Viral and fungal infections after liver transplantation — Part II

Abstract

Viral and fungal infections in liver transplant recipients are important to recognize and treat early because of their association with substantial morbidity and mortality. Some viruses, such as cytomegalovirus and human herpesvirus 6, have immunomodulatory properties and can facilitate other infections, including fungal infections. Cytomegalovirus has long been recognized as an important virus in transplantation, but in the past decade other viruses have also received attention in the medical literature because of their association with particular clinical syndromes. Although human herpesvirus 6 has been associated with fever, rash, and encephalitis, a direct cause-and-effect relationship is still lacking. Human herpesvirus 8 has been found to be the cause of Kaposi sarcoma. Molecular techniques (e.g., pp65 antigenemia and polymerase chain reaction) that have been introduced for routine diagnosis of viruses have facilitated the diagnosis of asymptomatic viral infections and the institution of preemptive therapy. Nonetheless, the diagnosis of invasive fungal infections in liver transplant recipients is often delayed and thus associated with high mortality. Despite the use of new antifungal agents in clinical practice and the reduced incidence of fungal infections because of antifungal prophylaxis regimens, mortality has not decreased. Future patient outcomes may improve with early identification of patients who have risk factors for invasive fungal infections and with the development of new molecular diagnostic techniques for early detection. Liver Transpl12:2–11, 2006. © 2005 AASLD.

This, the second of two review articles describing infectious complications in liver transplant recipients, is focused on viral and fungal infections. Although all infections are important in the immunosuppressed transplant recipient, and their early recognition may affect outcome, some are more common than others. Many infections have received attention in the medical literature in recent years because of their newly discovered association with clinical syndromes. This discussion focuses on those infections of greatest relevance: cytomegalovirus (CMV), Epstein-Barr virus (EBV), herpes simplex virus (HSV) and varicella zoster virus (VZV), adenovirus, human herpesvirus 6 (HHV-6) and 7 (HHV-7), human herpesvirus 8 (HHV-8), Candida, Aspergillus, and Cryptococcus neoformans.

VIRAL INFECTIONS

Cytomegalovirus

Cytomegalovirus is a member of the human herpesvirus (HHV) group and the largest virus known to infect humans. As an immunomodulator virus,1, 2 it contributes to the intensity of the immunosuppressant state of the patient, is associated with coexisting fungal infections in liver transplant recipients,3 and is associated with chronic rejection.4

BMT, bone marrow transplant; CMV, cytomegalovirus; D, donor; EBV, Epstein-Barr virus; HHV, human herpesvirus; HHV-6, human herpesvirus 6; HHV-7, human herpesvirus 7; HHV-8, human herpesvirus 8; HSV, herpes simplex virus; KS, Kaposi sarcoma; KSHV, Kaposi sarcoma-associated herpesvirus; PTLD, posttransplantation lymphoproliferative disease; R, recipient; VZV, varicella zoster virus.

Diagnosis

Infection with CMV may occur in the absence of symptoms (CMV infection), or it may result in symptomatic infection (CMV disease) with mild to life-threatening symptoms.5 CMV disease may manifest as a nonspecific, febrile viremic illness often associated with leukopenia, thrombocytopenia, or both (CMV syndrome). A second form of disease is organ-specific. CMV hepatitis, for example, is characterized by increased serum concentrations of hepatocellular enzyme. Pneumonia caused by CMV is often identified by a diffuse interstitial infiltrate. A biopsy specimen of the involved organ will demonstrate histopathologic changes suggestive of CMV infection, and intranuclear inclusions will provide strong support for the diagnosis of CMV infection. A diagnosis of CMV retinitis is usually made with a typical ophthalmologic examination by an experienced ophthalmologist without requiring virologic confirmation.5

The virus can be isolated from appropriate tissue specimens, but asymptomatic transplant recipients will commonly shed the virus in saliva and urine. Detection of the virus in the absence of symptoms is a manifestation of CMV infection rather than CMV disease. The laboratory isolation of CMV was originally conducted by cell culture using human fibroblasts and later on using the shell-vial culture method for rapid detection. In recent years, the diagnosis of CMV has become much more sensitive and at the same time very rapid, using the pp65 antigenemia assay and the polymerase chain reaction test.6–8

Risk Factors

The risk of acquiring CMV infection or disease after transplantation can be stratified by the serologic status of the donor and the recipient. The highest risk of developing CMV infection occurs in the seronegative recipient (R) of an organ from a seropositive donor (D) (referred to as D+/R−). CMV infection will eventually develop in most recipients in this category and a high percentage of them will become symptomatic. Endogenous CMV can be reactivated in seropositive recipients, or they can sometimes become superinfected with an exogenous de novo virus from the donor organ (D+/R+). Seronegative recipients who receive an organ from a seronegative donor (D−/R−) are considered at low risk, although the CMV infection can still develop from unscreened blood products. Other risk factors for CMV infection include specific posttransplantation immunosuppressive medications. The antilymphocytic agents [murine monoclonal antibody to the CD3 antigen of human T cells (Orthoclone OKT3), rabbit antithymocyte γ-globulin (Thymoglobulin), and horse antithymocyte γ-globulin (ATGAM)] are often associated with CMV reactivation.

Treatment

The antiviral medication with clinically significant anti-CMV activity is ganciclovir, which is available in intravenous and oral formulations. Although the bioavailability of oral ganciclovir is low, the oral valine ester prodrug of ganciclovir, valganciclovir hydrochloride, achieves serum levels comparable to doses administered intravenously.9 Other drugs with anti-CMV activity, foscarnet sodium and cidofovir, are nephrotoxic and therefore less suitable for use in transplant recipients.10 Ganciclovir inhibits CMV by competing with the viral enzyme DNA polymerase, and also by direct incorporation into the virus DNA and discontinuation of its chain elongation. For the medication to be active, phosphorylation to its ganciclovir-triphosphate form must occur. The initial phosphorylation is done by viral phosphotransferase enzyme; subsequent phosphorylation to the biphosphate and triphosphate forms is accomplished by cellular kinase enzymes. Although not commonly observed, viral resistance to ganciclovir may occur. A mutation in the viral UL97 gene that encodes the information of the viral protein kinase leads to ganciclovir resistance. A second mechanism of resistance results from a mutation that results in alteration of the virus DNA polymerase. Although infrequently encountered, drug-resistant CMV strains can be isolated from high-risk transplant recipients after multiple exposures to ganciclovir.11

Prophylaxis

Prevention of CMV infection is based on two strategies. The first entails administration of an antiviral agent to all transplant recipients immediately after transplantation (i.e., universal prophylaxis). Among 143 patients randomized either to intravenous ganciclovir for 2 weeks followed by high-dose acyclovir for 10 weeks or to 12 weeks of high-dose acyclovir alone,12 the incidence of CMV disease was lower in the ganciclovir-acyclovir group (9% vs. 28%). However, the effect in the subgroup of high-risk patients (D+/R−) was less pronounced (43% vs. 64%). In a later trial that compared oral ganciclovir (1,000 mg three times daily for 3 months) with placebo, the incidence of CMV disease in the ganciclovir and placebo groups was 4.8% and 18.9%, respectively (P < .001).13 A beneficial effect was observed in the high-risk group (14.8% and 44%, respectively; P = .02).13 A study in high-risk solid-organ recipients examined the efficacy and safety of a 3-month trial of oral valganciclovir compared with oral ganciclovir.14 During the 6-month observation, CMV disease developed in 12.1% of the valganciclovir group and in 15.2% of the oral ganciclovir group. However, a subgroup analysis demonstrated a higher incidence of CMV disease in liver transplant recipients who received valganciclovir (19% vs. 12%, respectively).14

The effect of CMV immune globulin in the prevention of CMV was studied by Snydman et al.15 in a randomized placebo-controlled trial. An overall reduction in incidence of CMV disease was observed in the CMV immune globulin group compared with the placebo group (19% vs. 31%, respectively), but a subanalysis of the high-risk group (D+/R−) showed no protection at all.

Preemptive Therapy

The second preventive strategy, which was introduced by Rubin,16 is preemptive therapy. Transplant recipients are carefully monitored for early evidence of viral replication by use of a sensitive laboratory method, usually CMV antigenemia assay or polymerase chain reaction for detection of viral DNA. At the first positive test result, antiviral preemptive therapy is initiated.8, 17, 18 Some investigators use a specific quantitative threshold value as an indication for initiation of antiviral medication in seropositive transplant recipients but begin treatment in seronegative recipients whenever any level of CMV antigenemia or viral DNA becomes detectable.17 In addition, many experts advise preemptive antiviral therapy whenever acute rejection is treated with augmented immunosuppression, especially when an antilymphocytic agent (e.g., Orthoclone OKT3) is used.19

Whether universal or preemptive prophylaxis is preferable remains controversial. Some authorities argue that universal prophylaxis hastens development of resistance to antiviral medication. However, the reliance of the preemptive strategy on frequent laboratory testing raises concern about the societal cost of medical care.

Despite the advances made in sensitive diagnostic methods (e.g., rapid CMV cultures and assays for antigen and CMV DNA tests) and the availability of effective antiviral drugs for prevention and treatment, CMV infection is still a major cause of morbidity, particularly in high-risk liver recipients (D+/R−).

Epstein-Barr Virus

Symptoms and signs of infection with EBV may resemble those of CMV infection. Like CMV, it may be transmitted from donor to recipient, particularly when the recipient is EBV seronegative. Reactivation of latent virus may also occur.

Fever, especially in association with leukopenia, atypical lymphocytosis, or thrombocytopenia, indicates the possibility of acute EBV infection (infectious mononucleosis-like syndrome). However, atypical symptoms and signs may occur in nearly half the cases of EBV in transplant recipients.20 Most primary EBV infections occur in children. Because as many as 90% of adults are EBV seropositive (often from a previous subclinical infection), reactivation is presumed to be the predominant pathophysiologic process in adults with active EBV infection.

Posttransplantation Lymphoproliferative Disease

Although most cases of PTLD are associated with EBV infection, some are not. The immunologic response to EBV that primarily infects B cells and causes them to proliferate requires intact T cells. The immunosuppressive agents used in transplantation to prevent rejection alter the function of T-cell immunity, allowing this proliferation to continue and to cause PTLD. The disease is considered systemic with lymphocytic infiltration of both nodal and extranodal tissue.21 Lymphoproliferation may involve the transplant organ as well. Physical examination of patients with PTLD may not reveal palpable lymphadenopathy. Retroperitoneal or other deep-space involvement may be demonstrated by computed tomography or other imaging methods. In a series of 4,000 consecutive patients who had liver transplantation between 1981 and 1998 at the University of Pittsburgh Medical Center, PTLD was diagnosed a median of 10 months after transplantation.22 The incidence in children was 3-fold greater than that in adults. One-year patient survival was 85%; estimated actuarial 20-year survival was 45%. Risk factor analysis indicated that survival was not affected by detection of EBV but was higher with limited disease, with the polymorphic form of disease, in children, and under tacrolimus-based immunosuppression.22

Diagnosis

The definitive diagnosis of PTLD is histopathologic. Histologic findings of PTLD in kidney transplant recipients have been described by Frizzera et al.23 Although a polymorphic and polyclonal pattern has a generally more benign course than a monomorphic and monoclonal pattern, there are exceptions.24 A recent histologic classification differentiates between hyperplastic and neoplastic growths.21 Although most cases of PTLD are associated with EBV and their B-cell origin can be confirmed by the presence of CD20 receptors, there has been a recent increase in PTLD that is not associated with EBV, especially when EBV occurs more than a year after transplantation.24

Risk Factors

The fact that EBV seronegativity is a risk factor for development of PTLD explains the high incidence of PTLD in children. Excessive immunosuppression, specifically the use of a monoclonal antilymphocyte agent (e.g., Orthoclone OKT3) has been linked to PTLD in heart and liver transplant recipients.25, 26 Identified risk factors in EBV-seronegative adults include corticosteroid boluses, blood products (i.e., red blood cells and fresh frozen plasma), and prior CMV disease.27 Central nervous system involvement and mortality increased substantially when EBV seronegativity, CMV mismatch, and Orthoclone OKT3 coexisted.26

Treatment

Because few centers have experience with a large number of cases of PTLD, controlled treatment studies are nonexistent, and optimal therapy remains poorly defined. PTLD is a heterogenous disorder. Some patients improve simply by reducing their immunosuppression. In other patients, aggressive disease is refractory to decreased immunosuppression and additional treatment methods must be used (e.g., antiviral medication, interferon, immunoglobulins, anti–B-cell antibodies, anti-EBV cytotoxic T-cells, chemotherapy, surgery, and radiation).28

Reduction of immunosuppression is often the first approach to treatment of PTLD. Starzl et al.29 observed that, in many PTLD cases, reduction or discontinuation of immunosuppression may lead to dramatic clinical improvement. With this strategy, anticalcineurin agents are withheld, low-dose corticosteroids are prescribed, and signs of rejection are carefully monitored.

Antiviral medication (such as acyclovir or ganciclovir) is often part of the treatment of PTLD, with the EBV viral load monitored in some patients. Using quantitative polymerase chain reaction, Green28 followed children after small-bowel transplantation and preemptively administered ganciclovir and immunoglobulins to adjust immunosuppression on the basis of serum viral load levels of EBV. The rationale for antiviral use can be questioned because these agents are active only during the replicative phase of EBV, not during the latent phase, and because EBV associated with PTLD is found within transformed B cells in a nonreplicative form.

In bone marrow recipients with oligoclonal B-cell PTLD, specific anti–B-cell monoclonal antibodies (anti-CD21 and anti-CD24) led to complete remission of disease.30 In recent years, rituximab, a similar product that targets CD20 receptor-positive cells (a B-cell marker), has been used in solid-organ transplant recipients with some success. PTLD in bone marrow transplant (BMT) recipients originates from the donor, whereas PTLD in solid-organ transplant recipients originates from the recipient. Therefore, donor-derived EBV-specific cytotoxic T lymphocytes have been used successfully in pediatric BMT patients for prevention and sometimes for treatment of PTLD31) but not in EBV-seronegative solid-organ recipients.32

Herpes Simplex Virus and Varicella Zoster Virus

Infections with two members of the herpesvirus group, HSV or VZV, were common before the institution of routine prophylaxis with acyclovir in liver transplantation. Before routine antiviral prophylaxis, HSV infections, typically with oral or genital mucositis, developed in nearly one third of transplant recipients.33 Three patients in this early series had visceral disease (e.g., hepatic, intestinal, or disseminated), and the diagnosis was established only at autopsy.33 Hepatitis and pulmonary involvement with either HSV or VZV has been reported,34–36 some without any skin manifestations.

With use of routine acyclovir prophylaxis, the incidence of this infection has declined substantially. Two oral antiviral agents with good bioavailability, valacyclovir hydrochloride and famciclovir, are used for prophylaxis and treatment of HSV and VZV infections and are well tolerated.37, 38 Susceptible transplant recipients who are exposed to VZV are usually given postcontact prophylaxis with varicella zoster immunoglobulin. Varicella vaccine is an attenuated live virus, so it can be administered to susceptible transplant candidates prior to but not after transplantation.

Adenovirus

Adenovirus is a DNA virus that causes infection in pediatric and, less commonly, adult transplant recipients. In transplant recipients and other immunocompromised patients, the virus may cause colitis, hepatitis, pneumonitis, hemorrhagic cystitis, and encephalitis. Adenoviral infection may be confused with CMV because of its similar clinical and laboratory features: fever, leukopenia, negative bacterial cultures, and presence of intranuclear inclusion bodies. Some of the more than 50 known adenoviral serotypes have been associated with specific clinical syndromes. For example, serotypes 1, 2, and 5 have been associated with hepatitis. One report described 5 of 224 (2%) pediatric liver transplant recipients in whom acute or fulminant hepatitis developed, leading to 2 deaths.39 Four of these five patients had received antilymphocytic agents for steroid-resistant rejection. The patients presented with a high fever a mean of 73 days after transplantation.

Asymptomatic infection can occur, just as with CMV infection. After liver transplantation in pediatric patients, the virus may be isolated from urine, throat, and stool specimens in 8 to 10% of recipients.40, 41 There is no approved treatment for adenoviral infection, although anecdotal reports suggest possible benefit from cidofovir or ribavirin.

Human Herpesvirus 6 and 7

CMV, HHV-6, and HHV-7 all belong to the β-subfamily of the Herpesviridae. In recent years, HHV-6 and HHV-7 have been associated with several clinical syndromes. Like CMV, they are believed to be immunomodulator viruses. Several investigators have shown that immunosuppression may induce reactivation of latent CMV, HHV-6, and HHV-7. These viruses may interact and facilitate each other's activity.42, 43 Because much more clinically relevant information concerns HHV-6 than HHV-7, this discussion will focus on HHV-6.

HHV-6 is a frequent cause of infection in childhood, and it is the cause of roseola infantum. It also frequently causes subclinical infection, so that by 2 years of age, almost all children are seropositive.44 Of the two recognized variants of HHV-6 (A and B), variant B is believed to be the cause of most infections in transplant recipients, principally by reactivation of the virus 2 to 8 weeks after transplantation.45 HHV-6 incidence in liver transplant recipients varies from 14 to 82%.45 Symptoms and signs attributed to HHV-6 include fever, rash, cytopenia (mainly leukopenia), interstitial pneumonitis, encephalitis, and liver dysfunction or hepatitis,45–48 but a definite cause-and-effect relationship with HHV-6 is often lacking. The virus may have immunosuppressive effects through interaction with CMV or by affecting the immune system directly. An association between invasive fungal infections and HHV-6 has been suggested.49

The diagnosis of HHV-6 infection requires traditional viral culture, but isolation is laborious and time consuming. A more rapid method using shell-vial culture requires only 72 hours. The serologic tests available are of limited value because of high seroprevalence in the general population and also because transplant recipients receiving immunosuppressive medication may not mount a serologic response. Sensitive polymerase chain reaction-based detection of HHV-6 may detect latent virus and overestimate the presence of infection. Detection of HHV-6 antigen in peripheral blood mononuclear cells (antigenemia) may be useful, but experience with this technique is limited.46 Antiviral agents active against CMV are also active against HHV-6, including ganciclovir, foscarnet, and cidofovir.

Human Herpesvirus 8

In addition to being the putative cause of Kaposi sarcoma (KS), HHV-8, which is also called KS-associated herpesvirus (KSHV), has been associated with Castleman disease and primary effusion lymphoma.50 Both EBV and HHV-8 belong to the γ-subfamily of Herpesviridae and cause tumors in transplant recipients.50 Before the AIDS epidemic, KS was rare in the United States. It is characterized by purplish, nodular skin lesions but also may occur in visceral organs. KS has been described primarily in kidney transplant recipients of Mediterranean descent, but some cases have been reported in liver transplant recipients receiving cyclosporine51 or tacrolimus.52 A presentation similar to that of patients with PTLD has been reported in liver transplant recipients who have persistent HHV-8 viremia prior to diagnosis of visceral KS.51 Visceral KS may involve the transplanted liver.

The precise route of transmission of HHV-8 is often unknown. It is presumed to occur by direct contact with saliva, semen, or other body fluids, but transmission can also occur through organ transplantation or transfusion of blood products. Regamey et al.53 reported transmission to seronegative transplant recipients after transplantation of seropositive-donor kidneys in Switzerland, a country with a relatively low seroprevalence for HHV-8. Seroprevalence one year after transplantation increased from 6.4 to 17.7%.53 KS developed in 2 of 25 patients (8%) who had seroconversion within 1 year after transplantation; both patients received antilymphocytic agents for steroid-resistant rejection. In addition to primary infection from de novo acquisition of the virus, HHV-8 infection may reactivate with immunosuppression. Reactivation occurs mostly in countries with a high prevalence of HHV-8 seropositivity (e.g., southern Italy).54 The transfusion of blood products may also serve as a method of transmission.55 A study from the University of Pittsburgh Medical Center identified increased HHV-8 seropositivity with organ transplantation, from 5.3% before transplantation to 15.8% after transplantation. Many of the patients who had seroconversion had received organs from seronegative donors, which suggests that HHV-8 may have been acquired from another source (e.g., blood products).55

Treatment of KS has some similarities to treatment of EBV-associated PTLD. KS lesions, especially cutaneous ones, may regress with reduction or discontinuation of immunosuppressant medications.50 For patients with visceral KS, chemotherapy agents such as bleomycin sulfate, doxorubicin hydrochloride liposome, and vincristine sulfate have been used with some success.50 HHV-8 is sensitive in vitro to antiherpes agents such as ganciclovir, foscarnet, and cidofovir. There are also anecdotal reports of treatment of KS with a combination of antiviral and chemotherapy agents in patients who have not received organ transplants.56 However, whether antiviral agents have a role in treatment of HHV-8 in transplant recipients is not known.

FUNGAL INFECTIONS

Diagnosis and Antifungal Agents

Fungal infections after liver transplantation have historically resulted in high mortality, largely attributable to delayed diagnosis and limited therapeutic options. The diagnosis of any infection relies on identification of suggestive symptoms and signs, as well as on laboratory isolation of the causative microorganism. However, in the case of fungal infections in transplant recipients, symptoms are often few and subtle, and signs are usually not specific. Moreover, laboratory identification of a fungal pathogen is difficult because of the frequent isolation of some fungi as colonizers in the absence of disease (e.g., Candida) and because of the slow growth of other pathogens. Published definitions of invasive fungal infections have made comparisons more meaningful.57 The recognition of factors that place liver transplant recipients at risk for invasive fungal infections should result in improved diagnosis of infections and possibly in improved identification of patients who may benefit from antifungal prophylaxis.

The armamentarium of antifungal agents has expanded substantially in the past decade. The polyene antifungals (e.g., colloidal amphotericin B and the newer liposomal formulations of amphotericin) achieve fungicidal activity by binding to ergosterol, thereby disrupting the fungal cell membrane. The liposomal amphotericin preparations are less nephrotoxic than the parent compound and are generally better tolerated by patients.58 The triazole antifungals (e.g., fluconazole, itraconazole, and voriconazole) act by inhibiting C-14 demethylation of lanosterol, which results in ergosterol depletion and accumulation of aberrant sterols in the cell membrane. These agents can be administered both orally and intravenously. The triazoles are active against Candida (fluconazole), Aspergillus (itraconazole and voriconazole), and such molds as Fusarium and Scedosporium or Pseudallesceria. In addition, some triazoles under development possess substantial activity against Zygomycetes. Furthermore, a new family of antifungal agents, the echinocandins, is now available. Caspofungin acetate was approved for use in 2001 and micafungin sodium was approved in 2005. These compounds inhibit the integrity of fungal cell walls by interfering with β (1,3)-glucan synthase.59 The availability of these new antifungal agents has prompted some investigators to consider combination antifungal treatment of aspergillosis, a disease that produces extremely high mortality (≥ 80%) in transplant recipients.60

Incidence and Types of Infections

Although the incidence of invasive fungal infection has decreased at many transplant centers, fungal-associated mortality is still high. Reports of two series of invasive fungal infection from a single transplant program, the University of Pittsburgh Medical Center, illustrate the declining incidence of fungal infection. In the early series, invasive fungal infection occurred in 26 (42%) of 62 adults (30 episodes) who underwent liver transplantation between 1981 and 1983.61 Most of these episodes were caused by Candida (73%) or Aspergillus (20%). Fungal-associated mortality was 69%. In a later series from 1989 through 1992, invasive fungal infection developed in 55 (6.6%) of 834 adults who underwent liver transplantation (65% had Candida, 16% had Aspergillus, 16% had Cryptococcus, and 2% had Phaeohyphomycetes.62 Fungal-associated mortality was 54.5%.

The decreased incidence of fungal infection is most likely due to multifactorial causes. Over time, surgical methods and techniques have become increasingly sophisticated and postoperative care has improved. Advances in immunosuppressive management have permitted reduced use of corticosteroids or has even eliminated steroid use. Finally, the identification of risk factors for invasive fungal infection has led to preemptive use of prophylactic antifungal agents.

Risk Factors

Many studies have identified a number of risk factors associated with invasive fungal infections in liver transplant recipients. Collins et al.63 identified 4 principal risk factors for invasive fungal infection in 168 transplant recipients: serum creatinine of more than 3 mg/dL, an operation time of 11 hours or longer, retransplantation, and early colonization. The rate of invasive fungal infections increased in this cohort from 1% in transplant recipients who lacked predictors to 67% in those who had two or more risk factors. Other investigators have identified risk factors of lengthy surgery, retransplantation, and a long stay in the intensive care unit,64 renal failure,64, 65 treatment with antibiotics,64, 66, 67 need for transfusion products,65–67 and fungal colonization. Infection with immunomodulator viruses (e.g., CMV,3 HHV-6,49 or hepatitis C virus64) has been also cited as a potential predictor of invasive fungal infection. These risk factors should be taken into consideration when antifungal prophylaxis is being contemplated.

Antifungal Prophylaxis

Attempts to prevent fungal infections have used both universal and preemptive prophylactic strategies. Nystatin was initially used to prevent Candida mucositis. Fluconazole or other triazoles are routine prophylaxis in some transplant centers, whereas an amphotericin formulation may be used in others. After the importance of risk factors for development of fungal infection became known, physicians began to use targeted or preemptive antifungal prophylaxis. For example, candidates for antifungal prophylaxis might include patients whose transplantation surgery was complicated, those who received multiple blood transfusion products, and those affected by renal failure. However, although antifungal prophylaxis has reduced the incidence of fungal infections, it has not led to improvement in the overall mortality in most series.68, 69

A double-blind, placebo-controlled, randomized trial of itraconazole 5 mg/kg daily (maximum course, 56 days) in 71 consecutive patients found that an increased proportion of patients in the placebo group received treatment for suspected but unproven fungal infection.70 Another randomized trial with 188 transplant recipients who received either an oral itraconazole solution (200 mg twice daily) or oral or intravenous fluconazole (400 mg every 24 hours) for 10 weeks found an equivalent incidence of proven invasive fungal infection (7% and 3%, respectively) and no difference in mortality.71 A third randomized trial of 129 patients who received either liposomal amphotericin followed by oral itraconazole, intravenous fluconazole followed by oral itraconazole, or placebo found reduced fungal colonization in patients who received antifungal medication but a similar incidence of proven fungal infections among the 3 groups.72 A fourth randomized trial comparing a 10-week course of fluconazole (400 mg daily) with placebo showed that fewer superficial and invasive fungal infections developed in patients who received fluconazole than placebo (4% vs. 28% and 6% vs. 23%, respectively) but overall mortality did not differ.69

Targeted Prophylaxis

To evaluate the effectiveness of targeted antifungal prophylaxis, Fortun et al.68 administered liposomal amphotericin (cumulative dose, 1-1.5 g) to recipients with specific risk factors. Compared with historical controls, these patients had an overall reduced incidence of fungal infection (14% vs. 36%) and a specific reduction in invasive aspergillosis (5% vs. 23%).68 Similarly, a study of amphotericin-B lipid complex prophylaxis at doses of 1 to 5 mg/kg showed no proven invasive fungal infection in 30 high-risk transplant recipients, and the drug was tolerated well.73

Candida

Candidiais is the most common fungal infection in liver transplantation. Candida is a known colonizer of the gastrointestinal tract. When seven different locations of the alimentary canal were sampled endoscopically in 30 liver transplant candidates, all 30 patients were found to be colonized in at least 1 site.74Candida infection may arise at transplantation because biliary spillage may occur with anastomotic connection to the bowel.

In the past decade, many transplant centers have begun using routine antifungal prophylaxis. As a result, there has been an increased occurrence of infections with Candida species other than C albicans (e.g., C glabrata and C krusei).75 When Husain et al.76 studied risk factors for invasive candidiasis, patients with such infections were more likely to have received prophylaxis and to have had a worse outcome.75, 76 Risk factors for invasive candidiasis include use of antibiotics for prevention of spontaneous bacterial peritonitis, the need for posttransplantation dialysis, and retransplantation.75

Aspergillus

Aspergillosis is the second most common fungal infection after liver transplantation, and it has been associated historically with high mortality. Infection with Aspergillus is characterized by angioinvasion, resulting in tissue infarcts that may be impossible to eradicate with antifungal medication.

Inhalation of airborne spores results in pulmonary infection, with extrapulmonary dissemination to the central nervous system and virtually any other organ. Isolation of Aspergillus from the respiratory secretions of patients who are not immunosuppressed usually indicates colonization, but its presence in transplant recipients should not be ignored unless invasive disease can be excluded.77, 78 The diagnosis of aspergillosis can be elusive, with isolation of the fungus possible in one study in only 79% of patients with Aspergillus pneumonia.77 In a series of 26 cases of invasive aspergillosis, infection occurred a median of 17 days after transplantation, 27% of patients had retransplantation, and 85% of patients had renal insufficiency (54% required dialysis), but increased immunosuppression was less important.79 Mortality in this series of patients was 92%.79

Because of the high mortality rate associated with Aspergillus infection, antifungal therapy is often instituted empirically with any clinical suspicion of aspergillosis. Prompt administration of antifungal medication, particularly high-dose liposomal amphotericin, has decreased the resulting mortality.60 The newer triazoles (e.g., itraconazole and voriconazole) and the echinocandins also appear to be promising.

Cryptococcus neoformans

Cryptococcus neoformans, yeast that is tropic to the central nervous system, is the most common cause of meningitis in transplant recipients. This fungus has been associated with bird excrement, but in many transplant recipients the source is not identified. Inhalation of fungal spores presumably results in either symptomatic pneumonia or asymptomatic infection, with subsequent dissemination to other body sites, most commonly the central nervous system. Geography and environment are important factors. In the eastern United States, where Cryptococcus can be more readily isolated from the soil, the incidence of infection is higher than elsewhere.80 Cryptococcal infection has been reported to occur with a frequency of 12 cases per 1,000 transplant recipients.81 Symptoms developed a mean of 30 months (range, 1-146 months) after transplantation. At presentation, infected patients had pneumonia only (46%), meningitis only (36%), dissemination to multiple distant organs (11%), or involvement of another single organ (e.g., lymph node) (7%). Mortality was 25%.81

Cryptococcal infection may present as a “smoldering” or subacute infection. Patients with cryptococcal meningitis may not have nuchal rigidity or headache.80, 82 Although the serum cryptococcal antigen is quite helpful for diagnosis of meningitis or disseminated disease, its sensitivity in patients with pneumonia is only about 40%. Patients without overt central nervous system symptoms should undergo lumbar puncture because of the possibility of subclinical meningitis. Guidelines for treatment of cryptococcosis in human immunodeficiency virus-negative patients have been promulgated by the Infectious Diseases Society of America.83 Cryptococcal meningitis is usually treated with a combination of amphotericin B (conventional or liposomal) and flucytosine (5-FC) for 6 weeks, followed by maintenance treatment with fluconazole for 6 months. Other regimens are also acceptable.83

Other Fungi

Other fungi may cause infection in liver transplant recipients. Fungal culture is crucial for identification of these fungi. For example, a filamentous fungus may appear to be Aspergillus in biopsy specimens but prove to be Scedosporium apiospermum, a fungus often resistant to amphotericin B but susceptible to voriconazole. Moreover, patients in geographic regions with endemic fungi may have infections (e.g., coccidioidomycosis, histoplasmosis, and blastomycosis) that put them at risk. Transplant centers in Arizona, for example, report coccidioidomycosis among transplant recipients with clinically significant frequency.84

SUMMARY

Viral and fungal infections should be recognized and treated in liver transplant recipients not only because of their capacity to cause disease but also because of their frequent association with each other. Some viruses that infect transplant recipients (e.g., CMV or HHV-6) have immunomodulatory properties and thus affect overall immunosuppression and susceptibility for other infections (e.g., fungal). Although the diagnosis of a viral infection usually occurs early and permits preemptive therapy, the diagnosis of a fungal infection is still delayed and associated with high mortality. Outcomes may be improved, however, with early identification of patients with risk factors for invasive fungal infections and with development of new molecular diagnostic techniques.

Acknowledgements

Editing, proofreading, and reference verification were provided by the Section of Scientific Publications, Mayo Clinic. We thank Dr. Jerry Smilack, MD for critical review of the manuscript.