2004 - Patterns and Predictors of Pneumonitis after Chemoradiation Followed by Durvalumab

S. Appel¹, L. Wilk^2,3, O. Haisraely⁴, A. Saad⁵, C. Fink⁶, J. Bar^7,8, and Y. Lawrence⁹; ¹Radiation Oncology, Sheba Medical Center, Ramat Gan, Israel, ²Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel, ³Sheba Medical Center, The Arrow Project, Ramat Gan, Israel, ⁴Chaim Sheba Medical Center, Ramat Gan, Israel, ⁵Sheba Medical Center, Tel-Hashomer, Israel, ⁶Shamir Medical Center, Beer Yaacov, Israel, ⁷Sheba Medical Center, institute of Oncology, Ramat gan, Israel, ⁸Sackler School of Medicine, Tel Aviv University, TelAviv, Israel, ⁹Radiation Oncology Department, Sheba Medical Center, Ramat Gan, Israel

Purpose/Objective(s): Durvalumab (DURV) (Infinzi) given after chemoradiation for locally advanced non-small cell lung cancer (LANSCLC) improves oncologic outcome and is standard of care, with pneumonitis being the most prevalent side effect. Here we study patterns of durvalumab induced pneumonitis (DIP) and search for predictors of this side effect. Materials/

Methods: The database comprised of pts. with LANSCLC who were treated with chemoradiation to 60 Gy delivered with photon, VMAT at our center, and then consolidation DURV during the years 2017-2022. Medical charts were searched for occurrence of pneumonitis, graded according to CTCAE v.5. We analyzed patients, tumor and treatment variables predicting pneumonitis using COX regression. Radiation dose variables were divided into subgroups to find any significant cutoffs.
Results: Our cohort comprised of 119 patients, mean age 66 (SD8), 65% male, 58% stage IIB-IIIA, 51% adenocarcinoma. Pneumonitis was recorded in 35 pts. (29.4%): G1-12; G2-18; G3-3; G4-1; G5-1. DURV was interrupted due to pneumonitis in 24 pts. (20%) and was not resumed in 17 pts. (14.3%). Median time to pneumonitis was 2.6 months (range 0.3-11) and risk decreased over time (HR 0.93, p<0.001). Patient who suffered DIP received less DURV cycles (median 10 (IQR 4-20) vs. median 19 (IQR 10-24) (p<0.001)). Radiation induced pneumonitis prior to DURV administration was recorded in 9/119 cases, yet none of them suffered DIP. DIP did not correlate significantly with age, gender, weight loss at presentation, ECOG performance status, IHD, smoking, COPD, tumor histology or side, chemotherapy protocol or motion management. The risk of all grades DIP pneumonitis was increased with advanced tumor stage (HR 1.9, p=0.06) and adaptive replanning (HR 2.1, p=0.024). Dosimetric variables that correlated with all grades DIP were lung V20 = 27% (41% vs 24.7% HR 2, p=0.043), ipsilateral lung V5 =70% (43.3% vs. 25%, HR 2.2, p=0.027), mean lung dose (MLD) = 16 Gy (43.5% vs. 26%, HR 2.1, p=0.05), heart V40 = 6% (41.7% vs 25%, HR 2.0 p=0.045) and mean heart dose (MHD) = 6 Gy (39.7% vs 20%, HR 2.5, p=0.017). DIP grades 2-5 correlated with lungs V20 = 27% (35.3% vs. 13%, HR 3.2, p=0.005), ipsilateral lung V5 = 71% (32% vs. 16.5% HR 2.4, p=0.045), MLD = 15 Gy (30% vs. 13%, HR 2.5, p=0.03), heart V40 = 4% (30.6% vs 15.3%, HR 2.4 p=0.04) and MHD = 6Gy (27.6% vs. 12%, HR 2.8, p=0.03). The 3-year (and 95% CI) LC, PFS and OS were 68% (59-78%), 50% (40-60%) and 74% (65-82%).
Conclusion: Pneumonitis caused DURV interruptions in 20%. The risk of DIP correlated with advanced tumor stage and adaptive replanning. Radiation dose variables that correlated with G2-5 DIP were lungs V20 = 27%, ipsilateral lung V5 = 71%, MLD = 15 Gy, heart V40 = 4 % and MHD = 6 Gy. DIP did not occur in patient with radiation-induced pneumonitis prior to DURV administration. We recommend that strict lung and heart constraints be applied in patients scheduled to receive DURV.