(P144) Prospective Assessment of Patient-Specific Pulmonary Radiation Pneumonitis

Ngoc T. Pham, MD, Dmitriy Meshkov, BS, Richard Castillo, PhD, Edward Castillo, PhD, Samantha Aso, MD, Brian Hobbs, PhD, Thomas M. Guerrero, MD, PhD; University of Texas Health Science Center; UT MD Anderson Cancer Center

Purpose: Radiation pneumonitis (RP) is a significant toxicity following thoracic radiotherapy, with no method to predict individual risk. Our lab had previously shown that pulmonary metabolic response rate (PMRR) is highly predictive of RP in lung cancer. In this prospective study, we used PMRR as a surrogate marker to predict radiation-induced lung toxicity in lung cancer patients receiving thoracic radiotherapy.

Materials and Methods: Forty-one patients receiving thoracic radiation for lung cancer were enrolled in a phase II study. Each patient received respiratory surveys before RT initiation and at each weekly follow-up visit. In addition, patients also received restaging positron emission tomography (PET)/CT imaging 30–60 days after completion of RT. The PMRR was determined from the slope of the normalized standardized uptake value (SUV) vs the dose. Deviation of the dose response from a linear model was tested for each case. St. George Respiratory Questionnaire (SGRQ) was used to monitor pneumonitis toxicity. An increase of respiratory efforts in four different activities compared with baseline was considered to be symptomatic. The follow-up endpoint was 1 year after completion of RT, or death. The demographics, dosimetric factors, and PMRRs were evaluated for correlation with RP.

Results: Thirty-eight patients completed the trial. Two patients did not complete the study, and three patients did not qualify due to a restaging PET of more than 3 months. Mean baseline lung function of all patients was 41.46 on the SGRQ. Twenty-two (58%) patients developed RP. Mean ΔSGRQ score for symptomatic patients was +19.93, and mean ΔSGRQ score for asymptomatic patients was –6.225. No clinical factors (age, tumor size, tumor location, tumor history, or dosimetry parameters) were associated with symptoms. The median PMRR for symptomatic patients was 0.014 (range: 0.002–0.076) and 0.007 for asymptomatic patients (P < .002). Receiver operating characteristics (ROC) analysis yielded an area under the curve (AUC) of 0.79 (95% confidence interval [CI], 0.64–0.94), with a PMRR threshold at 0.0108 providing sensitivity of 64% and specificity of 88%.

Conclusions: The radiation dose response on fluorodeoxyglucose (FDG) PET/CT imaging exhibited a linear relationship on statistical modeling. Patients who developed symptomatic RP had a significantly higher dose-response slope (PMRR) than asymptomatic patients.