Treatment Breaks Are Equally Impactful Upon HPV+ and HPV- Patients Undergoing Definitive Radiotherapy: A Multi-Institution Analysis

Abstract

HPV-associated oropharyngeal carcinoma (HPV+ OPC) is a distinct disease relative to its historic counterpart (HPV- OPC). It is possible that the time to treatment initiation (TTT) and the time on treatment (TOT) are of variable importance within the different diseases. We tested this hypothesis among a multi-institution cohort. After IRB and legal approval, data from matched retrospective head and neck cancer registries at two tertiary referral cancer centers and regional satellites were merged. Patients with AJCC 7th Edition stage I-IVB, non-metastatic squamous carcinoma of the oropharynx with pathologically confirmed HPV status treated from 2006-2017 were included. All patients were managed non-operatively with definitive radiotherapy with or without chemotherapy and completed the prescribed treatment. TTT was defined as days elapsed between biopsy and initiation of radiotherapy. TOT was defined as days between the first and last fraction of radiation. Progression-free survival (PFS) was defined as the time between the start of radiation and any recurrence or death. The impact of the two candidate variables was investigated using Cox regression. Nonlinear effects were explored using penalized smoothing splines with 3 degrees of freedom. Interaction terms were investigated and data reduction was used for the construction of the multivariable regression model. From a combined database of 1,232 oropharynx patients, 681 were included that met the criteria above. HPV status was positive in 585 (86%) and negative in 96 (14%). Chemotherapy was delivered in 89%. The median TTT was 34 days (range 3-1,084) and median TOT was 49 days (range 32-95). TTT of ≥45 days occurred in 27% and ≥60 days in 12%. TOT of ≥50 days occurred in 40%, 55 days in 13% and 60 days in 3%. The median clinical follow-up was 41 months (range 1.3-146). On univariate analysis, TTT was not found to be associated with PFS in either the HPV+ or HPV- cohorts. TOT was associated on univariate analysis with PFS in all patients (HR 1.028, 95% CI 1.007-1.049, p=0.013); this effect size was similar between both the HPV+ (HR 1.022, 95% CI 0.996-1.048, p=0.093) and HPV- (HR 1.023, 95% CI 0.989-1.059, p=0.193) cohorts. Non-linear effects were not observed for either TTT or TOT. Multivariable adjusted analysis demonstrated no statistically significant interaction between HPV status and TTT or TOT after adjusting for comorbidity, smoking status, stage, and chemotherapy use. No evidence was demonstrated to support a differential effect of radiation therapy timing between HPV+ and HPV- OPC. Equal weight on early initiation of radiotherapy and minimizing treatment breaks can be maintained between the two diseases.