The incidence of occult endocrinopathies in patients with cancer undergoing immunotherapy in community practice.

Abstract

e14571 Background: Immune checkpoint blockade (IO) can induce inflammation of the pituitary, thyroid or adrenal glands. This usually results in non-specific symptoms such as headache, low-energy, nausea and vomiting, which can be difficult to differentiate from symptoms associated with cancer and therapy-related symptoms. Therefore, the exact incidence of endocrinopathies is exceedingly difficult to estimate in community practices. Also, the variable methods of assessment, diagnosis, and monitoring used in different clinical trials make it challenging to precisely measure the incidence of endocrinopathies. Methods: This is a single-center retrospective chart review of patients diagnosed with cancer receiving immunotherapy for cancer treatment who had routine hormone levels checked during their treatment. Data collected includes tumor types and the types of IO agents used. Laboratory data collected included thyroid-stimulating hormone (TSH), testosterone level, follicular stimulating hormone (FSH), luteinizing hormone (LH), cortisol levels. Results: In total, 75 patients were included in the study. The primary indication for IO was lung cancer in 43 patients (57%), genitourinary tumors (12%), melanoma (12%) and head & neck cancers (5.3%). Single-agent nivolumab (39 patients) was the most common IO agent used followed by single-agent pembrolizumab (22 patients), ipilimumab (11 patients), atezolizumab (3 patients), avelumab (1 patient) (There was one patient who got nivolumab initially and then pembrolizumab). Nine patients were treated with ipilimumab/nivolumab combination. The mean number of cycles received was 9.1. The total number of patients who developed at least one abnormal hormone level was 57(76%), with 33 out of 74 (45%) patients had at least one abnormal TSH, 29 out of 44 (66%) patients had at least one abnormal testosterone level, 10 out of 49 (20.4%) patients had at least one abnormal FSH and/or LH level, 36 out of 52 (69%) patients had at least one abnormal cortisol level. The mean number of days from starting IO to develop the first abnormal laboratory result was 106 days. Conclusions: The incidence of endocrinopathy was significantly high in patients receiving IO in this study, which is higher than what is reported in previous clinical trials. This could be due to frequent testing in asymptomatic patients and strict laboratory cut-off values which is not always clinically meaningful. This finding may highlight the importance of routine monitoring of the endocrine function during IO treatment. Routine measurement of hormone levels can detect asymptomatic endocrinopathy which may warrant further work-up and treatment. These findings should be validated in a larger prospective study.