MODEL score predicts 1-year mortality after liver transplantation in patients with ACLF
Click Here to Manage Email Alerts
WASHINGTON — The Multi-Organ Dysfunction and Evaluation for Liver transplantation, or MODEL, score predicted high probability of mortality in transplant patients with severe acute-on-chronic liver failure, according to research.
“Acute-on-chronic liver failure (ACLF) is common and associated with high short-term mortality,” Ruben Hernaez, MD, MPH, PhD, associate professor of gastroenterology and hepatology at Baylor College of Medicine, said at The Liver Meeting. “We note that the global prevalence using the EASL study data is 35% prevalence. When we are looking at severe ACLF — that is, the presence of two, three or more organ failures — then the prevalence is about 56%.”
He added, “These are very sick patients — the mortality at 28 days for ACLF-2 is 38% and ACLF-3 72%. At this moment, liver transplantation is probably the only lifesaving procedure that can be done to save these patients’ lives.”
He noted that, despite “good outcomes,” LT is underutilized among patients with severe ACLF.
Aiming to develop a tool that predicted mortality within 1 year after LT in patients with severe ACLF, Hernaez and colleagues evaluated data from 15 LT centers across the United States from 2014 to 2019. Of 735 patients (median age, 55 years; median MELD-Na, 40) who underwent LT in the ICU, 70.8% had severe ACLF (ACLF-2, n = 237; ACLF-3, n = 284), and 19.9% of patients with severe ACLF died within 1 year of LT.
Using stepwise regression, researchers created the MODEL score, with a range of 0.08 to 1.14, and included several variables: age older than 50 years (OR = 2.24), one vasopressor at LT (OR = 1.8), more than one vasopressor at LT (OR = 4.05), respiratory failure (OR = 2.03) and multidrug-resistant bacterial infection before LT (OR = 1.9). The combined c-statistic of training and validation sets for the MODEL score was 0.71.
Researchers reported that a MODEL score greater than 0.43 predicted mortality at least 30% with a specificity above 87% within 1 year after LT, and a score greater than 0.66 predicted mortality at least 40% with a specificity above 94%.
Discrimination analysis showed the MODEL score outperformed the MELD-Na, MELD-lactate, donor-MELD, delta MELD and balance of risk scores at predicting short-term mortality among patients with severe ACLF following LT.