Noninvasive monitoring of liver fibrosis is an unmet critical need in cystic fibrosis (CF) liver disease. Liver biopsy is infrequently used in CF due to the patchy nature of liver involvement and is not suitable for longitudinal studies of progressive liver disease in CF. Indeed, unlike many other liver diseases, synthetic dysfunction is rare in CF liver disease. Portal hypertension due to hepatic fibrosis mediates most if not all of the complications seen in CF liver disease.

The Longitudinal Assessment of Transient Elastography in Cystic Fibrosis (ELASTIC-CF) study assessed if combining FibroScan® measurement of liver stiffness transient elastography with ultrasound would improve the prediction of the development of a nodular liver on ultrasound and development of portal hypertension over time in children and young adults with CF. Study visits occurred at baseline and for at least two annual follow-up visits.

This study is an ancillary study to the CFLD PUSH study. If requesting access to the ELASTIC-CF data, it is recommended to also request access to the CFLD PUSH data to obtain comprehensive information collected on study participants.

The specific aims and hypotheses for this study were:

1. To determine if transient elastography (TE), when combined with ultrasound (US) pattern characterization can improve the prediction of progression to a nodular pattern on US
• Hypothesis 1: TE results can differentiate children with a heterogeneous or homogeneous patterns on grayscale liver US who will progress to a nodular pattern on US
2. To confirm the feasibility of obtaining TE measurements in children with CF
• Hypothesis 2a: Valid TE measurements will be obtained in more than 90% of children with CF
• Hypothesis 2b: To determine the variability of TE measurements taken at different liver sites in the same patient
3. To prospectively assess whether TE data are associated with conventional laboratory markers of hepatic fibrosis
• Hypothesis 3: TE based values will correlate with conventional biomarkers of fibrosis/liver disease severity in CF: APRI, Fib-4, platelet count, and spleen size
4. To determine if TE with fat content can improve assignment of grade in situations where there is a dichotomy of radiology grades (i.e., 2HTG, 2NL)
• Hypothesis 4a: The combination of grayscale liver US image pattern assessment and TE values with fat content will differentiate between those individuals with dichotomous radiology grades (2HTG/2NL, 2HTG/2HMG, or 2HTG/2CIR) and risk of progression to a nodular pattern on US
• Hypothesis 4b: Higher fat content will be associated with a lower risk of progression to a nodular pattern on US
5. To determine if TE can predict the development of portal hypertension and its complications in children and young adults with CF and a nodular pattern on US
• Hypothesis 5: Among those children with a nodular pattern on US, TE results will correlate with clinical findings of portal hypertension (splenomegaly, thrombocytopenia, variceal hemorrhage, and/or ascites)
6. To pilot the correlation of TE and hepatic fibrosis and fat content in children and young adults with CF who have a clinically indicated liver biopsy
• Hypothesis 6: TE assessment of liver stiffness and fat content will correlate with steatosis and fibrosis staging on liver biopsy in CF

The primary outcome measure was liver stiffness measurement (LSM) obtained via transient elastography using FibroScan. The secondary outcome measure was liver steatosis obtained via transient elastography and controlled attenuation parameter (CAP) to quantify liver steatosis.

Inclusion criteria:

• Enrolled in the CFLD PUSH study at a center with transient elastography capability

Exclusion criteria:

• Presence of significant ascites
• Active medical device implant
• Open wound near sensor application site
• Pregnancy
• Unable or unwilling to give informed consent or assent
• Participant unable or unwilling to tolerate the TE measurement procedure

The use of LSM and CAP has demonstrated excellent feasibility, reproducibility between two anatomical sites, and strong associations with biochemical and nodular ultrasound findings of liver disease. Most of the ELASTIC-CF participants (54.1%) had a normal (NL) ultrasound pattern, and 13.5%, 12.8%, and 19.5% had heterogenous (HTG), homogenous (HMG), and nodular (NOD) ultrasound patterns, respectively.

Participants with a NOD ultrasound pattern had a significantly higher median LSM compared with NL, HTG, and HMG patterns. Furthermore, the study further expands on a previously noted association between increased LSM and decreased platelet count among participants with CF. This study showed that LSM was associated with an increase in ultrasound spleen-size-for-age z-score (SSAZ) and decreased platelet count, suggesting that an elevated LSM may be used as a surrogate for portal hypertension. In addition, the study found that CAP was significantly associated with ultrasound patterns, consistent with previous reports. Median CAP in the HMG group was significantly higher and above the previously reported pediatric steatosis cutoff.