Radionuclide ventriculography phase analysis for risk stratification of patients undergoing cardiotoxic cancer therapy

• Published in: Journal of nuclear cardiology 2022-04, Vol.29 (2), p.581-589
• Main Authors: Jones, K.A., Small, A.D., Ray, S., Hamilton, D.J., Martin, W., Robinson, J., Goodfield, N.E.R., Paterson, C.A.
• Format: Article
• Language: English
• Subjects: Breast cancer; Breast Neoplasms - diagnostic imaging; Breast Neoplasms - drug therapy; Cancer therapies; Cardiology; Cardiotoxicity; diagnostic and prognostic application; dyssynchrony; Ejection fraction; Entropy; Female; Heart Diseases; Humans; image analysis; Imaging; Medical diagnosis; Medicine; Medicine & Public Health; Nuclear Medicine; Original; Original Article; Radiology; Radionuclide Ventriculography; Retrospective Studies; Risk Assessment; RNA: planar; Stroke Volume; Ventricular Dysfunction, Left - chemically induced; Ventricular Dysfunction, Left - diagnostic imaging; Ventricular Function, Left
• ISSN: 1071-3581; 1532-6551
• DOI: 10.1007/s12350-020-02277-z

Accurate diagnostic tools to identify patients at risk of cancer therapy-related cardiac dysfunction (CTRCD) are critical. For patients undergoing cardiotoxic cancer therapy, ejection fraction assessment using radionuclide ventriculography (RNVG) is commonly used for serial assessment of left ventricular (LV) function. In this retrospective study, approximate entropy (ApEn), synchrony, entropy, and standard deviation from the phase histogram (phase SD) were investigated as potential early markers of LV dysfunction to predict CTRCD. These phase parameters were calculated from the baseline RNVG phase image for 177 breast cancer patients before commencing cardiotoxic therapy. Of the 177 patients, 11 had a decline in left ventricular ejection fraction (LVEF) of over 10% to an LVEF below 50% after treatment had commenced. This patient group had a significantly higher ApEn at baseline to those who maintained a normal LVEF throughout treatment. Of the parameters investigated, ApEn was superior for predicting the risk of CTRCD. Combining ApEn with the baseline LVEF further improved the discrimination between the groups. The results suggest that RNVG phase analysis using approximate entropy may aid in the detection of sub-clinical LV contraction abnormalities, not detectable by baseline LVEF measurement, predicting a subsequent decline in LVEF.