Premortem Heparin Administration and Location of Withdrawal of Life-Sustaining Interventions in DCD

The meta-analysis by Cao and colleagues1 pools data from retrospective cohort studies to assess the importance of location of withdrawal of life-sustaining therapies (WLST) (intensive care unit [ICU] vs operating theater [OT]) and premortem heparin administration in liver transplantation after donation after cardiocirculatory determination of death (DCD). The authors assert that “withdrawal in the OT and premortem heparin administration improve DCD liver transplant outcomes.” These conclusions are overstated and not sufficiently reflective considering the low quality of data: None of the studies actually compared outcomes based on location of WLST or premortem heparin administration. Rather, they were single-center comparisons of transplantation outcomes in DCD versus donation after brain death (DBD). Studies were performed over a span of longer than 30 years, in multiple countries, at centers with varying degrees of experience with DCD, and with substantial heterogeneity in cold and warm ischemic time (WIT). An association from observational studies does not prove causation, particularly in an unadjusted univariate analysis. Such data are, at best, hypothesis-generating. Outcomes were consistently superior with transplantation after DBD, irrespective of the location of WLST and use of heparin. Even if the odds of poor outcomes with DCD versus DBD were numerically greater when WLST occurred in ICU, or when heparin was withheld, no comparative analyses were presented in the article. For the data in Figures 31 and 51 (allograft and patient survival, respectively), we have generated mixed effects models to compare studies where WLST occurred in the ICU versus OT. In neither case were the results statistically significant (P = 0.58 and 0.13, respectively). Similarly, for data presented in Figure 71 (primary graft nonfunction), there was no significant difference based on whether or not premortem heparin was administered (P = 0.21). We have expanded the authors' analysis to assess allograft failure based on whether or not heparin was administered (data not presented in the manuscript). The odds ratio for graft failure in DCD versus DBD was 1.7 (1.3-2.3) with heparin and 1.6 (0.6-4.6) without (Figure 1; P = 0.92).FIGURE 1: Comparison of 1-year rates of allograft failure in liver transplantation following donation after cardiocirculatory death vs. donation after brain death, categorized by use of premortem heparin.Minimizing WIT is important, but existing literature does not demonstrate that WLST should occur in the OT. Although WIT may be slightly longer, the ICU environment is usually much better suited to provide end-of-life care to organ donors and their families. Despite the rationale for premortem anticoagulation, there is little human data to support the need for it. Outcomes have proven favorable in jurisdictions where premortem heparin is not used.2,3 The impact of anticoagulation in DCD may vary depending on the organ system.3–5 There is essentially no research assessing optimal dosing of heparin in DCD, which is typically much higher than what is required for therapeutic anticoagulation (>300 units/kg vs ≈ 80 units/kg). There is justifiable concern that such high-dose heparin could worsen intracranial hemorrhage in brain-injured patients. Heparin is commonly administered before WLST, even though a sizable proportion of potential donors do not die within the requisite time frame. To avoid this scenario, some centers wait to administer heparin until it is clear that the patient will actually become an organ donor. The work of Cao and colleagues highlights the need to better study these issues, rather than justifying a call for practice changes that are unsupported by existing data and may compromise optimal end-of-life care in DCD donors.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.