Aspergillosis is the most common invasive mycosis affecting heart transplant recipients and in spite of major advances achieved in the diagnosis and management of this complication, it still leads to significant morbidity and mortality. The overall incidence of invasive aspergillosis (IA) has decreased over the last few years. It reached 25 % before the cyclosporine A era and decreased afterwards to between 3.3 % and 14 % (Stinson et al., 1971; Guillemain et al., 1995; Grossi et al., 1992; Denning and Stevens, 1990). In a recent multicenter study of pneumonia in heart transplant recipients (Cisneros et al., 1998) Aspergillus sp. was the second most common etiologic agent and the resulting incidence of Aspergillus pneumonia was 4.2 % in the first year post-transplant. In a nation-wide study including 4,338 solid organ transplant recipients performed in Spain, 113 suffered IA (2.6 %). Heart transplant recipients accounted for 41/113 cases and the incidence in this population was 4.4 % (Muñoz et al.,1996).
Invasive aspergillosis (IA) is usually a relatively early complication. Muñoz et al. (1996) found that episodes of IA were diagnosed a median of 63 days (6-618) after heart transplantation and Cisneros et al. (1998) found that Aspergillus lung infections appeared a median of 36 days (19-139) after transplantation.
Increased immunosuppresion is commonly present before an episode of IA. Rejection has been observed in 64 % of patients some time in the 3 months before an episode of IA and in 26 % of them at the same time as IA developed (Muñoz et al., 1996). Previous (56 %) or concomitant (24 %) bacterial infections were also common, most frequently pneumonia, bacteremia, or surgical site infections (Muñoz et al., 1996).
Muñoz et al. (1996) found that more than half of their 41 patients had experienced CMV disease before aspergillosis (47 %). In 54 % of these patients CMV was present at the time of Aspergillus diagnosis. Neutropenia was detected in only 9.7 % of their patients, reintervention in 22 % and allograft dysfunction in 15 %.
Increased environmental contamination is one of the most important risk factors for IA and outbreaks have been related to the contamination of air ducts (Gurwith et al., 1971). Finally, it has been suggested that toxic levels of cyclosporine A may increase the risk of IA by reducing the antifungicidal capacity of granulocytes especially when administered with steroids (Roilides et al., 1994).
Muñoz et al. (1996) observed that IA was localized in 76 % of their infected HT patients and disseminated in the remaining 24 %. The lungs were affected in 39/41 cases, and the CNS in 7. Other locations were: kidney 3, thyroid gland 2, heart 3, skin 1, stomach 2, arteries 1, pericardium 1, trachea 1, bronchi 1 and paranasal sinuses 1. Denning and Stevens (1990) observed 19/55 (35 %) cases of disseminated aspergillosis at presentation. Dissemination occurred later in two other patients and all patients with disseminated infection died (Denning and Stevens,1990).
Muñoz et al. (1996) found that 88 % of their patients with aspergillosis had respiratory symptoms. These consisted of cough, pleuritic pain, low fever, hemoptysis and occasionally dyspnea. 17 patients (41 %) required mechanical ventilation. Lung infiltrates were alveolar in 27 cases, interstitial in 5 and mixed in 5. 19 patients (41%) had a pulmonary nodule and infection involved both lungs in 51 % of patients. Cavitation was present in 46 % of the cases, and pleural effusion in 34 % (Muñoz et al., 1996). Grossi et al. (1992) described 22 heart transplant recipients with IA. The lung was involved in 19 of them and all presented as nodular infiltrates.
Muñoz et al. (1996) noted that ten (24 %) of their heart transplant patients with aspergillosis had neurological symptoms including altered mental status, focal manifestations, seizures and headache. Out of 182 cardiac transplant recipients seen at Stanford university up to 1980, 10 developed cerebral aspergillosis and died (Britt et al., 1981). Most of these cerebral infections had spread from a pulmonary source causing areas of focal meningo-encephalitis or brain abscesses.
Mediastinitis and pericarditis are uncommonly related to Aspergillus though they have been described in heart transplant recipients resulting in the rupture of an aortic pseudoaneurysm (Byl et al., 1993; Loire et al., 1994).
Aspergillosis should be considered in the differential diagnosis of every lung infection in a HT patient. Aspergillus fumigatus is the most commonly recovered species, although other species have also been found (De Vuyst et al., 1992). More recently, Fakih et al. (1995) reported the first case of Aspergillus granulosus infection in a cardiac transplant patient.
Muñoz et al. (1996) observed that 78 % of their HT patients with aspergillosis were diagnosed alive, a median of 6 days after symptoms appeared. The most useful technique was found to be bronchoalveolar lavage followed by needle aspiration of nodular lesions. Sputum cultures have a limited value in the diagnosis of IA since they may be positive in the absence of invasive disease and vice versa. In spite of this, the diagnosis of IA should always be considered if any Aspergillus is isolated from a HT recipient.
The positivity rates of different diagnostic techniques performed by Cisneros et al. (1998) were: sputum examination 45 %, BAL 83 %, bronchial washing 100 %, transbronchial biopsy 50 %, open lung biopsy 50 % and transthoracic needle aspiration 50 %.
Conventional amphotericin B (C-AMB), liposomal preparations of AMB and itraconazole are the therapeutic options for IA. Adjunctive therapy with rifampicin should not be used (Tucker et al., 1992; Modry et al., 1985). In most studies, half of the patients for whom conventional amphotericin B therapy has failed, respond to itraconazole or a lipid-associated AMB preparation. No data to the contrary are currently available and there are some indications that therapy with liposomal AMB or itraconazole is associated with a lower crude mortality than that associated with C-AMB.
It is not possible to make an evidence-based therapeutic recommendation, since data about too few HT recipients with IA have been published in the literature and comparative studies are not available. In especially severe cases, immunosuppression may be reduced, although this approach should be done very carefully as it may lead to fatal rejection episodes.
The timing of drug administration is also important. Once appropriate samples have been obtained, therapy must be initiated as soon as possible and it may be discontinued if the diagnosis is not proven subsequently. Solid organ transplant recipients should receive systemic antifungals when any of the following circumstances are present:
1) Pulmonary infiltrate and isolation of Aspergillus spp. from respiratory secretions or BAL.
2) Bronchoscopic evidence of tracheobronchial ulcerations or bronchitis plus visualization of hyphae in the BAL sample.
3) Histological or cytological evidence of hyphal invasion in any tissue.
4) Any unexplained intracranial process, including vascular strokes, with or without fever (British Society for Antimicrobial Chemotherapy Working Party on Fungal Infection, 1993).
Muñoz et al. (1996) found that conventional amphotericin B was used in 31 % of their HT patients with aspergillosis, liposomal amphotericin B in 46 % and itraconazole in 5 %. No therapy could be given in 18 % of the cases. Mortality was significantly lower in the group of patients treated with liposomal AMB, which appeared as a statistically independent protective factor. However, prospective studies are needed to confirm this issue.
As mentioned, conventional amphotericin B (0.7-1.5 mg/kg/d) is the standard treatment of IA in most hospitals. The use of low doses of this drug is especially attractive for patients on cyclosporine. A recent study which analyzed this option showed better results in transplant recipients than in neutropenic patients (10 of the 26 transplant recipients successfully treated received <30 mg of AMB daily) (Denning and Stevens, 1990). However, the safety of this approach has not been conclusively proven and it cannot be recommended for all cases.
Endocavitary administration of AMB has also been described, although it is not very common (Camarata et al., 1992) and should not be the sole route of therapy.
Lipid preparations of amphotericin B permit the use of higher doses of the drug with a low nephrotoxicity. Therapeutic successes have been reported with each of the three modalities. However, data on solid organ transplant recipients are often mixed with those of other population groups and specific data on heart recipients are not provided in any of the studies.
Further prospective and comparative studies are needed to assess the use of lipid amphotericin B preparations in HT patients. Considering a) the very high price of these preparations and b) the lack of data on these compounds, their current indication in most centers is the treatment of systemic mycosis in patients refractory to conventional amphotericin or presenting toxicity. Lipid formulations may be especially attractive for patients with invasive aspergillosis taking cyclosporine A, for example heart transplant recipients. In my opinion, lipid formulations of AMB should be the first choice of antifungal in this setting.
A multicentre study on the use of Ambisome in the therapy of invasive mycoses in immunocompromised patients (including 17 solid organ transplant recipients), found that responses were satisfactory in 41% of Aspergillus patients (Ringdén et al., 1991).
A multicentre study using 58 patients with a probable or proven mycosis including 19 transplant patients, (two of whom were heart recipients with pulmonary aspergillosis - one previously treated with C-AMB) found that 59 % of aspergillosis patients had a satisfactory response; both HT recipients survived (Ng and Denning, 1995).
Out of 5 HT recipients with invasive aspergillosis or mucormycosis treated with liposomal amphotericin B (mean daily dose 3.1 mg/kg), 4 were cured. No significant interference with cyclosporine was detected and tolerance was very good (Berenguer et al., 1994). Liposomal amphotericin B has also been used with good tolerance in a heart transplant recipient with a cutaneous mucormycosis associated with a peripheral catheter. The patient survived although amputation was required (Baraia et al., 1995).
The safety and antifungal efficacy of amphotericin B lipid complex (Abelcet) has been recently evaluated in an open-label, single-patient, emergency-use study of patients who were refractory to or intolerant of C-AMB. Out of 130 cases of aspergillosis, 27 occurred in solid organ transplant recipients. There was a complete or partial response to ABLC in 42 % of aspergillosis cases (Walsh et al., 1998).
Amphotericin B colloidal dispersion (ABCD) was administered to 10 solid organ transplant recipients with proven or probable aspergillosis (White et al., 1997). No patient obtained a complete response in this group, but 4 (40 %) obtained a partial response.
Itraconazole has been used in the treatment of heart transplant patients with invasive aspergillosis. Several studies reported succesful results (Viviani et al, 1989; Denning et al., 1989; Sánchez et al.,1995; Faggian et al., 1989; Grossi et al., 1992; Gamba et al., 1991).
The use of Itraconazole in the treatment of invasive aspergillosis in patients with severe diseases has also been investigated (Denning et al., 1994). Solid transplant recipients showed a 7 % failure rate (44 % failure rate in HIV positive patients in the same study). At the end of therapy 6/14 transplant patients had a complete response, 2 a partial response, 2 were stable and 4 were considered as therapeutic failures. Median length of therapy was 322 days and therapy had to be interrupted in 29 % of the patients. No specific data on HT patients was provided. Besides its use as initial therapy, itraconazole has also been used as maintenance therapy in patients previously treated with amphotericin.
The recommended administration schedule includes loading doses (600 mg daily for 4 days) and subsequent doses of at least 400 mg daily (Denning, 1998; Denning et al., 1994; Caillot et al., 1997). The new formulation of itraconazole in cyclodextrin is better absorbed (Prentice et al., 1997) and should be used initially in the same dose. Cyclosporine A levels should be reduced to half the same day itraconazole is prescribed (Kramer et al., 1990) and then monitored. Itraconazole serum concentrations should be measured within 7 days of starting therapy. When itraconazole serum concentrations are low (e.g. <1mg/l by HPLC, or <5 mg/l by bioassay) the dose of itraconazole should be increased to 600 mg or 800 mg daily (in divided doses of 200 mg) or therapy changed. Capsules should be administered with food or with an acid beverage. Itraconazole syrup is better absorbed on an empty stomach. Itraconazole is thus a therapeutic option that should be considered in solid organ recipients with mild or moderately severe invasive aspergillosis, able to take oral medication and not receiving drugs that reduce itraconazole levels.
Muñoz et al. (1996) noted that four of their HT patients underwent surgical intervention for the treatment of aspergillosis. The main indications for surgery are haemoptysis, poor clinical response, isolated lesions, endocarditis, endophthalmitis, sinusitis, peritonitis and osteomyelitis. The surgical resection of foci of aspergillosis may reduce the need for antifungal therapy and can prevent potentially fatal hemorrhages.
Heart transplant recipients with invasive aspergillosis usually have lower mortality rates (50 %; range 11-78 %) when compared with other host groups (Denning, 1996) and the prognosis is even better for cardiac patients who have isolated pulmonary infections. Muñoz et al. (1996) found that 27 of their HT patients died (65.8 %) and death was directly attributed to the IA in 22 cases (53.6 %); one patient relapsed.
Primary prophylaxis against aspergillosis is not recommended. Some centers with a very high incidence, including ours, administer itraconazole during the first months or in periods of high immunosuppression. I have implemented this approach in more than 70 patients with satisfactory results. However, more studies are needed before this approach can be recommended. Secondary prophylaxis is not indicated.
Dr Patricia Muñoz
Servicio de Microbiologia y Enfermedades Infecciosas
Hospital General Universitario Gregorio Marañón
Doctor Esquerdo 47
28007 Madrid, Spain