Discussion
ECMO technology, patient selection and its implementation has improved dramatically since the early studies. This has occurred due to better circuits, improved oxygenator technology, more experienced teams, and more informed decision-making regarding timing of implementation. This increase in the deployment of ECMO is evident in ELSO reporting, with studies increasing regarding factors impacting survival in different subsets of ECMO populations.
Interestingly, our data has shown no correlation between survival, and several factors that had previously been thought to contribute to mortality; the type of operation, operative times (including cross-clamp and cardiopulmonary bypass times), open chest status, reoperation for bleeding, the presence of frozen mediastinum and the timing of cannulation. Our data has shown that patients who are centrally cannulated may have a trend toward ability to wean from ECMO, with a non-significant trend toward short and long-term survival as well. One possible consideration may be more efficient venous drainage and cardiac offloading via central cannulation, as the timing of cannulation was not a key factor in survival based on our analysis.
One critique of the study could be the inclusion of VV patients with a VA ECMO cohort. We included these patients as they represented severe organ impairment in the postoperative period, and the majority of these patients were supported with RVAD with in-line oxygenator representing RV failure, a known indicator of mortality in postcardiotomy patients[7]. Though a small sample size, no statistical significance was noted between groups. Of note, analysis was also conducted for every comparison with and without the VV cohort included and there were no noted statistically significant differences in the data at short or long-term follow-up.
The use of ECMO in the post-cardiotomy population has been reported to have high morbidity and mortality with ability to wean from ECMO <50% and hospital survival ranging between 25-40%, and minimal data on long-term outcomes available. Our experience has demonstrated that PC-ECMO utilization may be associated with a reasonable rate of long-term survival. Overall survival data in our PC-ECMO group was: 61% were weaned, 55% survived hospitalization, and 45% were alive one month after discharge. This compares favorably to data from the ELSO registry (59% weaned, 43% discharged), and superiorly to ECPR support (41% weaned, 29% discharged).
This data also offers a glimpse into long-term survival for PC-ECMO patients. At 12 months, 40% of PC-ECMO patients were still alive, and at almost 3 years after discharge, 33% of patients were still alive. The estimated 12 month survival for patients surviving their index hospitalization was 66%, and survival was 60% at both 24 and 36 months, indicating a reasonable prognosis for long-term survival if patients reach hospital discharge. For reference, in renal failure patients receiving hemodialysis, the median survival at 3 years is approximately 57%[8-9]. Data exists for many chronic medical processes with five-year survival rates to use as a comparison (Figure 8). Heart failure, gastric and esophageal cancer, lung cancer and pancreatic carcinoma all have median survival at 5 years <30%[10]. More specifically, localized pancreatic adenocarcinoma carries a median 5-year survival of 34%, which is comparable to the overall survival of our PC-ECMO cohort[10]. Certainly, medical and surgical oncologists may advocate for the surgical resection of localized pancreatic adenocarcinoma, provided that the patient is otherwise an appropriate surgical candidate[11].
More robust data is needed to determine predictors for quality of life following PC-ECMO, which may yield improved decision-making strategies for ECMO implementation based on these individual patient characteristics, allowing physicians to predict which patients stand to benefit the most from ECMO cannulation from a quality of life standpoint.
In closing, our results demonstrate that PC-ECMO is a reasonable strategy to salvage patients who experience refractory post-cardiotomy cardiogenic shock, and has a 1-year and 3-year survival comparable to other surgically-treated diseases. Further data is needed to help characterize which patients may benefit most from ECMO implementation, but our experience has shown that post-cardiotomy ECMO support is no longer a bridge to nowhere.