Early Toxicity and Patient Reported Outcomes From a Phase 2 Trial of FDG-PET Response-Based De-Escalated Definitive Radiotherapy for p16+ Oropharynx Cancer

Abstract

<h3>Purpose/Objective(s)</h3> Pre- and mid-treatment FDG-PET parameters are prognostic of treatment response in p16+ oropharynx cancer (OPC). Our prospective Phase 2 de-escalation study utilizes FDG-PET response criteria to select patients eligible for de-escalated radiotherapy and hypothesizes that early swallow function and quality of life will improve with lower dose radiation. <h3>Materials/Methods</h3> Eligibility criteria stipulated stage I-II (AJCC 8) p16+ OPC with FDG-avid disease. Patients with matted lymph nodes or history of head and neck surgery were excluded. All patients received concurrent weekly carboplatin/paclitaxel. FDG-PET was obtained at baseline and at fraction 10. Patients were de-escalated to 54 Gy in 27 fractions if both pretreatment metabolic tumor volume (MTV) 50% of SUVmax was ≤22cc and interval MTV above SUV 2.5 decreased ≥50% from baseline. Those not meeting FDG-PET response criteria received 70 Gy in 35 fractions. Planned interim analysis examined early toxicity and patient-reported outcomes (PROs: xerostomia questionnaire, UWash QoL, FACT-HN, PRO-CTCAE) after accrual of 48 of 75 evaluable patients. Change from baseline scores of PROs was assessed with a mixed-effects model regarding +/- 0.5 standard deviations as clinically significant, consistent with prior work. <h3>Results</h3> Overall population (n=48) characteristics were: median age 59.5 years, 85% male, 44% ever smoker with 20 pack-year median, 90% T1-2, and 81% N1 and similar between cohorts. Pre-treatment GTV size was not different (30.8 cc 54 Gy vs 33.1 cc 70 Gy cohort, p=0.54). Exactly 50% of patients (n=24) met the de-escalation cohort criteria with median decrease in interval MTV_2.5 of 63% (IQR 56-82%) vs median decrease of 26% (IQR 4-40%) among the 70 Gy cohort. The dose to the larynx, constrictors, oral cavity, and salivary glands in the 54 Gy cohort was 70-78% of that received by the 70 Gy cohort (all comparisons p<0.01). De-escalated patients lost significantly less weight as percent of baseline (7.0% vs 11.5%, p<0.001) and fewer required feeding tube placement during treatment (1 vs 5, p=0.08). The 70 Gy cohort also had significantly worse videofluoroscopic swallowing function at 3 months compared to baseline with average increase in Penetration Aspiration Scale of 2 points (vs 0 points in the 54 Gy cohort, p=0.023). Across all 4 PRO surveys, while 54 Gy patients reported faring numerically better than the 70 Gy patients at 1 month post-treatment, this did not reach clinical or statistical significance. <h3>Conclusion</h3> Mid-treatment FDG-PET response adaptation allows for approximately half of stage I-II p16+ OPC patients to be de-escalated to 54 Gy. This results in an approximately 25% reduction in dose delivered to organs known to affect toxicity and quality of life with significantly better objective measures of toxicity and numerically improved PROs.