Secondary Logo

Journal Logo

Original Article: Pancreatology

Extended Screening for Cystic Fibrosis-related Liver Disease Including Elastography in Children and Adolescents

Højte, Christine; Jørgensen, Marianne Hørby; Jensen, Flemming; Katzenstein, Terese L.§,||; Skov, Marianne

Author Information
Journal of Pediatric Gastroenterology and Nutrition: November 2020 - Volume 71 - Issue 5 - p 663-668
doi: 10.1097/MPG.0000000000002872
  • Free

Abstract

What Is Known/What Is New

What Is Known

  • Cystic fibrosis-related liver disease is the third most frequent cause of death among cystic fibrosis patients.
  • No specific test for diagnosis of cystic fibrosis-related liver disease exists.
  • Cystic fibrosis-related liver disease is primarily believed to be a biliary disease.

What Is New

  • Only including gamma glutamyl transferase as a good marker for cystic fibrosis-related liver disease.
  • Ultrasound examination of cystic fibrosis patients should be combined with FibroScan or Shear Wave elastography.
  • FibroScan is easy to perform and could be preferred for follow-up.

Cystic fibrosis-associated liver disease (CFLD) is the third most frequent cause of death among CF patients (1). The pathogenesis remains unclear and why only some CF patients develop cirrhosis, whereas others with similar mutation are unaffected, is still an enigma. Focal biliary fibrosis is clinically the most significant hepatic lesion associated with CF (2). Some reports have shown an emerging burden of milder liver disease in the adult population (3,4) whereas other longitudinal studies report a cumulative incidence ranging from 27% to 35%, without incident cases after the age of 18 years (1,5), thus, screening for liver disease (LD) in children and adolescents is crucial.

Histological assessment, which is considered gold standard for diagnosing liver disease, is of questionable accuracy in CFLD because of the focal distribution of the fibrosis (6). The classic recommended diagnostic tests for CFLD (Debray and Colombo) include at least 2 of the following: 1, clinical hepatomegaly (confirmed by ultrasound); 2, abnormal liver enzymes (transaminase above upper limit at 3 consecutive determinations over 12 months); and 3, ultrasound (US) abnormalities other than hepatomegaly (1). A recent broader definition including signs of cirrhosis, portal hypertension, or elevated GGT has been suggested (7). An echographic scoring system has been added for patients with CF (8) but not sufficiently validated and never included in the criteria. Elastography is based on ultrasound and assess stiffness of the liver as an indicator of liver fibrosis (9,10). Recent studies of elastography in paediatric patients with either Vibration-Controlled Transient elastography (FibroScan) or Real-Time Shear Wave elastography (SWE) are promising. Combining biochemical markers with either SWE (11) or FibroScan (12) may improve early identification of CFLD.

Our aim was to study CFLD in a pediatric population using an extended screening program of liver biochemistry and noninvasive methods in order to suggest a prioritized program in future screening of CFLD.

PATIENTS AND METHODS

Study Population

Between March 2014 and November 2015, all CF patients ages 1 to 18 years treated at The Copenhagen CF Center (95 out of 187 pediatric CF patients in Denmark) were invited to participate. Children younger than 1 year were excluded as execution of FibroScan and SWE require cooperation. The CF diagnosis was based on characteristic clinical features, positive sweat test, and CFTR genotype (13). Patients were classified as pancreatic sufficient or insufficient and presence of CF-related diabetes (CFRD) was recorded.

Methods

The patients were screened for liver disease using physical examination screening for signs of cirrhosis, biochemical tests, conventional US, and supplementary to the previously used program, FibroScan and SWE. All examinations were performed within a period of 6 months in each patient.

Biochemical tests included haematology tests (haemoglobin, transferrin, transferrin saturation, leucocytes, platelets), liver tests (alanine aminotransferase [ALT], aspartate aminotransferase [AST], alkaline phosphate [ALP], gamma glutamyl transferase [GGT], albumin, bilirubin, INR), and total bile acid. The AST-to-platelet ratio index (APRI) was calculated as described by Leung et al (14). Patients were screened for alpha-1-antitrypsin deficiency (A1AT-deficiency) on both the Z and S allele of SERPINA1. They were further screened for autoimmune diseases (antinuclear antibody [ANA], anti-neutrophil cytoplasmic antibody [ANCA]) and cancer (α-fetoprotein). Ursodeoxycholic acid (UDCA) treatment was paused at least 14 days before blood tests. Reference values were according to our laboratory (15).

Conventional ultrasound was performed using a Philips Healthcare Epiq 7G ultrasound system with a C1-5 Broadband Curved Array transducer by standardised method. Ultrasonographic evaluation describing liver size, hepatic texture, steatosis, gallbladder abnormalities, biliary sludge, and splenomegaly was obtained. Presence of regenerative liver nodules and enlargement of liver segment 1 (caudate lobe) were noted. The liver was sonographically classified by the Williams Ultrasound Scoring System (WUSS), which evaluates 3 characteristics of CFLD; coarse liver parenchyma, nodularity of the liver edge, and increased periportal echogenicity (8). Patients were allocated into 3 groups: no liver disease (NLD) = score 3; intermediate liver disease (ILD) = scores 4 to 6; and liver disease (LD) = scores 7 to 9 (8).

Doppler analysis was performed to confirm patency of the splanchnic vessels and hepatic veins to evaluate the blood flow within them.

The SWE was done using the integrated Philips Healthcare Elastography on an EPIQ 7 ultrasound machine (Bothell, WA; ElastPQ) with the C1-5 Broadband Curved Array Transducer. In total, 10 validated measurements were taken and mean, standard deviation, minimum and maximum elastography values were calculated by the system and given in m/s.

Vibration-Controlled Transient elastography measurements were performed using the FibroScan 502 (Echosens, Paris, France) [for technical details see Nobili et al (16)] and probes S1, S2, M were selected according to the child's age (<6 years S1, 6–14 years S2, >14 years M) and chest circumference (<45 cm S1, 45–75 cm S2, > 75 cm M) (9). Only procedures with a success rate of 75% and a ratio of the interquartile range and median (IQR/median) ≤25% were considered reliable.

Intra-and Interobserver Variability

All physical examinations and FibroScans were performed by the same observer and all measurements of SWE and WUSS were performed by 1 of 2 experienced senior radiologists. The observers were blinded to other results (scans or tests) at the time of examination.

Statistics

Biochemical results were compared using linear regression. Differences in age, z-height, z-weight, z-BMI, and FEV1% were analysed by an independent 2-sample t-test of unequal variances (F-test was performed to determine variances). Biochemical and sonographic results were compared by linear regression if both variables were continuous; with the variables being categoric 1-way ANOVA was used. Forward multiple linear regression was used to investigate if a combination of explanatory variables can predict 1 dependent variable. Only variables that raise significance in the univariate analysis were included in the multiple regression analysis.

Proportions of sonographic abnormalities, abnormal liver function test, and manifest CFLD were compared using a 2-proportion-z-test or Fisher exact test for small samples. P values <0.05 were considered as significant.

Ethics

Written informed consent was obtained from parents/caretakers and patients above the age of 15 years. The study protocol was approved by the Danish National Committee on Health Research Ethics. Protocol no.: H-4-2011-042.

RESULTS

Eighty-five consecutive patients were included (Table 1). One patient was excluded because of liver transplantation at ages 11 years in 2014. Some of the younger patients below 2 years did not complete either SWE (N = 6) or ultrasound (N = 1). Six patients were not tested for alpha 1-antitrypsin deficiency because of the lack of DNA. All patients underwent physical examination and FibroScan.

T1
TABLE 1:
Patient characteristics

No patients had abnormal alfa-fetoprotein or positive ANA or ANCA. None were homozygous for the ZZ-allele indicating A1AT-deficiency. Three patients were heterozygote for PIZ and 5 patients were heterozygote for PIS.

Liver Decompensation

No patients showed clinical stigmata of liver decompensation. According to biochemical and US results (Table 2), none showed evidence of biochemical decompensation. Five patients had elevated levels of INR (>1.2) but all were below 1.5. One patient had an elevated bilirubin (24 μmol/L) without association between either INR or bilirubin and other biochemical results (linear regression).

T2
TABLE 2:
Biochemical and sonographic results

Portal Hypertension

No patients showed clinical, sonographic, or biochemical signs of PTH, nor clinically evident splenomegaly. Four patients had splenomegaly on US of whom 1 had marginally decreased platelets (135 × 109/l). Maximal portal venous velocity in the hilum was above 20 cm/s in all patients. Thrombocytopenia was seen in 5 patients, (but all >130 × 109/l) and without a reported history of gastrointestinal bleeding. Albumin was low in 61 patients (72.6%) and was negatively associated with INR (linear regression, r = 0.32, P = 0.003). One patient had low hemoglobin but without clinical signs of bleeding or of occult bleeding monitored by ferritin, transferrin, and transferrin saturation (data not shown).

Liver Fibrosis and Inflammation

Biochemical Results

At least 1 of the 3 liver-specific enzymes ALT, ALP, and GGT were above the upper limit of normal in 22 patients (26.2%) (Fig. 1A). Ten patients (11.9%) had ≥2 elevated liver enzymes and in 6 patients (7.1%), all 3 liver enzymes were elevated. High GGT was seen solely in patients with ≥2 abnormal biochemical results.

F1
FIGURE 1:
Comparison of results. (A) Biochemical results. (B) Ultrasonographic results (WUSS >3). (C) Ultrasonographic results (WUSS >6). (D) Biochemical and ultrasonographic results.

A skewed distribution of elevated total bile acid was observed with 82% in ages 1 to 3 years and 33% in ages 16 to 18 years. The majority of patients (80%) with ≥2 abnormal liver enzymes had elevated total bile acid.

APRI was elevated in 33 patients (40.7%) when using the diagnostic thresholds (APRI >0.264) reported by Leung et al (14). No biochemical results were associated with z-height, z-weight, z-BMI, or FEV1%.

Sonographic Results

Changes consistent with liver disease (WUSS 7–9) were shown in 7 patients (8.4%). Either liver disease or intermediate liver disease was seen in 26 patients (31.3%). One patient had a WUSS of 7 to 9 but showed no other signs of liver disease.

Abnormal SWE (1.58--2.54 m/s) was shown in 9 patients (12%) as well as abnormal FibroScan, which was shown in 9 patients (10.7%).

Age difference between patients with abnormal versus normal test was neither seen in WUSS (11.2 vs 9.6 years; P = 0.2), SWE (13.2 vs 9.7 years; P = 0.053) nor FibroScan (12.9 vs 9.7 years; P = 0.075).

Twenty-seven patients (32.1%) had an abnormal result of ≥1 sonographic method, 9 patients had ≥2, and 8 patients had ≥3 (Fig. 1B). If including only patients with WUSS 7 to 9, 8 patients had an abnormal result of ≥2 methods, and 6 patients had ≥3 (Fig. 1C).

Neither z-height, z-weight, z-BMI, nor FEV1% was related to any of the sonographic results.

A linear association was found between FibroScan and SWE (r2 = 0.44, P < 0.001), and a highly significant difference in proportions of abnormal SWE results was detected between patients with abnormal FibroScan versus normal FibroScan (P < 0.001). Both FibroScan and SWE was abnormal in 9 patients and only 1 in each group failed to have an abnormal result of both (see Fig. 1B). Proportion tests were made to compare FibroScan to WUSS (P < 0.001), and SWE to WUSS (P < 0.001). Significant differences were found indicating association between FibroScan and WUSS, and SWE and WUSS.

Combined Biochemical and Sonographic Results

In univariable analysis, we found a significant higher ALT, AST, GGT, and APRI score and lower thrombocytes in patients with the highest WUSS score (1-way ANOVA, P < 0.001 for all 5 variables), the same pattern was seen with SWE and Fibroscan with a positive association with ALT, AST, GGT, and APRI (linear regression, all P values less than 0.004) and negative association with thrombocytes (linear regression P < 0.001 for both analyses). In multiple regression with these 5 variables as independent variables and either WUSS, Fibroscan, or SWE as the dependent variable, only GGT came out significant for Fibroscan (r2 = 0.41, P < 0.001), but both GGT and APRI came out significant for SWE and WUSS (SWE: r2 = 0.42, P = 0.006 for GGT and P = 0.01 for APRI, WUSS: r2 = 0.38, P = 0.002 for GGT and P = 0.034 for APRI).

Nine patients (10.7%) had ≥2 abnormal sonographic results and 10 patients (11.9%) had ≥2 abnormal biochemical results (Fig. 1D). Eight patients (9.5%) had both ≥2 abnormal results of sonographic methods and ≥2 abnormal biochemical results and were defined as having manifest CFLD (described in Table 3). Patients with manifest CFLD were significantly older (13.7 vs. 9.7 years; P = 0.04). Mean value of total bile acid did not differ significantly, but 7 out of the 8 manifest CFLD patients did have an elevated result. All manifest CFLD patients had significantly higher mean value of APRI as compared with the remaining patients (0.89 vs. 0.25; P = 0.006). Sonographic splenomegaly was seen only in manifest CFLD patients (4 patients). Regeneration noduli were detected in 8 patients (9.6%), 6 among the manifest CFLD patients.

T3
TABLE 3:
Patients with manifest cystic fibrosis-related diabetes

Only 1 patient in the defined manifest CFLD group was heterozygote for PIS, the remaining 8 patients who were heterozygote for A1AT deficiency did not show any signs of liver disease.

DISCUSSION

CFLD is primarily believed to be a biliary disease (17). Bilirubin and total bile acids, however, showed no specific diagnostic value, when comparing with sonographic results. Nevertheless, 52.4% of all patients and 87.5% of the manifest CFLD patients did have elevated total bile acid. Although no correlation between total bile acid and age was found, abnormal results were mainly detected in the younger age group with exception of the manifest CFLD patients, who were mainly in puberty. ALP was elevated in 10 patients of whom 7 were among the 8 manifest CFLD patients.

Ten patients (12%) had elevated GGT and all had ≥2 abnormal biochemical results of the selected markers in this study (ALT, ALP, GGT) (Fig. 1A). Of the manifest CFLD patients, 87.5% had elevated GGT confirming the findings of other studies, where GGT has been reported to be associated with CFLD (18,19). We found in multiple regression with ALT, AST, GGT, APRI score and thrombocytes as independent variables and either WUSS, Fibroscan, or SWE as the dependent variable only GGT to be significant for both Fibroscan, SWE, and WUSS. This indicates that GGT might be the best marker for ultrasonic changes.

Both ALP and GGT, but not ALT, were found to be abnormal in significantly older patients. This confirms CFLD as being primarily a bile duct disease, as well as severe cases predominantly being seen in the second decade of life (20).

The usefulness of abnormal transaminase has often been questioned in the diagnosis of CFLD, as they may be both transient and insufficiently sensitive and specific and often correlate poorly with severity of liver damage (6,17,20). We found biochemical signs of liver inflammation in 11.9% of patients defined as ≥2 abnormal results of liver specific markers (ALT, ALP, GGT). We found elevated ALT in 21.7% of patients with the majority (80%) also having elevated AST, thus not a useful supplementary marker of liver damage.

Colombo et al (6) demonstrated in a prospective study that 56% of 48 patients diagnosed with CFLD had abnormal results of at least 2 liver enzymes (AST, ALT, GGT). Of patients, who never developed CFLD, 41% occasionally had elevation of at least 2 liver enzymes. In spite of a different study design and choice of liver markers, these results by far exceed our 11.9%. This may reflect our choice of the more bile duct specific marker ALP. Occasional biochemical abnormalities are, however, common in CF, likely caused by antibiotics, steatosis, or concurrent infections (21). To further complicate the diagnosis of CFLD, several studies have found that children with advanced fibrosis may have normal biochemistry (22,23).

Our cross sectional study did not allow direct comparison with the Debray-Colombo criteria (1) or the new ESPGHAN criteria (7).

We did not find any signs of chronic liver failure, as INR were only slightly elevated in 5 patients. This could easily be because of vitamin K deficiency. This was, however, not tested.

Hypoalbuminemia was seen in a major part of patients and was negatively associated with INR. This may not only reflect liver function but also nutritional status.

Previous studies have found FibroScan to correlate with fibrosis scores from liver biopsies (10,24) and/or sonographic evaluation (25,26). As early fibrotic changes in CFLD are often focal, biopsies may consequently be false negative (27). In comparison, FibroScan and SWE examine a larger proportion of the liver (26) and were, therefore included. In general, ultrasound is associated with poor sensitivity and specificity in detecting and staging fibrosis as steatosis and fibrosis appear similar (20). Both are common lesions in CF and especially multilobular nodularity is demonstrated with US resulting in a high WUSS. SWE and FibroScan are useful in distinguishing between no fibrosis and advanced fibrosis, but are less sensitive in detecting mild or moderate fibrosis (16). An abnormal WUSS despite normal FibroScan and SWE could, therefore, still indicate degrees of fibrosis and close follow-up is recommended.

SWE and FibroScan have been reported to have similar outcome (28) and the agreeing results in this study between especially FibroScan and SWE show that these 2 methods are compatible. We found FibroScan easier to perform than SWE, especially in younger children, because of the need of clear sonographic visualization and the child's capacity of sitting still and holding the breath during SWE measurements.

Unfortunately, the noninvasive methods cannot be related to histological findings of liver disease as liver biopsies were not included in the present study.

Carrier of PIZ variant for A1AT deficiency has recently been shown to be associated with liver disease in adults (29). We were not able to show the same association, despite 8 children being heterozygote for either PiZ or PiS; however, the numbers may be too small. It will be interesting to investigate if adults with cystic fibrosis who develop liver disease later in life are more likely to be carriers for either PiZ or PiS.

Categorizing only 9.5% of our patients as having CFLD is probably an underestimation compared with other studies where prevalence of CFLD ranged from 25% to 41% and prevalence of cirrhosis ranged from 7.8% to 13% (6,30–32), corresponding better with our findings. This difference could be because of our strict diagnostic criteria with inclusion of several imaging modalities and supplementary bile duct specific biochemical tests. One of our patients had abnormal results of WUSS, SWE, and FibroScan but had only 1 elevated liver enzyme (ALT). This patient obviously has changes consistent with liver fibrosis and should be followed closely. In our study those with abnormalities on ultrasound (WUSS), but not on elastrography might have been overlooked. Elastographic measurements are, therefore, suggested as a supplement and not a substitution for conventional US.

Strength and Limitation

A major strength of this study is the relatively large sample of a pediatric CF population with a high inclusion rate from 1 year of age.

The ability to determine the true sensitivity and specificity of biochemical and sonographic tests or the true prevalence of CFLD in our study population is limited as liver biopsies were not done and a concluding diagnosis of cirrhosis therefore not made. Secondly, the cross sectional study design makes comparison with the Colombo-Debray criteria impossible.

CONCLUSIONS

In conclusion, diagnosis of CFLD is complicated as reflected in the variable reported prevalence. After using an extended screening program for CFLD including elastographic measurements we suggest WUSS and either FibroScan or SWE, combined with GGT as the most specific diagnostic markers for CFLD.

Identifying early signs of CFLD remains a challenge and further longitudinal studies are needed.

Acknowledgment

Morten Dunoe, cand. scient. Head of Molecular Genetics laboratory, Department of Clinical Genetics, Rigshospitalet is thanked for analyses of A1AT-deficiency alleles.

REFERENCES

1. Debray D, Kelly D, Houwen R, et al. Best practice guidance for the diagnosis and management of cystic fibrosis-associated liver disease. J Cyst Fibros 2011; 10 Suppl 2:S29–S36.
2. Westaby D. Bush A, Alton E, Davies J, Griesenbach U, Jaffe A. Cystic fibrosis: liver disease. Basel, Progress in Respiratory Research. Karger:2006.
3. Koh C, Sakiani S, Surana P, et al. Adult-onset cystic fibrosis liver disease: diagnosis and characterization of an underappreciated entity. Hepatology 2017; 66:591–601.
4. Toledano MB, Mukherjee SK, Howell J, et al. The emerging burden of liver disease in cystic fibrosis patients: a UK nationwide study. PLoS One 2019; 14:e0212779.
5. Stonebraker JR, Ooi CY, Pace RG, et al. Features of severe liver disease with portal hypertension in patients with cystic fibrosis. Clin Gastroenterol Hepatol 2016; 14:1207–1215.
6. Colombo C, Battezzati PM, Crosignani A, et al. Liver disease in cystic fibrosis: a prospective study on incidence, risk factors, and outcome. Hepatology 2002; 36:1374–1382.
7. Debray D, Narkewicz MR, Bodewes F, et al. Cystic Fibrosis-related liver disease: research challenges and future perspectives. J Pediatr Gastroenterol Nutr 2017; 65:443–448.
8. Williams SG, Evanson JE, Barrett N, et al. An ultrasound scoring system for the diagnosis of liver disease in cystic fibrosis. J Hepatol 1995; 22:513–521.
9. Goldschmidt I, Streckenbach C, Dingemann C, et al. Application and limitations of transient liver elastography in children. J Pediatr Gastroenterol Nutr 2013; 57:109–113.
10. de Ledinghen V, Le Bail B, Rebouissoux L, et al. Liver stiffness measurement in children using FibroScan: feasibility study and comparison with Fibrotest, aspartate transaminase to platelets ratio index, and liver biopsy. J Pediatr Gastroenterol Nutr 2007; 45:443–450.
11. Calvopina DA, Noble C, Weis A, et al. Supersonic shear-wave elastography and APRI for the detection and staging of liver disease in pediatric cystic fibrosis. J Cyst Fibros 2019; 19:449–457.
12. Lewindon PJ, Puertolas-Lopez MV, Ramm LE, et al. Accuracy of transient elastography data combined with APRI in detection and staging of liver disease in pediatric patients with cystic fibrosis. Clin Gastroenterol Hepatol 2019; 17:2561–2569.
13. Rosenstein BJ, Cutting GR. The diagnosis of cystic fibrosis: a consensus statement. Cystic Fibrosis Foundation Consensus Panel. J Pediatr 1998; 132:589–595.
14. Leung DH, Khan M, Minard CG, et al. Aspartate aminotransferase to platelet ratio and fibrosis-4 as biomarkers in biopsy-validated pediatric cystic fibrosis liver disease. Hepatology 2015; 62:1576–1583.
15. Hilsted L, Rustad P, Aksglaede L, et al. Recommended nordic paediatric reference intervals for 21 common biochemical properties. Scand J Clin Lab Invest 2013; 73:1–9.
16. Nobili V, Vizzutti F, Arena U, et al. Accuracy and reproducibility of transient elastography for the diagnosis of fibrosis in pediatric nonalcoholic steatohepatitis. Hepatology 2008; 48:442–448.
17. Kobelska-Dubiel N, Klincewicz B, Cichy W. Liver disease in cystic fibrosis. Prz Gastroenterol 2014; 9:136–141.
18. Van Biervliet S, Verdievel H, Vande Velde S, et al. Longitudinal transient elastography measurements used in follow-up for patients with cystic fibrosis. Ultrasound Med Biol 2016; 42:848–854.
19. Bodewes FA, van der Doef HP, Houwen RH, et al. Increase of serum gamma-glutamyltransferase associated with development of cirrhotic cystic fibrosis liver disease. J Pediatr Gastroenterol Nutr 2015; 61:113–118.
20. Mueller-Abt PR, Frawley KJ, Greer RM, et al. Comparison of ultrasound and biopsy findings in children with cystic fibrosis related liver disease. J Cyst Fibros 2008; 7:215–221.
21. Pereira TN, Walsh MJ, Lewindon PJ, et al. Paediatric cholestatic liver disease: diagnosis, assessment of disease progression and mechanisms of fibrogenesis. World J Gastrointest Pathophysiol 2010; 1:69–84.
22. Colombo C, Apostolo MG, Ferrari M. Set al. Analysis of risk factors for the development of liver disease associated with cystic fibrosis. J Pediatr 1994; 124:393–399.
23. Walsh MJ, Lewindon PJ, Shepherd RW, et al. Detection and follow-up of hepatic fibrosis in cystic fibrosis: a role for diagnostic liver biopsy and serum markers in evaluating outcomes of cystic fibrosis liver disease. J Gastroenterol Hepatol 2009; 50: (Suppl 4): A759.
24. Breton E, Bridoux-Henno L, Guyader D, et al. Value of transient elastography in noninvasive assessment in children's hepatic fibrosis. Arch Pediatr 2009; 16:1005–1010.
25. Menten R, Leonard A, Clapuyt P, et al. Transient elastography in patients with cystic fibrosis. Pediatr Radiol 2010; 40:1231–1235.
26. Witters P, De Boeck K, Dupont L, et al. Non-invasive liver elastography (Fibroscan) for detection of cystic fibrosis-associated liver disease. J Cyst Fibros 2009; 8:392–399.
27. Stewart L. The role of abdominal ultrasound in the diagnosis, staging and management of cystic fibrosis liver disease. J R Soc Med 2005; 98: (Suppl 45): 17–27.
28. Bota S, Herkner H, Sporea I, et al. Meta-analysis: ARFI elastography versus transient elastography for the evaluation of liver fibrosis. Liver Int 2013; 33:1138–1147.
29. Strnad P, Buch S, Hamesch K, et al. Heterozygous carriage of the alpha1-antitrypsin Pi∗Z variant increases the risk to develop liver cirrhosis. Gut 2019; 68:1099–1107.
30. Lamireau T, Monnereau S, Martin S, et al. Epidemiology of liver disease in cystic fibrosis: a longitudinal study. J Hepatol 2004; 41:920–925.
31. Gaskin KJ, Waters DL, Howman-Giles R, et al. Liver disease and common-bile-duct stenosis in cystic fibrosis. N Engl J Med 1988; 318:340–346.
32. Lindblad A, Glaumann H, Strandvik B. Natural history of liver disease in cystic fibrosis. Hepatology 1999; 30:1151–1158.
Keywords:

cystic fibrosis; gamma glutamyl transferase; transaminases; transient elastography; ultrasonography

Copyright © 2020 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition