Lifestyle modifications are the first line approach for lipid management whereas statins are the mainstay of treatment in those who do not achieve the target with non-pharmacological approaches (1). Statin treatment is safe and the benefit for CV prevention goes beyond the mere reduction of cholesterol encompassing their anti-inflammatory, antithrombotic and antioxidant actions (2). However, safety concerns particularly related to muscle side effect are among the most significant determinants of poor statin adherence and therapy withdrawal [statin intolerant (INT) patients] (3). On the contrary, a considerable number of patients may still have high cholesterol level despite the maximum tolerated stating dose [statin inadequate responder (IR) patients] and require additional lipid-lowering therapies (LLT). Randomised controlled trials (RCTs) evaluating combination of statins with ezetimibe demonstrated that this combination is adequate in patients with low to moderate risk but still at least one-third patients with high or very high CV risk will require the addition of a third compound to achieve the LDL-c target. The approval of proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors was a milestone in the history of lipid management, however it posed economic issues with regard to the broad use in the large population of statin-IR and statin-INT (4). Therefore, the use of new cost-effective LLTs associated with reduced muscular side effects can definitely improve the success of treatment.

Bempedoic acid (BA), formerly ETC-1002, is a small-molecule first-in-class of inhibitors of ATP citrate lyase (5). In 2016 the key papers by Pinkosky et al. demonstrated that BA led to low density lipoprotein (LDL) receptor upregulation, LDL-cholesterol (c) decrease, attenuation of atherosclerosis, reduction of hepatic lipids and body weight and improved glycaemic control (6, 7). BA effects on LDL-c can be detected when given both alone and as add-on therapy to other LLTs and it also shows anti-inflammatory effects by decreasing high-sensitivity C-reactive protein. BA seems a promising approach for optimal lipid management, however its effect on serum uric acid (SUA) levels raised safety concerns. BA is known to be associated with modest SUA elevation and the putative mechanism is the competition between the BA glucuronide metabolite and UA for the same renal transporters involved in the excretion of these compounds (8). SUA is a recognised CV risk factor (9) and prognostic SUA thresholds have been identified with regard to CV mortality and CV events such as cerebrovascular diseases (10–12). Therefore, not only it seems contradictory to use a drug able to control one CV risk factor while triggering another one, but the evidence of a higher incidence of gout induced by BA in the recently published CLEAR Outcome large RCT (13) added a layer of complexity. In fact, since the target population of BA treatment is often multimorbid and shows high to very CV risk, is it imperative to clarify the safety profile of this drug to avoid overestimation of an adverse event whose incidence may be minimised or even prevented. On this basis, we aimed to explore in detail the available data on SUA and gout from RCTs assessing BA and clarify their strength and clinical relevance.

A systematic literature search was performed on PubMed, Embase, Web of Science, Cochrane Library, Emcare, Academic Search Premier and Google Scholar from inception to 4 May 2023 according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines (Supplementary Table S1). The search strategy was developed by an expert librarian (JWS) based on the Population/Intervention/Comparator/Outcome (PICO) framework (Supplementary Text S1). Title and abstract screening, as well as the article full-text review, was conducted in duplicate independently by two authors (EC and FB) with any disagreements resolved by discussion. The following inclusion criteria were used: English language Phase 2 and Phase 3 RCT were included if they enrolled patients undergoing treatment with BA regardless of the study outcomes. Data extraction was performed independently in duplicate using a pilot-tested data extraction form. Data pertaining to the article's identifying information, methods, population, interventions, efficacy and safety outcomes were extracted. The risk of bias of included articles was not assessed as we were mostly interested in reviewing methodological aspects of the studies. Likewise, meta-analysis and related assessments (e.g., Cochran's Q for heterogeneity) were not performed. However, to facilitate comparison of studies, effect size (unadjusted odds ratios (OR) or standardised mean differences and 95% confidence intervals (CI)) was computed using the data available in the included articles.

A total of 745 references were retrieved (317 after deduplication) of which 23 were eligible for full text assessment and 16 were included in the review (Supplementary Figure S1). Six of the 16 articles were phase 3 RCTs (13–18), 1 article was the open-label extension (OLE) of one RCT (19) and 9 articles were phase 2 trials.

BA is a promising LLT to be used alone or as combination therapy in statin-INT and statin-IR patients, however its SUA-increasing effect as well as its putative effect on gout raised safety concerns (27–29). However, previous SLRs were conducted before the publication of the large CLEAR Outcome study and the Authors identified the relatively low patient number as a limitation to derive meaningful conclusions. Furthermore, they highlighted a possible selection bias related to baseline CV risk and renal function. We demonstrated that currently available data from phase 2 and phase 3 RCTs do not allow to confirm a clear association between BA and gout and that several aspects remain unclear. In particular, the paucity of information about SUA levels at baseline and/or at end of follow-up do not allow to quantify the effect sizes for BA-induced SUA elevation. As stated in the 2021 ESC guidelines on CV disease prevention in clinical practice, the achievement of recommended LDL-c goals in patients with high and very high CV risk is of paramount importance (30). Although statins are overall safe and efficacious, muscle side effects severely impact on treatment adherence and retention rate, and a considerable number of patients may not achieve the treatment goal despite the maximum tolerated statin dose. Therefore, combination therapy or switch to other LLTs or is required for an optimal lipid management. The combination of statins and ezetimibe, two oral and relatively low-cost drugs, proved to be adequate in patients with low to moderate risk but still at least one-third patients with high or very high CV risk will require the addition of a third compounds (31). PCSK9 inhibitors revolutionized lipid management but their elevated cost may be a barrier to large-scale use in particular clinical setting (4). Therefore, novel efficacious and less expensive oral therapies are eagerly awaited. In phase 3 RCTs BA addition to background LLT in statin-IR and statin-INT was superior to placebo and allowed to achieve both a more pronounced reduction of LDL-c (14–18) and a better prevention of CV events even in patients at high/very high CV risk (13). Nonetheless, safety data related to its ability to increase SUA levels/trigger gout are still a matter of debate and may limit its implementation in clinical practice. The clinical relevance of SUA goes well beyond the simple association with gout and/or nephrolithiasis (32). In fact, hyperuricemia (with or without gout) is in an independent CV risk factor and based on the association between SUA levels and mortality (both CV and all-cause) and CV events, it is conceivable that CV damage begins with levels of SUA lower than the generally accepted threshold of 6 mg/dl. The Uric Acid Right for Heart Health (URRAH) study identified a SUA threshold value of 4.7 mg/dl for all-cause mortality and 5.6 mg/dl for CV mortality (10). In addition, SUA is an independent risk factor for cerebrovascular events after adjusting for potential confounding variables, including arterial hypertension, and a valid prognostic cut-off value (>4.79 mg/dl) has been identified. Concordantly, SUA levels >5.34 mg/dl (sensitivity 52.32, specificity 63.96, p < 0.0001) were the univariate prognostic cut-off value for all heart failure, whereas SUA levels >4.89 mg/dl (sensitivity 68.29, specificity 49.11, p < 0.0001) for fatal heart failure (11). Our analysis of BA safety data from 6 phase 2 RCTs and 6 phase 3 RCTs highlighted some aspects deserving to be discussed in light of the future use of BA in clinical practice. First of all, despite the well-known pharmacokinetic characteristics of BA, namely the competition of its glucuronide metabolite and UA for renal excretion, baseline SUA levels are not considered as inclusion/exclusion criteria. By looking at the mean (standard deviation) SUA levels provided by two studies (14–17) it becomes clear that people with baseline hyperuricemia have been enrolled. Furthermore, it is not known for how long these patients have had hyperuricemia or whether they ever received/were receiving any ULT at study enrolment. In addition, we should keep in mind that also patients with normal SUA levels at baseline (with or without ongoing ULT) may still have history of hyperuricemia and/or gout. In this regard, only one study provided information on history of gout and only in patients developing gout during the study period rather than in the entire study cohort. They reported that almost half of the patients developing gout during the study period already had a diagnosis of gout, hence these events should have been classified as gout flares rather than new-onset gout (14). In addition, one of the studies providing baseline SUA levels recruited statin-INT patients with high CV risk (14) and the mean SUA values at baseline imply that SUA management in this population was not optimal despite its clear role as CV risk. Another facet of this complex scenario pertains to time of follow up. The only 2 studies where BA effect size on gout was significant are those with the longest follow up. In particular, the effect size was weakly significant at 12 months (19) but became more consistent up to a median follow up of 40 months (13). This further reinforces the importance of an accurate SUA assessment at baseline but also underlines the need of SUA monitoring overtime. In fact, as nicely shown in the OLE of the CLEAR Harmony study, SUA increase reaches its peak within the first 12 W of treatment and then it remains stable until drug withdrawal (19). Therefore, a careful monitoring of SUA levels may allow to optimise SUA management at baseline, adjust SUA management overtime as needed and ultimately prevent flares/new onset of gout. In conclusion, there is still a knowledge gap on whether BA triggers gout and based on existing data it is reasonable to speculate that this effect may be, at least in part, the result of a selection bias not considering SUA levels and history of gout when starting BA treatment. Real-life data may be instrumental to shed some light on this matter and improve the safety profile of BA in statin-INT and statin-IR patients.