Baseline factors affecting closure of venous leg ulcers

Objective The objective of this study was to characterize factors associated with closure of venous leg ulcers (VLUs) in a pooled analysis of subjects from three randomized clinical trials. Methods Closure of VLUs after treatment with HP802-247, an allogeneic living cell therapy consisting of growth-arrested human keratinocytes and fibroblasts, vs standard therapy with compression bandaging was evaluated in three phase 3 clinical trials of similar design. Two trials enrolled subjects with VLUs ranging from 2 cm² to 12 cm² in area with 12-week treatment periods; the third trial enrolled subjects with VLUs between >12 cm² and ≤36 cm² with a 16-week treatment period. The first trial went to completion but failed to demonstrate a benefit to therapy with HP802-247 compared with placebo, and because of this, the remaining trials were terminated before completion. On the basis of no differences in outcomes between groups, subjects from both HP802-247 and control groups were pooled across all three studies. Cox proportional hazards regression analysis was employed to evaluate factors associated with VLU closure. Results This analysis included data from 716 subjects with VLU. Factors evaluated for association with healing included age, gender, race, diabetes, glycated hemoglobin level, body mass index, treatment (HP802-247 vs compression alone), and ulcer characteristics including location and area and duration at baseline. In an initial model including all of these putative factors, the following were significant at the P <.10 level: diagnosis of diabetes mellitus, gender, wound location (ankle or leg), baseline wound area, and wound duration at baseline. In a final model including only these factors, all but diabetes mellitus were significant at the P <.05 level. Effect sizes were as follows (hazard ratio [95% confidence interval]): female gender (1.384 [1.134-1.690]), wound location on the leg (1.490 [1.187-1.871]), smaller wound area at baseline (0.907 [0.887-0.927]), and shorter wound duration at baseline (0.971 [0.955-0.987]). Conclusions Factors associated with VLU lesions including location, area, and duration were important predictors of healing. Women were more likely than men to achieve wound closure. Factors including body mass index, the presence of diabetes mellitus, and higher concentrations of glycated hemoglobin were not significant independent predictors of wound closure in this analysis.