Positive Predictive Value of PET CT in HPV Associated Oropharyngeal Cancers Treated with Definitive Proton Therapy

Abstract

Proton therapy (PT) offers attractive dosimetric advantages but there is emerging literature suggesting relative biological effectiveness (RBE) differences at the distal end of the Bragg peak may result in increased posttreatment FDG avidity. While the utility of PET CT is well established, lingering posttreatment inflammation from PT may affect the interpretation of posttreatment scans (PTS). We seek to characterize the positive predictive value (PPV) and negative predictive value (NPV) for residual disease/locoregional recurrence in patients undergoing definitive PT with HPV associated oropharyngeal cancer. We performed a single institution retrospective analysis of patients with HPV associated oropharyngeal cancer treated with PT between 2016-2022. Patients were included if they had both pretreatment and initial posttreatment restaging PET/CT available for analysis. SUVmax was recorded in both the primary tumor and involved lymph nodes. Patients were considered to have a positive PET CT if there was an SUV decrease of less than 65% from the baseline exam or if there was SUV > 4.0 based on thresholds defined by previous studies. Chi-square analysis was used to compare patients who had a positive PTS to those who did not. Sixty-two patients were included for analysis with a median age of 62. The entire cohort had a median follow up of 21 months (Range 2-71 months) of whom 86% received chemotherapy. Eleven patients (17%) had PTS defined as positive, of whom only one (PPV 9%) was found to have residual/locoregionally recurrent disease within an area of FDG avidity on PTS. Another patient developed a new primary tumor 9 months later in an area outside of FDG avidity on PTS. The remainder were followed with serial clinical exam with or without the compliment of repeat imaging. They remained without evidence of residual/locoregionally recurrent disease at the time of last follow up. Fifty-one patients had negative posttreatment scans, with only one patient having residual/locoregionally recurrent disease (NPV 98%). On Chi-square analysis T1-2 vs T3-4 tumor, use of CTV margin vs no margin, chemotherapy vs no chemotherapy, smoking history, and tonsil vs base of tongue subsite were not statistically significant in association with positive PTS. PET/CT for post-treatment evaluation of patients treated with PT had an expectedly high NPV of 98% for residual/locoregionally recurrent disease; however, the PPV was 9% (1/11) which is much lower than would be expected with photon-based treatment. These findings may be the result of lingering acute inflammation caused by RBE differences at the distal end of the Bragg peak. Our findings are hypothesis generating and require validation with larger well-matched patient cohorts.