Share this article on:

Comparison of Acoustic Radiation Force Impulse/Serum Noninvasive Markers for Fibrosis Prediction in Liver Transplant

Pinto, Joana*; Matos, Hugo*; Nobre, Susana; Cipriano, Maria A.; Marques, Margarida§; Pereira, João M.S.§; Gonçalves, Isabel; Noruegas, Maria J.*

Journal of Pediatric Gastroenterology & Nutrition: March 2014 - Volume 58 - Issue 3 - p 382–386
doi: 10.1097/MPG.0000000000000226
Original Articles: Hepatology and Nutrition

Objectives: Fibrosis, related to several causes, can be diagnosed in children and adolescents’ liver grafts that are >1 year old. At present, liver biopsy is the gold standard for assessing liver damage in the posttransplant setting. We aimed to evaluate the accuracy of noninvasive biomarkers of fibrosis, namely, acoustic radiation force impulse (ARFI), aspartate-to-platelet ratio index, and aspartate aminotransferase/alanine aminotransferase (AST/ALT) ratio index, either alone or in combination, for predicting fibrosis in pediatric patients submitted to liver transplantation.

Methods: We prospectively assessed liver fibrosis in 30 children/adolescents with liver transplant through biopsy (liver transplant follow-up during 12 months). ARFI with Virtual Touch Software (Acuson 2000) was performed, and blood samples were taken to determine liver function and platelet count. Two groups were analyzed according to the histopathologic stage of fibrosis, namely, none/mild (F0–F1) versus significant fibrosis (F2-4).

Results: The mean age of the 30 patients was 11 years (3–18 years), with a mean posttransplant period of assessment of 6.5 years. Twenty-four patients (80%) presented stage F0–F1 fibrosis and 6 patients (20%) presented stage F2-4. The area under the curve using receiver operating characteristic analysis for ARFI, aspartate-to-platelet ratio index, and AST/ALT ratio index for significant fibrosis was 0.76 (P = 0.052), 0.74 (P = 0.066), and 0.69 (P = 0.162), respectively. Through multivariate logistic regression analysis, the only independent predictor of significant fibrosis was ARFI (odds ratio 10.7, 95% confidence interval 1.2–95.7; P = 0.045). The combination of ARFI and AST/ALT ratio index presented a good diagnostic accuracy of fibrosis (area under the curve of 0.83; P = 0.013).

Conclusions: ARFI may serve as a potential method for assessing significant fibrosis in pediatric patients with liver transplant, particularly in combination with AST/ALT ratio index.

*Medical Imaging Department

Hepatology Department, Pediatric Hospital of Coimbra

Pathology Department

§Laboratory of Biostatistics and Medical Informatics, University Hospitals of Coimbra, Coimbra, Portugal.

Address correspondence and reprint requests to Joana Pinto, MD, Medical Imaging Department, Pediatric Hospital of Coimbra, Av. Afonso Romão, 3000-602 Coimbra, Portugal (e-mail:

Received 19 October, 2013

Accepted 19 October, 2013

The authors report no conflicts of interests.

Liver transplantation has become the standard treatment for children with end-stage liver disease. In contrast to adults, most pediatric liver diseases resulting in transplant do not have a high tendency of recurrence in the graft (1). Fibrosis can be identified in hepatic grafts after 1-year posttransplant, and can be related to de novo autoimmune hepatitis (2), chronic rejection (3), cryptogenic hepatitis (4), azathioprine toxicity (5), and chronic biliary or venous outflow obstruction. Whether this fibrosis will subsequently progress to clinical graft dysfunction or increased rates of graft loss is unclear. At present, the diagnosis is made using center-specific criteria, and there is no consensus for a standardized set of criteria. Availability of standardized criteria would enable centers to compare and pool results, improve therapy, and better understand pathophysiologic disease mechanisms. Liver biopsy (LB) is the reference standard for assessing liver injury in the posttransplant setting (2); however, it is associated with complications such as bleeding and pain. Noninvasive tests for screening fibrosis are clearly essential to triage those cases that require biopsy and eventually manipulation of immunosuppression.

Some studies have enhanced the importance of acoustic radiation force impulse (ARFI) and serum markers in detection of fibrosis in pediatric patients with chronic liver disease (6–8). ARFI elastography is the most recent noninvasive procedure used for the assessment of liver fibrosis (9–11). There are only 2 recent reports that have attempted to stage liver fibrosis in pediatric patients with liver transplant by noninvasive markers (12,13). One studied the use of transient elastography (12) and the other focused on ARFI technology (13).

The aim of our study was to evaluate the accuracy of ARFI alone and in combination with other noninvasive biomarkers of fibrosis such as the aspartate-to-platelet ratio index (APRI) and the serum aspartate aminotransferase/alanine aminotransferase (AST/ALT) ratio index for the prediction of fibrosis in a pediatric population presenting for liver transplantation.

Back to Top | Article Outline



Our institutional ethics committee approved the study and all of the patients were enrolled after informed consent was received. Between May 2010 and December 2012, we enrolled 30 consecutive children and adolescents with a liver transplant follow-up during 12 months and who had undergone LB. For each child, ARFI with Virtual Touch Software (Acuson S2000; Siemens Medical Solutions, Mountain View, CA) was performed and blood samples were taken to determine liver function and platelet count.

Back to Top | Article Outline


Liver stiffness measurements using ARFI elastography were performed the day before planned LB using an ultrasound device (Acuson S2000) equipped with a 4-MHz transducer. Examinations were performed by a single physician with more than 2 years of experience with ARFI elastography. Scans were performed through intercostal and/or subxiphoid approaches. During real-time B-mode imaging, a square region-of-interest cursor was moved on to the vessel-free graft liver parenchyma (Fig. 1). As previous studies concluded that the right lobe was more accurate for diagnosing liver fibrosis than the left lobe (14), measurements were performed in the right lobe, in cases of whole-liver grafts. To ensure quality, a mean of 9 successful acquisitions were performed in each patient, 3 per equivalent profundity, and a mean value was considered to represent the elastic modulus of the liver. Results are expressed as meters per second.

Back to Top | Article Outline


LBs were performed as part of a follow-up protocol or for suspected rejection. Ultrasound-guided liver biopsies were performed under anesthesia, using a free-hand technique. A 16-gauge automatic needle was used and liver fragments were collected through a subxiphoid or intercostal approach. Specimens were fixed in formalin and embedded in paraffin. Sections were assessed by an experienced pathologist with 18 years of experience in liver histopathology. Specimens were semiquantitatively evaluated according to the scoring system of Batts and Ludwig (15). Fibrosis was staged on the F0–F4 scale: F0—no fibrosis; F1—enlarged, fibrotic portal tracts; F2—periportal or portoportal fibrosis but intact architecture; F3—fibrosis with architectural distortion, but no obvious cirrhosis; F4—probable or definite cirrhosis. Children were divided into 2 groups according to histopathologic staging of hepatic fibrosis: none or mild (F0–F1) versus significant fibrosis (≥F2).

Back to Top | Article Outline

Laboratory Measurements

Blood samples were collected within 4 months following LB, whereupon laboratory data, namely, AST, ALT, and platelet counts were determined. Coagulation blood tests were performed the day of LB. The upper limit of normal (ULN) for ALT and AST was defined as 50 IU/L. APRI was calculated using the following formula (16):

Back to Top | Article Outline

Statistical Analysis

For statistical analysis, SPSS version 21.0 was used (SPSS Inc, Chicago, IL). The χ2 test was used to compare categorical variables, expressed as percentages. Continuous variables, expressed as means, were compared using the Student t test for those with a normal distribution, or the Mann-Whitney U test otherwise. Multivariate stepwise logistic regression analysis was used to identify independent predictors of significant fibrosis. Receiver operating characteristic (ROC) curves, and area under the ROC curve (AUC) were applied to assess and compare the diagnostic accuracy of each noninvasive hepatic fibrosis test and their associations. All P values were 2-sided, and P ≤ 0.05 was considered to indicate statistical significance.

Back to Top | Article Outline


Between January 2010 and December 2012, a total of 30 patients met the inclusion criteria for the study. There were 19 boys (63%) and 11 girls (37%), with a mean age of 11 years (range 3–18 years). The mean posttransplant period was 6.8 years (range 1–11 years). There was no significant difference between patients in the F0–F1 and ≥F2 groups regarding sex, age, and time after hepatic transplantation (Table 1). The underlying reasons for hepatic transplant are presented in Table 2. The majority of the patients were transplanted for metabolic disease (n = 14; 46.7%) or biliary atresia (n = 12; 40%). Twenty-four patients (80%) presented no or mild fibrosis (stage F0–F1) and 6 patients (20%) presented significant fibrosis (stage ≥F2). Table 3 shows the distribution of patients according to the histopathologic stage of fibrosis. There were no patients in stage 4 of fibrosis. Most (n = 27; 90%) LBs were performed as part of a follow-up protocol. The remaining children (n = 3; 10%) underwent LB for suspicion of rejection that was confirmed in 2. Most patients (n = 25; 83.4%) collected blood sample the day of LB. Four (13.3%) had their liver function test 1 month or less after LB and only 1 child (3.3%) collected blood sample 4 months after LB.

Back to Top | Article Outline

ARFI Imaging

Mean shear wave velocity (SWV) was 1.43 m/s in the F0–F1 group (range 1.00–2.43 m/s; median 1.28; standard deviation [SD] 0.40) and 1.91 m/s in the ≥F2 group (range 1.00–2.40 m/s; median 2.08; SD 0.48). The 2 groups demonstrated a nonnormalized distribution. There was a statistically significant difference in SWV between the 2 groups (Table 1) (Mann-Whitney U = 34.4; P = 0.05). Multivariate logistic stepwise regression analysis showed that ARFI was an independent predictor of significant hepatic fibrosis (odds ratio 10.7; 95% confidence interval 1.2–95.7; P < 0.05). The AUC for diagnosis of significant fibrosis (≥F2) was 0.76 (Table 4, Fig. 2), with borderline significance of P = 0.052. The optimal value for the diagnosis of significant fibrosis was 1.57 m/s with a sensitivity of 83%, a specificity of 71%, a positive predictive value (PPV) of 42%, and a negative predictive value (NPV) of 94%. Two children presented discrepant results with extremely high SWV values (2.37 and 2.43 m/s) and no fibrosis in the biopsy specimen. One child presented moderate necroinflammatory activity upon LB and the other presented cholestasis in relation to reflux in the afferent small intestine loop; if these children were excluded from the study, the AUC for the diagnosis of significant fibrosis with ARFI elastography would be 0.82.

Back to Top | Article Outline


The mean for APRI was 0.58 in the F0–F1 group (range 0.20–4.80; median 0.31; SD 0.92) and 1.00 in the ≥F2 group (range 0.29–2.65; median 0.68; SD 0.88). Using the Mann-Whitney U test for unpaired samples, there was no statistically significant difference in APRI values between the 2 groups (P = 0.065, Table 1). The AUC for diagnosis of significant fibrosis (≥F2) was 0.74 (Table 4, Fig. 2), although this was not statistically significant (P = 0.066). The most discriminant cutoff value for significant fibrosis in ROC analysis was 0.4, with a sensitivity of 83%, a specificity of 58%, a PPV of 31%, and a NPV of 94%.

Back to Top | Article Outline

AST/ALT Ratio Index

The mean AST/ALT ratio index value was 1.37 in the F0–F1 group (range 0.66–3.50; median 1.28; SD 0.65) and 0.94 in the ≥F2 group (range 0.80–1.10; median 0.93; SD 0.10). Regarding the AST/ALT ratio index, there was no significant difference between patients with no/mild fibrosis and those with significant fibrosis (Mann-Whitney U, P = 0.162, Table 1).

The AUC for the diagnosis of significant fibrosis (≥F2) with the AST/ALT ratio index was 0.69 (Table 4, Fig. 2). This value was, however, not significant (P = 0.162). The optimal cutoff AUC value for the diagnosis of significant fibrosis was 1.1, with a sensitivity of 100%, a specificity of 63%, a PPV of 40%, and a NPV of 100%.

Back to Top | Article Outline

Combinations of ARFI and Serum Markers

We evaluated paired combinations of ARFI with either APRI or the AST/ALT ratio index, as well as all 3 markers together for the diagnosis of significant fibrosis. ROC analyses of individual noninvasive tests and in combination are shown in Table 4 and Figure 2. The best diagnostic value was obtained with the combination of ARFI and the AST/ALT ratio index, with a significant AUC of 0.83 and a P value of 0.013.

Back to Top | Article Outline


To our knowledge, this is the second study to evaluate the diagnostic accuracy of noninvasive tests in assessing significant liver fibrosis in pediatric patients with liver transplant. The present study shows the diagnostic value of ARFI, a newly developed method to estimate liver stiffness compared to and in combination with common serum markers such as APRI and the AST/ALT ratio index, for the assessment of significant liver fibrosis (≥F2).

After liver transplantation, LB is still the reference tool for assessing hepatic fibrosis and disease progression. To reduce the number of liver biopsies, the development of noninvasive tests to assess hepatic fibrosis has been an active area of research in recent years. The diagnostic performance of ARFI has been shown to be efficient and superior to that of APRI and AST/ALT ratio index. When combining ARFI with the other serum markers, the most accurate diagnosis was achieved with ARFI+AST/ALT ratio index, which not only presented a higher AUC but also was clearly significant.

In our study, the AUC of ARFI test for significant fibrosis (0.76) was slightly lower than that in the study of Noruegas et al (6) that demonstrated an AUC of 0.82. This difference could be explained by a greater number of patients in stage F3 of fibrosis and patients with cirrhosis (F4), resulting in an increase in ARFI diagnostic performance for significant fibrosis. Nevertheless, ARFI was the only independent predictor of significant hepatic fibrosis through multivariate logistic stepwise regression analysis.

The optimal value for the diagnosis of significant fibrosis (1.57 m/s) was intermediate between that demonstrated in the studies by Noruegas et al (1.39 m/s) (6) and Hanquinet et al (2 m/s) (13). Up until now, the only study to evaluate ARFI diagnostic performance for revealing fibrosis in children with liver transplant was carried out by Hanquinet et al (13); however, this was not a clear indication of the effectiveness of the method because only approximately 60% of the patients had undergone liver transplants. Further studies using larger samples of pediatric patients with liver transplant are needed to refine SWV cutoff points for significant fibrosis.

The elevated sensitivity (83%) and NPV (94%) suggest that ARFI could be reliably used for first-line prebiopsy evaluation, thus avoiding LB in certain patients, despite its borderline significance value (P = 0.052), which can be explained by the limited number of subjects in this study. The different types of approach (subxiphoid vs intercostal) in ARFI data acquisition were not compared. The choice of approach was based on individuals to obtain the best acoustic window. There are no published studies in patients who have undergone liver transplant concerning this subject, but it could be an interesting area to investigate.

Furthermore, the influence of the graft type (left split, right split, or whole) on SWV values was also not evaluated. In the literature, the only established finding is that measurements in the right hepatic lobe are more accurate for diagnosing liver fibrosis because the left lobe is more susceptible to respiratory and cardiac movements. This was observed both in adult and pediatric patients without transplant (14). So, in general, liver graft measurements should be performed in the right lobe.

Two children presented discrepant results with high SWV values, but no fibrosis in the LB specimen. One patient presented moderate necroinflammatory activity upon pathologic liver analysis, and frank elevation of ALT values in relation to acute graft rejection. This is in line with the studies that reported elevated values of liver SWV, assessed by ARFI elastography, associated with high degrees of necroinflammatory activity (17,18). Authors of these studies concluded that ARFI specificity decreased in the presence of necroinflammatory activity. Extensive inflammatory infiltration, hepatocyte swelling, and tissue edema were evoked as the confounding factors. The other patient presented a history of cholestasis in relation to biliary reflux in the afferent loop, mild elevation of AST and ALT levels, and normal LB. In the literature, some studies have been carried out with Fibrotest (Biopredictive, Paris, France) that showed that extrahepatic cholestasis increased liver stiffness, irrespective of fibrosis (19,20). This was because of an augmentation of hydrostatic pressure in the liver with extrahepatic cholestasis. The same process could, therefore, be observed with ARFI elastography, but further studies are required to confirm this.

Necroinflammatory activity and cholestasis are definitive confounding factors for the accuracy of ARFI in the diagnosis of fibrosis. These results draw attention to the relatively weak specificity of ARFI, and the requirement for biopsy in cases of high SWV values. If these patients were excluded from the study, the diagnostic performance of significant fibrosis with ARFI elastography would be even greater, with an AUC of 0.82.

The usefulness of APRI as a noninvasive test of liver fibrosis in pediatric patients with chronic liver disease is controversial. Yang et al (21) concluded that in children with nonalcoholic fatty liver disease, APRI revealed a significant difference between patients with mild and significant fibrosis. Conversely, McGoogan et al (8) concluded that APRI was not a good marker of fibrosis in children with chronic viral hepatitis. Although we did not find a significant difference between the 2 groups (F0–F1 and ≥F2), the AUC (0.74) was acceptable and similar to that of previous studies on children with chronic hepatitis B (22). The lack of patients with cirrhosis in our study could explain the absence of a significant difference between groups. The pathobiology of hepatic chronic disease shows that there is an early increase in AST levels and a late decrease in platelet count in disease progression.

The AST/ALT ratio index used in isolation was a weak predictor of fibrosis. This is in line with previous studies in children with nonalcoholic fatty liver disease (17), in which an AUC of 0.53 was observed. In our study, the AST/ALT ratio index failed to distinguish significant hepatic fibrosis from no/mild fibrosis in pediatric patients with liver transplant with the lowest AUC of 0.69, nonsignificant. Among patients with altered liver function tests, the majority of false-positive cases (n = 9, 69%) presented signs of acute or chronic rejection or necroinflammatory activity upon LB. In addition, we observed that some patients with normal liver function tests had significant fibrosis (n = 2, 12%); however, the AST/ALT ratio index in combination with ARFI showed improved sensitivity and specificity. This could constitute a good diagnostic tool for avoiding LB in children with normal SWV values and liver function. Although steatosis was not evaluated in the present study, Motosugi et al (23) suggested that fat deposition did not affect liver stiffness measurements by ARFI.

There are, however, some limitations to be considered in the interpretation of our study. This was a study conducted at a single hospital, and as such, both identified and unidentified confounding factors may have influenced the results. The present study is also limited because of its small sample size; in particular, there was a small number of patients with significant fibrosis (n = 6) and none with cirrhosis. If there were a larger sample size, results may be more expressive. The fact that the time interval between biopsy and blood sampling was extended up to 4 months constitutes, in our opinion, a relative limitation, because the development of liver fibrosis is usually slow after liver transplantation. Furthermore, this was observed only in a minority of the patients (16.6%). Because only 10% of patients underwent LB for clinical reasons, we believe that the potential confounding effect from the presence of rejection is low. Nevertheless, future studies performed with protocol biopsies only are preferable to eliminate all of the possible confounding factors.

According to our study, ARFI is a promising screening test for detecting significant liver fibrosis (≥F2) in pediatric patients with liver transplant. In future years, further multicentric prospective studies with larger series of patients, preferably 5 or more years after transplantation, and with available protocol biopsies are needed to define clear cutoff ARFI values and confirm the effectiveness of this method in diagnosing liver fibrosis in these patients.

Back to Top | Article Outline


1. Zhang KJ, Tung BY, Kowdley KV. Liver transplantation for metabolic liver diseases. Clin Liver Dis 2007; 11:265–281.
2. Kerkar N, Hadzic N, Davies ET, et al. De-novo autoimmune hepatitis after liver transplantation. Lancet 1998; 351:409–413.
3. Neumann UP, Langrehr JM, Neuhaus P. Chronic rejection after human liver transplantation. Graft 2002; 5:102–107.
4. Evans HM, Kelly DA, McKiernan PJ, et al. Progressive histological damage in liver allografts following pediatric liver transplantation. Hepatology 2006; 43:1109–1117.
5. Martin SR, Russo P, Dubois J, et al. Centrilobular fibrosis in long-term follow-up of pediatric liver transplant recipients. Transplantation 2002; 74:828–836.
6. Noruegas MJ, Matos H, Gonçalves I, et al. Acoustic radiation force impulse-imaging in the assessment of liver fibrosis in children. Pediatr Radiol 2012; 42:201–204.
7. Shima H, Igarashi G, Wakisaka M, et al. Noninvasive acoustic radiation force impulse (ARFI) elastography for assessing the severity of fibrosis in the post-operative patients with biliary atresia. Pediatr Surg Int 2012; 28:869–872.
8. McGoogan KE, Smith PB, Choi SS, et al. Performance of AST to platelet ratio index (APRI) as a noninvasive marker of fibrosis in pediatric patients with chronic viral hepatitis. J Pediatr Gastroenterol Nutr 2010; 50:344–346.
9. Friedrich-Rust M, Wunder K, Kriener S, et al. Liver fibrosis in viral hepatitis: noninvasive assessment with acoustic radiation force impulse imaging versus transient elastography. Radiology 2009; 252:595–604.
10. Takahashi H, Ono N, Eguchi Y, et al. Evaluation of acoustic radiation force impulse elastography for fibrosis staging of chronic liver disease: a pilot study. Liver Int 2010; 30:538–545.
11. Goertz RS, Zopf Y, Jugl V, et al. Measurement of liver elasticity with acoustic radiation force impulse (ARFI) technology: an alternative noninvasive method for staging liver fibrosis in viral hepatitis. Ultraschall Med 2010; 31:151–155.
12. Goldschmidt I, Stieghorst H, Munteanu M, et al. The use of transient elastography and non-invasive serum markers of fibrosis in pediatric liver transplant recipients. Pediatr Transplant 2013; 17:525–534.
13. Hanquinet S, Rougemont AL, Courvoisier D, et al. Acoustic radiation force impulse (ARFI) elastography for the noninvasive diagnosis of liver fibrosis in children. Pediatr Radiol 2012; 43:545–551.
14. Toshima T, Shirabe K, Takeishi K, et al. New method for assessing liver fibrosis based on acoustic radiation force impulse: a special reference to the difference between right and left liver. J Gastroenterol 2011; 46:705–711.
15. Batts KP, Ludwig J. Chronic hepatitis. An update on terminology and reporting. Am J Surg Pathol 1995; 19:1409–1417.
16. Lackner C, Struber G, Liegl B, et al. Comparison and validation of simple noninvasive tests for prediction of fibrosis in chronic hepatitis C. Hepatology 2005; 41:1376–1382.
17. Chen SH, Li YF, Lai HC, et al. Effects of patient factors on noninvasive liver stiffness measurement using acoustic radiation force impulse elastography in patients with chronic hepatitis C. BMC Gastroenterol 2012; 12:105.
18. Arena U, Vizzutti F, Corti G, et al. Acute viral hepatitis increases liver stiffness values measured by transient elastography. Hepatology 2008; 47:380–384.
19. Harata M, Hashimoto S, Kawabe N, et al. Liver stiffness in extrahepatic cholestasis correlates positively with bilirubin and negatively with alanine aminotransferase. Hepatol Res 2011; 41:423–429.
20. Millonig G, Reimann FM, Friedrich S, et al. Extrahepatic cholestasis increases liver stiffness (FibroScan) irrespective of fibrosis. Hepatology 2008; 48:1718–1723.
21. Yang HR, Kim HR, Kim MJ, et al. Noninvasive parameters and hepatic fibrosis scores in children with nonalcoholic fatty liver disease. World J Gastroenterol 2012; 18:1525–1530.
22. Lebensztejn DM, Skiba E, Sobaniec-Lotowska M, et al. A simple noninvasive index (APRI) predicts advanced liver fibrosis in children with chronic hepatitis B. Hepatology 2005; 41:1434–1435.
23. Motosugi U, Ichikawa T, Niitsuma Y, et al. Acoustic radiation force impulse elastography of the liver: can fat deposition in the liver affect the measurement of liver stiffness? Jpn J Radiol 2005; 29:639–643.

acoustic radiation force impulse; aspartate aminotransferase/alanine aminotransferase index; aspartate-to-platelet ratio index; noninvasive diagnosis of liver fibrosis; pediatric liver transplantation

© 2014 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology,