Does Modern Mediastinal Irradiation Cause Acute Subclinical Cardiac Toxicity? The Final Results of the MEDICATE Study (Mediastinal Irradiation and Cardio-Toxic Effects)

Abstract

Radiation-induced heart disease (RIHD) is a late effect of cardiac irradiation (RT), and has been demonstrated in Hodgkin and non-Hodgkin lymphoma (HL and NHL), as well as in other thoracic cancers. There is no established measurement tool to detect acute cardiac damage, however high sensitivity troponin I and T (HsTnI and HsTNT) and imaging modalities such as echocardiograms have shown promise in some studies. We conducted a prospective trial to characterize acute and subacute cardiac damage in patients with thoracic cancers receiving cardiac radiation therapy. Patients were recruited from Hematology or Lung clinics at our institution if they were to receive cardiac RT of at least 30Gy to 5% of cardiac volume or a mean dose of 4Gy. Following consent and CT simulation, conformal (3D) and intensity modulated radiation therapy (IMRT) plans were created and cardiac structures were contoured. Patients had blood drawn for C-reactive protein (systemic inflammatory marker), HsTnI, HsTnT prior to RT and at 2 and 4 weeks into radiation therapy and completed 3D echocardiograms (3D) prior to RT and one year post-RT. Serum markers levels, imaging results and cardiac RT dose-volume data were analyzed for association. Twenty patients were recruited (including one who withdrew consent). The median value for “mean LV dose” was 3.1 Gy, and the median value for “mean total cardiac dose” was 8.6 Gy. Statistically significant positive associations between hsTnI and hsTnT at all time points, with an apparent decrease in mean value over time, and therewas a positive association between hsTnI increase between week 2 to 4 and maximum cardiac LV dose (ρ= 0.66, p=0.027). Although overall the mean change in LVEF for the whole cohort was non-significant (57.57 pre-RT and 56.4 post –RT), the mean dose to the left ventricle was associated with decrease in ejection fraction from pre-RT to one year echocardiograms (ρ= -0.46, p=0.054). An association was also seen between increase in left ventricular strain (LVS) and cardiac mean dose (ρ= 0.46, p=0.058), and cardiac dose to maximum 5% volume (ρ= 0.48, p=0.043). Our study suggests that hsTnI and hsTnT are intimately related, but detection of acute cardiac damage was not demonstrated potentially due to limitations of these markers, or low radiation therapy doses using modern techniques. Our results also suggest 3D echocardiogram findings may detect subacute damage at one year depending on dose to cardiac volumes. Given limitations of our study, we suggest that techniques to measure subacute cardiac damage including serum biomarkers and imaging modalities require further characterization to optimize their use. Identification early damage could facilitate the ability to closely monitor and intervene in patients at risk for RIHD. The current study is reassuring that modern quality-assured radiation techniques can prevent acute cardiac damage, even after anthracycline chemotherapy.