Heart Dosimetry is Correlated With Risk of Radiation Pneumonitis After Lung-Sparing Hemithoracic Pleural Intensity Modulated Radiation Therapy for Malignant Pleural Mesothelioma

International Journal of Radiation Oncology, Biology, Physics 2017 September [Epub 2017 April] [Link]

Yorke ED, Jackson A, Kuo LC, Ojo A, Panchoo K, Adusumilli P, Zauderer MG, Rusch VW, Shepherd A, Rimner A


To determine clinically helpful dose-volume and clinical metrics correlating with symptomatic radiation pneumonitis (RP) in malignant pleural mesothelioma (MPM) patients with 2 lungs treated with hemithoracic intensity modulated pleural radiation therapy (IMPRINT).
Treatment plans and resulting normal organ dose-volume histograms of 103 consecutive MPM patients treated with IMPRINT (February 2005 to January 2015) to the highest dose ≤50.4 Gy satisfying departmental normal tissue constraints were uniformly recalculated. Patient records provided maximum RP grade (Common Terminology Criteria for Toxicity and Adverse Event version 4.0) and clinical and demographic information. Correlations analyzed with the Cox model were grade ≥2 RP (RP2+) and grade ≥3 RP (RP3+) with clinical variables, with volumes of planning target volume (PTV) and PTV-lung overlap and with mean dose, percent volume receiving dose D (VD), highest dose encompassing % volume V, (DV), and heart, total, ipsilateral, and contralateral lung volumes.
Twenty-seven patients had RP2+ (14 with RP3+). The median prescription dose was 46.8 Gy (39.6-50.4 Gy, 1.8 Gy/fraction). The median age was 67.6 years (range, 42-83 years). There were 79 men, 40 never-smokers, and 44 with left-sided MPM. There were no significant (P≤.05) correlations with clinical variables, prescription dose, total lung dose-volume metrics, and PTV-lung overlap volume. Dose-volume correlations for heart were RP2+ with VD (35 ≤ D ≤ 47 Gy, V43 strongest at P=.003), RP3+ with VD (31 ≤ D ≤ 45 Gy), RP2+ with DV (5 ≤ V ≤ 30%), RP3+ with DV (15 ≤ V ≤ 35%), and mean dose. Significant for ipsilateral lung were RP2+ with VD (38 ≤ D ≤ 44 Gy), RP3+ with V41, RP2+ and RP3+ with minimum dose, and for contralateral lung, RP2+ with maximum dose. Correlation of PTV with RP2+ was strong (P<.001) and also significant with RP3+. CONCLUSIONS: Heart dose correlated strongly with symptomatic RP in this large cohort of MPM patients with 2 lungs treated with IMPRINT. Planning constraints to reduce future heart doses are suggested.