Topical Treatment of Hypertensive Leg Ulcers With Platelet-Derived Growth Factor-BB: A Randomized Controlled Trial

Objective To determine the healing effect of topical becaplermin gel vs hydrogel dressing on hypertensive leg ulcers.

Design Randomized, double-blind, parallel-assignment, controlled study.

Setting Ambulatory or hospitalized patients from 17 dermatology departments.

Patients Among 64 consecutive randomized patients with 1 or more hypertensive leg ulcers who fulfilled all inclusion criteria, 59 received the allocated intervention, and findings were analyzed.

Intervention Becaplermin gel (human recombinant platelet-derived growth factor-BB, 0.1%, in hydrogel) or hydrogel dressing was applied, both in doses of 1 cm/cm², once daily for 8 weeks. Follow-up continued for 4 weeks beyond the final gel application.

Main Outcome Measures The primary end point was complete wound closure rate after 8 weeks of treatment. Secondary end points were percentages of patients with complete wound closure at week 12; changed ulcer area after treatment vs baseline; and changed ulcer-related pain and health-related quality of life during the study.

Results Complete wound closure rates were comparable after 8 weeks for becaplermin (5 of 28 patients) and hydrogel (3 of 31 patients) (8 percentage-point difference; 95% confidence interval, –10% to 26%). No statistically significant differences were observed between the 2 groups for percentages of complete closure at week 12, changed ulcer area at week 8, or changed ulcer-related pain and quality of life during the study (P   >  .05 for all comparisons).

Conclusions Topical becaplermin gel is not superior to hydrogel dressing for hypertensive leg ulcer wound closure. Surgical management by grafting remains the most promising treatment strategy but requires further evaluation.

Trial Registration clinicaltrials.gov Identifier: NCT00970697

Hypertensive leg ulcers (HLUs), first described in the 1940s, were renamed Martorell HLUs or necrotic angiodermatitis by American and European dermatologists.^1,2 Hypertensive leg ulcers are extremely painful, superficial, rapidly spreading, necrotic wounds on the dorsolateral part of the leg and have red purpuric margins.^3-7 Several case series have confirmed the highly characteristic and almost invariable clinical presentation. Skin biopsy specimens of the wound border extending from healthy surrounding skin to the area of necrosis show subcutaneous arteriosclerosis and should be taken to distinguish the diagnosis of HLU from atypical differential diagnoses.^6-11 Patients with long-standing hypertension and/or diabetes (usually patients older than 60 years) without peripheral arterial occlusive disease develop HLUs. The pathophysiologic characteristics of HLUs include dermal and subdermal vessel arteriosclerosis, inappropriate local vasoconstriction, but no significant involvement of the large deeper vessels.^3,5,12 They may represent up to 15% of all treated skin ulcers in specialized departments.³Medical management of HLU is currently symptomatic: controlling hypertension and diabetes, wound debridement, and application of the usual dressings.¹³ Surgical management with grafts may achieve rapid pain relief and healing,^{2,3,8-11,14,15} possibly by releasing growth factors into the wound.¹⁶ However, because only case series have been published, the evidence level of any HLU treatment is limited.

Becaplermin is a recombinant human platelet-derived growth factor (rhPDGF-BB), which plays a pivotal role in wound healing and is the only growth factor now approved by American and European authorities to heal chronic wounds such as diabetic ulcers while allowing for adequate peripheral circulation.¹⁷ The present study was designed to determine the effect of topical becaplermin gel on HLU healing.

This multicenter, randomized controlled trial was conducted in 17 French centers from March 2004 to June 2009 and performed in accordance with the Declaration of Helsinki. Informed written consent was obtained from each patient before study inclusion. The ethics committee of the Piti é-Salp êtri ère Hospital, Assistance Publique –H ôpitaux de Paris (AP-HP), and regulatory authorities approved the study protocol (No. 3403), registered in clinicaltrials.gov (NCT00970697).

We enrolled consecutive ambulatory or hospitalized patients with 1 or more HLUs. Eligibility criteria were age 18 years or older, ability to give informed consent and follow the treatment procedure, and presence of 1 or more HLUs of 1 to 30 cm² in total area. Inclusion criteria were presence of an HLU defined by clinical criteria alone (typical location, superficial or deep skin necrosis surrounded by a characteristic purpuric margin, peripheral progression, and intractable pain) in (1) a patient who met the World Health Organization criteria¹⁸ for hypertension, treated or not; and/or (2) a patient with diabetes treated with an oral agent, insulin, or diet without clinical signs of severe chronic venous insufficiency (eg, skin hyperpigmentation, lipodermatosclerosis); and (3) a patient without significant peripheral arterial occlusive disease, defined as the presence of peripheral pulses or an ankle brachial index of 0.8 or higher.^4-7

Exclusion criteria were concomitant cutaneous vasculitis (eg, palpable or necrotic purpura); ongoing systemic disease known to be associated with pyoderma gangrenosum or necrotizing vasculitis (such as rheumatoid arthritis), other autoimmune diseases, or cryoglobulinemia; known allergy to hydrogel or becaplermin gel; prior or evolving cancer; evolving systemic disease (eg, cardiac or renal failure, hepatic insufficiency, thrombotic disease, connective tissue disorder); serum creatininemia (creatinine level, >2.26 mg/dL uncontrolled diabetes (fasting glucose level, >250 mg/dL under treatment); bone or joint exposure in the wound; and concomitant treatment with corticosteroids, immunosuppressive or cytotoxic drugs, and/or iloprost during the 3 months preceding the study. (To convert creatinine to micromoles per liter, multiply by 88.4; to convert glucose to millimoles per liter, multiply by 0.0555.)

Eligible participants were randomly assigned by facsimile though a central automated system designed by the Clinical Research Regional Department (AP-HP) to receive either topical becaplermin gel (rhPDGF-BB, 0.01%) (Regranex gel; Ethicon Division of Johnson & Johnson Wound Management, Issy-les-Moulineaux, France) or hydrogel (Duoderm Hydrogel; Convatec, Rueil-Malmaison, France). A computer engineer not responsible for data acquisition prepared the assignments. All participants and investigators were blinded to assigned treatment. Duoderm Hydrogel is a commercially available hydrogel dressing, formulated as a sodium carboxymethylcellulose aqueous-based gel, similar in composition to the becaplermin excipient and identical in color, consistency, and odor to becaplermin gel.

Masked 15-g tubes of either becaplermin or hydrogel, identical in color, shape and size, were provided in blister packs by the AP-HP central pharmacy the week after the baseline visit. During that consultation, a medical history was obtained and a physical examination performed. When several HLUs were present, the most recent was considered the reference ulcer.

For both groups, treatment began 1 week after randomization, during the second visit (week 0), and all patients received the same daily local care: wound irrigation with normal saline, application of a continuous thin layer of gel on the wound, covering with a moist saline gauze and a bandage. Each reference wound surface was estimated on treatment day 1 by measuring wound length and width to determine the appropriate becaplermin gel or hydrogel volume to apply that was maintained throughout the study. This dose is similar to that approved by the US Food and Drug Administration and European authorities for becaplermin gel use on diabetic ulcers: a single 15-g tube is enough to treat a 5-cm² wound for 6 weeks (1 cm of gel/cm²/d). The study treatment was first applied at the study center by a nurse not involved in the trial and completely blinded to patient assignment. The patient or his/her caregiver was instructed on proper wound care, gel application, and wound dressing, which was continued until complete healing or for a maximum of 8 weeks, shorter than that recommended for diabetic ulcers but comparable to the time required for complete HLU healing reported in retrospective studies using cutaneous autografts.^2,3,14

When the study treatment was stopped at week 8, each investigator selected the replacement local wound care, depending on the wound stage¹³ and, if the ulcer was not healed, decided if a pinch graft was appropriate. All patients were observed through week 12.

The primary outcome was the complete wound closure rate at treatment week 8. The secondary outcomes were (1) percentages of patients with complete wound closure at week 12; (2) changed ulcer area after treatment week 8 vs baseline; (3) changed ulcer-related pain assessed with a pain visual analog scale (VAS; range, 0-100 mm); (4) health-related quality of life under treatment, evaluated with the Skindex scale¹⁹; and (5) adverse events. To calculate ulcer area, we used simple wound measurements previously described as being highly correlated with planimetric wound measurements.²⁰ An investigator blinded to patient assignments and not applying the gel collected outcome measures at baseline and at every subsequent study visit (weeks 0, 2, 4, 8, and 12).

We used a sequential method with a restricted procedure²¹ to allow early termination for a large size effect with a maximum of 30 patients per group; the main criterion was analyzed after every 4 consecutive patients to ensure that type I and II errors would be maintained at their desired values, despite multiple analyses during the trial (PEST software, version 3; University of Reading, Reading, England). The statistician gave no information to the clinicians about observed between-group differences. For secondary criteria, we used t tests for Gaussian variable comparisons and Mann-Whitney tests for non-Gaussian quantitative variable comparisons. Time-dependent pain or quality-of-life changes were compared by analysis of variance for repeated measures. The χ² or Fisher exact probability tests were used for qualitative parameters.

Analyses were based on intention to treat. For missing data, the last observation was carried forward. All tests for secondary criteria were 2 sided, α-risk set at 5%, and computed with SAS version 9.2 (SAS Institute Inc, Cary, NC).

Among the 64 consecutive patients enrolled, 59 received their allocated interventions, and findings were analyzed (Figure). At baseline, the treatment and control groups were comparable (Table): 95% (56 of 59) and 40% (23 of 59) of study patients, respectively, had hypertension and/or diabetes. Six patients, 2 assigned to becaplermin gel and 4 controls, had a history of cancer (1 breast, 1 cervix, 2 prostate, and 2 bladder), which was in complete remission at inclusion. Among the 59 patients without palpable peripheral pulses, 4 had an ankle brachial index higher than 1.3 (range, 1.38-1.73) related to age and/or diabetes medial calcinosis.

Complete wound closure rates for becaplermin and hydrogel, respectively, were 18% (5 of 28 patients) and 10% (3 of 31 patients), respectively, at week 8 (an 8 percentage-point difference; 95% confidence interval [CI], −10.3 to 26.0) and 36% (10 of 28 patients) and 26% (8 of 31 patients) at week 12 (10 percentage-point difference; 95% CI, –13.6 to 33.4).

During the study, no significant becaplermin effect on health-related quality of life (P  =  .98) or pain (P  =  .71) was observed by analysis of variance for repeated measures. No cancer relapse was observed.

Topical becaplermin, compared with hydrogel dressing, did not improve the complete wound closure rate (primary outcome measure) after treatment week 8 and had no significant effect on quality of life, pain, or median wound area. A limitation of the study is the failure of the inclusion criteria to include a biopsy specimen with supportive histologic analysis. We included only typical HLUs, excluding patients with cutaneous vasculitis or the usual systemic diseases associated with necrotizing vasculitis or pyoderma gangrenosum. None of the included lesions evolved during follow-up into a vasculitis or a pyoderma gangrenosum, making a misclassification bias improbable.

Complete HLU healing is more accurate than wound-area reduction because HLU-associated pain is known to be disproportionate to the lesion size,^6,8,11 as supported by our findings. Indeed, even though the median wound area tended to be lower for the becaplermin group, no significant effect of becaplermin was observed on the health-related quality of life during the study, as evaluated with the Skindex Scale score, compared with the hydrogel dressing group. At the same time, mean VAS scores tended to decrease in both groups during the study, probably owing to the medical management of pain and local wound care by atraumatic gels beginning at baseline.

The severity of our inclusion criteria could account for the homogeneity of our enrolled patient characteristics and their small number. However, they are representative of the HLU population described by Graves et al⁶ and in the largest published case series, including in terms of healing time.^2,4,5,11

The becaplermin excipient has hydrogel properties and performs at least comparably to good wound practice alone to treat diabetic ulcers.²² Therefore, using a hydrogel dressing as the control was logical, even though a between-group difference may be more difficult to demonstrate.

Becaplermin biological activity includes chemotactic recruitment and proliferation of cells involved in cutaneous wound repair. Except for becaplermin treatment of diabetic ulcers, randomized controlled trial results using topical growth factors for chronic wounds have been disappointing,¹⁷ suggesting that growth factors might be rapidly degraded by wound proteases or that wound cells have decreased proliferative ability. To date, no medical management of HLU has emerged,²³ and surgical management by grafting, to achieve complete healing, remains the most promising treatment strategy but should be evaluated in a randomized controlled trial.

Correspondence: Patricia Senet, MD, Service de Dermatologie, H ôpital Tenon, 4 rue de la Chine, 75970 Paris CEDEX 20, France (patricia.senet@tnn.aphp.fr).

Accepted for Publication: January 31, 2011.

Published Online: April 11, 2011. doi:10.1001/archdermatol.2011.84

Author Contributions: Drs Senet, Vicaut, and Chosidow had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Senet, Vicaut, Debure, Lok, and Chosidow. Acquisition of data: Senet, Vicaut, Beneton, Debure, Lok, and Chosidow. Analysis and interpretation of data: Senet, Vicaut, Debure, and Chosidow. Drafting of the manuscript: Senet, Vicaut, and Chosidow. Critical revision of the manuscript for important intellectual content: Senet, Vicaut, Beneton, Debure, Lok, and Chosidow. Statistical analysis: Vicaut, and Chosidow. Obtained funding: Senet, Vicaut, and Chosidow. Administrative, technical, and material support: Senet, Vicaut, Beneton, Debure, Lok, and Chosidow. Study supervision: Senet, Vicaut, and Chosidow.

Funding/Support: This study was supported in part by AP-HP, the French Society of Dermatology, and the AFSSAPS (Agence Fran çaise de S écurit é Sanitaire des Produits de Sant é). Johnson & Johnson France provided the becaplermin gel (Regranex Gel).

Role of the Sponsors: The sponsors had no role in the design and conduct of the study; in the collection, analysis, and interpretation of data; or in the preparation, review, or approval of the manuscript.

Additional Contributions: We are indebted to the investigators, members of the French Angio-Dermatology Group of the French Society of Dermatology (Nathalie Baudot, MD, Edouard Begon, MD, Christophe Bedanne, MD, PhD, Jean-Marie M. Bonnet Blanc, MD, PhD, Guillaume Chaby, MD, Jacqueline Chevrant Breton, MD, PhD, Olivier Dereure, MD, PhD, Vincent Descamps, MD, PhD, S égol ène Fays-Bouchon, MD, Juliette Fontaine, MD, Philippe Humbert, MD, PhD, Isabelle Kupfer, MD, PhD, Laurence Lecleach, MD, Philippe Modiano, MD, PhD, Saskia Oro, MD, Barbara Roth, MD, Fran çois Truchetet, MD); and to AP-HP personnel for their inestimable technical support (Fabien Maugard and Saliha Djane, MD, from the Direction de la Recherche Clinique; Sophie Gerard, Mounir Aout, and Veronique Jouis from the Unit é de Recherche Clinique; Brigitte Lehman from the Agence G én érale des Equipements et Produits de Sant é).