Peripheral arterial disease (PAD) results in significant morbidity and mortality in patients with kidney failure.[1,2] The traditional risk factors for PAD in the general population include age, gender, diabetes mellitus (DM), hyperlipidemia (HL), hypertension and smoking. In patients with kidney failure, there may be additional interplay, such as elevated homocysteine levels, with vascular calcification from chronic inflammation and mineral bone disease. Currently, statin therapy is recommended in patients with PAD, and there exist ample data to demonstrate lipid-lowering by statins to confer cardiovascular (CV) protection. Robust clinical trials in the general population have provided irrefutable evidence for statin therapy in both primary[6,7] and secondary prevention.[8,9] However, CV benefits were not observed in major clinical trials attesting statin therapy in kidney failure cohorts.[10–12] In the Study of Heart and Renal Protection (SHARP), for example, though a fixed dose of simvastatin and ezetimibe among >6000 predialysis CKD subjects and >3000 dialysis patients resulted in a 17% reduction in atherosclerotic events, as compared with placebo, the clear benefit in event reduction observed in predialysis patients was lost in patients on dialysis.
Based on these data, the 2013 Kidney Disease: Improving Global Outcomes Clinical Practice Guideline for Lipid Management has thus not advocated follow-up measurement of lipid levels for the majority of dialysis patients nor the initiation of statin therapy in prevalent dialysis patients. In another clinical practice guideline originating from Hong Kong, statin treatment in dialysis-dependent patients is not uniformly advocated but can be individualized. At present, there is a dearth of data regarding PAD prevention in the kidney failure population, and whether statin therapy may afford clinical benefit in patients with kidney failure is unknown. Moreover, with the exception of SHARP, which had analyzed for the outcomes of noncoronary revascularization, clinical trials had not evaluated PAD as a CV outcome in patients with kidney failure.
In this issue of Nephrology Dialysis Transplantation, Hsu et al. conducted a population-based cohort study to investigate the effect of statin therapy on PAD in Taiwanese hemodialysis (HD) patients with HL. From the National Health Insurance Research Database, a total of 3658 HD patients on statin treatment for HL were matched 1:1, via propensity score, with 3658 HD patients without statin treatment and followed for a mean of 4.18 years. The core finding observed by the investigators was an approximate 2.5% greater cumulative incidence of PAD in the statin group as compared with the nonstatin group, with a corresponding incidence of 16.9 and 12.5/1000 person-years and an adjusted hazard ratio (HR) = 1.34 [95% confidence interval (CI) 1.12–1.62]. Fluvastatin (HR = 1.88, 95% CI 1.12–3.14) and atorvastatin (HR = 1.6, 95% CI 1.24–2.08) were particularly associated with elevated risk. Multivariate analyses also demonstrated a dose-dependent relationship between statin dosage and the risk of PAD, which was higher among patients who received moderate-intensity statins.
These findings are not only consistent with current literature, regarding the lack of efficacy of statins to reduce CV outcomes in kidney failure, but also underscore their potentially harmful effects in this particular cohort. It has previously been reported that statin treatment is associated with coronary artery calcification progression in kidney failure, by hampering vitamin K metabolism. In this regard, the promotion of vascular calcification by statin therapy on the background of a uremic milieu may underlie the increased association with PAD.
Although well performed in a homogeneous population, using a propensity score matching system, which incorporated a comprehensive list of covariates, the investigation has numerous limitations some of which are inherent to studies of this kind. First, there were incomplete data regarding baseline demographics, medication history, modality of HD and biochemistry results. Subgroup analyses were, therefore, hindered. Furthermore, statin compliance could not be ascertained and their effect on lipid levels was unknown. This is of particular disadvantage given that HL is an independent risk factor for PAD incidence. Similarly, lifestyle and smoking habits were unknown. Second, the diagnosis of PAD in this study was only based upon ankle–brachial index (ABI) measurements, calculated by dividing the systolic blood pressure at the ankle by that in the upper arm. Falsely normal or high ABI values can occur with medial arterial calcification (noncompressible vessels) associated with DM and albuminuria, resulting in underdiagnosis. Inclusion of pulse wave velocity measurements and abdominal aortic calcification scores by computed tomography could markedly improve the sensitivity of the diagnosis. Third, the duration of CKD and the degree of albuminuria prior to the initiation of dialysis are other confounders as vascular age is known to be impacted upon by these factors. Finally, there could be differences in clinical attributes and risk factors that were not measured between the cohort that was treated with statins and the cohort that was not, such that even propensity score matching may not have eliminated bias by indication. The provocative data herein do, however, provide interesting observations to support future investigations looking into the mechanism of vascular calcification and its role in PAD development in the kidney failure population. However, given the negative results in overall CV outcome from major clinical trials, the utility of statin therapy may be gradually sidelined.
Ultimately, this study provides further damaging evidence to the use of statins in kidney failure. Despite their efficacy to lower lipid levels, it remains evident that a more complex pathophysiology resulting in CV morbidity and mortality is at play in the kidney failure population, such that traditional statin therapy appears unfruitful. The current study would lend support to opening a new channel of investigation looking into CV and hopefully also PAD outcomes of newer lipid-lowering therapies, such as proprotein convertase subtilisin/kexin Type 9 (PCSK9) inhibition. PCSK9 is a secreted serine protease that binds to the extracellular domain of the hepatocyte low-density lipoprotein (LDL) receptor and promotes its lysosomal degradation, thus increasing concentrations of circulating LDL cholesterol. Monoclonal antibodies that act as PCSK9 inhibitors induce an opposite effect by sequestering PCSK9 and thereby preventing LDL receptor catabolism, leading to an increase in its density on hepatocytes. In a post hoc analysis of a PCSK9 antibody, alirocumab, among the 315 patients with CKD with Stage 3 CKD from eight pooled trials, the safety and efficacy on renal and lipid parameters were similar in those with CKD. However, these data excluded patients with advanced CKD and included those with kidney failure; therefore, it remains unclear if this class of drugs will reduce CV morbidity and mortality in dialysis patients. While statins may only have a peripheral role to address vascular disease in patients with kidney failure, future studies to explore new lipid-lowering or other strategies to reduce vascular risk in kidney failure are long overdue.
Supported by an Endowment Fund established for the 'Yu Professorship in Nephrology' at the University of Hong Kong awarded to S.C.W.T., Mrs Rita T. Liu SBS of L&T Charitable Foundation Ltd & Indo Café and Mr Winston Leung.
Nephrol Dial Transplant. 2020;35(10):1645-1647. © 2020 Oxford University Press