Study population and setting
This study retrospectively investigated the prevalence of acute liver injury (ALI) and associated clinical characteristics in all patients tested for SARS-CoV-2 infection with RT-PCR between March 8 and April 14, 2020 at three hospitals in the New York Presbyterian network (n=2273 with positive test and n=1108 with negative test). Study patients were a median age of 65 years old and 57% male, with a racial and ethnic composition of 50% Hispanic/Latino ethnicity, 23% white, 21% Black, 0.9% Asian, and 56% other/unknown race. The prevalence of specific comorbidities within the cohort included hypertension (60%), diabetes (39%), and chronic kidney disease (21%).
Summary of Main Findings
Patients who tested positive had significantly higher rates of ALI across severities of ALI with 45% (vs. 26% in test negative) having mild (peak ALT > upper limit of normal [ULN]), 21% (vs. 12%) having moderate (peak ALT > 2 to 5 times ULN), and 6.4% (vs. 5%) having severe (peak ALT > 5 times ULN) liver injury. Among participants who tested positive for SARS-CoV-2, severe livery injury (SLI) was more common in younger and male patients but was not associated with elevated bilirubin or alkaline phosphatase. Using multivariable logistic regression, SLI among patients with a positive SARS-CoV-2 test was associated with elevated inflammatory markers including peak ferritin (OR 2.40; 95% CI: 1.90-3.08) and peak interleukin-6 (OR 1.45; 95% CI: 1.10-1.93), but not d-dimer, c-reactive protein, procalcitonin, creatinine, or troponin. A multivariable model controlling for BMI, hypertension, and renal replacement therapy use suggested that peak ALT (OR 1.14; 95% CI: 1.00-1.30), older age (OR 1.07; 95% CI: 1.07-1.10), diabetes (OR 1.30; 95% CI: 1.01-1.68), and intubation (OR 4.77; 95% CI: 3.49-6.55) were associated with severe COVID-19 outcomes (death or discharge to hospice)
This study included a racially and ethnically diverse study population representative of groups heavily impacted by COVID-19 in urban settings. The thresholds for mild, moderate, and severe ALI were clearly defined, standardized, and assessed. This study also utilized regression analysis to control for many potential confounders.
This study looked at an exclusive inpatient population at three hospitals in a single hospital network located in one city at a notable peak in the epidemic, which may limit generalizability to outpatient settings, persons with milder disease, or other geographical settings. Extrahepatic contributions to ALT were not able to be characterized, which may result in an overestimate of ALI prevalence related to SARS-CoV-2 infection. This study is non-randomized, therefore unassessed confounders may exist. Patients without available alanine aminotransferase (ALT) test results were excluded (n=3532), which likely induced selection bias as individuals with available ALT results available are different from patients without ALT test results and availability of ALT testing is likely related to SARS-CoV-2 testing.
This study is among the first to elucidate the prevalence and clinical associations of acute liver injury in a large representative cohort.
This review was posted on: 20 November 2020