Evaluation of radiation-induced pleural effusions after radiotherapy to support development of animal models of radiation pneumonitis

Document Type



Pulmonary and Critical Care


Not all animal models develop radiation-induced pleural effusions (RIPEs) as a form of radiation-induced lung injury (RILI). Such effusions are also not well characterized in humans. The purpose of this study is to identify occurrences of RIPE in humans, provide justification for development of relevant animal models, and further characterize its risk factors in cancer patients. We also aim to identify dose thresholds for cardiopulmonary toxicity in humans to shed light on possible pathogenic mechanisms for RIPEs. We carried out a retrospective review of medical records of 96 cancer patients receiving thoracic irradiation (TRT) at our institution. Fifty-three (53%) patients developed a new pleural effusion post TRT; 18 (19%) had RIPE; and 67% developed RIPE ipsilateral to the site irradiated. None developed "contralateral only" effusions. Median time to development was 6 mo (IQR; 4-8 mo). Of 18, 8 patients (44%) had concomitant asymptomatic (radiographic only) or symptomatic radiation pneumonitis and pericardial effusion. Dosimetric factors, including combined and ipsilateral mean lung dose (MLD), were significantly associated with increased risk of RIPE. Angiotensin converting enzyme inhibition, steroids, or concurrent chemotherapy did not modify incidence of RIPE. Our results substantiate the occurrence and incidence of RIPEs in humans. In cancer patients, a median time to development of effusions around 6 mo also supports the onset of RIPEs concurrent with radiation pneumonitis. Future work needs to include large populations of cancer survivors in whom delayed RIPEs can be tracked and correlated with cardiovascular changes in the context of injury to multiple organs.


Health Physics