Varied phenotypes and management of immune checkpoint inhibitor-associated neuropathies

Abstract

Objective: To describe the spectrum, clinical course, and management of neuropathies associated with immune checkpoint inhibitors (ICIs).

Methods: Patients with ICI-related neuropathy (irNeuropathy) were identified and their clinical characteristics compared to neuropathy attributed to cytotoxic agents.

Results: We identified 19 patients with irNeuropathies. ICIs included anti-programmed death-1 (PD1), 9; anti-cytotoxic T-lymphocyte-associated antigen-4 (CTLA4), 2; and combination of anti-CTLA4 and anti-PD1, 8. Median number of ICI doses prior to neuropathy onset was 4. Rate of neuropathies following ICI therapy was 0.7%. Underlying malignancies included melanoma (n = 15), lung adenocarcinoma (n = 3), and cholangiocarcinoma (n = 1). Neuropathy phenotypes were cranial neuropathies with or without meningitis (n = 7), nonlength-dependent polyradiculoneuropathies with and without cranial nerve involvement (n = 6), small-fiber/autonomic neuropathy (n = 2), ANCA-associated mononeuritis multiplex (n = 1), sensory neuronopathy (n = 1), length-dependent sensorimotor axonal polyneuropathy (n = 1), and neuralgic amyotrophy (n = 1). Immune-related adverse events involving other organ systems were common (58%). Corticosteroid use for management of neuropathy was associated with improvement in median modified Rankin Scale score (1 vs 0, p = 0.001) and Inflammatory Neuropathy Cause and Treatment Disability score (2 vs 0.5, p = 0.012) (Class IV). Significantly higher proportion of irNeuropathies had acute or subacute and nonlength-dependent presentations (p < 0.001) and rate of hospitalization for irNeuropathy was also higher (p = 0.002) compared to toxic neuropathy from chemotherapy.

Conclusion: Neuropathy is a rare complication of ICIs that often responds to immunosuppression. Recognition of its wide phenotypic spectrum and distinct clinical characteristics and prompt management with corticosteroids may lead to favorable outcomes.