Reply to: "Liver tests abnormalities in COVID-19: trick or treat?"

To the Editor:

We thank the Humanitas COVID-19 Task Force for their interest in our manuscript and for their thoughtful comments.

In their letter, Vespa et al., pointed out that our finding of a 41% prevalence of elevated liver function tests (LFTs) at admission appears to be higher than rates reported throughout Asia and particularly, the rates reported in their cohort. We reviewed and summarized the articles mentioned in their letter (Table 1 ).[1], [2], [3], [4], [5], [6] We found that the prevalence of elevated LFTs at admission in our study was significantly lower than that reported by Vespa et al. : alanine aminotransferase (ALT, 54/417 vs. 78/292, p <0.001), gamma-glutamyltransferase (GGT, 68/417 vs. 106/292, p <0.001) or alkaline phosphatase (ALP, 20/417 vs. 28/292, p = 0.012) and higher for total bilirubin (TBIL, 97/417 vs. 31/292, p <0.001), but comparable with other patient series from Asia.[3], [4], [5], [6] In fact, due to the differences in the cut-off values of elevated LFTs, underlying liver disease, and demographic factors, the prevalence of elevated LFTs in patients with COVID-19 at admission may vary widely. Of note, it has been observed that the prevalence of elevated LFTs would increase in the course of COVID-19, as seen in our study, which found that the time from disease onset to hospitalization significantly influenced the elevated rate of LFTs at admission.

Table 1

The prevalence of elevated LFTs reported in different cohorts of patients with COVID-19.

	Sample size	Elevated LFTs (%)	Elevated LFTs, cut-off value, %					Overall prognosis
			ALT	AST	TBIL	GGT	ALP
Cai et al.1	417	41%	>40 U/L, 13.0	>40 U/L, 18.3	>17.1 μmol/L, 23.2	>49 U/L, 16.3	>135 U/L, 4.8	91 (21.8%) developed severe disease during hospitalization.
Vespa et al.2	292	n/a	>50 U/L, 26.7	>50 U/L, 18.5	>1.2 mg/dl, 10.6	>55 U/L, 36.2	>150 U/L, 9.6	27 (9.2%) patients were transferred to ICU, and 56 (19 .2%) died.
Fan et al.3	147	37.2%	>40 U/L, 18.2	>35 U/L, 21.6	>20.5 μmol/L, 6.1	>45 U/L, 17.6	>100 U/L, 4.1	1 patient died.
Zhang et al.4	115	n/a	>50 U/L, 9.6	>40 U/L, 14.8	>21 μmol/L, 6.1	>57, 13.0	>120, 5.2	31 cases (26.96%) were evaluated as severe cases, 7 (6.09%) patients progressed to critical condition and 1 died (0.87%).
Chen et al.5	99	43.4%	>50 U/L, 28.3	>40 U/L, 35.4	>21 μmol/L, 18.2	n.a.	n.a.	17 (17%) patients developed acute respiratory distress syndrome and 11(11%) patients died.
Guan et al.6	1099	n/a	>40 U/L, 21.3	>40 U/L, 22.2	>17.1 μmol/L, 10.5	n.a.	n.a.	173 (15.7%) developed severe disease during hospitalization.

Comparison of categorical variables was done using the χ2 test or the Fisher exact test, if the cell counts were small. ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; GGT, gamma-glutamyltransferase; ICU, intensive care unit; LFT, liver function test; n.a., not available; TBIL, total bilirubin.

Because of the limited medical resources available during the beginning of the pandemic, many hospitals in high pandemic areas may have only received patients with severe disease, which would result in seeing patients with more severe liver injury. In our study, all laboratory-confirmed patients with COVID-19 who were admitted to the only referral hospital in Shenzhen, China from January 11 to February 21, 2020 were included. Elevated LFTs were defined as the following criteria: AST or ALT values >40 IU/L, GGT >49 U/L, ALP >135 U/L, and TBIL >17.1 mol/L. Of the 417 patients in our studies, 21 (5.04%) had underlying liver disease, 123 (32%) were overweight, and 41 (10.7%) were obese at admission. The median time from disease onset to hospitalization was 4 days (IQR 1–5).

So far, older age has been acknowledged as one of the most important risk factors for deterioration due to COVID-19. Others at highest risk of severe COVID-19 are people with certain underlying conditions, including hypertension, cardiovascular disease, diabetes, chronic respiratory disease, cancer, renal disease, and obesity. Some studies have also reported that abnormal LFTs were associated with overall prognosis. , Fan et al., for example, reported that baseline abnormal liver function was associated with prolonged hospital stay. In a recent systematic review, Kunutsor and Laukkanen conducted a meta-analysis of 16 retrospective cohort studies based in China and concluded that elevated levels of liver injury markers, particularly aminotransferases, may be associated with progression to severe disease or death. As seen in Table 1, the cohorts with higher mortality usually had more severe LFT changes.

In the study by Vespa et al., elevated ALP above 150 IU/L was associated with deterioration (p = 0.048) and became non-significant after adjustment for age and sex (p = 0.13). In our study, GGT was substantially elevated at admission, whereas the increase in ALP was not pronounced in patients with COVID-19. This finding was consistent with other cohorts.[3], [4], [5], [6] The patients with COVID-19 showed elevated GGT levels and normal ALP, which may controvert the hypothesis that SARS-CoV-2 directly injures cholangiocytes. Therefore, we agree that ALP or GGT elevation may just be a marker of patient frailty or represent a stronger systemic inflammatory response to SARS-CoV-2 infection, rather than a sign of cholestatic liver injury. Liver and lung injuries share the same underlying mechanism, which may be directly or indirectly induced by SARS-CoV-2 infection. Laboratory liver function tests should thus be interpreted with caution and always in the context of a complex, multi-organ disease.

Financial support

This work is funded by the National Infectious Diseases Clinical Research Center, Funds for the construction of key medical disciplines in Shenzhen, the Sanming Project of Medicine in Shenzhen (SZSM201612014), and the Bill and Melinda Gates Foundation.

Authors' contributions

J Chen designed the study, received the grant support, had full access to all data in the study, and takes responsibility for the integrity of the data and the accuracy of the data analysis. QX Cai and L Xu contributed to the writing and statistical analysis of the report. All authors contributed to data acquisition, data analysis, and data interpretation, and reviewed and approved the final version of the manuscript.

Conflict of interest

The authors declare no conflicts of interest that pertain to this work.

Please refer to the accompanying ICMJE disclosure forms for further details.