Introduction: Intestinal Failure-Associated Liver Disease (IFALD) is cholestatic and currently defined as Type 1 (raised alkaline phosphatase (ALP) X 1.5 above upper limit; Type 2 (+ raised BILI (50 - 100); Type 3 (raised ALP,  BILI>100 and signs of end stage liver disease). This seems to underestimate actual liver dysfunction in practice, particularly as the ALP is not liver-specific.

Methods: Retrospective review of surgical infants receiving PN for ≥28 days (i.e. with intestinal failure, IF) between Jan. 2004 to Dec. 2015. Serial liver biochemistry (including aspartate aminotransferase -AST and γ-glutamyl transpeptidase -GGT) was evaluated to 12 months. Abnormal GGT was defined as >2 upper limit (i.e.>110 IU/L). An intestinal failure index (IFI) was calculated as enteral calories/total calories. Data are described as mean (SEM) and correlated using non-parametric tests. A P value of < 0.05 was regarded as significant.

Results: 64 infants [gastroschisis (n=35), intestinal atresia (n=7) and necrotizing enterocolitis (n=22)] had IF. Median gestational age was 35 (23-41) weeks and birth weight 2.1 (600g-3.7) kg. At 12 months 6/64 (9%) were still on home PN due to short gut syndrome and 2 (3%) had died of sepsis. Initial BILI at 1 month (mean – 78 (7.7) µmol/L) correlated with birth weight and gestational age (rs = -0.42, P = 0.0004; rs = -0.47, P < 0.0001) but not IFI. (P = 0.39). There was variable correlation with other liver enzymes (ALP rs = 0.36, P = 0.001; AST rs = 0.63, P < 0.0001) and none with GGT, P = 17). Using current definition: IFALD at 1 month was “diagnosed” in only 5 (8%) (Type 1, n = 1; Type 2 n=4); whereas raised (>50) BILI was noted in 36 (56%). IFALD at 6 months was “diagnosed” in 1 (1.5%) (Type 2,  n = 1) whereas raised (>50) BILI was noted in 10 (16%). At 12 months there was no IFALD in surviving infants (n = 62).  Abnormal GGT at 1, 6 and 12 months was seen in 38 (59%), 8 (12%) and 0 respectively. Substituting GGT for ALP leads to an incidence of “new” IFALD at 1 month of [Type1, n=21 (16%) and Type 2, n=19 (30%)] and at 6 months [Type 1, n =1 (1.5%); Type 2, n=6 (9%) and Type 3, n = 1(1.5%)].

Conclusion: The prevalence of IFALD, as currently defined, underestimates real liver dysfunction in this susceptible population; raised ALP being too stringent. Substitution of GGT for ALP in the definition is a more realistic representation.