METHODS: One hundred and nine breast cancer patients monitored for cardiotoxicity underwent 2D-echo, followed by pNaI and CZT MUGA scans on the same day. LVEF for CZT camera was processed using both automatic and manual processing methods, thus yielding four methods for the LVEF analysis.
RESULTS: Significant correlation (p<0.01) was seen among all four methods, with varied correlation strengths. Moderate correlation was seen between 2D-echo and both pNaI (r=0.56) and CZT cameras (automatic r=0.54, manual r=0.56). Strong correlation was registered between pNaI and CZT camera (automatic r=0.72, manual r=0.71). Bland-Altman limits of agreement among the three scans were wide and suboptimal. The widest limits were -21.1 to +16.2 (37%) between 2D-echo and CZT auto-processing.
CONCLUSION: Any one of the modalities can be used to measure LVEF, however, their results should not be used interchangeably. The same method of measurement is advised for serial scans.