Comparison of impedance cardiography and cardiac magnetic resonance imaging for the evaluation of cardiac function in early-stage breast cancer patients.

Objective Breast cancer treatment can negatively impact cardiac function in some breast cancer patients. Current methods (MUGA, echocardiography) used in clinical practice to detect abnormal cardiac changes as a result of treatment suffer from important limitations. Use of alternative techniques that would offer safe, inexpensive and non-invasive cardiac function assessment in this population would be highly advantageous. The aim of this study was to examine the agreement between impedance cardiography (ICG) and cardiac magnetic resonance imaging (CMR) in quantifying stroke volume (SV), cardiac output (CO) and end-diastolic volume (EDV) in this population. Approach Sixteen breast cancer patients underwent ICG and CMR assessments at three time-points (before treatment, immediately after chemotherapy, and four months after chemotherapy). Bland-Altman analysis was used to quantify the accuracy and precision of ICG (relative to CMR) in estimating absolute values of SV, CO and EDV. Four methods (concordance rate, polar concordance rate, clinical concordance rate and trend interchangeability rate) were also used to assess ICG performance in tracking changes in these variables. Main results Bland-Altman analysis showed that the accuracy of ICG relative to CMR was -3.1 ml (SV), 0.2 L·min-1 (CO) and -26.0 ml (EDV) and precision was 13.2 ml (SV), 1.1 L·min-1 (CO) and 20.1 ml (EDV), respectively. Trending ability assessment showed that 1) the concordance rate was 87% (SV), 73% (CO) and 73% (EDV), 2) the polar concordance rate was 67% (SV), 53% (CO) and 33% (EDV), 3) the clinical concordance rate was 33% (SV), 40% (CO) and 20% (EDV) and 4) the trend interchangeability rate was 29% (SV), 43% (CO) and 17% (EDV), respectively. Significance Our findings show that, although ICG showed good accuracy for absolute SV and CO measurements and for CO and EDV changes, precision was poor for all variables in terms of both absolute measurements and trend tracking performance. This suggests that ICG cannot be used interchangeably with CMR in breast cancer patients.