Symptom Progression in Peripheral Artery Disease: Sounding the Alarm for Cardiovascular Risk

Peripheral artery disease (PAD) is a malignant atherosclerotic phenotype associated with a heightened risk of cardiovascular (CV) events, limb-related events, and death.¹ Many patients with PAD have additional comorbidities, such as diabetes mellitus or atherosclerosis in additional vascular beds, which further amplify this risk.² In the last two decades, the landscape of PAD management has dramatically changed with the emergence of new risk factor modifying drugs and advances in endovascular therapies specifically targeting this high-risk patient population. Understanding how disease progression impacts the full spectrum of cardiovascular and limb outcomes is critical to help identify patients who will benefit from escalation of their secondary prevention therapies, some of which, such as antiplatelet and antithrombotic agents, also increase bleeding risk.

Despite a growing body of research, the natural history of PAD is poorly defined, in part due to the heterogeneity of the PAD phenotype. PAD is traditionally defined by an ankle-brachial index (ABI) ≤0.90.³ Prior data demonstrated that an ABI decline >0.15 over time is associated with more than a two-fold increased risk of all-cause and cardiovascular mortality at 3-years.⁴ However, in a separate analysis, while plaque progression over a 12-month period occurred in 35% of individuals with PAD, there was no change in ABI over this same time frame, suggesting a disconnect between dynamic atherosclerotic plaque and arterial flow disturbances in the limb.⁵

Because many epidemiological cohorts lack serial ABI measures, most estimates of PAD incidence and prevalence rely on identification of intermittent claudication (IC). A meta-analysis of 94,772 patients with PAD across 35 studies showed a progression to worsening IC or critical limb ischemia (CLI) in 21% (12–29%) over 5 years.⁶ This study included asymptomatic individuals, and it showed that important risk factors, including tobacco use, diabetes, concurrent CV disease, increasing age, low and high ABI, and male sex, were associated with disease progression. Unfortunately, the age of these studies limits our understanding of PAD progression in a contemporary setting with appropriate utilization of secondary prevention strategies, and both heterogeneity among the studies and lack of a unified grading scale of PAD progression further limit its utility.

Given this data gap in our understanding of PAD, the analysis from Rymer et al. is a welcome addition.⁷ Using data from the Examining Use of Ticagrelor in Peripheral Artery Disease (EUCLID) trial, the authors present one of the largest analyses of PAD disease progression using a real-world classification schema with robust follow up data. EUCLID was a large randomized placebo-controlled trial examining the effect of ticagrelor versus clopidogrel in patients with symptomatic PAD.⁸ At enrollment, PAD symptoms were assessed using the Rutherford classification with repeat assessments occurring throughout the trial. The Rutherford classification describes seven stages of PAD, including three categories of IC severity while also capturing ischemic rest pain and limb tissue loss.

In the present analysis, the authors examined changes in Rutherford classification for 12,759 trial participants over a one-year period. Patients with mild and moderate IC were combined into one category for analysis, and change was defined as progression of at least 1 Rutherford stage. Based on Rutherford stage improvement, stability, or progression, baseline characteristics were then correlated with the risk of future CV and limb events. The primary endpoint was major adverse cardiovascular events (MACE), a composite of CV death, myocardial infarction, or ischemic stroke. Secondary endpoints included all-cause mortality, major amputation, any amputation, and acute limb ischemia.

At one year, the majority of patients 8,132 (63.7%) had no change is symptoms, while 3,240 (25.4%) were classified as improved and 1,387 (10.9%) were classified as worsened. While 2,451 (19.2%) patients were initially asymptomatic, 26% of these subjects had clinically worsening PAD at follow up. In the 10.9% of patients with worsening Rutherford classification, 94% were either initially asymptomatic (46.5%) or presented with mild/moderate claudication (47.7%). Factors that were associated with disease progression were a decrease in ABI, age >65 years, diabetes, prior aspirin or clopidogrel use, and increasing number of diseased vascular beds.

There was a significant association between change in symptom classification and subsequent adverse outcomes, even after adjustment for confounders. Worsened symptom classification at 12 months was associated with a higher risk of MACE (adjusted hazard ratio [HR] 1.30; 95% confidence interval [CI] 1.05–1.62), all cause death (adjusted HR 1.29; 95% CI 1.03–1.62) and major amputation (adjusted HR 4.12; 95% CI 2.46–6.88). This serves as a reminder that atherosclerosis is a systemic disease, and disease activity in the limb should heighten concern for polyvascular events.

This unique work provides important insight into patients with Rutherford stage 0–3 disease. Only a small number of patients had CLI at baseline (Rutherford stage ≥4), which may limit generalizability, although it is difficult and perhaps unethical to document the natural history of CLI since these patients typically warrant revascularization. The combined classification of mild and moderate claudication (Rutherford 1 and 2) limits our understanding of the progression of these groups individually, though the clinical distinction between these groups is somewhat arbitrary, and the majority (70.9%) experienced no disease progression. Notably, only 2.5% of patients with more severe IC (Rutherford 3) had disease progression. It is unclear whether this truly reflects underlying disease pathobiology or simply a decrease in mobility to palliate IC, thus further highlighting the inherent limitations of the Rutherford classification and the heterogeneity of the PAD phenotype.

Though the majority of this cohort had IC, 2451 (19.2%) were initially asymptomatic. This group demonstrated an alarming disease progression rate of approximately 26% in one year. Although IC is commonly viewed as the typical symptom of early stage PAD, prior data show the majority of patients with an abnormal ABI have atypical limb symptoms or are asymptomatic.⁹ The Rutherford classification does not account for this heterogeneity, and patients and providers should not be reassured by the absence of IC. Future studies will hopefully collect additional symptom descriptions and functional measures, such as the 6-minute walk test, to better characterize individuals with “asymptomatic” PAD at risk of adverse outcomes.

Perhaps most alarming, worsening PAD symptoms was associated with the same magnitude risk of MACE (HR 1.30; 95% CI 1.05–1.62) as all-cause mortality (HR 1.29; 95% CI 1.03–1.62). Within EUCLID, the 1-year rate of MACE among subjects with worsened PAD was 8.4%, while the 1-year rate of MACE among all subjects with Rutherford data was 6.2%. By contrast, the 1-year rate of MACE among subjects with stable coronary artery disease (CAD) in the Reduction of Atherothrombosis for Continued Health (REACH) Registry was 4.5%.¹⁰ This reemphasizes the severe risk to both life and limb of PAD, and the vigorous attention and aggressive medical management often received by those with stable CAD should similarly be directed toward our PAD population.

There are several limitations to the current study worth noting. Although the Rutherford classification system is widely used, the cutoffs between some stages are arbitrary and highly subjective. It also does not account for subjects with atypical limb symptoms or other functional limitations from PAD. Residual confounding in this post-hoc analysis remains possible, particularly given the different sets of covariates chosen by backward selection. Similarly, although ticagrelor did not reduce the risk of MACE compared to clopidogrel in the overall EUCLID trial, this analysis does not adjust for study arm assignment.

Nonetheless, the authors should be applauded for this important addition to our understanding of PAD. Clinicians should view this as a call to arms to aggressively treat risk factors with the same fervor that we approach CAD, not only in those with progressive disease but also in the often-overlooked “asymptomatic” patient population. We hope that future studies will expand upon this analysis to include functional assessments of PAD as well as more nuanced phenotyping of limb symptoms as they relate to CV risk.