Multiparametric evaluation of coronary flow predicts longterm outcome in heart transplantation: from coronary flow velocity reserve to its newly introduced companion

Background: Coronary microvascular dysfunction (CMD) leads to a worse prognosis in heart transplantation (HT) patients. Coronary flow velocity reserve (CFVR) estimates the physiologic impact of allograft disease on the coronary circulation. Purpose: Our aim was to determine the prognostic role of CFVR and its companion (CFVRC) on long-term survival of HT patients with a follow-up of 28 years. Methods: 134 HT patients, surviving at least 5 years after HT, with normal systolic ventricular function and no evidence of angiographic allograft vasculopathy or symptoms/signs of rejection were included. The enrolled population underwent echocardiographic evaluation of microvascular function by the assessment of both the ratio of hyperemic to rest diastolic peak velocity (DPVh and DPVr). These measurements yield CFVR and its associated companion, defined as CFVRC=v{(DPVr)²+(DPVh)²}, as well as basal and hyperemic coronary microvascular resistance (BMR and HMR). A CFVR=2.5 was considered abnormal; the median value of DPVh (75 cm/s) and CFVRC (80 cm/s) were utilized to dichotomize the population. Results: Based on CFVR and DPVh, HT patients can be assigned to four groups: group 1 (n=32), discordant with preserved CFVR (3.1±0.4); group 2 (n=60), concordant with preserved CFVR (3.4±0.5); group 3 (n=31), concordant with impaired CFVR (1.8±0.3) and group 4 (n=11), discordant with impaired CFVR (2.0±0.2). Survival for each patient group is presented in the Figure (panel A). Specifically, survival was similar in group 1 when compared to group 3 (p=0.8), but significantly lower when compared to group 2 (p=0.03). Therefore, a normal CFVR (>2.5) may not be able to predict the unfavourable long-term outcome. CFVR in fact is an incomplete dimensionless ratio; if the paired velocities are low with high BMR and HMR (group 1), the use of CFVR alone may miss some events, that are yet captured by CFVRC. Differences between survivors and no survivors are presented in the Table. At multivariable survival analysis, CMD, DPVh<75 cm/s, CFVRC<80 cm/s were independent predictors of mortality in HT patients. Consequently, we evaluated the added role of the CMD, DPVh<75 cm/s and CFVRC<80 cm/s to prognostic models including the clinical (Figure, panel B) predictors of mortality. The inclusion of CFVRC<80 cm/s to model with clinical predictors of mortality permitted better prediction of survival in HT patients, compared to only adding CMD or DPVh<75 cm/s. Conclusions. This study is the first to demonstrate that the CFVR alone, even representing a determinant of survival in long-term HT patients, is not sufficient to completely predict long-term survival in HT patients. In comparison to CMD and DPVh, the CFVRC provides a significant improvement in survival prediction in long-term HT patients. Thus, the proposed multiparametric approach offers a more comprehensive evaluation of prognosis in HT patients, just by applying available data without the need to perform additional measurements.