Transient elastography assessment of liver allograft fibrosis in pediatric transplant recipients

Abstract

TE measures liver stiffness to assess fibrosis. Its use in post‐transplant patients was reported in few small pediatric studies. We evaluated TE ability to predict liver graft fibrosis in a large cohort while comparing it to the performance of APRI and FIB‐4. We also investigated the effect of graft type on LSMs. Patients at Boston Children's Hospital who underwent LT and LSM ≤ 1 year from biopsy (2007‐2018) were eligible. Ninety‐four patients (45%M) aged 1‐21 years (89% < 18 years; 13% < 2 years) were eligible. Median time between transplant/biopsy and LSM was 5.1 years and 52 days, respectively. Thirty‐nine percent received whole‐liver grafts, 54% TV grafts, and 6% as part of MV. At LSM, median ALT was 25 [IQR 16‐33] IU/L. Twenty‐one percent had METAVIR ≥ F2. LSM was statistically higher among those with significant fibrosis (METAVIR ≥ F2) compared to those with METAVIR F0/F1 (median [IQR] 7.5 [4.6, 13.6] vs 5.1 [4.0, 6.4] kPa, respectively) (P = .005 by Wilcoxon rank‐sum test). APRI and FIB‐4 distributions were not different across METAVIR stages. The AUROC for LSM was 0.71 (95% CI 0.56‐0.85) with an optimal cut‐point of 6.5 kPa. Graft type had no influence on the AUROC for LSM. TE is useful for assessing significant graft fibrosis in children and young adult LT recipients and performs better than APRI and FIB‐4. TV grafts demonstrate similar correlation with histology as whole‐liver grafts.

Diagnostic performance of 2D‐shear wave elastography and serum fibrosis markers for evaluation of hepatic graft fibrosis in pediatric liver‐inclusive transplant recipients: A prospective pilot study

“…27 Consistently, our finding also confirmed its poor diagnostic performance, with an AUROC of 0.71 when it was used as a single test. Furthermore, in contrast with the only study examining FIB4 in LT children that showed a poor AUROC (0.62) for FIB4 to predict significant graft fibrosis, 28 we have shown for the first time the usefulness of FIB4 in the pediatric LT setting, especially using the hepatic graft's age instead of the patient's age to calculate graft FIB4. We demonstrated that graft FIB4 had a good AUROC of 0.85 to distinguish children with significant hepatic graft fibrosis from those with no-mild fibrosis.…”

“…Nevertheless, 2D‐SWE was able to distinguish children with significant hepatic graft fibrosis (≥F2) from those with no‐mild fibrosis (F0‐F1) with a good AUROC of 0.80, and its diagnostic performance further improved when combined with serum markers APRI and/or FIB4. The LS cut‐off value of 1.2 m/s (4.2 kPa) was lower than that reported in previous studies that used TE (6.5–7.25 kPa) 28,34 or point‐SWE (1.5–1.57 m/s) 18,27 to detect significant fibrosis in LT children. Notably, the reference values for normal LS reportedly vary by elastography techniques as well as patient populations.…”

“…The serum‐based fibrosis markers AST/ALT ratio, APRI, and FIB4 have been examined in pediatric LT recipients 12,24,25,26,27 …”

“…Whether the normal LS range obtained by 2D‐SWE in LT children differs from that in healthy children warrants further investigation with a larger cohort that includes a control group of healthy children in addition to LT children. Moreover, a common limitation of elastography techniques across pediatric studies has been a high NPV but a lower PPV 25,27,28,34 ; therefore, strategies to combine LS measurement with other markers might help boost the PPV. Furthermore, a specific LS cut‐off should be determined based on the context in which the elastography technique is intended to be used, taking into account the false positive and false negative rates.…”

“…The serum-based fibrosis markers AST/ALT ratio, APRI, and FIB4 have been examined in pediatric LT recipients. 12,24,25,26,27 The advantage of using these biomarkers is that serum transaminases and platelet count are part of routine laboratory evaluation with no additional cost. Among the examined biomarkers, APRI was the most commonly used marker but has shown a heterogenous diagnostic performance and a wide range of cut-offs in predicting significant hepatic graft fibrosis.…”