Powikłania związane z pomostowaniem do przeszczepienia płuc – systematyczny przegląd i metaanaliza

Introduction

Lung transplantation (LuTx) is a life-saving surgery performed in patients with end-stage respiratory diseases, such as chronic obstructive pulmonary disease, cystic fibrosis, or pulmonary fibrosis. It is considered as a last-chance treatment when there are no benefits from conventional therapy [1]. Being on the waiting list is associated with deconditioning, which in some cases may require mechanical ventilation (MV). However, MV may provide insufficient gas exchange and it increases the risks of ICU-related complications [2]. Deconditioning may lead to death or worse outcomes after transplantation, due to chest muscle weakening, infections, or other complications. The introduction of extracorporeal membrane oxygenation (ECMO) may prevent deconditioning and ensure better outcomes after transplantation (possible physical therapy and enteral nutrition). ECMO supports cardiac or lung function, depending on the cannulation strategy, and may be applied in veno-venous or veno-arterial modes. Despite the difficulties of conducting ECMO support and its costs, the use of this support device is associated with several complications, such as bleeding, clotting or air bubbles in the cannulas, or acute kidney injury [3].

Aim of the research

The aim of this meta-analysis is to determine the frequencies of ECMO-related complications during bridging to lung transplantation.

Material and methods

This analysis was conducted using the PRISMA protocol (preferred reporting items for systematic reviews and meta-analyses) [4]. A thorough literature search was performed in the Medline and Google Scholar databases. The following terms were used in the search: “Bridging to lung transplantation” and “Bridging to lung transplantation – complications”. Inclusion criteria were composed of retrospective or prospective articles in English presenting experience with extracorporeal life support as a bridge to lung transplantation. Studies with or without a control group were included, while complications occurring during the bridging procedure had to be reported. Two independent reviewers screened the titles and abstracts, and the data were extracted manually. Forest plots were generated using Open Meta Analyst software, and heterogeneity was assessed with I2 statistics.

Results

After a primary search, we identified 1141 results. Ultimately, 13 studies met the inclusion criteria. The search strategy is presented in Figure 1. The characteristics of the included studies together with bridging and short-term LuTx outcomes are presented in Table 1. The included studies were published between 2011 and 2020. Out of 601 patients enrolled to a bridging procedure, 461 underwent lung transplantation. The mean time of extracorporeal life support was 12.4 days. Bleeding and thrombotic complications occurred in 24.6% of the included patients. Cerebrovascular events happened in 4.5% of patients, and acute kidney injury occurred in 13.2% of the included patients (Figures 2, 3).

Discussion

Veno-venous ECMO (VV ECMO) supports respiratory function only while veno-arterial EMCO (VA ECMO) cannulation supports both cardiac and lung function. It can be used as a rescue therapy, for instance, in severe COVID-19 [18]. Another use of this device is to bridge patients to lung transplantation. Deterioration while on a waiting list may cause death before matching with an appropriate organ. The first case of ECMO used as a bridge to lung transplantation was performed in 1977. However, after initial trials it was considered that ECMO is associated with worse outcomes and increased risk of infections and bronchial ischaemia. The first successful reports of bridging to LuTx were performed in the early 1990s [19]. Pulmonary deterioration with hypoxia or hypercapnia despite medical treatment, haemodynamic failure, or pulmonary hypertension are indications to place a patient on ECMO support. On the other hand, not qualifying for LuTx, bacteriemia, or failure of other major organs, among others, are considered as absolute contraindications [20] (Table 2).
The use of ECMO itself is associated with several complications. Exposure of the blood to the non-biologic surfaces of the cannulas may lead to adherence of blood cells to the surfaces, platelet activation, and fibrinogen deposition. These outcomes may lead to thrombotic complications, in turn leading to a cerebrovascular event. Therefore, the surfaces of the cannulas are covered in heparin, together with significant intravenous amounts of heparin introduced in anticoagulation protocol. An adequate anticoagulation strategy allows for the balance between thromboembolic and bleeding risks. However, use of heparin may result in heparin-induced thrombocytopaenia (HIT), which develops in approximately 0.5% to 5% of patients. The diagnosis of HIT while on ECMO support is challenging due to platelet consumption and various factors that may contribute to a platelet drop while on a circulatory support device, which might mimic HIT [21]. Alternatively, direct thrombin inhibitors may be introduced (bivalirudin) to prevent HIT. Anticoagulation is monitored using the actual clotting time (ACT), aPPT, PT, or anti-Xa. ACT is available at the bedside and is the most frequently used option with a target range of between 150 and 170 s. However, the results of ACT may be influenced by heparin-independent factors such as hypothermia, thrombocytopaenia, or anaemia [22]. ACT was designed to monitor heparin anticoagulation during cardiopulmonary bypass support, and aPTT has been used to monitor heparin in prophylactic ranges. aPTT is considered as more reliable, but studies comparing ACT and aPTT during ECMO support are inconclusive [23]. According to a recent meta-analysis conducted by Willems et al., anti-Xa-factor versus ACT and aPTT during ECMO is associated with fewer incidents of bleeding, while no increase in thrombotic events was observed [24]. Use of ECMO may also contribute to the development of renal injury and the need for renal replacement therapy. Constant inflammatory processes that take place while on ECMO or due to underlying disease, together with coagulation abnormalities, use of nephrotoxic substances, and ischaemia reperfusion injury may lead to acute kidney injury. Additionally, other complications that occur during ECMO, such as bleeding or thrombosis, might also increase the risk of kidney failure [25]. Renal replacement therapy is thought to prevent metabolic acidosis and hypervolaemia [26]. The occurrence of some complications depends on the cannulation strategy. The harlequin effect is a rare complication that might appear in peripheral VA ECMO support when blood from the native pulmonary circuit returns to the aorta together with oxygenated blood reinfused by the centrifugal pump of ECMO. Consequently, saturation in the lower parts of the body would be normal, while in the right hand it can be as low as 40%. Therefore, hypoperfusion of the upper body together with cerebral and coronary hypoperfusion may develop [27]. Limb ischaemia is another complication that may develop in femoral cannulation. The potential mechanisms include catheter thrombosis, emboli, or obstruction of flow by cannula [28]. Despite such complications, use of extracorporeal support devices is beneficial for some patients. As a rescue therapy, withholding ECMO would cause further deterioration, delisting from lifesaving LuTx and death. As seen in Table 1, most bridging procedures were successful and allowed for subsequent transplantation. This study cannot be considered without several limitations. Firstly, 11 out of 13 of the included studies were retrospective, which introduces the risk that some of the complications could have been underreported. Secondly, patients were not divided into separate groups based on underlying diseases. In addition, in some patients VV ECMO had to be changed to VA ECMO, which includes full anticoagulation (Figure 4).

Conclusions

Bridging to lung transplantation is a procedure that allows critically ill patients to undergo a lifesaving procedure. Our analysis proves that the frequencies of complications that may develop during bridging are non-negligible and may lead to death instead of surgery. Further studies should be focused on improving anticoagulation guidelines.

Conflict of interest

The authors declare no conflict of interest.

References