Vasoplegia after implantation of a continuous flow left ventricular assist device: incidence, outcomes and predictors

Abstract

Background: Vasoplegia after routine cardiac surgery is associated with severe postoperative complications and increased mortality. It is also prevalent in patients undergoing implantation of pulsatile flow left ventricular assist devices (LVAD). However, less is known regarding vasoplegia after implantation of newer generations of continuous flow LVADs (cfLVAD). We aim to report the incidence, impact on outcome and predictors of vasoplegia in these patients.

Methods: Adult patients scheduled for primary cfLVAD implantation were enrolled into a derivation cohort (n = 118, 2006-2013) and a temporal validation cohort (n = 73, 2014-2016). Vasoplegia was defined taking into consideration low mean arterial pressure and/or low systemic vascular resistance, preserved cardiac index and high vasopressor support. Vasoplegia was considered after bypass and the first 48 h of ICU stay lasting at least three consecutive hours. This concept of vasoplegia was compared to older definitions reported in the literature in terms of the incidence of postoperative vasoplegia and its association with adverse outcomes. Logistic regression was used to identify independent predictors. Their ability to discriminate patients with vasoplegia was quantified by the area under the receiver operating characteristic curve (AUC).

Results: The incidence of vasoplegia was 33.1% using the unified definition of vasoplegia. Vasoplegia was associated with increased ICU length-of-stay (10.5 [6.9-20.8] vs 6.1 [4.6-10.4] p = 0.002), increased ICU-mortality (OR 5.8, 95% CI 1.9-18.2) and one-year-mortality (OR 3.9, 95% CI 1.5-10.2), and a higher incidence of renal failure (OR 4.3, 95% CI 1.8-10.4). Multivariable analysis identified previous cardiothoracic surgery, preoperative dopamine administration, preoperative bilirubin levels and preoperative creatinine clearance as independent preoperative predictors of vasoplegia. The resultant prediction model exhibited a good discriminative ability (AUC 0.80, 95% CI 0.71-0.89, p < 0.01). Temporal validation resulted in an AUC of 0.74 (95% CI 0.61-0.87, p < 0.01).

Conclusions: In the era of the new generation of cfLVADs, vasoplegia remains a prevalent (33%) and critical condition with worse short-term outcomes and survival. We identified previous cardiothoracic surgery, preoperative treatment with dopamine, preoperative bilirubin levels and preoperative creatinine clearance as independent predictors.

Keywords: Cardiac Vasoplegia syndrome; Incidence; Mechanical circulatory support; Morbidity; Mortality; Outcome; Prediction.

Fig. 1

Stratification of vasoplegic patients according to the definition (Fig. 1a) and time frame of vasoplegia (Fig. 1b). Figure 1a. Overlap of the number of vasoplegic patients according to various definitions. Figure 1b. Time line of vasoplegia development in patients (numbers(%)) according to various definitions and time frames

Fig. 2

Patient survival after cfLVAD implantation in all patients in the derivation cohort (Fig. 2a), in patients with and without vasoplegia in the derivation cohort (Fig. 2b) censoring at heart transplantation or device removal (p <  0.01) and in patients with and without vasoplegia and stratified to Intermacs score (Fig. 2c)

Fig. 3

C-statistics of the final model and the temporal validation using the unified definition. Figure 3a. The predictive value of the final model, including previous cardiothoracic surgery, preoperative dopamine use, preoperative bilirubin level and creatinine clearance, calculated as area under the curve (AUC = 0.80, 95% CI 0.71–0.89, P <  0.01). Figure 3b. The predictive value of the temporal validation 0.74 (95% CI 0.61–0.87, p <  0.01)