Liver Involvement in Children With SARS-COV-2 Infection

Two Distinct Clinical Phenotypes Caused by the Same Virus

Adriana Perez MD; Amanda Cantor MD; Bryan Rudolph MD, MPH; Jonathan Miller MD; Debora Kogan-Liberman MD; Qi Gao PhD; Bernardo Da Silva; Kara G. Margolis MD; Nadia Ovchinsky MD, MBA; Mercedes Martinez MD


Liver International. 2021;41(9):2068-2075. 

In This Article


A total of 291 patients were included in the study: 220 (76%) and 71 (24%) were diagnosed with COVID-19 and MIS-C respectively (Figure 1). Elevated ALT was detected in 36% (n = 105) of all children, 31% (n = 69) of children with COVID-19, and 51% (n = 36) of children with MIS-C. Severe injury was noted in 8% and 4% of children with COVID-19 (Figure 1; Table S1) and MIS-C respectively. One child with MIS-C developed acute liver failure and recovered.

Figure 1.

Flow-chart of study cohort demonstrating the differences of ALI in the COVID-19 and MIS-C cohorts; A total of 291 patients were included in the study: 220 (76%) and 71 (24%) were diagnosed with COVID-19 and MIS-C, respectively. ALI was detected in 31% (n = 69) of children with COVID-19 and 51% (n = 36) of children with MIS-C. Severe injury defined as ALT > 200 U/L was noted in 8% of children with COVID-19 and 4% of children with MIS-C, respectively. Abbreviations: ALI, acute liver injury; ALT, alanine aminotransferase; COVID-19, Coronavirus disease 2019; MIS-C, system inflammatory syndrome in children

Liver Involvement in COVID-19

Children with E-ALT were significantly older, with a median age of 16 vs 11 years (P =.001). No differences were observed in race, gender, or ethnicity. E-ALT was associated with obesity (P <.001), immunocompromised status (P =.04), presence of any malignancy (P =.01), and chronic liver disease (CLD) (P =.01). Underlying CLD was present in 13% of children with E-ALT. The etiologies of CLD in patients in whom prior medical history could be discerned were non-alcoholic fatty liver disease (NAFLD) (six patients), biliary atresia (1), liver transplant recipient (1), and glycogenic hepatopathy (1) (Table 1 and Table 2; Tables S1 and S2).

Liver Involvement in MIS-C

Children with E-ALT were also significantly older, with a median age of 9.5 vs 6 years (P = .01). E-ALT was associated with Black race (P = .01) and male gender (P = .001). Most children had no prior medical history and no differences in BMI were noted (Table 1 and Table 2; Tables S3).

Liver Involvement in MIS-C Compared to COVID-19

Liver involvement in both cohorts was characterized by statistically significant elevation of median ALT above 2x ULN with mild cholestasis. MIS-C hepatitis was associated with decreased albumin (median 2.7, P < .001) and thrombocytopenia (mean 134, P = .01). Children with E-ALT and MIS-C were more frequently Hispanic (P = .04) and of Black race (P < .001). Cardiac dysfunction (P < .001), kidney injury (P = .01), elevated troponin (P = .03), higher pro-B natriuretic peptide (P < .001), higher median D-dimer levels (P = .01), and higher peak interleukin-6 (P = .04) were more frequently recorded in the MIS-C cohort with E-ALT (Table 1 and Table 2; Tables S4).

Children with E-ALT in both cohorts had a more severe disease course with a greater prevalence of multiorgan dysfunction (P = .001 in both cohorts), respiratory failure (P = .001 and 0.01 in COVID-19 and MIS-C, respectively), longer hospitalization (P < .001 and .001), and longer ICU stay (P = .01 and 0.04). No significant association was found between E-ALT and mortality in COVID-19 (see Table S6 for details), and no deaths were recorded in the MIS-C cohort. In multivariable analysis, E-ALT in the COVID-19 cohort was associated with an increase in c-reactive protein (CRP) level (OR 1.08 with 95% confidence interval [CI] [1.03–1.15]) after adjustment for age, gender, race, obesity, CLD, and ICU admission (Table 3). In the MIS-C cohort, male gender and Black race were associated with more than 5- and 4-fold increased E-ALT odds, respectively, when adjusted for age and CRP (Table 3). The association with gender was consistent in all multivariable regression models; Black race was also associated with E-ALT when adjusted for obesity but did not reach statistical significance in other models (Table S5).

Children with MIS-C had 2.3× increased odds of elevated ALT levels than children with COVID-19 after adjusting for age and race (Table 4).