Objective: To examine patient and wound variables presumed to influence healing outcomes in the context of therapeutic trials for chronic venous leg ulcers.

Methods: This double-blind, vehicle-controlled study was conducted with randomized assignment to one of four cell therapy dose groups (n = 46, 43, 44, 45) or vehicle control (n = 50). A 2-week run-in period was used to exclude rapid healers and those with infection or uncontrolled edema. This was a multicenter (ambulatory, private, hospital-based and university-based practices, and wound care centers in North America) study. Adults ≥ 18 years old with chronic venous insufficiency associated with an uninfected venous leg ulcer (2-12 cm(2) area, 6-104 weeks' duration) were included in the study. Excluded were pregnant or lactating women, wounds with exposed muscle, tendon or bone, patients unable to tolerate compression bandages, or patients who had exclusionary medical conditions or exposure to certain products. Exclusion during run-in included patients with infection, uncontrolled severe edema or with healing rates ≥ 0.349 cm/2 wk. Screen fail rate was 37% (134/362), and the withdrawal rate was~10% (23 of 228). Growth-arrested neonatal dermal fibroblasts and keratinocytes were delivered via pump spray in a fibrin sealant-based matrix, plus a foam dressing and four-layer compression bandaging. Treatment continued for 12 weeks or until healed, whichever occurred first. Patient demographic and wound-related variables were evaluated for influence on complete wound healing in all patients, as well as the subsets of treated and control patients.

Results: Wound duration (P = .004) and the presence of specific quantities of certain bacterial species (P < .001) affected healing in the vehicle group, while healing in the cell-treated groups was influenced by wound duration (P = .012), wound area (P = .026), wound location (P = .011), and specific quantities of certain bacterial species (P = .002). Age, sex, race, diabetes, HbA1C, peripheral neuropathy, and serum prealbumin did not significantly affect healing. Body mass index was positively associated with healing in cell-treated patients.

Conclusions: Wound duration is a quantifiable surrogate for one or more undefined variables that can have a profound negative effect on venous leg ulcer healing. Although cell therapy overcame barriers to healing, the only specific barrier identified was the presence of certain bacterial species. Interventional trials of potentially effective new therapies can be most informative when patients with suspected barriers to healing are included. The specific measurement of candidate barriers such as microbial pathogens, wound inflammatory state, and fibroblast function should be considered in future randomized trials to improve our understanding of the basis for chronicity.