Montefiore Einstein Comprehensive Cancer Center Bronx, NY, United States
N. Ohri1, S. Jolly2, B. Cooper3, R. Kabarriti1, W. R. Bodner III1, C. Guha1, S. Viswanathan1, E. Shum3, J. K. Sabari3, H. Cheng1, R. Gucalp1, E. Castellucci1, A. Qin2, S. M. Gadgeel4, and B. Halmos1; 1Montefiore Einstein Comprehensive Cancer Center, Bronx, NY, 2University of Michigan Rogel Cancer Center, Ann Arbor, MI, 3Perlmutter Cancer Center, NYU Grossman School of Medicine, New York, NY, 4Henry Ford Cancer Institute/Henry Ford Health System, Detroit, MI
Purpose/Objective(s): Initial reports from the Selective Personalized RadioImmunotherapy for Locally Advanced NSCLC Trial (SPRINT) indicated that patients with high PD-L1 expression can be treated safely and effectively with immunotherapy and radiotherapy (RT). Here we compare patient-reported outcomes (PROs) from SPRINT to PROs from patients who were treated with standard concurrent chemoradiotherapy on contemporary prospective trials to test the hypothesis that the SPRINT approach yields favorable patient-reported outcomes. Materials/
Methods: SPRINT participants with PD-L1 tumor proportion score = 50% were treated with “PembroRT”: three 21-day cycles of induction pembrolizumab, a four-week course of risk-adapted and de-intensified RT (55 Gy to tumors or lymph nodes measuring 20 cc or larger and 48 Gy to smaller lesions, all in 20 fractions and without concurrent chemotherapy), and up to 13 cycles of consolidation pembrolizumab. SPRINT participants with lower PD-L1 expression and participants on the control arm of NCT03481114 were treated with “ChemoRT”: a standard RT dose of 60 Gy in 30 fractions with concurrent weekly carboplatin and paclitaxel and standard adjuvant therapy (e.g., durvalumab). Participants in both trials completed a 12-item PRO-CTCAE survey at baseline, every two weeks during RT, and at follow-up visits. Study participants who completed the planned RT course and completed at least two PRO-CTCAE surveys were included in this analysis. Peak PRO-CTCAE grades recorded up to 90 days after RT completion were compared across treatment cohorts using Kruskal-Wallis testing. Multivariable logistic regression models were utilized to evaluate predictors of grade 2+ patient-reported toxicity. Kaplan-Meier curves and logrank testing were utilized to compare progression-free survival (PFS) and overall survival (OS) across treatment cohorts. Results: The PembroRT cohort included 19 patients, and the ChemoRT cohort included 26 patients. Mean esophagus and lung doses were significantly lower in the PembroRT cohort, with median values approximately 5 Gy lower than in the ChemoRT group. Average PRO-CTCAE grades were numerically higher in the ChemoRT group for all 12 symptoms, with statistically significant differences observed for cough, wheezing, dermatitis, dizziness, anxiety, and depression. Multivariable models demonstrated that patient-reported toxicity was generally not associated with chemotherapy receipt, while several patient-reported adverse events were associated with organ at risk doses. For example, grade 2+ dysphagia incidence was associated with mean esophageal dose. The PembroRT cohort had superior 3-year rates of PFS (61% v. 23%, p=0.033) and OS (88% v. 49%, p=0.013). Conclusion: Compared to standard chemoradiotherapy, the SPRINT treatment approach, which utilizes induction immunotherapy to reduce target volumes and risk-adapted RT, can improve patient-reported outcomes without compromising clinical outcomes.
