WIfI Classification Predicts Risk in Diabetic Limb Ischemia

In a large prospective subgroup analysis of the PRIME-WIFI registry, researchers found that the Wound, Ischemia, and foot Infection (WIfI) classification system retains significant prognostic value for diabetic patients with chronic limb-threatening ischemia (CLTI) who undergo endovascular revascularization. The study, published in Frontiers in Cardiovascular Medicine, evaluated 707 diabetic patients with 730 infrapopliteal limbs to determine how well this widely used clinical tool predicts major adverse events, amputation-free survival, mortality, and major amputation in this specific high-risk population. The findings suggest that while WIfI staging is effective for broad risk stratification, its ability to distinguish between intermediate risk levels is limited, particularly between stages 2 and 3.

Chronic limb-threatening ischemia represents the terminal stage of peripheral artery disease, characterized by chronic ischemic rest pain, non-healing ulcers, or gangrene resulting from severe arterial insufficiency. The prognosis at this stage is generally poor, with reported one-year major amputation and mortality rates each exceeding 20%. Diabetes mellitus is both the most prevalent and potent risk factor for CLTI, present in over 50% of cases. Diabetic patients face significantly higher risks of limb loss and adverse clinical outcomes compared to non-diabetic individuals, driven by distinct pathophysiologic features such as combined macrovascular and microvascular involvement, severe medial arterial calcification, and impaired immune response.

The WIfI classification system, introduced by the Society for Vascular Surgery in 2014, stratifies limb threat based on three domains: wound extent, severity of ischemia, and foot infection status. These domains are aggregated into four clinical stages, ranging from stage 1 (very low risk) to stage 4 (high risk) of major amputation within one year under medical therapy. Despite its widespread adoption and incorporation into international guidelines for diabetic foot ulcers and CLTI, questions remained regarding its performance specifically in diabetic patients after revascularization. Diabetes may introduce complexity to WIfI interpretation, as ischemia grading based on the ankle-brachial index (ABI) can be compromised by non-compressible, calcified vessels common in diabetic limbs.

The study was a sub-analysis of the PRIME-WIFI registry, a prospective, real-world, multicenter cohort derived from a Chinese peripheral arterial disease database. Between December 2020 and June 2023, consecutive patients were enrolled from 10 tertiary hospitals in mainland China. The current analysis included diabetic patients with CLTI (Rutherford category 4–6) who had angiographically confirmed infrapopliteal atherosclerotic disease and underwent endovascular revascularization. Patients were excluded if they were managed without endovascular therapy, underwent primary surgical revascularization, or lacked key baseline data. The mean age of the diabetic subgroup was 70.72 years, and 65.63% of patients were male.

At baseline, the limb characteristics indicated severe disease, with a predominance of Rutherford category 5. Limbs were distributed across all four WIfI stages, with a clear predominance of higher-risk stages 3 and 4. An ABI greater than 1.40 was recorded in 32 of the 730 limbs, reflecting the common issue of non-compressible vessels. Preprocedural runoff scores indicated severely compromised distal perfusion. Endovascular treatment strategies included plain old balloon angioplasty (POBA), drug-coated balloon angioplasty, stenting, or atherectomy, determined by treating operators based on clinical presentation and angiographic findings. The procedural success rate was 98.08%, with early complications within one week occurring in 3.84% of cases.

The primary endpoint was 12-month freedom from major adverse events (F-MAE), defined as a composite of all-cause mortality, major amputation, or clinically driven target-limb reintervention (CD-TLR). Secondary endpoints included amputation-free survival (AFS), freedom from major amputation, overall survival, and freedom from CD-TLR. The study also assessed clinical improvement and changes in the VascuQoL quality of life score at 12 months. Statistical analyses included Kaplan–Meier methods, multivariable Cox proportional hazards regression, and Harrell’s C-index to evaluate discriminatory ability. In the full PRIME-WIFI cohort, diabetes status showed a borderline interaction with WIfI stage for MAE, with a P for interaction of 0.083.

One-year freedom from MAE decreased across WIfI stages, with rates of 89.5% for stage 1, 77.4% for stage 2, 78.0% for stage 3, and 70.1% for stage 4. WIfI stage independently predicted MAE, with each one-stage increase corresponding to a higher hazard ratio of 1.31. Similarly, one-year amputation-free survival was 94.7% for stage 1, 83.3% for stage 2, 83.3% for stage 3, and 74.2% for stage 4. Higher WIfI stage was independently associated with worse AFS, with a hazard ratio of 1.53 per stage increase. Overall survival also declined with increasing stage, from 94.7% in stage 1 to 80.3% in stage 4, with WIfI stage independently associated with mortality (HR: 1.52).

Freedom from major amputation was 100.0% for stage 1, 95.3% for stage 2, 96.8% for stage 3, and 92.3% for stage 4. Higher WIfI stage was independently associated with major amputation, with a hazard ratio of 1.62 per stage increase. In contrast, freedom from clinically driven target-limb reintervention did not differ significantly across stages, with a log-rank P-value of 0.60. WIfI stage was not independently associated with CD-TLR, suggesting that factors beyond baseline limb threat, such as anatomical disease burden and vessel durability, influence the need for reintervention. Clinical improvement tended to decrease with increasing WIfI stage but did not reach statistical significance, with a P-value of 0.053.

Quality of life, assessed using the VascuQoL questionnaire, improved from baseline to 12 months across all WIfI stages. However, there was no significant difference in the magnitude of improvement among stages, as indicated by a time-by-WIfI interaction P-value of 0.247. Baseline VascuQoL scores differed significantly across stages, but the lack of differential improvement suggests that while higher-risk patients start with lower quality of life, the trajectory of recovery is similar across the spectrum. The Harrell’s C-index for WIfI stage across various endpoints ranged from 0.561 to 0.665, indicating modest discriminatory power, particularly for distinguishing between intermediate stages.

The study highlights that while WIfI staging provides a useful global risk-stratification tool in diabetic patients with infrapopliteal CLTI, its discriminatory granularity is reduced among intermediate-risk groups. The clearest separation in outcomes was observed between the extremes of the WIfI spectrum, with stage 1 showing the most favorable outcomes and stage 4 the poorest. Stages 2 and 3 showed substantial overlap in survival curves and limited pairwise differences. This pattern suggests that WIfI may be more effective for broad risk stratification than for providing granular discrimination across all four stages in this population.

Several limitations should be acknowledged. As a subgroup analysis derived from a prospective multicenter registry, the study may be subject to residual confounding and reduced statistical power. The 13.01% loss to follow-up may have introduced attrition bias. Additionally, glycemic control could not be adequately assessed because HbA1c was not systematically collected. Ischemia grading relied predominantly on ABI, which may underestimate ischemic severity in diabetic patients, although toe pressure and transcutaneous oxygen pressure were collected selectively. The exclusive inclusion of tertiary vascular centers in a single country may limit the generalizability of the findings to other healthcare settings.

These findings refine the overall PRIME-WIFI results by defining WIfI performance specifically in diabetic patients. They support the use of WIfI for broad risk stratification while highlighting the need for complementary assessment in diabetic CLTI. The study underscores that while baseline limb-threat severity is a significant predictor of outcomes, the relationship between this severity and recovery after revascularization is complicated by diabetes-related alterations in tissue perfusion and healing. WIfI should be interpreted as one component of longitudinal assessment in multidisciplinary limb-salvage pathways rather than a standalone framework for comprehensive post-revascularization risk assessment.

Subject of Research: Prognostic performance of the WIfI classification in diabetic patients with infrapopliteal CLTI after endovascular revascularization: a subgroup analysis of the PRIME-WIFI registry

Article Title: Prognostic performance of the WIfI classification in diabetic patients with infrapopliteal CLTI after endovascular revascularization: a subgroup analysis of the PRIME-WIFI registry

Article References: Zheng, H., Guo, J., Zhang, F., Tong, Z., Feng, Z., Wu, Z., Ye, M., He, C., Li, Q., Fang, X., & Guo, L. (2026). Prognostic performance of the WIfI classification in diabetic patients with infrapopliteal CLTI after endovascular revascularization: a subgroup analysis of the PRIME-WIFI registry. Frontiers in Cardiovascular Medicine, 13, Article 1984270. https://doi.org/10.3389/fcvm.2026.1984270

Image Credits: AI Generated

DOI: 10.3389/fcvm.2026.1984270

Keywords: Prognostic, performance, WIfI, classification, diabetic, patients, infrapopliteal, CLTI, endovascular, revascularization, subgroup, analysis