Sensitivity analysis
Patients with an underlying diagnosis of COPD, emphysema, bronchiectasis, lymphangioleiomyomatosis and pulmonary hypertension (n = 107) were excluded, since none of them was placed on VV-ECMO (to avoid aliasing between diagnosis and ECMO support), resulting in 21 BTT and 169 non-BTT patients in the sensitivity analysis (Table 5). Perioperative characteristics (Table 5) and postoperative outcomes (Table 6) in BTT and non-BTT patients were similar to those in the entire cohort. After matching, both subgroups were well-balanced with respect to the age, low platelet count, serum creatinine and prevalence of CF, while imbalance remained regarding the proportion of men, BMI, and preoperative hemoglobin levels (Table 7). Intraoperatively, CPB was used less often, while the use of ECMO was considerably higher in BTT compared with non-BTT patients (Table 7). Need for postoperative ECMO (62.0% vs. 8.7%), delayed chest closure (16.5% vs. 5.6%), tracheostomy (50.8% vs. 34.4%), chest infection (60.8% vs. 41.2%) and AKI requiring RRT (45.7% vs. 30.5%) were more common in BTT than in non-BTT patients, while the two subgroups were similar in respect to 30-day mortality, surgical re-exploration, chest drainage within 24 hours, sepsis, stroke and 1-year mortality (Table 7). With further adjustment for the unbalanced covariates (gender, BMI and hemoglobin level), VV-ECMO as a BTT was associated with 12.8-fold higher odds of need for postoperative ECMO and with 6-fold higher odds of tracheostomy, but it did not appear associated with any other early and mid-term outcome (Table 8).
DISCUSSION
The use of VV-ECMO as a BTT can allow patients with decompensated end-stage lung disease to remain eligible for LTx and offer a viable strategy for improving their post-transplant survival outcomes. In this study, we reported our single-center experience with 297 transplanted patients, 21 (7.1%) of whom were bridged to LTx with VV-ECMO. The most common diagnosis in both BTT and non-BTT recipients was CF. One of the reasons is that there is a well-established CF Unit in our institution which attracts tertiary referrals from the whole country. In the primary analysis, both 30-day and 1-year posttransplant mortality were considerably higher in patients requiring VV-ECMO as a BTT than in non-BTT patients. In addition, the incidence of the most important early postoperative complications, including need for ECMO, delayed chest closure, surgical re-exploration and AKI requiring RRT, was significantly increased in the bridged patients.
To minimize potential effects of selection bias and decrease variability of both groups, we performed further analysis comparing matched groups which were well-balanced in terms of preoperative recipients’ baseline characteristics. Importantly, after matching, we observed a similar 30-day mortality between the BTT and non-BTT patients (4.6% vs. 6.6%,p =0.083) despite a higher incidence of early postoperative complications (need for ECMO, delayed chest closure, AKI requiring RRT), while the 1-year mortality was even lower in the BTT patients (8.0% vs. 15.6%, p =0.238). Furthermore, when evaluating the effect of preoperative VV-ECMO on postoperative outcomes, it did not appear associated with 30-day or 1-year mortality. Moreover, in the sensitivity analysis, the two subgroups were similar in respect to 30-day (BTT 7.8% vs. 6.5%, p =0.048) and 1-year mortality (12.5% vs. 18%,p =0.154). The clinical condition of patients who were bridged to LTx with VV-ECMO is usually more critical than that among the rest of the patients who were not bridged, and this may negatively influence their outcomes. However, in our experience, post-transplant survival in bridged patients was comparable to that in patients who did not have pre-transplant VV-ECMO. Therefore, VV-ECMO has been demonstrated to be a valuable supportive strategy to prolong life in these critically ill patients while increasing the waiting period for suitable organs. Our early and mid-term results are in general consistent with previous reports that have shown no significant difference in post-transplant survival among BTT and non-BTT patients, especially in high-volume centers.4,10-12,24-29 Surprisingly, we have found that 1-year mortality was even lower in the BTT group but this might be related to several other factors. One of the reasons could be that the average duration of pre-transplant support with VV-ECMO in our cohort was relatively short (8 days) and this could positively affect the outcomes. As recently reported by Crotti et al. , patients who underwent LTx after a waiting period longer than 14 days had significantly higher rates of post-transplant mortality and morbidity.30 Furthermore, shorter waiting times after urgent listing have likely contributed to these favorable outcomes. In addition, we have observed more commonly intraoperative ECMO than CPB among BTT patients when compared to the non-BTT group, and it is well known that the intraoperative use of ECMO might have several advantages.31,32 We believe that the improved survival among BTT patients can be also related to an increased experience with this strategy, early ambulation of these patients, advancement in the perioperative care, and development of an experienced ECMO and multidisciplinary team.
On the other hand, Schechter et al. have reported a decreased 1-year post-transplant survival among patients requiring preoperative support including ECMO with MV.3 However, they have demonstrated in a multivariable analysis that ECMO alone was not associated with decreased 1-year survival.3 In our study, 38.1% of patients were supported using both VV-ECMO and MV before LTx, but the sample size was too small to perform a further analysis whether MV could have had any effect on postoperative outcomes. Furthermore, Mason et al. have reported that survival after LTx is markedly worse (1-month and 1-year post-transplantation survival were 72% and 50%, respectively) when preoperative mechanical support is necessary, although they suggested that additional risk factors for mortality should be considered when selecting patients for LTx in order to improve survival.2 In addition, Fischer et al. have reported that the perioperative mortality of LTx after preoperative ECMO can be even up to 60%.9
As expected, need for postoperative ECMO, delayed chest closure, tracheostomy, chest infection, and AKI requiring RRT were more common among BTT patients. This can be related to the common and well-known risks related to the use of ECMO such as bleeding complications, systemic inflammatory response, acute kidney injury and thromboembolic complications.33-35 However, the rate of these complications was lower than in some of the previous reports that demonstrated an incidence of tracheostomy in up to 77%36, delayed chest closure in 50%37 and stroke in 8%36 of recipients. Still, in our study it seems that both 30-day and 1-year survival have not been negatively affected by the increased incidence of early postoperative complications.
The strength of this study is the comparison of two cohorts of patients (BTT and non-BTT) that were matched. However, we acknowledge several study limitations. First, the analysis was performed retrospectively and designed as a single-center study, although the study period was up to 7 years and included moderate sample size with 1-year follow-up. The present study also lacks donor data as we were not able to collect these data for the whole study period. In addition, it would be interesting to expand the research and study primary graft dysfunction and rejection rate as we did not have this data. Further studies with long-term follow-up would be useful in order to analyze occurrence of late complications. Lastly, we were not able to extend our analysis including patients bridged with other devices (VA-ECMO, Novalung) due to a small sample size and different clinical characteristics some of these patients.