Musaddiq Awan, MD
Medical College of Wisconsin
Head and Neck Cancer
Purpose/Objective(s):To develop a nomogram for overall survival (OS) for patients receiving chemoradiation for non-HPV related head and neck squamous cell carcinomas (HNSCCs).
Materials/Methods:Patients eligible for this study had hypopharyngeal, laryngeal, or p16-negative oropharyngeal HNSCCs and were enrolled on RTOG 0129 or 0522. De-identified data were extracted from the database as of the most recent publications of these trials (1/3/13 for RTOG 0129 and 6/28/12 for RTOG 0522). Extracted data included age, gender, Zubrod status, baseline hemoglobin, pack-year smoking history, head and neck subsite, clinical T-stage, N-stage, delivered chemotherapy dose, delivered RT fractions, treatment duration, and delivered RT dose. Patients receiving < 60 Gy or with unknown smoking pack-year history were excluded. To consolidate the effect of different RT regimens, time-equivalent biologically effective dose was calculated. The RTOG 0129 dataset was used for nomogram development and the RTOG 0522 dataset was used for nomogram validation. Significant predictors of OS were obtained using univariate analysis on the developmental dataset. The nomogram was built using Cox proportional hazards regression with significant predictors from univariate analysis. This nomogram was internally validated using calibration plots with and without optimism correction and c-statistics on the developmental dataset. This nomogram was then independently validated on the validation dataset using c-statistics. To calculate the bias-corrected 95% confidence interval (CI) for each c-statistic, the c-statistic was bootstrapped using 1,000 resamples for each nomogram for both the developmental and validation datasets. A p ≤ 0.10 was considered significant for all analyses.
Results:241 patients from RTOG 0129 and 289 patients from RTOG 0522 met the criteria for this analysis. Predictors of worse OS on univariate analysis were older age (p = 0.002), male gender (p = 0.07), Zubrod 1 (p = 0.04), T4 disease (p = 0.001), N2-N3 disease (p < 0.001), hypopharyngeal primary site (p = 0.06), and lower cumulative cisplatin dose (p = 0.003). On multivariable analysis, age (p = 0.02), T4 disease (p = 0.02), N2-N3 disease (p < 0.001), and lower cumulative cisplatin dose remained significant (p = 0.07). The OS nomogram showed excellent correlation both with and without optimism correction. Internal validation of the OS nomogram yielded c=0.67 (95% CI: 0.63-0.72). External validation yielded c=0.65 (95% CI: 0.60-0.70). Nomogram scores successfully stratified patients into low, intermediate and high-risk groups for mortality (p < 0.001).
Conclusion:For non-HPV-related HNSCCs, a nomogram using age, gender, Zubrod, T-stage, N-stage, primary site and chemotherapy dosage was developed and successfully validated for OS outcomes using national clinical trial data.
Medical College of Wisconsin
The asset you are trying to access is locked. Please enter your access key to unlock.