Abnormalities in liver enzyme levels during Salmonella enteritidis enterocolitis

A. González-Quintela, J. Campos, R. Alende, A. López-Soto, S. Tomé, E. Otero and J. A. Torre

Department of Internal Medicine. Hospital Clínico Universitario. Santiago de Compostela. A Coruña. Spain

 

ABSTRACT

Objective: to evaluate the prevalence, associated factors, and time-course changes of abnormal liver enzyme serum levels in adult patients with Salmonella enteritidis enterocolitis.
Methods: the clinical records of 104 patients (age range 15-86 years, 46.2% males) admitted to hospital because of S. enteritidis enterocolitis were reviewed. The prevalence of abnormal liver enzyme levels was evaluated, as well as its possible relationship to data of systemic inflammatory response, severe sepsis, and bacteremia. In addition, time-course changes in serum levels of liver enzymes were studied in 16 cases with available follow-up after hospital discharge.
Results: in patients without a pre-existing cause for liver enzyme abnormalities (n = 84), the prevalence of serum AST elevation was 23.0% (95% CI 15.4-34.5%), of serum ALT elevation was 17.9% (95% CI 0.6-20.0%), and of GGT elevation was 19.0% (95% CI 11.6-29.3%). The prevalence of abnormality for any of these enzymes (AST, ALT, or GGT) was 35.7% (95% CI 25.7-46.8%). The prevalence of altered serum alkaline phosphatase was lower. Alteration in liver enzyme serum levels was moderate in the majority of cases, and was found in association with the presence of fever. Serum enzyme levels decreased during the convalescence period after hospital discharge.
Conclusions: abnormalities in liver enzyme levels are frequent during severe enterocolitis due to S. enteritidis in adult patients. These abnormalities are moderate and self-limited.

Key words: Salmonella. Salmonella enteritidis. Enterocolitis. AST. ALT. GGT.

---

González-Quintela A, Campos J, Alende R, López-Soto A, Tomé S, Otero E, Torre JA. Abnormalities in liver enzyme levels during Salmonella enteritidis enterocolitis. Rev Esp Enferm Dig 2004; 96: 559-566.

---

Recibido: 16-10-03.
Aceptado: 12-.01-04.

Correspondencia: Arturo González-Quintela. Servicio de Medicina Interna. Hospital Clínico Universitario. 15706 Santiago de Compostela. e-mail: mearturo@usc.es

 

INTRODUCTION

Salmonella spp. infections represent a common health problem worldwide, particularly in developing countries (1). Salmonella typhi infection induces a protracted febrile condition known as enteric or typhoid fever (1). Enteric fever is frequently associated with abnormalities in liver enzyme serum levels (2,3). Furthermore, a small subset of patients with enteric fever develop a severe liver disease with jaundice, commonly known as Salmonella hepatitis (2-4). Among non-typhoidal salmonellae, Salmonella enteritidis (S. enteritidis) infection usually induces a self-limited diarrheal disease (acute enterocolitis) which may be associated with bacteremia and systemic manifestations, and is typically acquired from contaminated eggs in food (1). Reports of possible abnormalities in liver enzyme levels during enterocolitis due to S. enteritidis are scarce. In a series of patients with acute enterocolitis of varying causes, abnormally high transaminase levels were found in 10% of cases due to Salmonella spp. infection (5). In addition, a case of severe acute hepatitis in a patient with multi-organ system failure after S. enteritidis sepsis has been reported (6). The present study was aimed at evaluating the prevalence, associated factors, and time-course changes of abnormal liver enzyme levels in adult patients admitted to hospital because of acute enterocolitis due to S. enteritidis.

PATIENTS AND METHODS

Study design

This is an analytical cross-sectional (prevalence) study. In a subset of patients, a longitudinal follow-up was performed (retrospective cohort study). Data were obtained after a review of computerized clinical records.

Setting

Patients admitted to a short-term stay Unit of the Internal Medicine Department (Hospital Clínico Universitario, Santiago de Compostela, Spain) with a diagnosis of enterocolitis due to S. enteritidis from January 1995 to July 2003 were included. Our center is a reference university hospital covering an area with approximately 500.000 inhabitants. Patients admitted to the short-stay unit have medical problems calling for an anticipated stay not longer than five days. Patients with enterocolitis are admitted to hospital in the presence of: a) evidence of severe sepsis; b) inability for oral rehydration; c) renal failure; or d) additional metabolic alterations. Routine therapy includes fasting, intravenous rehydration, and antimicrobial chemotherapy with either ciprofloxacin or ceftriaxone in the majority of cases.

Study population

A total of 107 patients were admitted to the Unit because of S. enteritidis enterocolitis during the study period. Of them, cases with at least one determination of liver enzyme levels during hospital admission (n = 104) were included in the study. All patients included fulfilled the following criteria: a) acute diarrheal illness; and b) microbiological or epidemiological evidence of S. enteritidis infection. Microbiological evidence included stool culture positivity (present in 87 of 101 studied cases, 86.1%), or blood culture positivity (present in 24 out of 100 studied cases, 24.0%). Epidemiological evidence included a diarrhea episode occurring during a microbiologically-proven S. enteritidis infection outbreak (four cases). Median age of included patients was 52 years (range, 15-86 years). Forty-eight patients (46.2%) were males.

In 20 patients (19.2%) there was a pre-existing liver disease or a known cause for it. This subgroup included 17 heavy drinkers, a patient with chronic hepatitis B, a patient with chronic hepatitis C, and a patient with hereditary hemochromatosis. For an estimation of the prevalence of abnormal liver enzyme levels, only patients without such a history of liver disease (n = 84) were included.

In five patients (4.8%) there was associated morbidity, which included a case of Crohn' disease, a case of non-Hodgkin' lymphoma, an HIV-infected case, a case with Behçet' disease, and a case with celiac disease.

All patients were discharged from hospital without significant complications after a median stay of four days (range 1-10 days).

Main outcome measures

Serum aspartate-aminotransferase (sAST), alanine-aminotransferase (sALT), gamma-glutamyl-transferase (sGGT), alkaline phosphatase (sAF) and bilirubin were assayed in a Hitachi analyzer (Boehringer-Mannheim, Mannheim, Germany). Serum samples were obtained during the first 72 hours of hospital admission in all cases. Upper reference levels for sAST, sALT, sGGT and sAF are ≤ 25 IU/L, ≤ 29 IU/L, ≤ 38 IU/L, and ≤ 195 IU/L, respectively. In patients younger than 18 years (five cases) the sAF determination was not considered when altered. An additional determination of liver enzyme levels performed after hospital discharge (median 24 days, range 13-60 days) was available for 16 patients with baseline alterations in liver tests. This determination was compared to baseline values in order to investigate time-course changes in liver enzyme levels after the acute period of disease.

The study investigated the possible relationship of abnormalities in liver enzyme levels with: a) detection of bacteremia; b) criteria of systemic inflammatory response (fever, tachycardia, altered white blood cell count); and c) criteria of severe sepsis (arterial hypotension, renal failure) (7). An axillary temperature higher than 38 ºC was considered to be fever. A heart rate higher than 90 bpm was considered tachycardia. The presence of leukocytosis (more than 12,000 cells/mm³), leukopenia (less than 4000 cells/mm³) or left shift (more than 10% of band forms) were considered alterations in white blood cell counts. A systolic blood pressure lower than 90 mmHg was considered arterial hypotension (7). A serum creatinine higher than 1.3 mg/dL was considered kidney failure.

Statistical analysis

The prevalence of abnormalities in liver enzyme levels were calculated with their corresponding 95% confidence intervals (CI). Numerical variables were represented as median and range values due to their non-normal distribution in the Kolmogorov test. The Chi-square test (with continuity correction, when needed) was used to compare proportions. The Mann-Whitney test was employed to compare numerical variables. Wilcoxon' test was employed for comparison of paired samples. Two-tailed p-values lower than 0.05 were considered statistically significant.

Ethical considerations

The study conformed to the Helsinki Declaration of 1975.

RESULTS

Prevalence of abnormalities in liver enzyme levels

In the sample studied, median (and range) liver enzymes were as follows: sAST 17 UI/L (7-426 UI/L), sALT 16 UI/L (6-617 UI/L), sGGT 18 UI/L (5-629 UI/L), and sAF 115 UI/L (59-641 UI/L).

In the group of patients with neither pre-existing liver disease nor a cause for it, the most frequent abnormalities were found for sAST, sALT and sGGT. The prevalence of elevation was as follows: sAST 23.0% (95% CI 15.4-34.5%), sALT 17.9% (95% CI 10.6-20.0%), and sGGT 19.0% (95% CI 11.6-29.3%). The prevalence of abnormality in any of these parameters (sAST, sALT, or sGGT) was 35.7% (95% CI 25.7-46.8%). The individual levels of sAST, sALT, and sGGT in patients without pre-existing liver disease are represented in figure 1. In two patients (2.3%, 95% CI 0.4-9.1%) serum transaminase elevation reached the range of acute hepatitis (sAST or sALT higher than 10-fold the highest normal level). Abnormalities in sAF were less frequent. In patients without pre-existing liver disease the prevalence of abnormal sAF was 6.1% (95% CI 2.2-14.2%), with a median of 111 IU/L (range 59-141 IU/L). In all but two cases sAF increase was associated with sGGT increase. Clinically detectable bilirubin elevations were observed in no cases (median 0.7 mg/dL, range 0.2-2.6 mg/dL).

Association of abnormalities in liver enzyme levels with potential factors

Table I provides a comparison of demographic and clinical variables between patients with and without abnormalities in liver enzyme levels (sAST, sALT or sGGT) within the group of cases without pre-existing liver disease. Patients with abnormalities in liver enzyme levels presented a higher prevalence of fever than patients without abnormalities. Bacteremia was more frequent in patients with abnormalities in liver enzyme levels but the difference was not statistical significant. Serum creatinine levels were slightly higher in patients with abnormalities in liver enzyme levels that in patients without them (median 1.0 mg/dL and range 0.7-3.5 mg/dL vs median 0.9 mg/dL and range 0.6-4.3 mg/dL, respectively, p = 0.04), but the prevalence of renal failure was not significantly different between both groups (Table I). No differences were observed in the prevalence of arterial hypotension, tachycardia, altered white blood cell counts, or antibiotic administration between patients with abnormalities in liver enzyme levels and patients without them (Table I).

Time-course changes of liver enzyme serum levels during the convalescence period

Figure 2 represents the outcome of sAST, sALT and sGGT levels during the convalescence period after hospital discharge in 16 patients with some alteration in these enzymes during the acute phase. There was a trend towards decrease of all three liver enzymes (statistically significant for sAST and sALT). This trend was observed in patients with and without pre-existing liver disease, but was particularly significant in patients without it (p = 0.007 and p = 0.01 for sASL and sALT, respectively), and in patients with the highest enzyme levels during the acute phase (Fig. 2). In patients without pre-existing liver disease, sAST levels went back to normal in all nine patients with baseline elevation, and sALT levels went back to normal in six out of seven patients with baseline elevation.

DISCUSSION

The present study shows that a sizeable proportion of patients with enterocolitis due to S. enteritidis present abnormalities in liver enzyme levels, particularly sAST, sALT and sGGT. These abnormalities are mild in the majority of cases, and self-limited.

Previous references of liver abnormalities during infectious enterocolitis are scarce. In a retrospective study of 727 cases of enterocolitis, Tositti et al. found that rotavirus was the most common cause of sALT increase, and suggested that rotavirus infection should be suspected in adult patients with acute diarrhea and serum transaminase elevation (5). These same authors, however, found that sALT was increased in 13 out of 125 cases (10%) of enterocolitis due to Salmonella spp. in patients without previous liver disease (5), a figure similar to that found in the present study. Abnormalities in liver enzyme levels during S. enteritidis enterocolitis have biological plausibility. S. enteritidis infection in these cases is apparently limited to the gastrointestinal tract, but is frequently accompanied by manifestations of systemic inflammatory response. Abnormalities in liver enzyme levels, together with additional systemic manifestations of S. enteritidis infection may be caused by a leaking of either bacteria or their products to the circulation (6), with subsequent immune activation and production of cytokines with potential liver toxicity. Thus, abnormalities in liver enzyme levels during S. enteritidis enterocolitis may be included in the liver dysfunction syndrome during sepsis (8,9). No data are available to affirm whether or not these abnormalities are more frequent in S. enteritidis infections than in other infections. In the studied population, however, abnormalities in liver enzyme levels were associated with fever but not with additional evidence of systemic inflammatory response (tachycardia or white blood cell count alterations), data of severe sepsis (arterial hypotension, renal failure) or detectable bacteremia.

The methodological limitations of this kind of studies should be considered. As far as selection of cases is concerned, it should be taken into account that the studied population consisted of in-hospital patients. All of them had an indication for admission, thus meaning that enterocolitis may be classified as severe in all cases (10). For this same reason, the vast majority of patients received antimicrobial chemotherapy. The reference population to whom results could be extrapolated should therefore be patients with S. enteritidis enterocolitis of similar severity. In addition, prevalence studies are not optimal to establish causal relationships (S. enteritidis infection as a cause of abnormalities in liver enzyme levels in this case) due to temporal ambiguity. Although cases with pre-existing liver disease (or a cause for it) were excluded in prevalence studies, it is possible that some additional patients had abnormalities in liver enzyme levels prior to their enterocolitis episode. Specifically, there were no available data to evaluate overweight as a cause for such abnormalities. The most frequent cause of liver disease in the sample studied was alcohol consumption, which in addition may modify the severity of S. enteritidis infections (11,12).

Even with those limitations, the role of S. enteritidis infection in the development of liver enzyme abnormalities is supported by consistency with previous studies (5), analogy with liver abnormalities during different Salmonella spp. infections (2,3), and the already mentioned biological plausibility. The longitudinal study (retrospective cohort study), although performed in a short number of cases, demonstrated a significant decrease of liver enzyme levels during convalescence, further suggesting that S. enteritidis infection was the cause of abnormalities in liver enzyme levels. Such abnormalities are mild and transient, but are also frequent and should be taken into account in patients with severe S. enteritidis enterocolitis.

REFERENCES

1. Lesser CF, Miller SI. Salmonellosis. En: Braunwald E, Fauci AS, Kasper DL, Hauser SL, Longo DL, Jameson JL, eds. Harrison' Principles of Internal Medicine. 15th Ed. New York: McGraw-Hill, 2001. p. 970-5.        [ Links ]

2. El-Newihi H, Alamy ME, Reynolds TB. Salmonella hepatitis: analysis of 27 cases and comparison with acute viral hepatitis. Hepatology 1996; 24: 516-9.        [ Links ]

3. Pramoolsinsap C, Viranuvatti V. Salmonella hepatitis. J Gastroenterol Hepatol 1998; 13: 745-50.        [ Links ]

4. Rodriguez Cuartero A, Rodriguez Fassio MA. Hepatitis aguda tífica. Rev Esp Enferm Dig 1993; 84: 133-4.        [ Links ]

5. Tositti G, Fabris P, Romano L, de Lalla F. Liver injury during gastroenteritis in adults: a retrospective study in 727 cases. Am J Gastroenterol 2002; 97; 487-8.        [ Links ]

6. Retornaz F, Fournier PE, Seux V, Jacomo V, Soubeyrand J. A case of Salmonella enteritidis septicemia complicated by disseminated intravascular coagulation, severe hepatitis, rhabdomyolisis and acute renal failure. Eur J Clin Microbiol Infect Dis 1999; 18: 830-41.        [ Links ]

7. ACCP/SCCM (American College of Chest Physicians/Society of Critical Care Medicine) Consensus Conference Committee. Definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med 1992; 20: 864-74.        [ Links ]

8. Sikuler E, Guetta V, Keynan A, Neumann L, Schalaeffer F. Abnormalities in bilirubin and liver enzyme levels in adult patients with bacteremia. A prospective study. Arch Intern Med 1989; 149: 2246-8.        [ Links ]

9. Wang P, Chaudry IH. Mechanism of hepatocellular dysfunction during hyperdynamic sepsis. Am J Physiol 1996; 270: R927-38.        [ Links ]

10. Goodman LJ, Segreti J. Diarrheal disease and gastroenteritis. Infect Dis Emerg Med 1992; 2: 591-605.        [ Links ]

11. Bassa A, Parras F, Reina J, Villar E, Gil J, Alomar P. Non-typhi Salmonella bacteraemia. Infection 1989; 17: 290-3.        [ Links ]

12. Galofre J, Moreno A, Mensa J, Miro J, Gatell J, Almela M, et al. Analysis of factors influencing the outcome and development of septic metastasis or relapse in Salmonella bacteremia. Clin Infect Dis 1994; 18: 873-8.        [ Links ]