Preventive effect of goshajinkigan on paclitaxel-induced peripheral neuropathy: an open-label, randomized phase II study (OLCSG2101).

Abstract

Background: Chemotherapy-induced peripheral neuropathy (CIPN) is a dose-limiting toxicity of paclitaxel (PTX)-containing regimens, and no preventive strategy has been established. We evaluated goshajinkigan (GJG), a traditional Japanese herbal medicine (Kampo), for preventing CIPN during PTX-containing chemotherapy.

Materials and methods: In this open-label, randomized phase II trial across 7 institutions in Japan, patients with malignant tumors scheduled for ≥4 courses of carboplatin plus PTX (≥150 mg/m²) were randomized 1:1 to GJG or control, with minimization by cancer type and age. The primary endpoint was grade ≥2 CIPN incidence within 4 courses. The protocol-prespecified primary analysis used an unstratified 2-sided log-rank test (α = 0.10); a supportive stratified analysis accounted for the minimization factors.

Results: Sixty-five patients were randomized (GJG, 32; control, 33); 52 (80.0%) had lung cancer. Grade ≥2 CIPN occurred in 28.1% (9/32) of the GJG arm vs 51.5% (17/33) of the control arm; the protocol-prespecified unstratified log-rank P value was .013, with a corresponding HR of 0.34 (90% CI, 0.16-0.72). In the stratified analysis accounting for cancer type and age, the estimate remained in favor of GJG but was attenuated and less precise (HR, 0.48; 90% CI, 0.22-1.02; stratified log-rank P = .101). Grade 3-5 adverse-event incidence was similar between arms.

Conclusion: GJG was associated with fewer grade ≥2 CIPN events in this predominantly lung-cancer population. Given the open-label design and uncertainty in the stratified analysis, these findings represent a phase II preventive signal warranting confirmation in a double-blind, placebo-controlled phase III trial.

Clinical trial registration number: Japan Registry of Clinical Trials (jRCTs061210047).