Inhibition of the hedgehog pathway in patients with basal-cell nevus syndrome: final results from the multicentre, randomised, double-blind, placebo-controlled, phase 2 trial

Background Aberrant hedgehog signalling underlies the development of basal-cell carcinomas. We previously reported the interim analysis of a multicentre, randomised, double-blind, placebo-controlled, phase 2 trial in patients with the basal-cell nevus (Gorlin) syndrome indicating that the smoothened inhibitor vismodegib reduces basal-cell carcinoma tumour burden and prevents new basal-cell carcinoma growth in patients with basal-cell nevus syndrome. We report the final results of this 36 month trial. Methods In our multicentre, randomised, double-blind, placebo-controlled, phase 2 trial we enrolled patients aged 35-75 years with basal-cell nevus syndrome with at least ten surgically eligible basal-cell carcinomas at the Children's Hospital Oakland, Columbia University outpatient dermatology clinic (NY, USA) and a private practice outpatient dermatology office in Newport Beach (CA, USA). Patients were assigned to vismodegib or placebo (2: 1) according to a randomisation sequence generated by computer code. The primary endpoint of the trial of 41 patients was to compare the effect of oral vismodegib (150 mg/day) versus placebo on the incidence of new surgically eligible basal-cell carcinomas after 3 months of treatment. In the subsequent, open-label phase (n= 37) patients continued vismodegib at two sites for as long as month 36 (n= 25) and at the third site were monitored up to month 36 (n= 12). Additional endpoints for this phase were: whether continuous versus interrupted dosing differentially affected tumour burden; time to reach various levels of reduction in tumour burden; reduction in tumour size in patients who took less than 50% of the expected number of vismodegib tablets; reduction in the number of surgical excisions required per year before, during, and after treatment; and the effect of vismodegib on hedgehog target gene expression. We monitored patients at visits every 3 months for up to 36 months. The primary endpoint was analysed on a modified intention-to-treat basis. This trial is registered with ClinicalTrials.gov, number NCT00957229. Findings Between Sept 22, 2009, and Jan 24, 2011, 41 patients were monitored for a median of 36 months (IQR 36-36). Patients treated with vismodegib (n= 26) had a mean reduced rate of new surgically eligible basal-cell carcinomas compared with patients randomly assigned to placebo (n= 15; two [SD 0 . 12] new surgically eligible basal-cell carcinomas per patient per year vs 34 [1 . 32] new surgically eligible basal-cell carcinomas per patient per year, p< 0 . 0001). In the 11 patients initially assigned to placebo, mean cross over to vismodegib reduced the development of new surgically eligible basal-cell carcinomas compared with placebo (0 . 4 [SD 0 . 2] new surgically eligible basal-cell carcinomas per patient per year vs 30 . 0 [7 . 8] new surgically eligible basal-cell carcinomas per patient per year, p< 0 . 0001). Only three (17%) of 18 patients tolerated vismodegib continuously for the full 36 months. Fewer new surgically eligible basal-cell carcinomas developed in patients receiving vismodegib continuously than in those who interrupted dosing (mean 0 . 6 [0 . 72] new surgically eligible basal-cell carcinomas per patient per year vs 1 . 7 [1 . 8] new surgically eligible basal-cell carcinomas per patient per year, p< 0 . 0001). Treatment-related grade 3-4 adverse events included weight loss of 20% or more (n= 6) and muscle cramps (n= 2). Two patients died during the course of the trial, one each from laryngeal and metastatic prostate cancer, deemed probably unrelated to drug. Interpretation Vismodegib reduces basal-cell carcinoma tumour burden in patients with basal-cell nevus syndrome. Adverse events associated with vismodegib frequently led to interruption of treatment, which is followed by basal-cell carcinoma recurrence.