Abstract
IMPORTANCE: No guidelines at present describe when fludeoxyglucose F 18-labeled positron emission tomography and computed tomography (FDG PET-CT) should be used in the initial posttreatment period for evaluation of oropharyngeal squamous cell carcinoma treatment outcome and recurrence. OBJECTIVE: To compare accuracies of the initial posttreatment PET-CT between primary treatment groups and to define indicators of false-positive findings. DESIGN, SETTING, AND PARTICIPANTS: This retrospective cohort study identified adults with a new diagnosis of oropharyngeal squamous cell carcinoma who received treatment with curative intent from October 1, 2006, through November 30, 2016, using the Alberta Cancer Registry (n = 380). Patients who underwent PET-CT within 1 year of treatment completion were included (n = 190). Of these, 103 patients (54.2%) had PET-CT findings positive for residual or recurrent disease, and 61 (32.1%) had false-positive findings. Among the 61 patients, 42 (68.9%) had received chemoradiotherapy (CRT) and 19 (31.1%) had primary surgery. Forty-two patients had true-positive findings, indicating a prevalence rate of disease of 22.1%. Data were analyzed from July through October 2017. EXPOSURES: One of 2 primary treatment modalities (surgery with or without adjuvant therapy vs CRT). All patients had posttreatment FDG PET-CT. MAIN OUTCOMES AND MEASURES: Primary outcome measures included the diagnostic odds ratio, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of PET-CT for detecting residual and/or recurrent disease. A multivariate analysis determined indicators of false-positive findings. Discriminative ability was assessed using receiver operating characteristic curve analysis of maximum standardized uptake value (SUVmax) metabolic data. RESULTS: Of the 190 participants, 77.9% were men, with a mean (SD) age at diagnosis of 58.5 (8.5) years. The diagnostic odds ratio was 19.3 (95% CI, 5.7-65.1); pooled sensitivity, 93.3% (95% CI, 80.7%-98.3%); and pooled specificity, 57.9% (95% CI, 49.4%-66.0%). The PPV of detecting disease was 54.7% (95% CI, 38.8%-69.8%) for primary surgery and 31.1% (95% CI, 20.2%-44.4%) for CRT. The NPV was 100% (95% CI, 94.7%-100%) for primary surgery and 96.6% (95% CI, 89.5%-99.1%) for CRT. Multivariate analysis identified treatment type, p16 disease, and smoking status as indicative of false-positive findings. In the receiver operating characteristic curve analysis for primary tumors, the optimal cutoff SUVmax for indicating true- vs false-positive results was 5.1 for surgically treated patients (area under the curve, 0.729; 95% CI, 0.570-0.888) and 5.3 for patients treated with CRT (area under the curve, 0.844; 95% CI, 0.700-0.989). CONCLUSIONS AND RELEVANCE: The results indicate a higher specificity for FDG PET-CT for initial posttreatment surveillance imaging among patients treated with primary surgery compared with nonsurgical management. Both sets of patients with posttreatment FDG PET-CT findings with an SUVmax greater than 5.0 should undergo close evaluation for possible residual or recurrent disease.