Infections and their impact on waiting list survival in patients with end stage liver disease

L. Alferink, C.A.M Schurink, W.G. Polak, R.M. de Man, B.E. Hansen, H.J. Metselaar

Chair(s): dr. Marion J. Siebelink, programmamanager Transplantatie Centrum, UMC Groningen & dr. Raechel J. Toorop, transplantatiechirurg, UMC Utrecht

Thursday 10 march 2016

10:00 - 10:10h at Zaal 14

Categories: Parallelsessie (donatie)

Parallel session: Parallelsessie IX - Donatie en Allocatie

Background and aims:
Liver transplantation (LT) is the treatment of choice for patients with end-stage liver disease (ESLD). Unfortunately, there is a mismatch between donors and recipients, resulting in an increase of size of and mortality rate on the waiting list. In order to improve survival on the waiting list, better management of complications in ESLD leading to mortality, is needed. An important complication in cirrhotic patients is infection. We therefore assessed the impact of infections on survival on the waiting list in patients with ESLD, in context of waiting list dynamics.

Methods:
We performed a retrospective cohort study of all adult patients listed for LT from December 2006 to December 2013 in a single center in the Netherlands. We studied type, risk factors, impact and cumulative effect of infections on withdrawal (i.e. death or clinical deterioration) of patients from the waiting list. We therefore conducted several regression models, taking competing risks on the waiting list, LT and withdrawal, and infection as time-varying covariate into account.

Results:
We included 312 patients in this study. During follow-up 72/312 patients (23.1%) were withdrawn due to clinical deterioration or death. Infection occurred in 144/312 (46.2%) of all patients, some had multiple infections resulting in a total number of 317 infections. Infection resulted in withdrawal in 42/317 (13.2%). Site of infection, isolated micro-organism and primary liver disease were associated with infection-related withdrawal (p=0.001, p=0.004 and p=0.046 respectively). Hazard ratio (HR) for withdrawal in patients with infection versus no infection was 5.87 (95% confidence interval 3.56-9.66). The HR for withdrawal increased until the third infection. In a multivariate cox-proportional hazard analysis, with infection as covariate, etiology (comparing groups, p=0.043), age (HR1.07, p<0.001), MELD-score (HR1.1, p<0.001), serum albumin (HR0.97, p=0.001) and the presence of refractory ascites (HR2.21, p=0.013) were associated with withdrawal, taking competing risks into account.

Conclusions:
We show that infection is the number one cause for withdrawal from the waiting list, with a HR of almost 6, in patients with ESLD. There is a cumulative effect of having multiple infections until the third infection on withdrawal. Localization of infection and type of micro-organism in infections are associated with the risk for withdrawal, as is MELD-score, refractory ascites and primary etiology.