Longitudinal Analysis of the Utility of Liver Biochemistry as Prognostic Markers in Hospitalized Patients With Corona Virus Disease 2019

The association of liver biochemistry with clinical outcomes of severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2) infection is currently unclear, and the utility of longitudinally measured liver biochemistry as prognostic markers for mortality is unknown. We aimed to determine whether abnormal liver biochemistry, assessed at baseline and at repeat measures over time, was associated with death in hospitalized patients with COVID‐19 compared to those without COVID‐19, in a United Kingdom population. We extracted routinely collected clinical data from a large teaching hospital in the United Kingdom, matching 585 hospitalized patients who were SARS‐CoV‐2 real‐time reverse transcription‐polymerase chain reaction (RT‐PCR) positive to 1,165 hospitalized patients who were RT‐PCR negative for age, sex, ethnicity, and preexisting comorbidities. A total of 26.8% (157/585) of patients with COVID‐19 died compared to 11.9% (139/1,165) in the group without COVID‐19 (P < 0.001). At presentation, a significantly higher proportion of the group with COVID‐19 had elevated alanine aminotransferase (20.7% vs. 14.6%, P = 0.004) and hypoalbuminemia (58.7% vs. 35.0%, P < 0.001) compared to the group without COVID‐19. Within the group with COVID‐19, those with hypoalbuminemia at presentation had 1.83‐fold increased hazards of death compared to those with normal albumin (adjusted hazard ratio [HR], 1.83; 95% confidence interval [CI], 1.25‐2.67), while the hazard of death was ~4‐fold higher in those aged ≥75 years (adjusted HR, 3.96; 95% CI, 2.59‐6.04) and ~3‐fold higher in those with preexisting liver disease (adjusted HR, 3.37; 95% CI, 1.58‐7.16). In the group with COVID‐19, alkaline phosphatase (ALP) increased (R = 0.192, P < 0.0001) and albumin declined (R = −0.123, P = 0.0004) over time in patients who died. Conclusion: In this United Kingdom population, liver biochemistry is commonly deranged in patients with COVID‐19. Baseline hypoalbuminemia and rising ALP over time could be prognostic markers for death, but investigation of larger cohorts is required to develop a better understanding of the relationship between liver biochemistry and disease outcome.

Albumin is a nonspecific marker of liver function and has been less consistently assessed; it is typically reported in patient baseline characteristics, with limited investigation of its utility as a prognostic marker. However, a recent meta-analysis of 20 retrospective cohort studies from China reported lower baseline albumin levels in patients with severe COVID-19 compared to mild cases but with significant heterogeneity between studies. (14) Another meta-analysis demonstrated that hypoalbuminemia could be included in prognostic machine-learning models to predict severe COVID-19 or mortality. (15) Several studies have investigated potential associations between liver biochemistry and death in patients with COVID-19 (6,(8)(9)(10)(11)16,17) or included liver biochemistry in the development of predictive models. (18)(19)(20)(21) A report from Italy showed that ALP >150 U/L at hospital admission (without adjusting for relevant confounders) was associated with clinical deterioration in 292 patients with COVID-19, (10) and another study from the United States reported that peak ALT >5 times upper limit of normal (ULN) during admission was associated with death in a cohort of 2,273 patients with COVID-19. (11) However, another U.S. study reported that elevations in ALT and AST on admission were associated with length of stay, intensive care unit (ICU) admission, and intubation, but not death. (6) Studies from Wuhan, China, did not find associations of ALT (18,19) or AST (19) elevation on admission with death in patients with COVID-19, while several studies have reported associations of elevated total BR (16,19) and low albumin on admission (19)(20)(21) with risk of death. Other studies considered abnormality of liver tests as one composite variable (7) or a composite endpoint (ICU admission or death) as the primary outcome (8,9,17) without specifying which individual liver test abnormalities were associated with risk of death. Given these variable associations between liver biochemistry and COVID-19 outcomes, the prognostic value of liver biochemistry derangement in COVID-19 needs further evaluation.
Having established a clinical data pipeline through the National Institute for Health Research (NIHR) Health Informatics Collaborative (HIC), (22,23) our tertiary referral hospital in the United Kingdom is strongly placed to undertake analyses using electronic health data from hospitalized patients. Using this resource, we aimed to determine the prevalence of deranged liver biochemistry at baseline and over the disease course in patients with COVID-19 with comparison to a matched group of patients without COVID-19 admitted during the same period. We also aimed to determine whether baseline liver biochemistry derangement was associated with risk of death in patients with COVID-19 and to compare longitudinal changes in liver biochemistry between patients with COVID-19 who died and those who survived.

Data ColleCtion
We used routinely collected clinical data from Oxford University Hospitals (OUH) National Health Service (NHS) Foundation Trust, a large teaching hospital trust in the southeast of the United Kingdom that has ~1,000 in-patient beds. The data are collected by the local NIHR HIC team in Oxford, being drawn automatically from operational systems into a data warehouse and linked to produce a comprehensive record for each patient with a data validation process, as described in our HIC methods paper. (23) The management of the data set is governed by the NIHR HIC Data Sharing Framework. All the data used for this study were provided in anonymized form by OUH NHS Foundation Trust with the prior approval of the Trust Information Governance Team following the satisfactory completion of a data protection impact assessment.
The data extracted for this study included detailed information on demographics, body mass index (BMI), emergency admissions, blood test results, diagnostic codes, procedures, ICU admission, prescriptions, medicines administration, and discharge destination/ outcome for all patients admitted to OUH between January 1, 2020, and August 21, 2020.

inClusion anD eXClusion CRiteRia
To select eligible data for adults with/without COVID-19, the inclusion criteria were as follows: (a) at least one real-time reverse transcription-polymerase chain reaction (RT-PCR) nose/throat swab having been undertaken, which is the validated clinical test being deployed through our hospital diagnostic microbiology laboratory (patients with COVID-19 were defined by at least one positive result; patients negative for COVID-19 were defined by the absence of a positive result); (b) age ≥18 years when tested; (c) hospitalized patients; and (d) at least one episode of liver biochemistry recorded at the time or after the RT-PCR test (the set of liver biochemistry routinely tested by OUH clinical biochemistry laboratory comprises ALT, ALP, albumin, and BR; AST and GGT are not routinely tested). Exclusion criteria were as follows: (a) SARS-CoV-2 RT-PCR test results reported as invalid; (b) missing age; or (c) pregnancy (antenatal/delivery/postpartum). Patients were followed until death or until the last available clinical record.

DeFinitions
We defined baseline as the date of the first positive SARS-CoV-2 RT-PCR test for a patient with COVID-19 and the date of the first negative test for a patient without COVID-19 .  The normal ranges set by the hospital biochemistry laboratory are as follows: ALT, 10-45 IU/L; ALP,  30-130 IU/L; BR, 0-21 µmol/L; albumin, 32-50 g/L.  The reference ranges for other blood tests are provided in Supporting Table S1. We defined baseline liver biochemistry (ALT, ALP, BR, or albumin) as the liver biochemistry measured within 7 days of a SARS-CoV-2 RT-PCR test and baseline derangement as at least one abnormal result at this time point. We defined peak/nadir liver biochemistry derangement as abnormality of the highest/lowest value recorded, respectively, at any point during follow-up. We defined liver biochemistry recovery as normalization following derangement. The primary outcome was death during follow-up, and secondary outcomes included ICU admission and invasive ventilation.
Preexisting comorbidities were defined by a historic diagnosis of a disease before baseline; diagnosis codes retrieved are provided in Supporting Table S2. A full list of prescribed drugs searched for data extraction is provided in Supporting Table S3.

statistiCal analysis
Statistical analysis was performed using R version 4.0.2. All significance tests performed were two-sided. P < 0.05 was deemed statistically significant.

propensity score matching
We identified eligible adults and conducted a propensity score matching process to ensure the group with COVID-19 was comparable to the group without COVID-19 in terms of demographics and preexisting conditions. We used the following variables to calculate propensity scores: age, sex, ethnicity, and preexisting comorbidities (liver disease, diabetes mellitus [DM], hypertension [HTN], coronary heart disease [CHD], chronic kidney disease [CKD], and cancer). We performed propensity score matching using the package MatchIt, with the nearest neighbor method applied, the matching ratio and caliper size set as 1:2 and 0.1, respectively, and without replacement.

Comparison of patients With and matched patients Without CoViD-19
For continuous variables, we calculated median and interquartile range (IQR) or mean and SD and used the Wilcoxon test or t test for comparison. For categorical variables, we computed number and percentage and used the chi-square or Fisher's exact test for comparison. We used the Shapiro-Wilk test and graphical methods for normality checks. We further stratified patients with COVID-19 by the severity of disease into mild/moderate and severe/critical subgroups (using respiratory rate and oxygen saturation thresholds of ≤30 breaths/minute and ≥90%, respectively, for mild/ moderate illness; and thresholds of >30 breaths/minute or <90%, respectively, for severe illness, as established by the World Health Organization [WHO]). (24) We then compared liver biochemistries and outcomes of each severity subgroup with COVID-19 to those of the group without COVID-19.

investigation on Whether liver Biochemistry predicts outcomes in patients With CoViD-19
We compared the presence of clinical outcomes in the groups with and without COVID-19. We also performed Kaplan-Meier (K-M) analysis to compare the survival probability over time between the two groups. Within the groups with and without COVID-19, we compared demographics, BMI, comorbidities, and baseline and peak/nadir liver biochemistry between those who died versus those who survived. We then performed K-M analysis to compare the survival probabilities over time between subgroups with and without deranged baseline liver biochemistry. We used univariate and multivariate Cox proportional hazards models to investigate whether liver biochemistry predicted death, reporting hazard ratios (HRs) and 95% confidential intervals (CIs). Variables with P < 0.1 in univariate analysis and/or clinically important parameters (demographics, comorbidities, and drug use) were included for the final multivariate model. To investigate associations of additional patient characteristics with risk of death and the robustness of HRs to adjustment for additional confounders, we performed sensitivity/subset analysis whereby associations were investigated in a subset of patients who were not missing data for confounders. In addition, we performed receiver operating characteristic (ROC) analysis on baseline liver biochemistry to examine predictive ability for COVID-19 death, reporting the area under the curve (AUC), sensitivity, and specificity.

longitudinal analysis of liver Biochemistry in patients With CoViD-19
We compared liver biochemistry between the groups with and without COVID-19 at each time point (to examine differences between groups), and we investigated liver biochemistry changes over time by comparing liver biochemistry at subsequent time points to their baseline within each group. Within the group with COVID-19, we first examined the changing pattern over time of liver biochemistry in subgroups of patients who died and survived by fitting linear regression lines with 95% CIs and reporting Pearson's correlation coefficients and linear regression significance, as applied by other longitudinal analyses. (25) We then used mixed effects models (lme4, sjPlot packages) (26,27) to test if the change in liver biochemistry was significant by considering fixed effects (the follow-up time, subgroup, and subgroupby-time interaction) and random effect intercept for individuals.

iDentiFiCation, DemogRapHiCs, anD outComes oF gRoups WitH CoViD-19 CompaReD to gRoups WitHout CoViD-19
We identified 6,311 eligible patients (585 adults with SARS-CoV-2 infection and 5,726 without) according to the inclusion/exclusion criteria (Supporting Fig. S1). Based on our 585 patients positive for COVID-19, we matched a cohort of 1,165 patients negative for COVID-19. After matching, there were no significant differences in demographics and preexisting comorbidities between the two groups ( Table 1; Supporting Table S4). Median duration of follow-up was 58 (IQR, 14-104) days in the group with COVID-19 and 50 (IQR, 20-78) days in the matched group without COVID-19, with no significant difference in the monitoring duration of liver biochemistry (interval between first and last available liver biochemistry test dates) between the groups (Table 1). Within the group with COVID-19, median follow-up duration was 10 (IQR, 5-21) days for those who died and 79 (IQR, 36-115) days for those who survived. Admitting specialties are summarized in Supporting Table S5.

assessment oF liVeR BioCHemistRy in tHe gRoup WitH CoViD-19 CompaReD to tHe gRoup WitHout CoViD-19
Baseline liver biochemistry was available in 492 patients with COVID-19 and 974 patients without COVID-19. The median time interval between consecutive liver biochemistry measurements was 1 (IQR, 1-4) day and 2 (IQR, 1-7) days (P < 0.001) for the groups with and without COVID-19, respectively.
Over follow-up, the group with COVID-19 had more deranged liver biochemistries, with a higher median peak ALT (34 IU/L vs. 26 IU/L, P < 0.001),  Table S6). Patients with COVID-19 also had significantly higher median ALT and lower median albumin values at time points throughout follow-up (7, 14, 21, and 28 days) compared to the group without COVID-19 (all P < 0.05) ( Fig. 2A,B). In the group with COVID-19, median ALT increased at 7 and 14 days compared to baseline (both P < 0.05) (Fig. 2C) and median albumin decreased at 7 days (P < 0.0001) and remained at low levels at subsequent time points (Fig. 2D). We did not identify differences in ALP and BR over time between these groups other than at baseline or at 7 days follow-up (Supporting Fig. S2). Baseline vital signs were recorded for 423 patients with COVID-19 and used to stratify disease severity, with 65 severe/critical COVID-19 cases and 358 mild/moderate cases. As expected, a higher rate of death was observed in severe/critical cases compared to mild/moderate cases (43.1% vs. 27.1%, P = 0.01). A higher proportion (31% vs. 19%, P = 0.04) in the severe/critical group had elevated ALT at baseline. However, there were no other significant differences in liver biochemistry between mild/moderate and serve/critical cases at baseline or during follow-up ( Table 2).
Compared to the group without COVID-19, both the subgroups with mild/moderate COVID-19 and severe/critical COVID-19 had a significantly higher prevalence of abnormal liver biochemistry both at baseline and during follow-up, with significantly higher rates of mortality, ICU admission rate, and the proportion requiring invasive ventilation (Supporting Tables S7 and S8). These results were consistent with those from the comparison between the whole group with COVID-19 and the group without COVID-19.

DemogRapHiCs, ComoRBiDities, anD liVeR BioCHemiCal CHaRaCteRistiCs assoCiateD WitH moRtality in patients WitH CoViD-19
In the group with COVID-19, patients who died were significantly older than those who survived (median, 82 years vs. 66 years, respectively, P < 0.001); significantly more likely to have preexisting comorbidities, including liver disease, DM, CHD, and cancer; significantly more likely to have baseline hypoalbuminemia (72.6% vs. 52.9%, P < 0.001); and more likely to have nadir albumin below the lower limit of the normal (LLN) range during follow-up (96.2% vs. 72.7%, P < 0.001) ( Table 3). In patients who died, baseline ALP was higher (P = 0.007) and peak BR was more likely to be abnormal compared to those who survived (P = 0.016; Supporting Table S9). Equivalent data for the group without COVID-19 are provided in Supporting Table S10.
Survival curves and log-rank tests (unadjusted for relevant confounders) within the group with COVID-19 showed that hypoalbuminemia compared to normal albumin at baseline was significantly associated with an increase in mortality (Fig. 3A). Surprisingly an elevated baseline ALT compared to a normal ALT at baseline was significantly associated with an increase in survival (Fig. 3B). An elevated ALP or BR at baseline was not significantly associated with lower survival probability over time (Fig. 3C,D). In addition, for the subset of patients with COVID-19 who had prothrombin time measured, elevation of this parameter (and of international normalized ratio [INR]) at baseline was significantly associated with a lower survival probability (Supporting Fig. S3). In multivariate analysis (fully adjusted for demographics, comorbidities, and prescribed drug use before baseline) for the group with COVID-19, those with hypoalbuminemia (i.e., <32 g/L) at baseline had a 1.83-fold increased hazards of death compared to those with normal baseline albumin (adjusted HR, 1.83; 95% CI, 1.25-2.67). Those aged ≥75 years had a ~4-fold increased hazards of death compared to those aged <75 years (adjusted HR, 3.96; 95% CI, 2.59-6.04), and those with preexisting liver disease had a ~3-fold increased hazards of death than those without preexisting liver disease (adjusted HR, 3.37; 95% CI, 1.58-7.16) (Table 4). However, we found no significant association between baseline ALT and hazards of death (adjusted HR, 0.86; 95% CI, 0.53-1.38; P = 0.53) in the fully adjusted analysis (Table 4). A 1-unit (1 g/L) decrease in albumin at baseline was significantly associated with a 5% increase in hazards of death (adjusted HR, 1.05; 95% CI, 1.02-1.09), while an age increase by 10 years was associated with an 82% increase in hazards of death (adjusted HR, 1.82; 95% CI, 1.56-2.13) (Supporting Table S11). Similarly, a 1-unit decrease in nadir albumin during follow-up was significantly associated with a 7% increase in hazards of death (adjusted HR, 1.07; 95% CI, 1.04-1.10) (Supporting Table S12). In the group with COVID-19, baseline albumin was significantly negatively correlated with age in those that survived (R = −0.264, P < 0.0001) but not in those who died (R = −0.027, P = 0.75) (Supporting Fig. S4).
ROC analysis further demonstrated the prognostic value of baseline albumin for COVID-19 death, which had higher overall performance (AUC = 0.642) compared to that of baseline ALP, BR, or ALT (Fig.  4A). The combination of baseline liver biochemistry parameters did not significantly improve the performance (AUC = 0.659, P = 0.08) (Fig. 4A; Supporting Table S13). Considering the other predictors (age and liver disease) identified from the multivariate model, we found adding albumin can further improve prediction of death (AUC significantly improved from 0.711 to 0.752, P = 0.002) (Fig. 4B; Supporting Table S14).  Adding other baseline liver biochemistry parameters did not make further significant improvement (Fig. 4B).

taBle 2. CompaRison oF DemogRapHiCs, ComoRBiDities, outComes, anD Baseline anD peaK/ naDiR liVeR BioCHemistRies BetWeen CoViD-19 suBgRoups stRatiFieD By Disease seVeRity
For the subset of patients with COVID-19 who had BMI data (survived vs. died, 246 vs. 96), HRs did not change materially after additional adjustment for BMI in the multivariate analysis (Supporting Table S15). Baseline albumin was weakly correlated with BMI both in those who died and survived (R = 0.08, P = 0.42 vs. R = 0.099, P = 0.12, respectively) (Supporting Fig. S5).

longituDinal assessment oF liVeR BioCHemistRy patteRns in patients WHo DieD WitH CoViD-19 CompaReD to tHose WHo suRViVeD.
Within the group with COVID-19, patients who died during follow-up had significantly lower median albumin at baseline and at 7 and 14 days after a positive SARS-CoV-2 RT-PCR compared to the patients who survived (all P < 0.001) (Fig. 5A). There was no significant difference in ALT at any time point other than 7 days between those who died and survived (Fig.  5B). For patients who died, ALP was higher at baseline and 28 days and BR was higher at 7 days compared to those who survived (all P < 0.05) (Fig. 5C,D).
Consistently, the mixed effects model also showed that within the subgroup who died, albumin significantly decreased over time (β = −0.12; 95% CI, −0.14 to −0.10; P < 0.001) while ALP significantly increased over time (β = 0.50; 95% CI, 0.29-0.70; P < 0.001), but the change in ALT or BR was not significant (P = 0.817, P = 0.489, respectively) (Supporting Table  S16). Considering the interaction between subgroup and time, the mixed effects model further revealed that (a) changes in albumin were significantly different over time between subgroups that died vs. survived (P < 0.001), with a decrease in the subgroup that died (β_time + 1 × β_interaction = −0.11) and an increase in the subgroup that survived (β_time + 0 × β_interaction = 0.08) based on the coefficients of time and interaction; (b) changes in ALP were also significantly different over time for the subgroups (P = 0.005), increasing in the subgroup that died (β_time + 1 × β_ interaction = 0.50) while decreasing in the subgroup that survived (β_time + 0 × β_interaction = −0.07); (c) changes in ALT (P = 0.126) or BR (P = 0.356) were insignificantly different over time for the subgroups that died and survived (Table 5).
Among patients with COVID-19 who had ≥2 longitudinal data points for ALT, ALP, BR, and albumin, 468 had at least one liver biochemistry derangement during follow-up. Among them, 26 still had one or more abnormal liver biochemistries at the end of follow-up (Supporting Table S17).

Discussion
We used an automatic approach through an established NIHR HIC clinical data collating bioinformatic pipeline (22,23) to capture a complete record of relevant clinical and laboratory parameters from hospitalized patients within the time period assessed. This resource allows us to fully analyze liver  biochemistry abnormalities and outcomes on a large cohort of patients with COVID-19 and propensity score-matched controls without COVID-19 for the first time. To our knowledge, this is the first study to comprehensively i) conduct longitudinal analyses of liver biochemistry patterns over time in patients with COVID-19 compared to a matched cohort of patients without COVID-19 attending hospital in the United Kingdom; ii) investigate whether albumin, in addition to other liver biochemistries, at baseline and during follow-up is associated with death in patients with COVID-19; and iii) analyze longitudinal liver biochemistry patterns in patients with COVID-19 who subsequently die or survive. The group with COVID-19 exhibited a ~2-fold higher death rate with a significantly lower survival probability compared to the group without COVID-19. Among the group with COVID-19, a higher proportion of patients had at least one abnormal liver biochemistry compared to the matched group without COVID-19. Patients with COVID-19 who died showed a decline in albumin and a greater increase in ALP over time compared to those who survived, and baseline hypoalbuminemia was a significant predictor of death in patients with COVID-19 on multivariate analysis with adjusting for relevant confounders.
In our study, rates of baseline and peak ALT derangement between the groups with and without COVID-19 were significantly different, consistent with findings Analysis were performed on the 492 patients with COVID-19 who had baseline liver biochemistry data (survived vs. died, 346 vs. 146); the cut-off age was based on the median age in the cohort. Liver biochemistry variables with P < 0.1, demographics, and comorbidities in univariate analysis were included for multivariate analysis. In multivariate analysis, HRs were fully adjusted for drug use before baseline (including antiviral drugs, antibiotics, anticoagulants, acetaminophen, immunosuppressants, statins) to reduce confounding effects.  Table 4. Baseline albumin + ALP + BR + ALT in the figure indicates including all baseline liver biochemistry parameters in the ROC analysis.
from a large US cohort. (11) The increase in ALT between baseline and at 14 days follow-up in the group with COVID-19 is also consistent with previous results. (6) Patients with preexisting liver disease had an increased risk of mortality in COVID-19, which is consistent with the findings of published studies, (28,29) although numbers with preexisting liver disease in our study were small. Because some patients with COVID-19 in our cohort were admitted to hospital for non-COVID-19 illnesses, we analyzed baseline liver biochemistry at the time of the SARS-CoV-2 RT-PCR rather than date of admission. This may partially explain differences between our study and previous studies that analyzed liver biochemistry measured on hospital admission. (3)(4)(5)(6) Variable patterns of treatment between cohorts may also account for differences.
Although baseline hypoalbuminemia was significantly associated with hazards of death in COVID-19, ALP and BR were not. Recent findings from two multicenter studies support the prognostic association of albumin with death in COVID-19, (21,30) and a previous prospective study also reported an association of albumin with a composite endpoint (ICU admission or death). (17) However, other studies of liver biochemistry in COVID-19 did not investigate albumin. (6,10,11) Interestingly, a previous study found that hypoalbuminemia is a strong predictor of 30-day all-cause mortality in acutely admitted medical patients. (31) The mechanism of the effect of hypoalbuminemia is not certain; it may reflect the broad association between low albumin and critical illness or be a marker for characteristics of host (e.g., nutritional status, comorbidity) or disease phenotype (e.g., immune activation). (32) As albumin is a cheap and widely available test, it can be usefully employed as a prognostic biomarker.
Due to variable population settings in previous studies (e.g., demographics, comorbidities), it is important to understand the populations in which prognostic models are developed (15) to ensure such models are externally valid, and to also adjust for relevant confounders. In our cohort, a univariate Cox proportional hazards model and K-M curve analysis suggest that an elevation in baseline ALT is weakly associated with lower hazards of death from COVID-19. However, the univariate analysis is unadjusted and may be confounded by additional factors. Interestingly, in a US cohort of 60 hospitalized patients with COVID-19, ALT at admission was also higher in those who survived compared to those who died. (6) In our multivariate analysis, the association of baseline ALT and hazards of death was attenuated toward the null and became nonsignificant after full adjustment for demographics, comorbidities, and prescribed drug use before baseline. Therefore, larger cohorts are warranted to investigate this unexpected association with more confidence. In the sensitivity analysis of patients with COVID-19 stratified by disease severity, severe/ critical cases were more likely to have elevated ALT at baseline compared to mild/moderate cases and were more likely to die compared to mild/moderate cases, in line with other studies. (3,12,13) The observed association of raised ALT with lower hazards of death is intriguing as recent data have suggested that genetic predisposition to fatty liver disease, conferred by possession of the patatin-like phospholipase domain containing 3 [PNPLA3] I148M variant (which is highly prevalent in the UK population), (33) is associated with both elevated ALT and a concomitant reduction in systemic inflammation (as measured by CRP) as well as preservation of albumin levels during COVID-19. (34) As such, elevated ALT at baseline may be a surrogate for nonalcoholic fatty liver disease (NAFLD) in a subset of patients; those with NAFLD may have a lower risk of severe COVID-19 outcomes. However, additional studies are required to confirm this association, and the underlying mechanism is not understood.
A study from the United States (11) reported a positive association of ALT >5 times ULN with death; however, we were unable to replicate this investigation as only a small number of participants had ALT elevated to this level. Furthermore, the prevalence of comorbid disease and BMI >35 kg/m 2 in the US study population was much higher compared to our cohort. Similarly, a large primary care cohort study (35) reported that BMI >40 kg/m 2 was associated with an increased risk of COVID-19-related death, but we were underpowered to replicate this analysis.
Although derangements in liver biochemistry are common in patients with COVID-19, the reasons for the liver injury remain unclear but may include direct viral damage, drug-induced liver injury, hypoxia, immune-mediated injury, sepsis, or cytokine release. (36,37) Angiotensin-converting enzyme 2 (ACE2), a functional receptor for SARS-CoV-2, (38) is found abundantly in the gastrointestinal tract and liver in addition to presenting in alveolar type 2 cells (the major SARS-CoV-2-targeting cell type in lung). A recent study observed a higher expression of ACE2 in cholangiocytes (~60% of cells) compared to hepatocytes (<3%). (39) Given the hepatic distribution of the ACE2 receptor, SARS-CoV-2 may well cause damage of both bile ducts and liver. (40,41) Our stratification analysis by disease severity revealed that compared to patients without COVID-19, patients with COVID-19 were more likely to have abnormal liver biochemistry and severe outcomes. Alternatively, the liver may be a bystander, with deranged liver biochemistry reflecting systemic disease. (42) It is interesting that liver biochemistry parameters are not currently included in the existing risk stratification tools, such as International Severe Acute Respiratory and Emerging Infection Consortium (ISARIC) scores, for predicting clinical deterioration (43) or mortality risk (44) for COVID-19. These parameters (or albumin in particular) may be considered for inclusion in future scoring models or used for early ICU review.
Routinely collected liver biochemistry is not consistent between settings, and therefore the definitions of liver biochemistry derangement may vary across studies. Although AST and GGT have been investigated in previous studies, these parameters were not available for our population. It would be interesting to include GGT in future analysis, and correlation of liver function tests with blood gases would be an approach to determine the extent to which deranged liver biochemistry may be associated with liver hypoxia. Other information, like respiratory parameters, could be included if available in the propensity score matching and multivariate analysis for future analyses. We recognize that analysis can be influenced by missing data, but we have reported missing values and investigated peak/nadir values in addition to baseline. We also undertook sensitivity analysis in a subset of patients with complete BMI measurements to investigate its association with death. Treatment outcomes are important; however, as changing approaches have been made to treatment for COVID-19 and the drugs used as part of clinical trials were not recorded in the current data set, we have not undertaken an analysis of responses to treatment in this study. Our cohort in the southeast of the United Kingdom may not be representative of populations elsewhere, especially in terms of ethnic diversity, so caution should be applied in extrapolation of results.
Further longitudinal studies of COVID-19 outcomes in diverse patient groups, including those with preexisting liver disease, are needed. The NIHR HIC program will continue to benefit the field of COVID-19 research because data accumulated for large teaching hospitals can be used to expand this analysis, resulting in a more generalizable study population and increased statistical power. In conclusion, liver biochemistry derangement is common in patients with COVID-19 at the time of a SARS-CoV-2 RT-PCR test and during the clinical course of disease. Baseline hypoalbuminemia and rising ALP over time are prognostic markers for death in patients with COVID-19, but investigations of larger cohorts are required to develop a better understanding of the relationship between liver biochemistry and disease outcome.