Paediatric acute hepatitis of unknown aetiology: a national investigation and adenoviraemia case-control study in the UK
Sema Mandal,Ruth Simmons,Georgina Ireland,Andre Charlett,Monica Desai,Laura Coughlan,Annabel Powell,David Leeman,Christopher Williams,Claire Neill,Maureen C O'Leary,Clare Sawyer,Frances Rowley,Caroline Harris,Catherine Houlihan,Claire Gordon,Tommy Rampling,Helen Callaby,Katja Hoschler,Jade Cogdale,Erik Renz,Praveen Sebastianpilli,Catherine Thompson,Tiina Talts,Cristina Celma,Emma A Davies,Shazaad Ahmad,Nicholas Machin,Laura Gifford,Catherine Moore,Elizabeth M Dickson,Titus H Divala,David Henderson,Kathy Li,Philip Broadbent,Ines Ushiro-Lumb,Clare Humphreys,Tassos Grammatikopoulos,Jane Hartley,Chayarani Kelgeri,Sanjay Rajwal,Ifeanyichukwu Okike,Deirdre A Kelly,Malcolm Guiver,Ray Borrow,Renu Bindra,Alicia Demirjian,Kevin E Brown,Shamez N Ladhani,Mary E Ramsay,Declan T Bradley,Ardiana Gjini,Kirsty Roy,Meera Chand,Maria Zambon,Conall H Watson
DOI: https://doi.org/10.1016/S2352-4642(23)00215-8
Abstract:Background: An increase in acute severe hepatitis of unknown aetiology in previously healthy children in the UK in March, 2022, triggered global case-finding. We aimed to describe UK epidemiological investigations of cases and their possible causes. Methods: We actively surveilled unexplained paediatric acute hepatitis (transaminase >500 international units per litre) in children younger than 16 years presenting since Jan 1, 2022, through notifications from paediatricians, microbiologists, and paediatric liver units; we collected demographic, clinical, and exposure information. Then, we did a case-control study to investigate the association between adenoviraemia and other viruses and case-status using multivariable Firth penalised logistic regression. Cases aged 1-10 years and tested for adenovirus were included and compared with controls (ie, children admitted to hospital with an acute non-hepatitis illness who had residual blood samples collected between Jan 1 and May 28, 2022, and without known laboratory-confirmed diagnosis or previous adenovirus testing). Controls were frequency-matched on sex, age band, sample months, and nation or supra-region with randomised selection. We explored temporal associations between frequency of circulating viruses identified through routine laboratory pathogen surveillance and occurrence of cases by linear regression. SARS-CoV-2 seropositivity of cases was examined against residual serum from age-matched clinical comparison groups. Findings: Between Jan 1 and July 4, 2022, 274 cases were identified (median age 3 years [IQR 2-5]). 131 (48%) participants were male, 142 (52%) were female, and one (<1%) participant had sex data unknown. Jaundice (195 [83%] of 235) and gastrointestinal symptoms (202 [91%] of 222) were common. 15 (5%) children required liver transplantation and none died. Adenovirus was detected in 172 (68%) of 252 participants tested, regardless of sample type; 137 (63%) of 218 samples were positive for adenovirus in the blood. For cases that were successfully genotyped, 58 (81%) of 72 had Ad41F, and 57 were identified as positive via blood samples (six of these were among participants who had undergone a transplant). In the case-control analysis, adenoviraemia was associated with hepatitis case-status (adjusted OR 37·4 [95% CI 15·5-90·3]). Increases in the detection of adenovirus from faecal samples, but not other infectious agents, in routine laboratory pathogen surveillance correlated with hepatitis cases 4 weeks later, which independently suggested an association (β 0·06 [95% CI 0·02-0·11]). No association was identified for SARS-CoV-2 antibody seropositivity. Interpretation: We observed an association between adenovirus 41F viraemia and paediatric acute hepatitis. These results can inform diagnostic testing recommendations, clinical management, and exploratory in vitro or clinical studies of paediatric acute hepatitis of unknown aetiology. The role of potential co-factors, including other viruses and host susceptibility, requires further investigation. Funding: None.