Gelfand, Joel M; Hoffstad, Ole; Margolis, David J, et al.
The Journal of Investigative Dermatology. Date of publication 2002 Dec 1;volume 119(6):1420-1425.
Surrogate markers are endpoints that occur early in the course of treatment and are intended to predict the true, meaningful clinical endpoint. Surrogate markers have been used to study treatments for a wide range of diseases in which the true outcome is delayed. The evaluation of therapies for venous leg ulcers is challenged by the prolonged observation period necessary to reach the endpoint of healing. We have performed a large cohort study to examine wound healing characteristics as candidate surrogate markers of venous leg ulcer healing using the Curative Health Services population. A total of 58,038 wounds met our definition of venous leg ulcer; however, 1550 wounds were excluded based on size, depth, site, and/or involvement of tendon or bone, leaving 56,488 wounds in 29,189 patients for analysis. The median wound size was 189 mm2, with a median wound duration of 3 mo. Using a large cohort of diverse venous leg ulcer patients, we demonstrate that after only 4 wk of treatment the wound parameters log healing rate, log wound area ratio, and percentage change in wound area can be valid surrogate markers of healing at 12 or 24 wk of care. Based on the area under the receiver operator characteristic curve log rate, log area ratio, and percentage change in area can discriminate which patients will heal at 12 or 24 wk of care (receiver operator characteristic 0.72-0.80). These surrogates were further validated by demonstrating that established risk factors for not healing such as wound size and wound duration are also important risk factors for not achieving the surrogate endpoint. These surrogate markers for venous leg ulcer healing may allow for early clinical trials to be more efficient, and can allow clinicians to identify patients unlikely to heal early in the course of treatment in order to expedite referral to specialty centers or for the selection of stepped treatment algorithms.
