Influence of extracellular volume fraction on peak exercise oxygen pulse following thoracic radiotherapy

Abstract

Background: Radiation-induced myocardial fibrosis increases heart failure (HF) risk and is associated with a restrictive cardiomyopathy phenotype. The myocardial extracellular volume fraction (ECVF) using contrast-enhanced cardiac magnetic resonance (CMR) quantifies the extent of fibrosis which, in severe cases, results in a noncompliant left ventricle (LV) with an inability to augment exercise stroke volume (SV). The peak exercise oxygen pulse (O₂Pulse), a noninvasive surrogate for exercise SV, may provide mechanistic insight into cardiac reserve. The relationship between LV ECVF and O₂Pulse following thoracic radiotherapy has not been explored.

Methods: Patients who underwent thoracic radiotherapy for chest malignancies with significant incidental heart dose (≥5 Gray (Gy), ≥10% heart) without a pre-cancer treatment history of HF underwent cardiopulmonary exercise testing to determine O₂Pulse, contrast-enhanced CMR, and N-terminal pro-brain natriuretic peptide (NTproBNP) measurement. Multivariable-analyses were performed to identify factors associated with O₂Pulse normalized for age/gender/anthropometrics.

Results: Thirty patients (median [IQR] age 63 [57-67] years, 18 [60%] female, 2.0 [0.6-3.8] years post-radiotherapy) were included. The peak VO₂ was 1376 [1057-1552] mL·min^{- 1}, peak HR = 150 [122-164] bpm, resulting in an O₂Pulse of 9.2 [7.5-10.7] mL/beat or 82 (66-96) % of predicted. The ECVF, LV ejection fraction, heart volume receiving ≥10 Gy, and NTproBNP were independently associated with %O₂Pulse (P < .001).

Conclusions: In patients with prior radiotherapy heart exposure, %-predicted O₂Pulse is inversely associated markers of diffuse fibrosis (ECVF), ventricular wall stress (NTproBNP), radiotherapy heart dose, and positively related to LV function. Increased LV ECVF may reflect a potential etiology of impaired LV SV reserve in patients receiving thoracic radiotherapy for chest malignancies.

Keywords: Cardiorespiratory fitness; Extracellular volume fraction; Peak exercise oxygen pulse; Radiotherapy.

Fig. 1

Example of myocardial ECVF calculation from native T1, post-contrast T1, and blood pool inversion times. Abbreviations: ROI = region of interest; ECVF = extracellular volume fraction; ms = milliseconds

Fig. 2

Independent predictors of peak exercise %O₂ pulse. Legend: Panel A - correlation between LV ECVF and %O₂ pulse. Panel B - correlation between resting LVEF and %O₂ pulse. Panel C - correlation between cardiac V10Gy and %O₂ pulse. Panel D shows the inverse correlation between NTproBNP and %O₂ pulse. Abbreviations: %O₂ pulse = %-predicted oxygen pulse; ECVF = extracellular volume fraction; LVEF = left-ventricular ejection fraction; V10Gy = %volume of the heart receiving ≥10 Gray; NTproBNP=N-terminal pro-brain natriuretic peptide

Fig. 3

Differences between groups based upon percent-predicted peak exercise O₂ pulse or median LV ECVF. Legend: Box and whisker plots (Panels A-F) demonstrating median (horizontal line within rectangular box), interquartile range, whiskers, and range between groups based upon %O₂ pulse (< or ≥ 85%). Panel A: Lower SVI in those with peak O₂ pulse < 85% of predicted. Panel B: Lower LVEF in those with a peak O₂ pulse < 85% of predicted. Panel C: Lower peak VO₂ in those with a peak O₂ pulse < 85% of predicted. Panel D: Higher LV ECVF in those with a peak O₂ pulse < 85% of predicted. Panel E: Higher NTproBNP levels in those with a peak O₂ pulse < 85% of predicted. Panel F: Trend towards higher V10Gy in those with a peak O₂ pulse < 85% of predicted. Panel G: Lower resting LVEF in those with LV ECVF values >median. Panel H: Lower peak %O₂ pulse values in those with LV ECVF >median. Abbreviations: SVI = stroke volume index; O₂ = oxygen; LVEF = left-ventricular ejection fraction; VO₂ = oxygen consumption; ECVF = extracellular volume fraction; NTproBNP=N-terminal pro-brain natriuretic peptide; V10Gy = %volume of heart receiving ≥10 Gray