Randomized trial of prophylactic minocycline for erlotinib-associated skin rash in non-small cell lung cancer (PEARL trial)

Abstract

Background: Acneiform rash as an adverse event often affects the treatment by EGFR-TKIs. Since minocycline has been suggested to reduce the rash, we assessed the efficacy and safety of prophylactic administration of minocycline simultaneously during erlotinib treatment. Method(s): Patients of ECOG performance status 0-2 with advanced NSCLC, who had not been treated with EGFR-TKIs and would receive erlotinib treatment were randomized 1:1 into either group A, with minocycline or group B, without minocycline. The patients assigned to group A were started on minocycline 100mg/day orally for 8 weeks with erlotinib. Primary end point was the frequency of grade >2 rash acneiform by independent assessment in first 8 weeks. We expected the prophylactic minocycline decreased the incidence of grade >2 skin rash from 50% to 30%. The planned sample size was 280 patients with a = 0.025 (one-sided) andb = 0.10. Result(s): Patients accrual was started in March 2015 and ended in June 2018 because of slow accrual. Ninety-four patients were finally enrolled and 93 were full-analysis set. The median age of the patients was 71 years old (range 45 to 89),58 patients were female. EGFR mutation status positive/negative/unknown=78/13/2 patients. The frequency of grade >2 rash acneiformby independent assessment was 33.3% [95%C.I. 20.0-49.0%] in group A vs. 44.2% [95%C.I. 29.1-60.1%] in group B (p = 0.296). The frequency by physicians'evaluation was 31.3% [95%C.I.18.7-48.8%] and 45.5% [95%C.I. 30.4-61.2%] (p = 0.161). However, the frequency of grade > 2 rash acneiform on day 15 by physicians' evaluation was significantly decreased (4.4% [95%C.I. 5.3-14.8%] in group A vs. 25.0% 95%C.I.13.2-40.3%] in group B (p = 0.005)). As for toxicity, the incidence of any grade skin-related toxicity as pruritus (39.6% vs. 63.6%) and pain of skin (14.6% vs. 25.0%) was less common in group A. Hence, anorexia (41.7% vs 20.5%), nausea (25.0% vs 4.5%), and dysgeusia (22.9% vs 15.9%) occurred more frequently. Conclusion(s): Although the frequency of acneiform rash tended to decrease, prophylactic administration of minocycline is not recommended because of increasing gastrointestinal toxicity.