Peripheral artery disease (PAD) involves atherosclerosis of the lower extremity arteries and affects over 200 million people worldwide (1). Despite its significant association with limb loss and death, PAD remains undertreated (2). As a result, patients with PAD usually have worse long-term prognosis compared to patients with coronary artery disease (3). One reason for this is the lack of a validated prognostic biomarker for PAD (4). The ankle brachial index (ABI) is currently the only validated screening tool for PAD; however, it is limited by operator dependence, erroneous interpretation, and unreliability in patients with diabetes due to calcified vessels (5, 6). Furthermore, ABI's are rarely performed in the primary care setting due to generalists' lack of comfort with performing and interpretating this test (7). A recent survey found that 79% of primary care providers did not perform ABI's routinely in their clinical practice, citing time constraints, unavailability of skilled personnel, and complexity of result interpretation as major barriers (8). Most clinicians viewed alternative forms of diagnosis, such as a blood or urine test, as preferable to ABI's and would enhance diagnostic simplicity and efficiency (8). Therefore, the identification and validation of a biomarker for PAD may improve cardiovascular risk stratification through early initiation of aggressive medical management, close ambulatory follow up, and arterial interventions.

In addition, ABI's are poor predictors of PAD-related complications (9). Pasqualini et al. (2012) showed that both low and high ABI were associated with cardiovascular mortality (10). Furthermore, Hatmi et al. (2014) demonstrated that a low ABI only had a sensitivity of 64% for predicting adverse cardiovascular events (11). Identification of better prognostic markers for PAD-related complications would improve risk-stratification and allow clinicians to identify patients who require further diagnostic evaluation, close follow-up, and aggressive medical/surgical therapy. For example, the Cardiovascular Outcomes for People Using Anticoagulation Strategies (COMPASS) trial demonstrated a significant reduction in major adverse cardiovascular events for patients with atherosclerotic vascular disease taking low-dose rivaroxaban (2.5 mg oral twice daily) in addition in ASA (12). Therefore, identifying high-risk PAD patients using a novel prognostic biomarker can allow clinicians to better select patients for intensive medical therapy including COMPASS trial low-dose rivaroxaban.

Fatty acid binding protein 3 (FABP3) is an intracellular protein that is normally absent from plasma but becomes released into circulation and excreted into urine following skeletal muscle injury, which may occur due to ischemia from PAD (13). Previous groups have demonstrated that FABP3 is associated with mitochondrial dysfunction, metabolic syndrome, and cardiovascular disease (14, 15). We previously demonstrated that FABP3 is associated with both the presence of PAD and severity of disease (16, 17). Given that urine samples can be obtained non-invasively and require less health care resources/personnel to collect and test compared to blood specimens, we assessed the diagnostic and prognostic value of urinary FABP3 (uFABP3) in PAD.

Our single-center prospective case-control study demonstrates that uFABP3/uCr has diagnostic and prognostic value in PAD. We identified a cohort of 214 patients with and without PAD and demonstrated that uFABP3/uCr was elevated in the PAD group. Furthermore, higher levels of uFABP3/uCr predicted future PAD-related complications. Interestingly, most of the observed adverse events occurred in patients with moderate PAD based on ABI rather than patients with severe PAD, therefore our data suggests that ABI at baseline is not a good predictor of future adverse events, as corroborated by previous work by other groups (10, 11). This prompted us to risk stratified our cohort into low and high uFABP3/uCr based on ROC analysis. Our data demonstrated that patients with high uFABP3/uCr had a lower mean ABI and were more likely to be diagnosed with PAD. Finally, we showed that patients with high uFABP3/uCr were more likely to progress in their PAD disease state and require vascular intervention over a 2-year period. Overall, uFABP3/uCr had good discriminatory ability to predict our primary outcome of MALE, with an AUROC of 0.78. This is the first study validating a urinary biomarker for the prognosis of PAD in humans.

FABP3 belongs to a family of multi-gene fatty acid binding proteins (FABP's) and is primarily expressed in skeletal muscle and myocardial tissue (33). FABP3 is found predominantly intracellularly within the cytosol, where it mediates the uptake and transport of fatty acids (34). Previous studies have demonstrated the release of FABP3 from skeletal muscle in patients following exercise or muscle injury (13, 35, 36). In a similar context, patients with PAD experience muscle ischemia either transiently (claudication) or constantly (chronic limb threatening ischemia) (37, 38). This leads to FABP3 release into circulation and eventual excretion into the urine (39, 40). Others have demonstrated that FABP3 is associated with endothelial dysfunction, which may result from PAD (41). These are potential biochemical pathways by which FABP3 acts as a biomarker for PAD.

Previous studies have examined FABP's as biomarkers for other disease states. Ozawa et al. showed that FABP3 may be a potential mediator for diabetic nephropathy in murine models (42). Hayashida et al. demonstrated that serum and urine FABP3 may be early markers of myocardial injury in patients undergoing cardiac surgery (43). Others have investigated urinary FABP's as biomarkers for patients with acute kidney injury requiring renal replacement therapy (44). Given accumulating evidence demonstrating the value of FABP's in acting as biomarkers for life-threatening diseases, ongoing study in this area is warranted, particularly with a view to translate research findings into clinical practice to improve patient outcomes through early diagnosis and treatment. This is the first study to demonstrate the prognostic value of uFABP3/uCr in PAD patients.

There are several potential explanations for our findings. First, we demonstrated that PAD patients are more likely to have coronary artery disease (CAD) due to a greater burden of related cardiovascular risk factors including dyslipidemia and diabetes. This corroborates previous evidence demonstrating that individuals with PAD often have widespread arterial disease, with 61% of PAD patients being diagnosed with co-morbid CAD (45). As a result, they are at a 6-fold increased risk of cardiovascular death compared to patients without PAD (46). This highlights the importance of developing a widely applicable screening and prognostic tool for PAD. Second, we found that uFABP3/uCr levels were higher in PAD patients. This is likely because muscle ischemia secondary to PAD leads to release of uFABP3 into circulation, which then becomes excreted into urine. This provides a valuable biomarker that can be measured non-invasively in patients to screen for PAD. Third, patients with high uFABP3/uCr were more likely to develop MALE, require vascular intervention, and have worsening PAD over a 2-year period. There was also an association between every one unit increase in uFABP3/uCr and PAD-related complications. This suggests that uFABP3/uCr not only has diagnostic value, but also provides prognostic information regarding disease trajectory. Importantly, we demonstrated that high uFABP3/uCr was associated with worse PAD prognosis independent of CAD. Specifically, patients with active CAD were excluded from our analysis (acute coronary syndrome, congestive heart failure, uncontrolled arrhythmia, or elevated troponin within the past 12 months), which significantly reduced the risk of confounding. Even after adjusting for comorbidities including CAD and diabetes, there remained an independent association between uFABP3/uCr and MALE / worsening PAD status, which corroborates previous findings from our group (17). Therefore, this biomarker has the potential to risk stratify patients and guide healthcare providers with regards to aggressiveness of medical and surgical therapy, intensity of follow-up, and allocation of health care resources. Less than 5% of patients with PAD in our study were taking low-dose rivaroxaban, which has been demonstrated to provide significant cardiovascular benefit in patients with atherosclerotic vascular disease based on the COMPASS trial (12). Furthermore, only 62% and 87% of patients with PAD were taking ASA and statins respectively, despite these risk-reduction medications being Grade 1A recommendations for cardiovascular mortality and limb salvage benefit based on American Heart Association guidelines (47). Therefore, there are significant gaps in medical management of PAD, which may be addressed through a prognostic biomarker that can improve identification of high-risk patients that require more intensive therapy. Fourth, there was likely no difference in 2-year major amputation between groups due to the low incidence of this event (n = 3/214) as well as the fact that patients were followed closely in this prospective study and underwent early revascularization to prevent limb loss. Fifth, our subgroup analysis demonstrated a significant association between uFABP3/uCr and PAD severity based on ABI, which corroborates previous evidence from our group (17). Our analysis of outcomes based on PAD severity showed that patients with moderate PAD were more likely to develop adverse events compared to those with mild PAD or no PAD. Only a small number of patients with severe PAD (ABI < 0.5) were included in this study. This might be due to the fact that patients with acute on chronic limb ischemia who typically have an ABI < 0.5 were excluded from this study. These patients were excluded since they already possess some of the recorded events for the study, such as need for vascular intervention or limb amputation. Within this small subset of patients and relative to patients with moderate PAD, we noted that a small number of patients with severe PAD based on ABI developed MALE or worsening PAD. This information reflects the poor prognostic ability of baseline ABI in predicting future adverse events, as demonstrated by other groups (10, 11). However, larger clinical trials are needed to investigate the prognostic ability of ABI. Our study suggests that uFAB P3/uCr may be a better prognostic marker for adverse PAD-related events.

This study has several limitations. First, this was not a randomized study and there were differences in baseline characteristics between groups. However, our analysis adjusted for important baseline characteristics including age, sex, hypertension, dyslipidemia, diabetes, smoking, coronary artery disease, congestive heart failure, medications (statins, ACE-I/ARB, beta blocker, diuretic, ASA, antiplatelets other than ASA, and low-dose rivaroxaban), and serum creatinine. Second, 2-year outcomes are reported, and longer follow-up is needed to better understand the prognostic value of uFABP3/uCr given that PAD is generally a long-standing disease. Third, we excluded patients with eGFR <60, acute limb ischemia, or acute coronary syndrome, congestive heart failure, uncontrolled arrhythmia, or elevated troponin within the past 12 months. Therefore, the results may not apply to all patients with suspected PAD and further investigation of uFABP3 is needed to better understand how the biomarker can be effectively implemented as a screening and prognostic tool in clinical practice. Fourth, additional variables including body weight, lean body mass, percent body fat, lipid profiles, and echocardiogram data were not collected and may provide additional information regarding our patient population. The demographic and clinical variables reported in our study reflect clinically relevant indicators of the health of vascular surgery patients. Future study could collect more granular data to better understand the impact of baseline factors on PAD progression. Fifth, the major amputation rate was low in our study (n = 3/214) and likely contributed to statistically insignificant findings for this second ary outcome. We hypothesize that with longer follow-up and more events, our study would have detected a significant difference in major amputation rates based on uFAB P3/uCr levels.

Our study demonstrates three important findings. First, elevated uFABP3/uCr is associated with PAD. Second, patients with high uFABP3/uCr are more likely to develop MALE, require vascular intervention, and have worsening PAD even after adjusting for CAD history. Finally, uFABP3/uCr predicts MALE with an AUROC of 0.78. These results suggest that uFABP3/uCr has good diagnostic and prognostic value in PAD independent of CAD. The severity of PAD both at baseline and over the study period were captured in our study. Measurement of uFABP3/uCr can improve risk-stratification and identify patients at high risk for developing PAD-related complications who require additional diagnostic evaluation, close follow-up, and aggressive medical management, including prescription of low-dose rivaroxaban. Given that urine samples are easily collected and analyzed in the primary care setting, this biomarker has potential for widespread clinical implementation as a PAD screening and prognostic tool. Larger studies, particularly clinical trials, should be performed to confirm our findings.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

The studies involving human participants were reviewed and approved by Unity Health Toronto Research Ethics Board, University of Toronto. The patients/participants provided their written informed consent to participate in this study.

SJ, NJ, and MS: acquisition, analysis, and interpretation of data, revising the manuscript for important intellectual content, and approval of the final manuscript draft submitted for publication. BL and AZ: methodology, statistical analysis, data analysis and interpretation, writing—original draft, revising the manuscript for important intellectual content, and approval of the final manuscript draft submitted for publication. RA and MQ: study concept and design, methodology, data analysis and interpretation, writing—original draft, revising the manuscript for important intellectual content, and approval of the final manuscript draft submitted for publication. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.