Influences on evaluability of radionuclide ventriculography in chemotherapy patients.

Abstract

Radionuclide ventriculography (RNV) is an established method of evaluating cardiotoxic side-effects of chemotherapy. The image quality of RNV depends on labelling yields obtained after red blood cell (RBC) labelling with 99Tc(m)-pertechnetate and has an influence on the evaluation of the left ventricular ejection fraction (LVEF). Several drugs and certain parameters of RBC labelling are known to have a detrimental effect on the labelling yield, but often the reason for poor image quality remains unclear. The aim of this study was to determine the effect of chemotherapeutic agents on LVEF evaluation. The chemotherapeutic medications and RNV data of 116 patients were noted. The patients underwent 205 RNV examinations (up to 7 RNV follow-up examinations) consisting of rest and stress studies. Ten patients with a poor labelling yield after in vivo labelling received an additional RNV study after in vitro labelling. The effects of commonly used anticoagulants and chemotherapeutic drugs on labelling yields were also investigated in experiments on in vitro RBC labelling. In vitro labelling had the advantage of better detection of pathological alterations in left ventricular motility, but often improved evaluability only slightly. The administration of corticosteroids showed an unexpected positive correlation with image quality (Spearman correlation coefficient: prednisone, 0.42403, P = 0.0013; prednisolone, 0.45629, P = 0.0286) and labelling yield (prednisolone, 0.65466, P = 0.0024), whereas idarubicin showed a negative correlation with image quality (-0.53364, P = 0.005). A slight positive correlation of prednisolone with LVEF at rest (0.45425, P = 0.0197) was also noted. Using our evaluation software, the manual contour method was found to be superior to the automatic determination of the left ventricular contour. Cycle ergometry alone caused a significant deterioration in image quality. The in vitro results suggested a negative influence of epirubicine on labelling yields at very high concentrations (10(-3) M). Our main result was that a clinically adequate study is possible in patients with moderate image quality and labelling yields. Furthermore, the administration of corticosteroids had a positive impact on image quality.