Daniel M. Keller, PhD

November 28, 2012

BOSTON — A significant proportion of young children with cystic fibrosis (CF) have asymptomatic cirrhosis, according to the results of a large study.

Cirrhosis is an early event in CF, the researchers reported here at The Liver Meeting 2012: American Association for the Study of Liver Diseases 63rd Annual Meeting. They noted that it will be important to find early biomarkers to identify which children with CF are at risk for cirrhosis.

Cirrhosis is known to affect 5% to 10% of people with CF by the time they are 15 to 20 years of age, according to lead author Michael Narkewicz, MD, from the division of pediatric gastroenterology, hepatology, and nutrition at the University of Colorado School of Medicine and Children's Hospital Colorado in Aurora.

Researchers from the Cystic Fibrosis Liver Disease Network of 11 centers designed this prospective study to investigate whether ultrasound findings predict cirrhosis in children with CF. Dr. Narkewicz presented the baseline data from the study.

Three centrally trained radiologists, who were blinded to the clinical data, categorized the standardized Doppler ultrasound findings as normal; homogeneous (HMG) with a diffuse homogeneous increased echogenicity pattern; heterogeneous (HTG) with a heterogeneous echogenicity; cirrhosis in which the liver has a heterogeneous texture with nodular parenchyma and margins that include enlargement of the caudate lobe; or cirrhosis with portal hypertension.

Preliminary data suggest that in the absence of established predictive biomarkers for cirrhosis, the HTG pattern on abdominal ultrasound can predict risk.

The study population involved children 3 to 12 years of age with CF and pancreatic insufficiency. Children with known cirrhosis, Burkholderia positivity, or short bowel syndrome were excluded from the analysis.

Of 543 children with a final ultrasound read, 9.2% were categorized as HTG, 5.9% as HMG, 4.1% as cirrhosis, and 80.8% as normal. Severity of fibrosis increased with age; median age for HTG was 8.6 years and for both HMG and cirrhosis was 9.6 years (P < .001).

The 9.2% prevalence of the HTG pattern is lower than the 15% to 20% found in previous studies, Dr. Narkewicz noted.

The researchers did not find any genotype effect. Both the F508del-CFTR mutation (none, homozygous, or heterozygous) and the heterozygous G551D mutation in the CF transmembrane conductance regulator (CFTR) were about equally distributed among the normal, HTG, HMG, and cirrhosis groups. In other words, neither mutation appeared to affect the ultrasound pattern.

Also, there was no association between ultrasound pattern and meconium ileus at birth, Pseudomonas status, the method of diagnosis of pancreatic insufficiency, or the sweat chloride value.

Baseline for Longitudinal Follow-up

Children with HTG patterns were matched in a 1:2 ratio with those with normal patterns for age, Pseudomonas status, and center, and are being followed longitudinally to assess the development of cirrhosis. Children with HMG and cirrhosis patterns will be followed without matching. Most of the children who had a normal ultrasound pattern are not being followed.

Dr. Narkewicz reported some preliminary data from the patients who are being followed. This cohort (n = 142) consists of HTG children and their matched controls, plus HMG and cirrhosis children.

"There are no differences in genotype, weight, height, body mass index z-scores, nutritional status, or CF-related diabetes in these groups. However, they are very small numbers," Dr. Narkewicz said. "Interestingly, none of the individuals who were identified with a cirrhotic pattern on ultrasound had been told that they had any liver issue, but the group of individuals in the cirrhosis group were more likely to have higher [alanine aminotransferase] and/or [gamma-glutamyl transpeptidase] than the others." The HMG group was more likely to be told that they had more liver disease, he reported.

Dr. Narkewicz noted that this is the first large study of liver ultrasound patterns in children with CF to use standardized definitions and consensus radiology gradings. The lower prevalence of the HTG pattern in this study might be due to the "more consistent definitions and the actual requirement that we have for consensus," he explained.

Despite the lower rate, "CF cirrhosis is an early event for at least half of the children that we know will eventually develop cirrhosis in CF," he added.

That is the key message for clinicians from the study, he told Medscape Medical News. "Almost all of these occur before 10 years of age. We started screening at 3 years of age, and there are abnormalities as early as 4 [years]. If we're going to do something about this, either with an interventional trial or screening, it's going to have to be targeted to...children 2 to 6 years of age," Dr. Narkewicz advised. "If we wait until 10, the majority of them will have well-established cirrhosis with portal hypertension, and none of us think we can reverse that."

Because cirrhosis occurs in a minority of children with CF, a very good predictor of risk for a severe outcome is needed. That way, children at high risk can be enrolled in clinical trials of potential disease-modifying interventions. Liver ultrasound is one such potential predictive tool, Dr. Narkewicz said

Benjamin Shneider, MD, professor of pediatrics and director of the Pediatric Hepatology Center at the Children's Hospital of Pittsburgh, Pennsylvania, who was not involved in the study, told Medscape Medical News that this study used "a definition of ultrasound findings that they labeled as cirrhosis...[but that] cirrhosis per se is typically a histologic diagnosis that would require a biopsy." Therefore, he said, the ultrasound findings are suggestive of cirrhosis but not definitive.

Dr. Shneider noted that almost 5% of unselected children with CF had findings indicating they might have cirrhosis. "That's a worrisome finding," he said, adding that cirrhosis in these children is typically asymptomatic. "That's the problem with this disease," he explained. The first symptom might be a very severe problem, like gastrointestinal bleeding or varices.

The study provides a couple of clinical implications, according to Dr. Shneider. It is possible that significant liver disease will be found earlier. In addition, if you can identify children who are at increased risk for liver disease, you can select patients who are candidates for some therapeutic intervention, he said.

At this point, the study just provides the baseline for a possible predictive marker. There is no specific intervention beyond current standard therapy if such a marker is found and validated, but the information could be used to give "anticipatory guidance to families," such as watching for features of portal hypertension or for signs of gastrointestinal bleeding, he explained.

There was no commercial support for the study. Dr. Narkewicz reports owning shares in Merck, receiving research support from Novartis and Vertex, and being a consultant to Vertex. Dr. Shneider reports being a consultant to Ikaria.

The Liver Meeting 2012: American Association for the Study of Liver Diseases (AASLD) 63rd Annual Meeting. Abstract 29. Presented November 11, 2012.