HEP 4-20-0462 Longitudinal analysis of the utility of liver biochemistry in hospitalised COVID-19 patients as prognostic markers

Affiliations: 1. NIHR Oxford Biomedical Research Centre, Big Data Institute, University of Oxford, United Kingdom 2. Nuffield Department of Medicine, University of Oxford, United Kingdom 3. Oxford University Hospitals NHS Foundation Trust, United Kingdom 4. Department of Computer Science, University of Oxford, United Kingdom 5. Oxford Liver Unit, Translational Gastroenterology Unit, John Radcliffe Hospital, Oxford University Hospitals, United Kingdom 6. Cambridge Liver Unit, Addenbrooke's Hospital, Cambridge, United Kingdom 7. Department of Infectious Diseases and Microbiology, Oxford University Hospitals NHS Foundation Trust, United Kingdom

Within a year of the first case being reported there had been over 100 million confirmed cases of 2 Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) infection and ~2.5 million 3 deaths have been reported globally by the end of February 2020; the UK is one of the worst 4 affected countries with over 4 million confirmed cases and over 121,000 deaths reported in this 5 time period (1). The clinical syndrome caused by SARS-CoV-2, COVID-19, primarily affects the 6 respiratory system but other organs, including the heart, gastrointestinal tract and liver may be 7 affected, and a systemic sepsis syndrome may develop (2). 8 9 Data on liver biochemistry in COVID-19 patients have been reported from China, the USA and 10 Italy. . These studies report that 37-69% of patients with COVID-19 had at least one abnormal liver 11 biochemistry on hospital admission (3-9) while 93% had at least one abnormal liver biochemistry 12 over the course of disease (6). Specially, the prevalence estimates of elevated alanine 13 aminotransferase (ALT), aspartate aminotransferase (AST) and bilirubin (BR) in hospitalised 14 COVID-19 patients are 9%-28%, 14%-35% and 6%-23%, respectively (3-6, 10-12). Some studies 15 reported that liver biochemistry abnormalities are associated with longer hospital stay (4), or 16 clinical severity (3,12,13), whereas other studies have not found a relationship between liver 17 biochemistry and severity (10). The set of liver biochemistry tests reported for COVID-19 patients 18 varies: ALT, AST, and total BR are typically included, with alkaline phosphatase (ALP) and 19 Gamma-glutamyl transferase (GGT) less frequently reported.

21
Albumin is a non-specific marker of liver function, and has been less consistently assessed; it is 22 typically reported in baseline patient characteristics, with limited investigation of its utility as a 23 prognostic marker. However, a recent meta-analysis of 20 retrospective cohort studies from 24 China reported lower baseline albumin levels in patients with severe COVID-19 compared to mild 25 cases, but with significant heterogeneity between studies (14). Another meta-analysis 26 demonstrated that hypoalbuminaemia could be included in prognostic machine learning models 27 to predict severe COVID-19 or mortality (15).

29
Several studies have investigated potential associations between liver biochemistry and death in 30 COVID-19 patients (6, 8-11, 16, 17), or included liver biochemistry in the development of 31 predictive models (18)(19)(20)(21). A report from Italy showed that ALP >150 U/L at hospital admission 32 (without adjusting for relevant confounders) was associated with clinical deterioration in 292 33 COVID-19 patients (10), and another study from the USA reported that peak ALT >5 times upper 34 limit of normal (ULN) during admission was associated with death in a cohort of 2,273 COVID-19 35 patients (11). However, another USA study reported that elevations in ALT and AST elevation on Accepted Article (NIHR) Health Informatics Collaborative (HIC) (22, 23), our tertiary referral hospital in the UK is 48 strongly placed to undertake analyses using electronic health data from hospitalised patients.

49
Using this resource, we aimed to determine the prevalence of deranged liver biochemistry at 50 baseline and over the disease course in COVID-19 patients, with comparison to a matched group 51 of non-COVID-19 patients admitted during the same period. We also aimed to determine whether 52 baseline liver biochemistry derangement was associated with risk of death in 53 and to compare longitudinal changes in liver biochemistry between COVID-19 patients who died 54 and who survived.

58
Data collection 59 We used routinely collected clinical data from Oxford University Hospitals (OUH) National Health 60 Service (NHS) Foundation Trust, a large teaching hospital trust in the South East of the UK, with 61~1000 in-patient beds. The data are collected by the local NIHR HIC team in Oxford, being drawn 62 automatically from operational systems into a data warehouse and linked to produce a 63 comprehensive record for each patient with a data validation process, as previously described in 64 our HIC methods paper (23). The management of the dataset is governed by the NIHR HIC Data

198
Over follow-up, the COVID-19 group had more deranged liver biochemistry, with a higher median 199 peak ALT (34 IU/L vs. 26 IU/L, p<0.001), a higher proportion with peak ALT >ULN (37.9% vs.  Table S6). COVID-19 patients also had significantly higher median ALT and 203 lower median albumin values at time points throughout follow-up (7, 14, 21, and 28 days), 204 compared to the non-COVID-19 group (all p<0.05) (Figure 2A-B). In the COVID-19 group, 205 median ALT increased at 7 and 14 days, compared to baseline (both p<0.05) (Figure 2C), and 206 median albumin decreased at 7 days (p<0.0001) and remained at low levels at subsequent time Accepted Article points ( Figure 2D). We did not identify differences in ALP and bilirubin over time between these 208 groups, other than at baseline or at 7 days (Supporting Figure S2). with an increase in mortality ( Figure 3A). Surprisingly an elevated baseline ALT, compared to a 239 normal ALT at baseline was associated with an increase in survival ( Figure 3B). An elevated 240 ALP or bilirubin at baseline were not significantly associated with a lower survival probability over 241 time ( Figure 3C-D). In addition, for the subset of COVID-19 patients who had prothrombin time Accepted Article associated with a lower survival probability (Supporting Figure S3).

262
The ROC analysis further demonstrated the prognostic value of baseline albumin for COVID-19 263 death, which has higher overall performance (AUC=0.642) compared to that of baseline ALP, BR 264 or ALT ( Figure 4A). The combination of baseline liver biochemistry parameters does not 265 significantly improve the performance (AUC=0.659, p=0.08) ( Figure 4A, Supporting Tables   266  S13). Considering the other predictors (age and liver disease) identified from the multivariate 267 model, we found adding albumin can further improve prediction of death (AUC significantly 268 improved from 0.711 to 0.752, p=0.002, Figure 4B, Supporting Tables S14). Adding other 269 baseline liver biochemistry parameters did not make further significant improvement ( Figure 4B).  Figure S5).

275
Accepted Article Longitudinal assessment of liver biochemistry patterns in patients who died with COVID-277 19, compared to those who survived.

278
Within the COVID-19 group, patients who died during follow-up had significantly lower median 279 albumin at baseline, 7 and 14 days after a positive SARS-CoV-2 RT-PCR, compared to the 280 patients who survived (all p<0.001) ( Figure 5A). There was no significant difference in ALT at 281 any time point other than 7 days between those who died and survived ( Figure 5B). For patients 282 who died, ALP was higher at baseline and 28 days and BR was higher at 7 days compared to 283 those who survived (all p<0.05) (Figure 5C-D).

Accepted Article
This article is protected by copyright. All rights reserved  Accepted Article gastrointestinal and liver involvement in patients with COVID-19: a systematic review and meta-analysis.

Accepted Article
This article is protected by copyright. All rights reserved