Discussion
Our data showed that in patients affected by PCS following cardiac surgery, the duration of VA-ECMO support is associated with increased mortality. Patients weaned from VA-ECMO after 4 to 7 days had significantly lower mortality compared with patients with shorter or longer mechanical support, even when adjusted for confounding. In addition, VA-ECMO support longer than 7 days was associated with a significantly increased risk of complications, including re-exploration for bleeding, blood transfusion, renal failure requiring renal replacement therapy, deep sternal wound infection, bloodstream infection and pneumonia.
Evidence from the Extracorporeal Life Support Organisation (ELSO) registry also showed that short duration of VA-ECMO is associated with high mortality. In this large registry that included 2699 VA-ECMO patients, survival increased up to day 4 and then decreased from day 4 to 12, with no significant change thereafter [1]. However, the study encompassed a mixed cohort of patients with only a minority of them undergone cardiac surgery, a limitation shared with previous studies [1-3,8,9]. On the other hand, the absence of clear guidelines on VA-ECMO weaning is the testament that this aspect is still poorly addressed [1-3]. In addition, data on survival with longer VA-ECMO runs are limited [1,3,4,9]. To our knowledge, the present analysis is the largest to date in evaluating the impact of VA-ECMO duration in adult patients affected by PCS following cardiac surgery. Distelmaier et al. [4] firstly addressed the impact of VA-ECMO duration on survival in 354 cardiovascular surgery patients, observing that longer VA-ECMO runs were associated with higher mortality even 2-years after hospital discharge [4]. More recently, Wang et al. [3] enrolled 166 PCS patients following coronary bypass surgery. More than 60% of patients received VA-ECMO for 3–6 days and had significantly lower mortality than those who were supported by VA-ECMO for < 3 days or ≥ 7 days [3].
Our data are consonant with previous studies, suggesting that in PCS patients following cardiac surgery VA-ECMO support longer than 7 days can be challenged, considering the associated early and late higher mortality. In this cohort of patients, the risks of complications appear to overcome the cardiopulmonary advantage exerted by the VA-ECMO support. Bloodstream infections have been demonstrated to be associated with longer VA-ECMO runs, occurring in 27.7% of treated patients [11,12]. Therefore, it is not surprising that longer VA-ECMO runs are associated with a higher risk of bloodstream infection along with an increased rate of blood transfusions and organ failure [11,12]. Among other complications, administration of large volumes of blood transfusion and renal failure requiring renal replacement therapy are potentially fatal conditions in longer VA-ECMO runs, particularly in PCS patients with an underlying severe cardiac dysfunction [13,14]. Similarly, our data confirmed that shorter VA-ECMO runs (≤3 days) are also associated with significantly higher mortality. Although we did not detect a higher rate of lethal haemorrhage in this patient group that has been previously suggested as main cause of the increased early mortality [15,16], the underlying primary cardiac condition seemed to play a major role in the survival of those patients [3,17]. Cardiopulmonary failure leading to multiorgan failure appeared to predominate over ECMO treatment. The hyperlactatemia observed in patients under VA-ECMO ≤3 days suggest a significant metabolic derangement in these patients. In this context, arterial lactate level may be useful in guiding the appropriate timing of VA-ECMO discontinuation, thereby avoiding futile prolonged support [19].
The results observed in our series are relevant considering the unsolved issue of balancing a fruitful VA-ECMO duration against a vain support especially in light of the uniquely high level of resources involved [19]. In addition, due to the lack of defined guidelines and indications, the duration of ECMO support is often based on arbitrary limits [1]. Data derived from ELSO registry over a 10-year period indicates that 52% of patients on VA-ECMO are discontinued from support because of irreversible organ failure [1]. Therefore, when cardiopulmonary recovery cannot be successfully achieved within 7 days, other therapeutic options should be considered, including ventricular assist device implantation or heart transplantation [1,7,19].
Certainly, our study is not exempted from limitations. First, our series is subjected to the limitations of all observational analyses, including selection bias and unmeasured confounding. Second, the present analysis is conditional to in-hospital survival only, and our data do not allow an assessment of the outcomes after weaning and discharge from the hospital. Third, a trend in the survival of patients with very long ECMO duration (> 15 days) may not be fully detected due to the small number of remaining individuals, with insufficient statistical power. Lastly, we cannot account for the surgeon and anaesthetist’s experience as well as for the differences in local policies of ECMO weaning. Despite these limitations, our cohort is currently the largest in evaluating the impact of VA-ECMO duration in the PCS setting.
In conclusion, in PCS following cardiac surgery, patients weaned from VA-ECMO after 4 to 7 days of support had significantly lower mortality compared with those with shorter or longer mechanical support. The present data can contribute to identifying the most ideal duration of VA-ECMO support, supporting clinicians in deriving more accurate prognostic models and timely weaning strategies.