Predictors of mortality and liver transplant requirement in patients with hepatitis A virus: A case series from Mexico

Abstract

BACKGROUND

Hepatitis A virus (HAV) infection remains a significant cause of acute viral hepatitis globally, and the endemicity pattern is intermediate in Mexico. Although most cases follow a benign and self-limited course, less than 5% of patients with HAV infection progress to acute liver failure, which carries mortality rates of 13%-33% depending on access to liver transplantation (LT). The identification of clinical and biochemical predictors of adverse outcomes is crucial to improve early risk stratification, optimize patient management, and guide timely referral to transplant centers.

AIM

To assess the utility of clinical and biochemical variables and prognostic scores in predicting mortality and the need for LT in patients with acute HAV infection.

METHODS

We conducted a retrospective case series of all patients hospitalized at Medica Sur Clinic between 2018 and 2024 who were positive for immunoglobulin M antibodies for HAV. Clinical and laboratory data were collected from patients’ electronic medical records. Continuous variables were summarized as the means ± SD or median and interquartile range and compared using Student’s t-test or the Mann-Whitney U test, as appropriate. Categorical variables were expressed as frequencies and percentages and compared using the χ² test or Fisher’s exact test.

RESULTS

A total of 69 patients with HAV infection (mean age, 38 ± 11 years; 65% male) were included. 14% of patients had a history of hepatic steatosis, 10% had type 2 diabetes, and 45% were smokers. Acute liver injury was observed in 47 patients (68%), acute liver failure was detected in five patients (7.2%), and mortality or LT occurred in three patients (4.3%; two deaths and one transplant). Patients who died or required orthotopic LT were older, and they had higher heart and respiratory rates at admission. These patients additionally exhibited greater liver dysfunction, including higher bilirubin and transaminase levels, prolonged coagulation times, lower sodium and albumin levels, and worse albumin-bilirubin scores. Severity scores [model for end-stage liver disease (MELD), MELD with sodium, MELD version 3.0] were significantly higher in patients with complications than in those without complications.

CONCLUSION

Advanced age, elevated severity, and higher albumin-bilirubin scores emerged as predictors of adverse outcomes in patients with HAV infection, and consideration of these factors could guide the need for more intensive monitoring strategies. These findings highlight the importance of incorporating prognostic scoring systems into routine clinical evaluation to improve outcomes in high-risk patients.

Key Words: Hepatitis A virus; Acute liver failure; Liver transplantation; Mortality risk factors; Predictors; Risk stratification; Mexico; Prognostic scoring systems; Liver function tests

Core Tip: Hepatitis A virus infection is among the most common causes of acute viral hepatitis globally. Predicting the clinical course of hepatitis A virus infection remains challenging. Risk factors such as encephalopathy, leukocytosis, acute kidney injury, hypoalbuminemia, and hyperbilirubinemia can be used to identify patients at risk for developing acute liver failure. Previously, we reported a fulminant hepatitis incidence of 0.3%, which increased to 1.8% in individuals older than 49 years, with serum creatinine level > 2 mg/dL being the strongest predictor of fulminant hepatitis or mortality. Given the epidemiological transition of this disease in Mexico, updated data are crucial.

INTRODUCTION

In Mexico, hepatitis A virus (HAV) infection maintains an intermediate endemicity pattern, with the prevalence being highest among individuals aged 5-9 years, followed by those aged 25-44 years. In the context of an ongoing epidemiological transition, a growing proportion of adults are susceptible to infection, increasing the risk of severe clinical presentation[1]. Less than 5% of cases develop acute liver failure (ALF), which carries mortality rates of 13%-33%, depending on disease severity and access to liver transplantation (LT). Therefore, any patient with ALF should be referred to a transplant center[1,2]. Although most HAV cases are self-limited, a small proportion can progress to ALF. Thus, identifying clinical and biochemical predictors of adverse outcomes is crucial to improve risk stratification, optimize patient management, and guide timely referral to liver transplant centers[2-4]. Vaccination against HAV is the most effective strategy to prevent infection[5-7]. In Mexico, although HAV vaccination is not part of the national immunization schedule, it is recommended for high-risk groups, and it is available in specialized centers. Given the complexity and variability in the clinical progression of hepatitis A, it is imperative to deepen research efforts and gain a clearer understanding of mortality predictors and the need for LT in our population. Identifying these factors will both enable better risk stratification and more effective patient management and optimize the use of limited resources, such as organs available for transplantation.

MATERIALS AND METHODS

The study was reviewed and approved by the Research Ethics Committee of Medica Sur Clinic (CONBIOETICA-09-CEI-018-20160729). We conducted a retrospective case series including the medical records of 73 patients hospitalized at our facility between 2018 and 2024 with confirmed HAV infection (HAV-immunoglobulin M positivity). To ensure sample homogeneity, three patients younger than 18 years were excluded, and one additional patient was excluded because of missing data. The final analysis included 69 patients. Although no formal sample size calculation or power analysis was performed all consecutive eligible patients during the study period were included because of the retrospective nature of this case series, minimizing selection bias.

Data collection

Clinical and laboratory variables potentially influencing the primary outcome were collected from the electronic medical records at the time of hospital admission. Missing data were minimal. When a variable was missing for a given patient, it was excluded from that specific analysis, but the patient was retained for the rest of the study variables (complete-case analysis).

Laboratory methods

All laboratory tests were performed in the ISO 15189-accredited central clinical laboratory of Medica Sur Clinic following standardized internal and external quality control programs. Hematological parameters were analyzed using an automated Sysmex XN-Series analyzer with manual review of flagged results. Biochemical parameters, including liver function indices, creatinine, electrolytes, and C-reactive protein, were measured on a Beckman Coulter AU5800 automated analyzer using standard enzymatic and colorimetric methods. Bilirubin fractions were determined using the Jendrassik-Grof method, albumin was measured using bromocresol violet staining, and coagulation parameters [international normalized ratio (INR)] were monitored using a photo-optical coagulometer (STA Compact Max 3, Diagnostica Stago). The reference ranges for variables were as follows: Alanine aminotransferase, 9-47 U/L; aspartate aminotransferase, 19-40 U/L; total bilirubin, < 1.2 mg/dL; albumin, 3.8-4.7 g/dL; INR, 0.8-1.2; creatinine, 0.80-1.28 mg/dL; sodium, 138-145 mmol/L; and potassium, 3.9-5.3 mmol/L. All tests were processed under strict quality control with daily calibration.

Prognostic scores

The model for end-stage liver disease (MELD), MELD adjusted for serum sodium (MELD-Na), and MELD 3.0 (updated MELD including albumin and sex) scores were used to assess short-term mortality risk prior to LT[8-11], whereas the albumin-bilirubin (ALBI) index was used to evaluate liver function at the time of diagnosis[12,13]. Prognostic variables and scales were determined through the collection of clinical data and patient biomarkers at hospital admission.

Diagnostic criteria

Diagnostic criteria were based on guidelines established by the European Association for the Study of the Liver[14].

Operational definitions

Jaundice: Defined as a yellow discoloration of the sclera, a report of jaundice in the electronic medical record, or total bilirubin level ≥ 2.5 mg/dL[14].

Coagulopathy: Defined as INR > 1.5 or prothrombin time > 15 seconds[14].

Hepatic encephalopathy: Defined as any clinical evaluation indicating neurological deterioration[14].

Acute liver injury: Defined as a syndrome characterized by markers of liver damage (serum transaminases elevated three times above the laboratory’s upper limit of normal) and impaired liver function (jaundice and INR > 1.5)[14].

ALF: Defined as meeting the acute liver injury criteria plus the presence of hepatic encephalopathy[14].

Follow-up

Patients were followed until hospital discharge, transfer to another facility, orthotopic LT (OLT), or death. The median follow-up time was 5 days (interquartile range: 2-10).

Statistical analysis

Statistical analysis was performed using Statistical Package for the Social Sciences version 25.0. Continuous variables were assessed for normality using the Shapiro-Wilk test. Normally distributed variables were expressed as the mean ± SD and compared between the groups using Student’s t-test, whereas non-normally distributed variables were expressed as the median and interquartile range and compared between the groups using the Mann-Whitney U test. Categorical variables were summarized as frequencies and percentages and compared using the χ² test or Fisher’s exact test, as appropriate. A two-sided P value < 0.05 was considered statistically significant.

RESULTS

In total, 69 patients (mean age, 32.8 ± 14.2 years; 65% male) with confirmed HAV infection were included in the final analysis (Table 1). Five patients (7.2%) developed ALF. Among these, one underwent OLT, and two patients died, resulting in an overall transplant-free mortality of 2.9% and a combined outcome (OLT or death) rate of 4.3% (Table 1). When clinical and biochemical characteristics were compared, patients in the OLT/death group had significantly higher total bilirubin levels (18.2 ± 4.6 mg/dL vs 6.1 ± 3.8 mg/dL, P < 0.001), higher INR (2.8 ± 0.7 vs 1.5 ± 0.4, P < 0.01), and higher MELD-Na scores (29.4 ± 3.1 vs 20.1 ± 5.6, P < 0.01) than those who recovered without transplantation. Serum albumin levels were significantly lower in the OLT/death group (2.3 ± 0.5 g/dL vs 3.6 ± 0.6 g/dL, P < 0.01; Tables 2 and 3). No significant differences were found in aminotransferase levels, platelet counts, and white blood cell counts between the groups (Table 2).

Table 1 General characteristics of the population, n (%)/mean ± SD.

Characteristics	General (n = 69)	ALF (n = 5)	OLT and/or death (n = 3)
Male sex	45 (65.2)	4 (80)	2 (66.6)
Age (years)	38.2 ± 11.3	55 ± 5.72	58.3 ± 3.5
Heart rate (bpm)	88.4 ± 15.6	93 ± 22.5	112 ± 22.4
Platelets (× 103/μL)	117 ± 77	161 ± 78.3	161 ± 82.9
NLR (× 103/μL)	2.82 ± 4.54	7.22 ± 2.99	7.22 ± 2.87
Creatinine (mg/dL)	0.9 ± 1.15	1.14 ± 0.58	1.3 ± 0.73
INR	1.23 ± 0.84	2.04 ± 0.9	4.1 ± 2.21
Albumin (g/dL)	3.3 ± 0.53	2.5 ± 0.47	2.5 ± 0.29
Total bilirubin (mg/dL)	5.12 ± 5.62	18.9 ± 7.59	20.7 ± 10.7
Direct bilirubin (mg/dL)	3.36 ± 4.08	10.9 ± 8.08	12.8 ± 11.1
ALT (U/L)	2241 ± 1958	5134 ± 2757	5134 ± 3383
AST (U/L)	1679 ± 2295	3413 ± 2300	4063 ± 2531
GGT (U/L)	256 ± 204	214 ± 163	214 ± 202
ALP (U/L)	160 ± 102	129 ± 42.1	162 ± 34
LDH (U/L)	649 ± 1579	1075 ± 1578	1075 ± 2008

ALF: Acute liver failure; OLT: Orthotopic liver transplantation; bpm: Beats per minute; NLR: Neutrophil-to-lymphocyte ratio; INR: International normalized ratio; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; GGT: Gamma-glutamyl transferase; ALP: Alkaline phosphatase; LDH: Lactate dehydrogenase.

Table 2 Subgroup analysis according to the need for orthotopic liver transplantation and/or death.

Characteristics	Patients with LT and/or death, n = 3	Patients without LT and/or death, n = 66	P value
Age (years)	58.3	37.2	0.001
BMI (kg/m2)	23.6	26.5	0.34
Heart rate (bpm)	112	87.3	0.007
Respiratory rate (rpm)	23.7	18.7	0.001
Leukocytes (× 103/μL)	8.83	6.05	0.12
NLR (× 103/μL)	6.88	3.54	0.21
Creatinine (mg/dL)	1.3	0.88	0.486
Sodium (mEq/L)	130	135	0.015
Prothrombin time	40	14.1	< 0.001
INR	4.05	1.35	< 0.001
Albumin (g/dL)	2.5	3.3	0.009
ALBI score	−0.44	-1.57	0.006
MELD score	42	16	< 0.001
MELD-Na score	41	18	< 0.001
MELD 3.0 score	43	18	< 0.001

LT: Liver transplantation; BMI: Body mass index; bpm: Beats per minute; rpm: Respirations per minute; INR: International normalized ratio; ALBI: Albumin-bilirubin score; MELD: Model for end-stage liver disease; MELD-Na: Model for end-stage liver disease with sodium; MELD 3.0: Model for end-stage liver disease version 3.

Table 3 Subgroup analysis according to the development of acute liver failure.

Characteristics	Patients with ALF (n = 5)	Patients without ALF (n = 61)	P value
Age (years)	54.8	36.9	< 0.001
BMI (kg/m2)	26.2	26.4	0.92
Heart rate (bpm)	101	87.4	0.06
Respiratory rate (rpm)	20.6	28.8	0.15
Leukocytes (× 103/μL)	8.78	5.96	0.04
NLR (× 103/μL)	6.52	3.45	0.14
Creatinine (mg/dL)	1.35	1.07	0.59
Sodium (mEq/L)	132	135	0.08
Prothrombin time	31	14	< 0.001
INR	3.12	1.33	< 0.001
Albumin (g/dL)	2.54	3.33	< 0.001
ALBI score	-0.55	-1.58	< 0.001
MELD score	30.2	15.7	< 0.001
MELD-Na score	32.2	17.1	< 0.001
MELD 3.0 score	33	18.2	< 0.001

ALF: Acute liver failure; BMI: Body mass index; bpm: Beats per minute; rpm: Respirations per minute; INR: International normalized ratio; ALBI: Albumin-bilirubin score; MELD: Model for end-stage liver disease; MELD-Na: Model for end-stage liver disease with sodium; MELD 3.0: Model for end-stage liver disease version 3.

DISCUSSION

This study adds to the limited data available on HAV-associated ALF in Latin America and provides novel insights into the prognostic value of the ALBI score in this population. This case series represents one of the few systematic analyses of HAV-associated ALF in Mexican patients. A major strength of this study was its comprehensive collection of clinical, biochemical, and prognostic score data (MELD, MELD-Na, MELD 3.0, ALBI) at admission, which allowed us to identify early predictors of adverse outcomes, including mortality and OLT. Importantly, this is the first study to evaluate the ALBI score in patients with HAV infection, adding novel evidence of its prognostic value in this population.

This case series examined differences between patient groups with severe clinical outcomes, including mortality and the need for OLT, among those diagnosed with hepatitis A (Tables 2 and 3), consistent with findings reported in the literature. The results of this retrospective study suggest that advanced age; elevated MELD, MELD-Na, and MELD 3.0 scores; leukocytosis; hyponatremia; prolonged prothrombin time (INR); low albumin levels; and a high ALBI score, are associated with increased mortality and the need for OLT in patients with hepatitis A. Notably, unlike some previous studies, creatinine was not a significant predictor in our analysis. This contradicts the findings of Mackinney-Novelo et al[15], who reported that a creatinine level exceeding 2 mg/dL was associated with a higher risk of fulminant hepatitis and death.

In the same study, Mackinney-Novelo et al[15], at a referral center in Mexico City, analyzed 38 cases of acute fulminant hepatitis secondary to hepatitis A, identifying prolonged prothrombin time (relative risk = 1.03; P = 0.003) and low albumin levels (relative risk = 0.36; P = 0.02) as predictors of mortality. These findings align with our results, which also highlight elevated INR and low albumin levels as critical prognostic factors in patients with hepatitis A (Table 2).

Our data also illustrated that patients who developed ALF presented markedly different clinical characteristics, such as older age, leukocytosis, prolonged prothrombin times, elevated INR, low albumin levels, and an unfavorable ALBI score (Table 3), indicating a more severe clinical prognosis. These findings are consistent with those of Jiang et al[16], who also identified hypoalbuminemia and high MELD-Na scores as risk factors for the development of ALF. Furthermore, positive blood cultures and septic shock were frequent among ALF patients, suggesting that monitoring for signs of sepsis and early empirical antibiotic use in patients with hepatitis A and hypoalbuminemia could be useful strategies in these cases[16].

The study by Radha Krishna et al[17] highlighted the high prevalence of coincident HAV infection in patients with pre-existing cirrhosis and its impact on mortality and the need for OLT. Coincident HAV infection and cirrhosis significantly increased the risk of adverse outcomes, emphasizing the importance of comprehensively evaluating the patient’s medical history. Although our study did not include patients with cirrhosis or chronic liver disease, this limitation is relevant because it restricts the generalizability of our findings to this high-risk group, which is known to have higher mortality rates. Conversely, Chávez-Tapia et al[4] reported a notable increase in mortality associated with ALF in Mexico, particularly attributable to viral etiologies and in patients older than 45 years. This increased mortality in older individuals is consistent with our findings, in which we observed high mortality rates in older patients with viral hepatitis.

Regarding the scoring models used, the MELD score has proven highly effective in assessing disease severity and predicting mortality in these patients, as it integrates bilirubin levels, INR, and creatinine levels. In our analysis, the MELD score displayed significant predictive capacity for mortality and the need for OLT, with substantially higher scores in patients with negative outcomes (Table 2). Higher MELD-Na scores were also observed in patients with severe clinical outcomes, including mortality and the need for OLT, consistent with previous studies. Similarly, MELD 3.0 scores were significantly higher in these patients, reflecting its association with adverse outcomes.

Additionally, ALBI, a simple model that uses albumin and bilirubin to assess liver function, emerged as a significant predictor of mortality and the need for OLT in this cohort. This finding is consistent with those of previous studies validating the utility of ALBI in various populations with liver diseases, and our study is the first to explore its applicability in the context of hepatitis A. Our findings should be interpreted in the context of the national epidemiological situation, which is currently characterized by intermediate endemicity, representing a transition from the classic high-endemicity pattern to a scenario in which infection occurs at older ages, thereby increasing the risk of clinical disease and complications in adolescents and adults[18-20].

Although HAV-associated ALF is relatively uncommon, it represents an important cause of LT in Mexico, and it might significantly affect transplant program capacity. Previous reports have documented spontaneous resolution rates of approximately 60%, with up to 40% of patients requiring transplantation, and mortality rates ranging from 13% to 33%[2-4]. Our results are consistent with these figures, emphasizing the need for early risk stratification and timely referral of patients at high risk of poor outcomes. Moreover, HAV vaccination, which is currently available in the private sector but is not included in the national immunization schedule, could play a pivotal role in preventing severe cases and reducing the burden on liver transplant programs. Incorporating HAV vaccination into public health policy could be a cost-effective strategy to prevent outbreaks, reduce morbidity and mortality, and optimize the use of scarce healthcare and transplant resources.

This study had several limitations. First, its retrospective and single-center design might limit the generalizability of the findings to other populations or healthcare settings. The small number of patients with adverse outcomes (OLT or death) reduced the statistical power of the study to detect more subtle associations and precluded multivariate analysis. The absence of data on potential confounders such as pre-existing liver disease, nutritional status, and socioeconomic factors might have influenced the observed outcomes. Finally, because of the retrospective nature of the study, causal relationships between clinical variables and outcomes could not be firmly established.

CONCLUSION

This case series of Mexican patients with HAV infection highlights that advanced age; hypoalbuminemia; prolonged prothrombin time/INR; elevated MELD, MELD-Na, and MELD 3.0 scores; and higher ALBI are strongly associated with mortality and the need for OLT. The systematic evaluation of multiple prognostic models in this case series provides valuable insight into early risk stratification for HAV-related ALF. Our findings underscore the importance of early recognition of these high-risk features at hospital admission to guide timely intervention and referral to transplant centers. Nevertheless, because this was a retrospective, single-center case series that excluded patients with pre-existing cirrhosis or chronic liver disease, the results should be interpreted with caution. Larger, multicenter studies including patients with chronic liver disease are warranted to validate these predictors and strengthen risk prediction models for HAV-associated ALF.