Managing suspected myasthenia gravis and myositis induced by pembrolizumab in a Jehovah's Witness sarcoma patient.

INTRODUCTION: Immune checkpoint inhibitors (ICIs) can cause a spectrum of adverse events known as immune-related adverse events (irAEs) that resemble autoimmune responses. Immune-mediated myasthenia gravis (MG) is a rare and serious neurologic adverse event that has been associated with ICIs requiring prompt treatment. In the Jehovah's Witness population, typical management of these adverse events may not be options, and alternative treatment choices would be needed. CASE REPORT: 73-year-old Jehovah's Witness patient with high-grade undifferentiated pleiomorphic sarcoma who developed immune-mediated MG approximately 4 weeks after initiation of pembrolizumab. On the day of admission, the patient presented with a three-day history of worsening ptosis, right greater than left. He was later found to be seronegative for MG. MANAGEMENT AND OUTCOME: The patient required therapy with pyridostigmine, steroids, and agreed to plasma exchange (PLEX) prior to discharge. He achieved near resolution of his neurologic symptoms, and pembrolizumab was discontinued. He later underwent radical resection of the left thigh soft tissue sarcoma and superficial inguinal lymph node dissection. He is now on active surveillance. DISCUSSION: While neurologic adverse events typically present 6 weeks after initiation of ICIs, MG has been reported occurring as early as 4 weeks after initiation. This rare and serious adverse event requires prompt treatment, and clinicians need to be aware of the alternative treatment options in this unique patient population. Early conversations regarding blood products and factions must be had to develop a treatment plan in accordance with the patient's personal decisions.

Sensitivity and specificity of drug interaction databases to detect interactions with recently approved oral antineoplastics.

RATIONALE: Drug-drug interactions (DDIs) with oral antineoplastics (OAs) are of increasing concern given the rapid increase in OA approvals and use in cancer patients. A small pilot study of 20 DDIs with OAs showed significant variability in commonly used DDI screening databases in sensitivity of detecting potentially clinically relevant DDIs. This study builds upon that work by expanding the number of potential DDIs analyzed and including a specificity analysis. METHODS: Newly approved OAs from 2016 to May 2019 (n = 22) were included in this analysis. Prescribing information for each drug was reviewed. A list of explicit and theoretical drug interactions was created for each OA by the two investigators. A board-certified oncology pharmacist adjudicated all DDI pairs for potential clinical significance. In total, 229 DDI pairs were used to analyze sensitivity of 5 DDI databases (Lexicomp®, Micromedex®, Medscape, Eporactes®, & Drugs.com). Additionally, 64 "dummy" or false DDI pairs were created to analyze specificity. Sensitivity and specific were analyzed using Cochran's Qtest, while accuracy was analyzed using chi-square test. RESULTS: There was significant variability among the databases with regards to sensitivity (p < 0.0001), specificity (p < 0.0001), and accuracy (p < 0.0001). In terms of accuracy (max score = 400), Lexicomp®(355), Epocrates® (344), and Drugs.com (352) scored higher than MicroMedex® (270) and Medscape (280). CONCLUSIONS: Considerable variability exists among DDI screening databases with regards to OAs and potential drug interactions. Clinicians should be vigilant in both screening for DDIs with OAs and describing DDIs encountered in clinical practice.