patients.

In this context, several PCSK9 inhibitors, including alirocumab, evolocumab, inclisiran, tafolecimab, ongericimab, and lerodalcibep, are either in widespread clinical use or nearing regulatory approval. However, their clinical application faces challenges, particularly due to the substantial heterogeneity within the FH population. A key issue is the variability in clinical evidence supporting these agents. Alirocumab and evolocumab, both monoclonal antibodies, have demonstrated clinical efficacy in HeFH and are recommended by current guidelines [23,24]. Inclisiran, a small RNA molecule, has shown promise in reducing LDL-C by 50 % in phase III trials, but it lacks robust long-term outcome data and is not yet endorsed by major guidelines, including the 2025 ESC/EAS update [23,25].

A more fundamental challenge arises from the heterogeneity of FH, necessitating a nuanced understanding of treatment efficacy across its subtypes [26]. In HeFH, the critical clinical question has shifted from whether PCSK9 inhibitors are effective to which agent offers the optimal efficacy and safety profile. However, high-quality studies comparing the various PCSK9 inhibitors in HeFH are scarce, and systematic comparisons, especially involving emerging agents, are lacking. The situation is even more complex for HoFH patients. The 2025 ESC/EAS Guidelines highlight the limited efficacy of PCSK9 inhibitors in HoFH due to their LDL-receptor-dependent mechanism [23]. Paradoxically, real-world evidence and the 2023 EAS Consensus Statement suggest that a subset of HoFH patients with residual LDL receptor activity may benefit from PCSK9 inhibitors, with LDL-C reductions of ∼30 % [27]. This creates a critical evidence gap: while PCSK9 inhibitors are not first-line for HoFH, synthesizing available data is essential to assess their efficacy and clarify their potential role.

Network meta-analysis (NMA) offers a robust framework to address these gaps by combining direct and indirect evidence from randomized controlled trials (RCTs), enabling quantitative comparisons of multiple interventions even in the absence of head-to-head trials [28]. This study aims to conduct a Bayesian NMA to systematically assess the efficacy and safety profiles of various PCSK9 inhibitors in FH patients. Specifically, the analysis will compare the efficacy of all available PCSK9 inhibitors in HeFH and estimate the LDL-C reduction achievable with this class of drugs in HoFH. Subgroup analyses will explore the impact of genotype on treatment response. Additionally, the study will include PCSK9 inhibitors developed in China (tafolecimab, ongericimab) and an investigational agent (lerodalcibep) in a systematic comparison with established therapies. This NMA will provide evidence-based insights for personalized FH management, optimize treatment strategies for refractory HoFH, and support the clinical adoption of novel PCSK9 inhibitors.

2. Methods

This network meta-analysis was drafted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for reporting systematic reviews [29].

2.1. Data sources and search strategy

Several online databases, including PubMed, Embase, the Cochrane Library, Web of Science, and ClinicalTrials.gov, were searched from the inception of the study to May 15, 2025. The following medical subject heading (MeSH) terms and keywords were applied: "familial hypercholesterolemia", "hyperlipoproteinemia type II" and "randomized controlled trial". The detailed search strategies for these databases are provided in the appendix (Supplemental Tables 1–4).

2.2. Selection criteria

The following inclusion criteria were established: (1) RCTs published in English or Chinese; (2) adult patients (≥18 years) with FH, regardless of genotype, sex, or race; (3) the experimental group received a PCSK9 inhibitor; (4) the control group received placebo, statins, ezetimibe, other PCSK9 inhibitors, or combination therapies; (5) follow-up duration of ≥12 weeks; (6) reporting of at least one efficacy measure: LDL-C, apolipoprotein B (ApoB), or lipoprotein(a) (Lp(a)); and (7) reporting of at least one safety measure: risk of any adverse events (AEs), serious adverse events (SAEs), injection-site reactions (ISR), or alanine aminotransferase (ALT) levels exceeding three times the upper limit of normal (ULN).

The exclusion criteria were as follows: (1) conference abstracts, letters, reviews, commentaries, and case reports; (2) unpublished results or unavailable data; and (3) duplicate reports.

2.3. Data extraction and outcome assessments

Two investigators (WD and SY) independently screened the literature on the basis of the selection criteria and extracted data from eligible studies. Disagreements were resolved through discussion with a third investigator (LS). Data on the first author, publication year, study population, interventions, follow-up duration, baseline characteristics, outcomes, and methodological quality were systematically recorded via a predesigned electronic form.

The methodological quality of all included trials was independently appraised by two reviewers (WD and SY) via the Cochrane risk of bias tool. Disagreements were resolved through consultation with a third independent reviewer (LS). The key domains assessed included random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective outcome reporting, and other sources of bias. The risk of bias for each domain was categorized as low, high, or unclear. All assessments were documented via Review Manager software.

2.4. Statistical analysis

Network meta-analysis was performed within a Bayesian framework using the gemtc package in R software (version 4.5.0). Non-informative priors were assigned to all model parameters to ensure that posterior distributions were primarily driven by the observed data [30]. For each outcome, we employed three independent Markov chain Monte Carlo (MCMC) chains with 50,000 iterations following an initial burn-in period of 10,000. Model convergence was assessed through visual inspection of trace plots and the Gelman-Rubin diagnostic. Continuous variables were presented as mean differences (MD) with 95 % confidence intervals (CI), whereas dichotomous variables were reported as risk ratios (RR) with 95 % CI. A random-effects model under a consistency framework was used to synthesize both direct and indirect evidence. The consistency of treatment effects across studies with varying treatment groups was assessed via a design-by-treatment interaction model, and discrepancies between direct and indirect evidence were evaluated via node-splitting methods. Network evidence diagrams were created for each outcome measure. Subgroup analyses were conducted to examine differences in drug efficacy across specific populations. The cumulative ranking probabilities (surface under the cumulative ranking, SUCRA) for different interventions were calculated and visualized via a heatmap [[31], [32], [33]]. Publication bias was evaluated using funnel plots, and its presence was further assessed with both Egger's regression test and Begg's test [34].

The SUCRA values represent the probabilistic ranking of treatments based on relative efficacy, and should not be construed as direct measures of clinical efficacy or certainty. Given the observed funnel plot asymmetry, caution is warranted in interpreting these rankings.

3. Results

3.1. Study selection and characteristics

In the initial screening of this study, a total of 2241 articles were identified. After 319 duplicates and 1922 articles that did not meet the inclusion criteria were excluded, 14 articles comprising 15 RCTs were ultimately included. The PRISMA flowchart depicting the screening process is shown in Fig. 1.

Fig. 1.

Flow diagram of the study selection process.

A total of 15 RCTs involving 2954 patients were included in this study. Specifically, 1890 patients were in the treatment group, and 1064 were in the control group, consisting of 2714 HeFH patients and 240 HoFH patients. The treatment group included six types of interventions: alirocumab in six trials, evolocumab in three trials, inclisiran in two trials, lerodalcibep in two trials, and ongericimab in one trial. The control group received either placebo or evolocumab. The key characteristics of the included studies are summarized in Table 1.

Table 1.

Characteristics of the included studies.

Author Year	Register number	Phase	Country	Diagnosis	Follow up (weeks)	Total patients	Male (%)	Regimens	Average age Mean(SD)		LDL-C Mean(SD)
									E	C	E	C
Blom 2020 [35]	NCT03156621	Ⅲ	South Africa, Japan, Italy, Canada, the Netherlands, China-Taiwan, USA and 13 other countries	HoFH	12	69	46.7	Alirocumab 150 mg Q2W Placebo	49.8 (14.2)	52.1 (11.2)	295.0 (154.6)	259.6 (175.8)
Ginsberg 2016 [36]	NCT01617655	Ⅲ	Canada, USA, Netherlands, Russia, South Africa	HeFH	24	107	48.6	Alirocumab 150 mg Q2W Placebo	59.5 (9.2)	57.0 (10.5)	196.3 (57.9)	201.0 (43.4)
Moriarty 2016 [37]	NCT02326220	III	USA, Germany	HeFH	18	62	63.4	Alirocumab 150 mg Q2W Placebo	52.1 (12.9)	51.7 (12.3)	174.15 (50.31)	193.50 (65.79)
Kastelein FH Ⅰ 2015 [38]	NCT01623115	Ⅲ	North America, Europe, South Africa	HeFH	24	486	55.7	Alirocumab 75 mg Q2W Placebo	53.2 (12.9)	53.2 (12.5)	144.7 (52.04)	144.4 (47.24)
Kastelein FH Ⅱ 2015 [38]	NCT01709500	Ⅲ	Europe	HeFH	24	249	51.5	Alirocumab 75 mg Q2W Placebo	51.3 (7.7)	51.9 (9.6)	134.6 (41.23)	134.0 (41.4)
Stein 2012 [39]	NCT01266876	Ⅱ	USA, Canada	HeFH	12	30	60	Alirocumab 150 mg Q4W Placebo	52.9 (11.2)	51.9 (9.6)	167.07 (49.73)	151.26 (34.02)
				HeFH	12	31	56	Alirocumab 200 mg Q4W Placebo	54.3 (9.6)	51.9 (9.6)	170.24 (57.08)	151.26 (34.02)
				HeFH	12	30	47	Alirocumab 300 mg Q4W Placebo	56.3 (10.2)	51.9 (9.6)	140.94 (24.72)	151.26 (34.02)
				HeFH	12	31	81	Alirocumab 150 mg Q2W Placebo	52.6 (12.3)	51.1 (14.2)	147.52 (32.76)	151.26 (34.02)
Raal(1) 2014 [40]	NCT01763918	Ⅲ	Australia, Asia, Europe, New Zealand, North America, South Africa, North America, Europe, Middle East, South Africa	HeFH	12	166	40	Evolocumab 140 mg Q2W Placebo	51.9 (12.0)	46.8 (12.1)	162.54 (50.31)	151.93 (34.83)
				HeFH	12	165	42	Evolocumab 420 mg Q4W Placebo	30 (12)	32 (14)	154.80 (42.57)	151.93 (42.57)
Raal(2) 2014 [41]	NCT01588496	Ⅲ	North America, Europe, Middle East, South Africa	HoFH	12	50	52	Evolocumab 420 mg Q4W Placebo	47.6 (13.6)	49.3 (11.3)	356.04 (135.45)	336.69 (147.06)
Raal 2012 [42]	NCT01375751	Ⅱ	North America, Western Europe, Hong Kong, China Singapore, South Africa	HeFH	12	111	54.5	Evolocumab 350 mg Q4W Placebo	51.8 (13.0)	49.3 (11.3)	158.67 (46.44)	162.54 (42.57)
				HeFH	12	112	62.5	Evolocumab 420 mg Q4W Placebo	43.8 (13.4)	40.7 (12.1)	151.93 (34.83)	162.54 (42.57)
Raal 2024 [43]	NCT03851705	Ⅲ	Hong Kong, China, Israel, Russia, Serbia, South Africa, China-Taiwan, Turkey, Ukraine.	HoFH	21	56	37.8	Inclisiran 300 mg day 1, day 90, and every 6 months thereafter Placebo	54.4 (12.48)	55.0 (11.81)	294.0 (136.3)	356.7 (122.4)
Raal 2020 [44]	NCT03397121	Ⅲ	USA Canada, South Africa, Netherlands, Spain, Denmark, Sweden, Czech Republic	HeFH	72	482	46.3	Inclisiran 300 mg day 1, day 90, and every 6 months thereafter Placebo	28.7 (15.2)	29.8 (14.8)	151.4 (50.4)	154.7 (58.0)
Raal 2025 [45]	NCT04034485	Ⅲ	India, Israel, Norway, South Africa, Turkey, United USA	HoFH	24	66	52	Lerodalcibep 300 mg Q4W Evolocumab 420 mg Q4W	52.5	53.8	392.81 (171.44)	428.64 (167.18)
Raal 2023 [46]	NCT04797104	Ⅲ	USA, Israel, Turkey, South Africa, Norway, India, Spain, New Zealand	HeFH	24	478	48.6	Lerodalcibep 300 mg Q4W Placebo	44.8 (12.6)	47.4 (13.5)	152.19 (67.34)	146.67 (57.07)
Lin 2025 [47]	NCT05325203	Ⅲ	China	HeFH	24	68	45.5	Ongericimab 150 mg Q2W Placebo	51.3 (13.0)	46.6 (10.8)	139.32 (38.70)	150.93 (69.66)
				HeFH	24	67	31.8	Ongericimab 450 mg Q4W Placebo	48.7 (13.2)	48.0 (14.4)	131.58 (42.57)	147.06 (34.83)
Chai 2023 [48]	NCT04179669	Ⅲ	China	HeFH	12	75	55.8	Tafolecimab 150 mg Q2W Placebo	51.9 (10.7)	47.4 (14.2)	164.48 (43.06)	165.25 (37.93)
				HeFH	12	73	43.8	Tafolecimab 450 mg Q4W Placebo	49.8 (14.2)	52.1 (11.2)	163.93 (44.12)	167.18 (47.01)

3.2. Risk-of-bias assessment

Overall, the studies exhibited a low risk of selection, attrition, reporting, and other biases. Raal 2025 [45] was an open-label trial with unblinded participants, investigators, and outcome assessors, introducing a high risk of performance and detection bias, especially for subjective outcomes like patient-reported adverse events. However, efficacy outcomes, being objective, are less prone to bias. Despite this, the study was included for its valuable data on lerodalcibep in the rare HoFH population, and the potential impact of bias on the results was minimal due to the objective nature of the measures. The quality of the selected studies is illustrated in Fig. 2, Fig. 3.

Fig. 2.

Risk of bias graph of the included studies.

Fig. 3.

Risk of bias summary for the included studies.

3.3. Network geometry

All models achieved satisfactory convergence, with Gelman-Rubin statistics below 1.1 and stable diagnostic plots (Supplementary Table 5 and Fig. 1).

Closed loops were absent for the Lp(a) and ISR outcomes, precluding inconsistency testing. Loops were present for LDL-C, ApoB, AEs, SAEs, and ALT outcomes, necessitating inconsistency assessment. Global inconsistency tests yielded P > 0.05 for all outcomes, indicating strong agreement. Local inconsistency tests revealed no statistically significant differences (P > 0.05) between direct and indirect comparison estimates for any regimen pair, suggesting consistency between direct and indirect evidence. The network geometries are presented in Fig. 4.

Fig. 4.

Network geometries for efficacy and safety analyses.

3.4. Efficacy endpoints

Comparative efficacy data for LDL-C, ApoB, and Lp(a) across genotyping subgroups are summarized in Table 2. The corresponding forest plots and SUCRA-based treatment hierarchies are presented in Fig. 5, Fig. 6, respectively.

Table 2.

Results of the network meta-analysis on efficacy.

Placebo VS	Alirocumab	Evolocumab	Inclisiran	Lerodalcibep	Ongericimab	Tafolecimab
	LDL-C
HoFH	−35.45 (−83.20,9.91)	−30.67 (−75.20,13.70)	−1.98 (−51.65,48.53)	−25.35 (−89.18,38.57)	–	–
HeFH	−46.39 (−59.57,-32.02)	−55.49 (−75.34,-34.96)	−49.55 (−78.91,-20.68)	−58.49 (−87.53,-31.54)	−74.98 (−103.16,-45.65)	−59.47 (−90.28,-29.92)
OVERALL	−44.29 (−57.98,-29.65)	−50.59 (−68.10,-32.29)	−33.43 (−57.52,-4.57)	−52.49 (−77.82,-27.09)	−75.34 (−108.69,-40.58)	−59.85 (−93.53,-25.21)
	ApoB
HoFH	−29.86 (−68.82,9.13)	−23.34 (−62.00,15.17)	–	−12.02 (−68.50,44.51)	–
HeFH	−35.75 (−46.37,-23.68)	−45.12 (−61.98,-28.04)	−36.11 (−60.08,-10.63)	−45.78 (−70.32,-20.83)	−64.64 (−89.40,-38.90)	−58.93 (−85.14,-32.44)
OVERALL	−34.78 (−44.31,-24.33)	−41.07 (−53.97,-27.63)	−36.31 (−59.97,-13.00)	−39.68 (−58.32,-19.78)	−64.75 (−89.77,-40.69)	−58.51 (−83.78,-32.62)
	Lp(a)
HoFH	−28.31 (−66.31,9.84)	−7.08 (−44.11,30.47)	–	−7.91 (−61.50,46.29)	–	–
HeFH	−17.45 (−23.40,-8.20)	−28.92 (−38.11,-18.90)	–	–	−59.66 (−75.19,-44.02)	−39.23 (−57.59,-21.21)
OVERALL	−18.28 (−27.63,-7.59)	−24.58 (−35.08,-9.75)	–	−25.07 (−51.51,5.78)	−59.8 (−82.63,-36.69)	−39.11 (−63.29,-14.64)

Fig. 5.

Results of the network meta-analysis on efficacy. Forest plots show the mean difference (95 % CI) for LDL-C, ApoB, and Lp(a), stratified by HoFH and HeFH genotypes.

Fig. 6.

SUCRA-based rank probability heatmaps for efficacy outcomes. (A) FH patients, (B) HeFH patients, (C) HoFH patients. Treatments are listed on the y-axis, and outcome measures are displayed on the x-axis. Darker shading indicates a greater reduction in the measure and a higher probability of being the most effective treatment. A dash ("-") indicates missing data.

3.4.1. Low-density lipoprotein cholesterol

Fifteen RCTs [[35], [36], [37], [38], [39], [40], [41], [42], [43], [44], [45], [46], [47], [48]] reported LDL-C outcomes. All PCSK9 inhibitors significantly reduced LDL-C levels compared to placebo in the overall FH population. Subgroup analyses revealed substantial heterogeneity by genotype. In HeFH patients, all agents showed significant reductions, with ongericimab demonstrating the largest reduction (MD: −74.98 [95 % CI: −103.16 to −45.65]) and the highest SUCRA (68.7 %). However, this result is based on a single Phase 3 trial with a small Chinese sample, necessitating cautious interpretation and further validation in larger, more diverse cohorts. In HoFH patients, none of the interventions significantly reduced LDL-C, although alirocumab showed the largest numerical effect (MD: −35.45 [95 % CI: −83.2 to 9.91]). Funnel plot asymmetry for LDL-C suggests small-study effects, potentially inflating point estimates and ranking precision.

3.4.2. Apolipoprotein B

Fourteen RCTs [[35], [36], [37], [38], [39], [40], [41], [42],[44], [45], [46], [47], [48]] reported ApoB outcomes. All PCSK9 inhibitors significantly reduced ApoB levels compared to placebo in the overall FH population. Subgroup analyses revealed notable differences in treatment response. In HeFH patients, all agents were effective, with ongericimab showing the largest reduction (MD: −64.64 % [95 % CI: −89.4 to −38.9]) and the highest SUCRA (63.6 %). However, this finding is based on a single trial, and its generalizability requires validation in larger populations. In HoFH patients, none of the interventions significantly reduced ApoB, though alirocumab showed the largest numerical effect (MD: −26.86 % [95 % CI: −68.82 to 9.13]) and a 59.0 % probability of ranking first.

3.4.3. Lipoprotein (a)

Eleven RCTs [[35], [36], [37], [38],[40], [41], [42],45,47,48] reported Lp(a) outcomes. PCSK9 inhibitors significantly reduced Lp(a) levels in the overall and HeFH populations, but no significant reduction was observed in HoFH patients. In HeFH, ongericimab showed the greatest reduction in Lp(a) (MD: −59.66 % [95 % CI: −75.19 to −44.02]) with a SUCRA of 95.0 %. However, this result is based on a single trial in a Chinese population, and further studies are needed to confirm its generalizability. In HoFH patients, none of the agents significantly lowered Lp(a), although alirocumab showed the largest numerical effect (MD: −28.31 % [95 % CI: −66.31 to 9.84]) with a 74.2 % SUCRA probability.

3.5. Safety endpoints

Safety assessments of six interventions based on 15 RCTs [[35], [36], [37], [38], [39], [40], [41], [42], [43], [44], [45], [46], [47], [48]] demonstrated no significant differences in the incidence of AEs between the interventions and placebo (Fig. 7, Fig. 8). The SUCRA rankings indicated that lerodalcibep (45.1 %) had the highest likelihood of reducing AEs. For SAEs, inclisiran significantly reduced the incidence (RR: 0.56 [95 % CI: 0.33 to 0.94]) and had the highest SUCRA rank (43.9 %); no significant differences were observed for the other five interventions versus placebo. In terms of ISR, according to 12 RCTs, lerodalcibep significantly reduced the risk (RR: 0.09 [95 % CI: 0 to 0.77]), with the highest SUCRA value (98.5 %). Conversely, inclisiran significantly increased the risk of ISR (RR: 11.04 [95 % CI: 4.42 to 37.79]). The remaining three interventions were not significantly different. For ALT >3 × ULN, lerodalcibep significantly reduced the incidence (RR: 0.00 [95 % CI: 0 to 0.26]), with the highest SUCRA value (96.7 %). The other five interventions were not significantly different from the placebo.

Fig. 7.

Results of the Network Meta-Analysis on Safety. Forest plots present the results for four safety outcomes as RR (95 % CI).

Fig. 8.

SUCRA-based rank probability heatmaps for safety outcomes. Treatments are listed on the y-axis, and outcome measures are displayed on the x-axis. Darker shading indicates a greater probability of the agent being ranked as the safest agent and a stronger ability to reduce risk. A dash ("-") denotes unavailable data.

3.6. Publication bias

To assess publication bias and small-study effects, Egger's test and Begg's test, along with funnel plots, were employed for three efficacy and four safety endpoints (Supplementary Fig. 1). The funnel plot for LDL-C showed significant asymmetry (P_LDL-C = 0.017), suggesting potential publication bias, which was confirmed by Begg's test (P_LDL-C = 0.026). Funnel plots for the other endpoints (P_ApoB = 0.48; P_Lp(a) = 0.74; P_AEs = 0.92; P_SAEs = 0.87; P_ISR = 0.64; P_ALT = 0.086) were symmetrical, with Begg's tests showing no significant bias (P_ApoB = 0.59; P_Lp(a) = 0.86; P_AEs = 0.75; P_SAEs = 0.75; P_ISR = 1; P_ALT = 0.16), indicating a low risk of publication bias.

4. Discussion

This meta-analysis demonstrated that all six evaluated PCSK9 inhibitors significantly reduced LDL-C, ApoB, and Lp(a) levels in patients with FH, particularly in HeFH patients. However, monotherapy with any PCSK9 inhibitor had no significant effect on HoFH. The overall safety profile was favorable across all agents, with the exception of inclisiran, which was associated with a significantly higher incidence of ISR than placebo was.

The clinical significance of LDL-C, ApoB, and Lp(a) extends beyond conventional lipid parameters. LDL-C is a key driver of atherosclerosis [49]. Elevated LDL-C penetrates the vascular endothelium, undergoes oxidation, and activates macrophages, leading to the formation of foam cells that accumulate within the arterial wall. This process progresses to the development of atherosclerotic plaques, causing stenosis and impaired blood flow, which in turn increases the risk of myocardial infarction, stroke, and other cardiovascular events [50,51]. ApoB, which represents the total number of atherogenic lipoprotein particles (including LDL, VLDL, and remnants) [[52], [53], [54]], provides a more accurate measure of "bad lipoprotein" burden and associated cardiovascular risk than LDL-C does [[53], [54], [55], [56]]. Lp(a) exerts dual pro-atherogenic effects: its lipid component facilitates cholesterol deposition, while its protein component, apo(a), impairs fibrinolysis and heightens thrombosis risk [57,58]. Cardiovascular risk significantly increases when Lp(a) levels exceed 50 mg/dL. Given that Lp(a) is primarily genetically determined and largely unaffected by lifestyle factors such as diet and exercise [54], it has substantial predictive value in patients with early-onset coronary artery disease and FH [59]. The European Society of Cardiology (ESC) guidelines classify elevated Lp(a) as an “ASCVD risk-enhancing factor” and advocate for routine testing in high-risk populations [53].

While ongericimab and tafolecimab have demonstrated significant efficacy in reducing key efficacy markers, these results should be interpreted with caution. Their recent market introduction, limited clinical experience, and the phase III trial's restricted sample size (comprising only a Chinese population) constrain the generalizability of the findings. In contrast, alirocumab and evolocumab are well-established with extensive clinical evidence supporting their use. Inclisiran, a long-acting agent, showed more modest short-term lipid reductions in this analysis (LDL-C reduction: 33.43 %, SUCRA = 0.4 %; ApoB reduction: 36.31 %, SUCRA = 1.7 %) and limited efficacy in HoFH patients (LDL-C reduction: 1.98 %, SUCRA = 5.6 %). Its biannual dosing regimen is designed for sustained long-term effect, and the relatively short follow-up periods (21–72 weeks) in the included trials may not fully capture its steady-state efficacy compared to more frequently administered monoclonal antibodies. Additionally, its limited efficacy in HoFH aligns with its mechanism of action, which relies on residual LDL receptor function. Conversely, alirocumab and evolocumab, administered more frequently, showed robust, stable lipid-lowering effects, significantly reducing LDL-C by 44.29 % and 50.59 %, ApoB by 34.78 % and 41.07 %, and Lp(a) by 18.28 % and 24.58 %, respectively. Based on the short-term data, alirocumab and evolocumab appear to have more pronounced lipid-lowering effects. However, further investigation is needed to assess the long-term comparative efficacy and cardiovascular benefits of inclisiran.

PCSK9 inhibitors are effective in treating FH patients, but genotype-based analyses reveal significant variability in response, emphasizing the need for personalized therapy. For HeFH patients, who retain partial LDLR function, PCSK9 inhibitors enhance hepatic LDLR expression, leading to substantial lipid-lowering effects when used alone. However, the response in HoFH is more complex. A meta-analysis found no significant benefit from PCSK9 inhibitor monotherapy in HoFH, aligning with the caution noted in the 2025 ESC/EAS Guidelines [23]. This broad conclusion may obscure the influence of residual LDLR function on efficacy. According to the 2023 EAS Consensus [27], the response in HoFH depends on residual LDLR activity, but current RCTs fail to stratify patients by LDLR function, leading to inconsistent results. For HoFH, treatment requires a stepwise approach: statins, ezetimibe, and PCSK9 inhibitors for patients with partial LDLR function; LDLR-independent therapies like Evinacumab, often combined with low-density lipoprotein apheresis (LA), for those with complete LDLR loss [60]. As genetic testing becomes more accessible, mutation-specific genotyping will guide personalized treatment, enabling rapid reduction of cardiovascular risk in HeFH patients and a tailored approach for HoFH [61,62].

Additionally, sex is an important variable influencing treatment response. A real-world study [63] found that men experienced greater LDL-C reduction with evolocumab, while Lp(a) reduction was unaffected by sex. These findings suggest that both genetic and sex-related factors influence treatment outcomes, requiring a multifactorial approach to FH management.

This study has several limitations that warrant cautious interpretation. The inclusion of an open-label trial [45] introduces performance and detection bias, particularly for subjective outcomes like patient-reported adverse events. However, efficacy outcomes are objective measures, less susceptible to bias. Global inconsistency tests showed no significant discrepancies, indicating the trial's inclusion did not notably affect evidence coherence. Other limitations include funnel plot asymmetry for LDL-C, suggesting potential publication bias or small-study effects, which may overestimate treatment effects and SUCRA rankings. Uneven sample sizes across PCSK9 inhibitors, especially for newer agents (ongericimab, tafolecimab) and the HoFH subgroup, limit statistical power and generalizability. Reliance on short-term lipid measures (12–24weeks) restricts assessment of long-term cardiovascular outcomes, chronic safety, and the efficacy of agents like inclisiran with extended dosing intervals. Finally, baseline heterogeneity across studies may confound pooled estimates.

Future research should focus on large-scale, genotype-specific RCTs with long-term follow-up to enable direct comparisons and evaluate cardiovascular endpoints. Additional mechanistic studies are necessary to clarify the differential responses of FH genotypes to PCSK9 inhibition, including the exploration of sex-based differences in treatment response. Understanding these differences is crucial for personalizing therapy for both sexes. Moreover, large-scale, multicenter, real-world studies are essential to validate the effectiveness and safety of these therapies in clinical practice, providing robust evidence for personalized FH management.

5. Conclusion

PCSK9 inhibitors generally show favorable efficacy and safety in FH patients. However, comparative rankings and point estimates should be interpreted with caution due to funnel plot asymmetry for LDL-C and imbalances in trial data. Ongericimab demonstrated promising results in HeFH in a single trial, but further validation is required. Inclisiran's efficacy may be underestimated due to short-term follow-up. Notably, monotherapy with PCSK9 inhibitors has limited efficacy in HoFH patients, highlighting the need for combination therapies. These findings emphasize the importance of personalized treatment decisions based on genotype, comorbidities, and patient preferences, rather than hierarchical rankings [64].

CRediT authorship contribution statement

Weiwei Ding: Writing – review & editing, Writing – original draft, Visualization, Methodology, Formal analysis, Data curation, Conceptualization. Lingyao Sun: Writing – review & editing, Visualization, Data curation. Yun Shi: Writing – original draft, Data curation. Lei Tian: Writing – review & editing, Supervision, Conceptualization.

Funding sources

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Appendix A. Supplementary data

The following is the Supplementary data to this article.