How to best assess tumor response after curative chemoradiotherapy (CRT) for head and neck squamous cell carcinomas (HNSCC): RECIST or PET/CT?

Abstract

6023 Background: Response evaluation criteria in solid tumors (RECIST) has questionable value when applied to potentially curable, locally advanced HNSCC. Combined functional and anatomical imaging offers potential advantages in response assessment. We previously showed that combined FDG-positron emission tomography/computed tomography (PET/CT) has higher accuracy in depicting head and neck cancer than CT or PET alone (Branstetter et al. Radiology 2005). Methods: We reviewed records of patients (pts) with previously untreated, stage III/IV HNSCC treated with primary CRT on 5 clinical trials from 2004–2007. All pts underwent PET/CT prior to treatment and approximately 8 weeks post CRT. The presence of a complete response (CR) was assessed by clinical exam, CT portion of PET/CT (RECIST) and PET portion of PET/CT. CR by PET was defined as complete disappearance of FDG activity attributable to tumor, without regard to the degree of CT response, as assessed on fused PET/CT. We evaluated the agreement of different methods in assessing CR and the correlation of CR with pt outcome. Results: 52 pts were analyzed. Median age was 54 years (20–74). Males 41 (79%). Stage III, 17%; IV, 83%. Primary sites: oropharynx (52%), oral cavity (6%), larynx (21%), hypopharynx (8%), nasopharynx (6%), unknown primary (8%). CRT regimens: cetuximab/cisplatin, 30; cetuximab/pemetrexed, 11; docetaxel/erlotinib, 9; cisplatin,1; carboplatin, 1. In addition, 28 pts received induction. Concordance between methods in assigning response as CR: CT and PET (62%), clinical exam and CT (43%); clinical exam and PET (51%). With a median follow up of 15 months, 1/27 of pts with CR by PET recurred vs 9/25 of pts without CR (p=0.004); 1/13 of pts with CR by CT recurred vs 9/39 of pts without CR (p=0.4); 8/42 of pts with CR by exam recurred vs 1/9 of pts without CR (p=0.5) Of 39 pts who did not achieve CR by CT (RECIST), 0/15 who had CR by PET progressed but 9/24 who had less than a CR by PET progressed (p=0.007). Conclusions: There is poor concordance between various methods of response assessment post CRT for HNSCC. Combined PET/CT is superior to CT (RECIST) and exam in predicting recurrence. PET negativity after CRT is a powerful predictor of outcome in HNSCC, regardless of response using RECIST. No significant financial relationships to disclose.