Stress Testing Prior to Allogeneic Hematopoietic Stem Cell Transplantation

Hematopoietic stem cell transplantation (HSCT) exposes patients to significant physiologic stress. Risk stratification prior to HSCT is routine. Despite limited data supporting the prognostic benefit, assessment of left ventricular systolic function and cardiac stress testing are often performed. A retrospective analysis of 284 allogeneic HSCT performed from 1999-2000 in 260 patients, all of whom had a pre-HSCT stress test, is presented. Stress echos, 170, and stress MUGAs, 114, were performed at Memorial Sloan-Kettering or with outside physicians. Patients who underwent supine bicycle exercise multiple gated cardiac blood pool imaging (stress MUGA) or treadmill exercise echo were included, allowing for assessment of augmentation in left ventricular ejection fraction with exercise and functional capacity. Stress tests used standardized exercise protocols (Bruce, modified Bruce for stress echo or WHO for supine exercise bicycle). Stress tests were in reasonable temporal proximity to HSCT, 46±61 days prior for stress echo and 36±66 days prior for stress MUGA (not significantly different). Average age at transplant was approximately 55 years and was not statistically different for patients who had stress echo or stress MUGA, but was significantly older when compared patients who were transplanted over the same period and for whom only a rest echo could be identified in the medical record (54±8yrs vs 52±7yrs, p=0.01). Coronary artery disease or at least one major cardiac risk was identified in 69% of patients who underwent stress testing. Two thirds of patients who had stress tests had leukemia and one third lymphoma. As measured with stress echo, percent maximum predicted heart rate (%MPHR), workload (measured in METS), and percent augmentation in LVEF with exercise did not correlate with ICU admission, in-hospital death or death within one year. Decreased exercise time and decreased rest LVEF both correlated significantly with in-hospital death (ET 6±3min vs 7±2min, p=0.01 and rest LVEF 57±7% vs 62±7%, p=0.01), but neither correlated with ICU admission or death at one year. Correlation with stress MUGA was weaker with decreased workload (measured in WATTs) correlating with in-hospital death (59±25WATTs vs 76±30WATTs, p=0.03), but without correlation with ICU admission or death at one year. Other stress MUGA parameters (%MPHR, rest LVEF, percent augmentation in LVEF, exercise time) had no significant correlation with ICU admission, in-hospital death or death within one year of transplant. Most stress test parameters correlate poorly with outcomes of HSCT. Routine stress testing is unlikely to be of significant prognostic value. Stress echocardiogram is of somewhat more value than stress MUGA given that rest LVEF and exercise time correlate with in-hospital death, but a standardized evaluation of exercise capacity in the clinic and rest echo may be adequate substitutes.