Double-blind, placebo-controlled study of HGF gene therapy in diabetic neuropathy

Abstract

Objective: To evaluate the safety and efficacy of a plasmid (VM202) containing two human hepatocyte growth factor isoforms given by intramuscular injections in patients with painful diabetic neuropathy.

Methods: In a double-blind, placebo-controlled study, patients were randomized to receive injections of 8 or 16 mg VM202 per leg or placebo. Divided doses were administered on Day 0 and Day 14. The prospective primary outcome was change in the mean pain score measured by a 7 day pain diary. Secondary outcomes included a responder analysis, quality of life and pain measures, and intraepidermal nerve fiber density.

Results: There were no significant adverse events attributable to VM202. Eighty-four patients completed the study. Patients receiving 8 mg VM202 per leg improved the most in all efficacy measures including a significant (P = 0.03) reduction at 3 months in the mean pain score and continued but not statistically significant reductions in pain at 6 and 9 months. Of these patients, 48.4% experienced a ≥50% reduction in pain compared to 17.6% of placebo patients. There were also significant improvements in the brief pain inventory for patients with diabetic peripheral neuropathy and the questionnaire portion of the Michigan Neuropathy Screening Instrument. Patients not on pregabalin or gabapentin had the largest reductions in pain.

Interpretation: VM202 was safe, well tolerated and effective indicating the feasibility of a nonviral gene therapy approach to painful diabetic neuropathy. Two days of treatment were sufficient to provide symptomatic relief with improvement in quality of life for 3 months. VM202 may be particularly beneficial for patients not taking gabapentin or pregabalin.

Figures

Figure 1

Analysis populations in the study. Intent to treat (ITT) population: All subjects enrolled in the study for whom there were data (n = 103). One subject randomized to the high dose (HD) group withdrew consent prior to any treatment. Safety population: All subjects enrolled in the study who received any study injections. The safety population is the same as the ITT population (n = 103). Efficacy population: All subjects in ITT/safety population receiving all study injections, except for those subjects excluded by protocol in a blinded review after a database lock (n = 84). Exclusion criteria were as follows: does not meet the eligibility criteria or does not have 6 month data. Use of protocol prohibited COX1/COX2 inhibitors. Use of opioid medications. Two patients in the HD group (one for shoulder pain, one for chronic cough), and one patient in the LD group (hernia repair) were excluded for this reason.

Figure 2

Mean pain scores on the daily pain diary expressed as change from baseline. (A) Scores for all patients in the efficacy population. The means, standard deviations, medians, minimum scores, and maximum scores are given in Table1. The low-dose (LD) group differed significantly from the P group at 3 months (*P = 0.04) by ANOVA with Dunnett's post hoc test. (B) Mean pain scores for patients not taking gabapentin or pregabalin. The LD group differed significantly from the P group both at 3 months (*P = 0.007) and at 6 months (**P = 0.005).

Figure 3

(A) BPI-DPN pain interference scores expressed as change from baseline. The means, standard deviations, medians, minimum scores, and maximum scores are given in Table2. The LD group differed significantly from the P group both at 3 months (*P = 0.046) and at 6 month (*P = 0.046) by ANOVA with Dunnett's post hoc test. (B) Patient's global impression in change (PGIC) expressed as the % of patients who were much or very much improved. The LD group differed significantly from the P group (*P = 0.008) by the Generalized Estimating Equation for repeated measurements. BPI-DPN, brief pain inventory for patients with diabetic peripheral neuropathy; LD, low dose.

Figure 4

Monofilament test administered as part of the MNSI expressed as the percent of patients improved or normal. Scores are given for (A) right leg, (B) left leg, and (C) both legs. A significant number of subjects in the LD group at 9 months was experienced improvement in sensory threshold of their left leg as compared with the P group (*P = 0.05 by Fischer exact test). MNSI, Michigan Neuropathy Screening Instrument; LD, low dose.