Risk and Clinical Risk Factors Associated with Late Lower Cranial Neuropathy in Long-term Oropharyngeal Squamous Cell Carcinoma Survivors

Importance: Lower cranial neuropathy (LCNP) is a rare, but permanent, late effect of radiotherapy and other cancer therapies. Lower cranial neuropathy is associated with excess cancer-related symptoms and worse swallowing-related quality of life. Few studies have investigated risk and clinical factors associated with late LCNP among patients with long-term survival of oropharyngeal squamous cell carcinoma (OPSCC survivors). Objective: To estimate the cumulative incidence of and identify clinical factors associated with late LCNP among long-term OPSCC survivors. Design, Setting, and Participants: This single-institution cohort study included disease-free adult OPSCC survivors who completed curative treatment from January 1, 2000, to December 31, 2013. Exclusion criteria consisted of baseline LCNP, recurrent head and neck cancer, treatment at other institutions, death, and a second primary, persistent, or recurrent malignant neoplasm of the head and neck less than 3 months after treatment. Median survival of OPSCC among the 2021 eligible patients was 6.8 (range, 0.3-18.4) years. Data were analyzed from October 12, 2019, to November 13, 2020. Main Outcomes and Measures: Late LCNP events were defined by neuropathy of the glossopharyngeal, vagus, and/or hypoglossal cranial nerves at least 3 months after cancer therapy. Cumulative incidence of LCNP was estimated using the Kaplan-Meier method, and multivariable Cox proportional hazards models were fit. Results: Among the 2021 OPSCC survivors included in the analysis of this cohort study (1740 [86.1%] male; median age, 56 [range, 28-86] years), 88 (4.4%) were diagnosed with late LCNP, with median time to LCNP of 5.4 (range, 0.3-14.1) years after treatment. Cumulative incidence of LCNP was 0.024 (95% CI, 0.017-0.032) at 5 years, 0.061 (95% CI, 0.048-0.078) at 10 years, and 0.098 (95% CI, 0.075-0.128) at 15 years of follow-up. Multivariable Cox proportional hazards regression identified T4 vs T1 classification (hazard ratio [HR], 3.82; 95% CI, 1.85-7.86) and accelerated vs standard radiotherapy fractionation (HR, 2.15; 95% CI, 1.34-3.45) as independently associated with late LCNP status, after adjustment. Among the subgroup of 1986 patients with nonsurgical treatment, induction chemotherapy regimens including combined docetaxel, cisplatin, and fluorouracil (TPF) (HR, 2.51; 95% CI, 1.35-4.67) and TPF with cetuximab (HR, 5.80; 95% CI, 1.74-19.35) along with T classification and accelerated radiotherapy fractionation were associated with late LCNP status after adjustment. Conclusions and Relevance: This single-institution cohort study found that, although rare in the population overall, cumulative risk of late LCNP progressed to 10% during the survivors' lifetime. As expected, clinical factors associated with LCNP primarily reflected greater tumor burden and treatment intensity. Further efforts are necessary to investigate risk-reduction strategies as well as surveillance and management strategies for this disabling late effect of cancer treatment.