Herpes simplex virus-associated acute liver failure: A difficult diagnosis with a poor prognosis

Authors


Abstract

We report 5 cases of acute liver failure related to herpes simplex (HSV) infection in 1 immunocompetent and 4 immunosuppressed patients. One patient was too ill for liver transplantation indication. Three patients, among the 4 listed, underwent liver transplantation. Three patients died 11 days to 1 year after transplantation and 2 patients died 2 to 3 days after admission. All presented with fever and none with skin lesions. The diagnosis of HSV-related hepatitis was made antemortem in only 2 patients on the basis of positive blood cultures and/or immunohistochemic findings. In the remaining patients, HSV diagnosis was made retrospectively on further histologic and virologic investigations. Primary HSV infection was certain or likely in all cases, including an HSV2 superinfection of an anti-HSV1-positive patient and two HSV superinfections of hepatitis B virus (HBV)-related chronic liver disease. In these latter patients, HSV diagnosis was totally unsuspected, despite fever. HSV superinfection has significantly contributed to liver dysfunction aggravation and death. In conclusion, the diagnosis of HSV hepatitis is difficult to establish in the absence of specific clinical signs. This may suggest the need for early administration of acyclovir in patients with suspected HSV hepatitis, without waiting for virologic confirmation. Diagnosis methods providing fast results (real-time polymerase chain reaction [PCR]) should be implemented. (Liver Transpl 2005;11:1550–1555.)

The spontaneous prognosis of herpes simplex virus (HSV)-associated fulminant hepatitis is poor, even though an effective antiviral treatment is available. This may be explained by the fact that the diagnosis is usually delayed, so that initiation of treatment is too late. At an advanced stage, liver transplantation may improve prognosis.1 The delay in diagnosis is essentially due to the lack of specific symptoms. We describe a series of 5 patients presenting with fulminant hepatitis related to HSV, and report the occurrence of HSV superinfection of hepatitis B virus (HBV) chronic liver disease. We describe presentation and clinical outcome, with particular emphasis on virologic and histologic findings.

Abbreviations

HSV, herpes simplex virus; HBV, hepatitis B virus; PCR, polymerase chain reaction; Ig, immunoglobulin.

Patients and Methods

From 1990 to 2003, 360 patients with fulminant hepatitis or acute liver failure were admitted to our Liver Intensive Care Unit. The etiology of acute liver failure was viral (28%), drug and toxic-induced (41%), due to other causes (17%), and undetermined (14%). Of these 360 patients, 5 (3 female, 2 male), mean age 35.6 yr (range: 15 to 55) had a final diagnosis of HSV-associated hepatitis (1.4%). The main clinical data and outcome are given in Table 1.

Table 1. Characteristics of the 5 Patients on Admission
Patient12345
  • Abbreviations: IMS, immunosuppressant drugs; AIDS, acquired immune deficiency syndrome; RT, renal transplantation; FAP, familial amyloid polyneuropathy; F, fibrosis; N, necrosis; I, nuclear inclusions.

  • *

    The patient was treated by cyclosporin and mycophenolate mofetyl.

  • The immunosuppressant treatment was steroids, azathioprine, antilymphocyte serum, and cyclosporin. The patient received 1-gm steroid bolus at day 15 following an acute rejection.

Age/gender36/M24/F55/F15/F48/F
Immunity stateAIDSImmunocompetentIMS* after RTPrednisolone therapyIMS after LT for FAP
Clinical courses     
 Delay between onset of disease and ALF1 day7 days7 days3 days?
 Fever38°C39°C39°C38.5°C38.5°C
 Abdominal painYesYesYesNoNo
 White cell count (Nl: 4–10,000 giga/L)5,6002,7005,0009,0008,000
 AST/ALT (IU/L) (Nl < 32/<36)5,093/1,6501,170/4322,064/3555,000/4,50070/22
 Acyclovir treatment (days after onset of the disease)NoYes (2)Yes (5)NoNo
 LTNoYesNoYesYes
 Transmission mode?Sexual?Nosocomial?Nosocomial?
Virological findings [for detail see Fig. 2]
 HSV1 and HSV-2 serology at admission intensive care unitNegativeNegativeIgG: +; IgM: ++IgG: +; IgM +++ 
 SeroconversionD 7 Anti-HSV 2 seroconversion 
 Viremia at admission (PCR and/or culture)HSV 1 +; HSV 2–HSV 2 +; HSV 1–HSV 1 +; HSV 2–HSV 2 recovered from explanted liver cultures 
Histological [for detail see Fig. 2]
 MaterialAutopsy liver biopsyEntire native liverAutopsy liver biopsyEntire native liver 
 Underlying hepatopathyPortal fibrosis (F1)F–Septal fibrosis (F3)  
 Histological findings suggestive or nonsuggestive of herpetic hepatitis     
  Semiquantitative analysisNonsuggestive; N+; I–Suggestive; N+; I++Nonsuggestive; N +; I–Suggestive; N +; I ++ 
Immunohistochemical staining     
 HSV++++++++ 
 HBV (HBc, Pre-S1)HBc and Pre-S1 + 
Figure 2.

Evolution of liver tests, virologic and histologic findings in patient 5. LB, liver biopsy. Patient 5 had an auxiliary partial orthotopic liver transplantation with a left liver because of the small size of the graft. The right native liver was removed on day 30 because portal cavernoma shunted the portal flow of the graft, imparting a medium term risk of inducing graft dysfunction. 48 hours after the right hepatectomy, the patient developed acute liver failure and was retransplanted. HSV desoxyribonucleic acid (DNA) was not detected in serum on days 0 and 15 after auxiliary partial orthotopic liver transplantation and was not detected either at day 0 in the native liver or in the first liver graft. HSV DNA was negative in a liver graft biopsy sampled on day 15. HSV-2 DNA was detected in a serum sample on the day of the right hepatectomy, which was 30 days after auxiliary partial orthotopic liver transplantation. Green line = PT; black line-AST; blue line = ALT.

Patient Management

Fulminant hepatitis was defined as acute hepatitis complicated by acute liver failure and hepatic encephalopathy within 2 weeks after the onset of disease. Criteria for liver transplantation were the presence of grade 3 to 4 hepatic encephalopathy associated with either factor V level less 20% of normal for patients less than 30 years of age, or factor V level less than 30% of normal in patients more than 30 years of age.2–4 All patients were managed according to a standardized protocol.2

Virologic Methods

Patients were screened for usual causes of acute viral hepatitis: anti-hepatitis A virus IgM, hepatitis B surface antigen, and anti-hepatitis B core antigen IgM. Pretransplant virologic assessment included IgM and IgG serology for Epstein-Barr virus, cytomegalovirus, varicella-zoster virus, and HSV and viral blood cultures on human fibroblasts. Cultures were fixed after 48 hours for cytomegalovirus antigens by the immunoperoxidase technique. In the occurrence of a cytopathic effect, the presence of HSV antigens was confirmed by immunofluorescence.

When available, liver tissue culture was performed upon clinicians' request. In hepatitis B surface antigen-positive subjects, HBV viral load was quantified with the Cobas HBV Monitor technique (Roche Diagnostics, Emeryville, NJ). Type-specific HSV deoxyribonucleic acid amplification was performed by conventional polymerase chain reaction (PCR) on serum or liver biopsy samples upon clinicians' request. Type specific HSV serology was performed with the HSV-1 recombinant immunoglobulin (IgG) enzyme-linked immunosorbent assay (ELIT; Eurobio, Paris, France) and HSV-2 recombinant IgG ELIT.

Primary HSV infection was considered when HSV detection by culture or PCR was associated either with a negative HSV serology, or with an anti-HSV seroconversion. In the presence of IgG and IgM association, an HSV reactivation could not be ruled out, thus IgM titers, clinical data, and epidemiologic data were taken into account.

Histologic Methods

Histologic review was performed on embedded specimens from native livers in all patients (at the time of transplantation in patients 2 and 4, and at the time of right hepatic resection in patient 5, with postmortem liver biopsies in patients 1 and 3); samples were also taken from the allograft in patient 5 (liver biopsies and explant at the time of the retransplantation). Slides were reviewed by one pathologist (M.S.). Findings were regarded as consistent with HSV-related hepatitis in the presence of extensive areas of hepatocyte necrosis with adjacent congestion, few inflammatory infiltrates, purple nuclear inclusions with a clear halo in the hepatocytes, and the presence of multinucleated hepatocytes5 (Fig. 1A). Immunohistochemic staining to confirm the diagnosis of HSV was performed when histologic findings were suggestive of HSV-related hepatitis or when virologic tests were previously positive for HSV. The presence of viral antigens was demonstrated by an indirect 3-layer immunoperoxidase staining and by monoclonal antibodies against HSV antigens (Dakopatt, Copenhagen, Denmark) (Fig. 1B).

Figure 1.

(A) Characteristics findings of HSV infection with large foci of inflammatory necrosis, with hepatocyte nuclear inclusion and a clear halo in patient 2. (B) Immunostaining of HSV in patient 2. Note that the staining is nuclear and also cytoplasmic.

Results

Table 1 and Figure 1 show the baseline clinical data, virologic, and histologic findings for detection of HSV in the 5 patients.

Fulminant Hepatitis Due to HSV-2 (Patients 2, 4, and 5)

Acute liver failure occurred during a previous hospitalization in 2 patients: 12 days after neurosurgery for a benign brain tumor in patient 4 and 31 days after an auxiliary partial orthotopic liver transplantation with a left liver in patient 5. These 2 patients were considered as immunosuppressed due to immunosuppressant drugs and corticosteroid therapy. A third patient was an immunocompetent female with a recent overseas travel history (patient 2). She presented abdominal pain in days prior admission. The delay between symptoms onset and admission in our intensive care unit was 7 days.

These 3 patients presented with fever a few days prior to hepatitis onset. Leukopenia was observed in 2 of them. The third patient had no leukopenia nor lymphopenia. None had visible pharyngeal or genital mucocutaneous vesicles or ulcers. Common viral causes of liver failure were excluded. No vascular, toxic or drug-related cause for hepatitis was found. Patient 2 was treated with acyclovir IV (30 mg/kg/day) for 1 day without improvement of hepatic function. All were listed for liver transplantation on a superemergency basis because of deteriorating liver function and neurologic status and underwent liver transplantation (patients 2 and 4) or retransplantation (patient 5).

After liver transplantation, hepatic function returned to normal. However, all patients died either from hemothorax 29 days after liver transplantation (patient 2), from chronic rejection and immunosuppression-induced complications 1 year later (patient 4), or from acute respiratory distress syndrome and multiorgan failure 11 days later (patient 5).

HSV was recovered from blood cultures before transplantation in patient 2 and from explanted liver cultures in patient 4. Histologic examination and immunostaining of both explanted livers confirmed HSV hepatitis. In patient 5, however, HSV infection was considered only after autopsy on the basis of positive immunostaining with anti-HSV-2 antibodies (Fig. 2).

Further virologic investigations suggested or confirmed primary HSV infections. Patient 2 was seronegative before transplantation and anti-HSV-2 seroconversion occurred 7 days after transplantation. A sexual transmission was evidenced by interrogation and serologic testing of her recent sexual partner. Patient 4, a 15-yr-old female, presented with highly positive anti-HSV IgM at admission in our intensive care unit. Since she was in the hospital for 12 days, a nosocomial HSV transmission was suggested but the source was not found. Patient 5 was seropositive for HSV-1 before the first liver transplantation. HSV-2 seroconversion and HSV-2 viremia were retrospectively evidenced 30 days after the first transplantation with a half-graft. The recipient of the second half-graft did not seroconvert to HSV-2, suggesting that HSV was not acquired from the donor organ.

Fulminant Hepatitis Due to HSV-1 in HBV Chronically Infected Patients (Patients 1 and 3)

Two immunocompromised patients with chronic HBV infection were admitted for acute liver failure. One male patient was infected by human immunodeficiency virus (patient 1) and a female patient had long-term steroid therapy (patient 3). The first diagnosis considered was HBV reactivation and antiviral therapy by lamivudine was given on admission. Prior admission, 1 patient (patient 3) was also given acyclovir therapy (30 mg/kg/day) because of immunosuppressed status. HBV viral loads were over 8.0 log copies/mL for patient 1 and 7.1 log copies/mL for patient 3. Anti-hepatitis B core antigen IgM were negative. No skin or genital lesions were identified on physical examination. Both had a few days-history of fever but none had leukopenia nor lymphopenia. Patient 1 was listed for liver transplantation on a superemergency basis but was withdrawn from the list because of worsening respiratory distress and cardiovascular collapse. Patient 3 deteriorated very rapidly within hours after a percutaneous liver biopsy. Both died within 24 hours after admission from brain death (patient 1) or hemoperitoneum and intrahepatic hematoma (patient 3).

HSV hepatitis was considered only on autopsy material. HSV staining was positive for both patients and was associated with HBV (pre-S1 and hepatitis B core antigens)-positive staining only in the human immunodeficiency virus-positive patient.

Retrospective virologic investigations suggested for both patients a primary HSV-1 infection. HSV-1 viremia was detected by PCR on both sera. Patient 1 was seronegative for HSV and patient 3 and highly positive for anti-HSV IgM. The transmission mode of HSV could not be determined.

Discussion

The diagnosis of HSV-related hepatitis is difficult. Of the 52 HSV hepatitis cases reviewed by Kaufman et al.,6 the diagnosis was made in the antemortem period for only 23%. Our series of 5 patients with HSV-1 or 2-associated acute liver failure emphasizes the difficulty of the diagnosis, in a life-threatening context. In most cases, a positive diagnosis is made too late. However, some features should be emphasized: 1) fever was present in all cases; 2) herpetic lesions were absent; 3) leukopenia was present in only 2 out of 5 patients; 4) aminotransferase levels were increased in all but 1 case; and e) virologic and histologic results were concordant, but were available too late.

In the absence of a common cause of liver failure, histology and immunostaining, assessed on transjugular liver biopsy specimens, can establish or confirm the diagnosis of herpetic hepatitis.7, 8 This is so even in the presence of an underlying liver disease, such as HBV chronic liver disease, as shown in 2 of the 5 patients. This liver biopsy should be performed by the transjugular route, even if liver function does not contraindicate the percutaneous route. Indeed, secondary bleeding is always possible if liver function deteriorates, as it did in 1 patient in this series (patient 3).

This study demonstrates that virologic testing of serum could establish the diagnosis by a non invasive method, since HSV viremia was positive in all cases. However, virologic diagnosis is often delayed in such patients, partly because the diagnosis is not suspected by clinicians and partly because most virologic investigations are not made on an emergency basis. Furthermore, “classical” PCR is time consuming and does not provide fast results. However, the recently-developed real-time PCR should change the situation by providing results in less than 3 hours after sampling.9–11 We also show that serologic assays have very limited value for establishing an early diagnosis and their results do not always contribute,12–14 as shown by 2 seronegative cases in our series. Thus, the molecular methods appear as the most reliable for diagnosis.15

In the present series, HSV hepatitis presentation was classical only for 2 patients, the young immunocompetent female (patient 2) and the young female on steroid therapy after brain surgery. Indeed, clinical presentation, positive blood culture or positive PCR and typical histologic findings were consistent with this diagnosis. In the remaining patients, however, HSV hepatitis was unsuspected, but certainly contributed to death. HSV-2 superinfection of an HSV1-positive patient, as was the case for patient 5, was previously described.16 This superinfection may not have been implicated in liver failure since no massive HSV-induced necrosis was seen. Nevertheless, HSV disseminated infection certainly contributed to clinical deterioration after the second transplantation, with sequelae including acute respiratory distress syndrome, liver failure, and septic shock. In the last 2 patients, the diagnosis was misleading at early clinical presentation because of the concomitant chronic infection by HBV. However, the absence of anti-hepatitis B core antigen IgM, the absence of HBV staining in 1 patient, and the strong HSV staining on liver autopsy specimens suggest that HSV had largely contributed to liver function aggravation in association with the underlying HBV infection.17

HSV hepatitis has been described during both recurrent and primary HSV infection in immunocompromised hosts.18–21 In immunocompetent hosts, however, only primary infections have been associated with hepatitis.22–25 Primary Infection was certain in 3 cases. In 2 cases, HSV reactivation could not be totally ruled out due to the presence of both IgG and IgM and the absence of previous serum samples. However, high titers of HSV IgM in both cases, and the young age in 1 case made this hypothesis likely.

Both HSV types can induce severe or fulminant acute hepatitis21 although HSV-2 is more frequent, perhaps because HSV-2 primary infections are more common in adulthood as sexually transmitted diseases.23, 26 Among the 3 patients with HSV-2 primary infection in our series, only 1 (patient 2) had evidence of sexual transmission. In another 2 patients (patients 4 and 5), primary infection developed during hospitalization, suggesting nosocomial transmission. The transmission mode of the 2 HSV-1 primary infections was not ascertained.

The spontaneous prognosis of HSV-related hepatitis is poor, in spite of the existence of an antiviral treatment.27 Several case reports and extensive studies of acyclovir (30 mg/kg/day intravenously) for neonatal HSV infection have shown the effectiveness of this drug.28, 29 In addition, acyclovir is easy to administer and has few side effects; it requires only dose reduction in the presence of renal insufficiency.30 Unfortunately, only 2 patients in our series were treated with acyclovir and despite treatment 1 patient required transplantation and the other died rapidly.

Despite the rarity of this disease and in view of its very poor prognosis, we feel that the diagnosis of HSV-related hepatitis should be considered in immunocompetent or immunocompromised patients with high fever, with or without leukopenia, and with or without marked elevations of aminotransferase levels after exclusion of common causes of fulminant hepatitis. In such cases, systematic preemptive treatment with acyclovir should be given. This treatment could be stopped in the case of subsequent negative viremia or negative histology.7 Diagnostic PCR methods providing rapid results should be implemented and liver biopsy considered.

Ancillary

Advertisement