Serology versus nucleic acid amplification to diagnose acute hepatitis E, United Kingdom, 2014-18

Abstract

Objectives: Diagnosing hepatitis E infection usually involves specific IgM testing, but sensitivity/specificity concerns mean many guidelines and practices include confirmatory tests. We studied whether additional information confirmatory tests provide justifies their use.

Methods: We examined 9131 records of anti-hepatitis E IgM assays, 7615 of IgG assays, and 1726 of RT-PCR assays from our regional laboratory, spanning October 2014-October 2018. We paired 495 IgM assay results with a RT-PCR result. We examined whether IgM results predicted PCR results, reviewed discrepant pairs, and investigated the correlation between IgG and PCR results in patients with strongly reactive IgM assays.

Results: Anti-hepatitis E IgM titres are bimodal. A high cut-off value optimises prediction of RNA detectability. 7/404 low-IgM samples had detectable RNA, 6 from immunosuppressed patients. 26/91 high-IgM samples did not have detectable RNA. In high-IgM samples, RNA detectability was not associated with IgG titre (one-tailed Mann-Whitney U test, p = 0.14).

Conclusions: In immunocompetent patients, tests beyond IgM seldom add clinically useful information. In patients with immunocompromise, IgM and RNA could contribute information. Additional tests' extra costs/intervention delays cannot be justified. IgM assay cut-offs should reflect titres' bimodal distribution, with values standardised using international units.

Keywords: Hepatitis E; Hepatitis E virus; Polymerase chain reaction; Serologic tests.