Blood tests predict mortality following liver resection

Simple blood tests predict mortality following liver resection

Last Updated: 2010-11-23 13:35:26 -0400 (Reuters Health)

By Gabriel Miller

NEW YORK (Reuters Health) - A set of three simple preoperative tests may predict in-hospital death following hepatectomy, according to a study published in the November issue of Archives of Surgery.

Many previous studies have identified factors postoperatively that contribute to death after liver resection, the investigators say. However, their study has identified a set of tests done before surgery that predict mortality and can be used to decide who should -- and should not -- have the operation immediately.

"Our experience showed that preoperative liver function tests and coagulation as well as evaluation of liver capacity predict the mortality rate in patients requiring right-sided major or an extended major liver resection with extra-hepatic bile duct resection," lead author Dr. Elie Housseau-Oussoultzoglou at Hautepierre University Hospital in Strasburg, France told Reuters Health in an email. "All these biological parameters are easily available and measurable preoperatively in a simple blood sample."

Though the study found several pre-operative measures that were predictive of mortality, hepatic resections are extremely complicated procedures in a widely varying patient population. Considering the high mortality rate already established with these operations, it 's not surprising that a retrospective review would uncover a few significant measures, said Dr. William Jarnagin, chief of the Hepatopancreatobiliary Service at Memorial Sloan-Kettering Cancer Center in New York City.

"These are big operations and they're performed on patients with advanced cancers so there is a lot of morbidity," said Dr. Jarnagin, who was not involved in the study. "It's complicated to analyze (and) there are a lot of factors, so to pick one or two things, even though they might be significant, is not a simple thing at all."

Dr. Jarnagin added that several of the pre-operative tests the authors suggest are already done on a routine basis by many clinicians.

The study retrospectively reviewed the preoperative evaluation of 67 patients without cirrhosis undergoing liver resection of at least four contiguous liver segments.

More than half of the patients (n=35) had co-morbidities prior to their operation, most commonly hypertension and cardiovascular disease.

The main outcome measure was postoperative mortality, which included intraoperative death, death within 90 days after surgery and in-hospital death.

In the series, all of the postoperative deaths (7%) were related to liver failure, which was mainly associated with the extent of liver resection and the inability of the liver to regenerate, the investigators report`.

On univariate analysis, five factors were associated with higher postoperative mortality: serum alanine aminotransferase (ALT) level > 40 U/L, preoperative prothrombin (PT) ratio less than 70%, a preoperative indocyanine green retention rate at 15 minutes (ICGR-15) greater than 15%, preoperative biliary drainage, and extrahepatic bile duct resection.

Based on the results of their multivariate analysis, the authors propose that patients who have two or more risk factors based on ICGR-15, PT ratio, and ALT level should not have surgery. Surgery can proceed in patients with one risk factor, they suggest, although mortality rates will be higher than 5%.

"In patients with altered preoperative liver tests and function, the initially planned surgery should be postponed, until full recovery of the biological parameters," Dr. Elie Housseau-Oussoultzoglou said. "During this interval a palliative anti-tumoral therapy is administered to avoid disease progression.

"Considering the high expected risk of lethal liver failure it would be preferable to survive few months with a chronic disease instead (of dying) within a thirty-day hospital stay," Dr. Housseau-Oussoultzoglou added.

SOURCE: http://link.reuters.com/tab86q

Arch Surg 2010;145:1075-1081