LISBON – Conventional liver function tests may miss a diagnosis of nonalcoholic fatty liver disease in patients with type 2 diabetes, according to prospective study findings.
Data from the NAFLD substudy of the ongoing Edinburgh Type 2 Diabetes Study (ET2DS) found that although hyaluronic acid (HA) may be a reasonably good indicator of whether liver fibrosis was absent, standard enzyme tests missed a significant proportion of fibrosis cases.
"The literature that we have is biased by the fact that when researchers have looked at more advanced liver disease, it’s been in patients who have already earned themselves a liver biopsy for whatever reason," Dr. Rachel Williamson said in an interview at the annual meeting of the European Association for the Study of Diabetes (EASD).
"So what we add by this study is that this is an unselected population of patients who are essentially otherwise potentially well, from a liver point of view," said Dr. Williamson of the Western General Hospital in Edinburgh.
The primary aim of the ET2DS is to identify risk factors for memory problems in people with type 2 diabetes. The 5-year study is also providing the opportunity to study the natural history and risk factors for comorbid disease and complications associated with diabetes.
Liver assessment was performed in 939 of 1,000 individuals aged 60-75 years who were randomly selected from the Lothian Diabetes Register. The mean age of the cohort was 69 years, 52% were women, and 98% were white. Mean body mass index was 31.3 kg/m2, mean HbA1c was 7.2%, and the duration of diabetes was 9 years.
All patients in the liver cohort underwent abdominal ultrasound, standard liver function tests (LFTs), detailed screening of secondary causes of liver disease, HA level, platelet count, and alpha-fetoprotein measurements. Because HA is found in high concentrations in the synovial joints, patients’ history of joint disease was obtained.
"Liver disease is common in type 2 diabetes and this is mainly due to nonalcoholic fatty liver disease," Dr. Williamson said. "The prevalence of NAFLD in our population was 42%," she added, which is lower than the 70%-80% seen in some studies.
Despite the high prevalence of NAFLD in patients with diabetes, its detection is fraught with difficulty. Patients are often asymptomatic; conventional LFTs do not reflect changes in personal history; and liver biopsy, which remains the clinical standard for diagnosis, is invasive, expensive, and prone to sampling bias.
The quest to find reliable, noninvasive markers for advanced liver disease led Dr. Williamson and colleagues to examine the relationship between HA and the prevalence of hepatic fibrosis, portal hypertension, and hepatocellular carcinoma (HCC). Another noninvasive marker, the ratio of the platelet count to spleen size, was also used, and the usefulness of LFTs, alanine aminotransferase, aspartate aminotransferase, bilirubin, and gamma-glutamyltransferase were assessed.
An HA level greater than 50 ng/mL has previously been linked to liver fibrosis, with a higher cut-off of 100 ng/mL deemed to be more predictive of advanced liver disease. Using the lower threshold, 45% of the cohort had high HA levels, which could have been due to liver fibrosis in 24%. In all, 6% of the study population had HA in excess of 100 ng/mL. "We concluded that this 6% almost certainly had liver fibrosis," Dr. Williamson said.
The prevalence of portal hypertension, cirrhosis and HCC in the entire cohort was 1.1%, 0.4%, and 0.2%, respectively. Figures were slightly lower in patients who had a no secondary cause of liver disease (0.6%, 0.2%, and 0.3%).
Although mean levels of ALT, AST and GGT were highest in patients with liver cirrhosis, compared with those with raised HA (more than 50 ng/mL or more than 100 ng/mL) and no arthritis, steatosis, or normal liver scan, these levels remained within normal limits.
The positive predictive values of ALT and GGT above normal for predicting fibrosis were low (26% and 29%, respectively). The respective negative predictive values were 75% and 75%.
"The use of conventional liver function tests to screen for liver disease missed a significant proportion of cases of fibrosis predicted by raised HA levels, Dr. Williams concluded.
"A normal hyaluronic acid level can be quite reassuring," Dr. Williamson observed in the interview. She noted that although HA may only be measured in patients with known steatosis or abnormal LFTs, it could help to decide if referral to a gastroenterologist was needed, or if follow-up in the diabetes clinic was sufficient.
"I would routinely refer somebody with a significantly high hyaluronic acid onto the gastroenterologist, and we might then be considering doing liver biopsy or more invasive monitoring or further tests."