The effect of durvalumab consolidation after definitive radiochemotherapy for non-operable stage III non-small cell lung cancer on the dose effect relation for therapy related pulmonary infiltrates as a risk factor for pneumonitis.

Background: Consolidation therapy with the anti-programmed death-ligand 1 (PD-L1) antibody durvalumab, or other immune checkpoint inhibitors, has been associated with improved progression-free and overall survival in patients with stage III non-small cell lung cancer (NSCLC) as demonstrated in randomized clinical trials. The purpose of the present study is to evaluate the dose-response relationship for partial lung infiltrate volumes per dose bin after definitive radiochemotherapy as a sensitive end point to detect a durvalumab effect on the lung parenchyma in patients with subclinical or grade ≤2 pneumonitis.

Methods: Consecutive patients from a prospective registry with inoperable NSCLC stage III who developed no or pneumonitis grade ≤2 after definitive radiochemotherapy with or without durvalumab consolidation were included. Pulmonary infiltrates outside the planning target volumes were contoured in the follow-up computed tomography (CT) at the time of maximum infiltrate expression. Partial lung infiltrate volumes per dose bin were determined over the entire dose range. A mixed random and fixed effect model was used to fit dose response curves stepwise in dose bins of 5 Gy. The Akaike information criterion (AIC) was used for model comparison.

Results: Sixty patients with and 44 without durvalumab consolidation were analysed. The step model showed a significant dose response relationship for the pulmonary infiltrates (P<0.001, F-test) that was modified by the durvalumab effect (P<0.001, F-test). There was a significant dependence of the durvalumab effect on radiation dose (P=0.003). The durvalumab effect increased with dose from 0% in the lowest dose bin as reference to 5.2%±1.2% partial lung infiltrate volume in the highest dose bin. There was significant inter-individual heterogeneity of partial lung infiltrate volumes at each dose bin (P<0.001, F-test). The percentage of lung volume receiving at least 5 Gy (V5) was the most significant characteristic increasing risk of pulmonary infiltrates (P<0.001, F-test). Multivariable proportional hazards Cox-model showed that pulmonary infiltrates at 5-10 and 35-40 Gy dose bins were dominant factors determining the risk of pneumonitis grade 2.

Conclusions: The relationship between radiation dose and lung infiltrates observed by follow-up CT scans after radiochemotherapy is sensitive enough to detect factors that systematically increase the radiation dose response. In this case, the focus is on durvalumab consolidation and radiation dose to the lung. The dose-response relationship may help in the prediction of grade 2 pneumonitis. However, further research is needed to understand the biological factors that contribute to the large differences in response to radiation dose between individuals.