Challenges in the management of invasive aspergillosis in hematopoietic stem cell transplantation

While hematopoietic stem cell transplantation (HSCT) has saved numerous lives, infections, particularly those due to opportunistic fungi and viruses, continue to take a toll among allogeneic recipients. In the 1990s, the frequency of systemic candidiasis sharply declined while the incidence of invasive aspergillosis (IA) rose from 6 to 11% during allogeneic HSCT [1]. In recent years, large surveillance networks reported IA as the most common invasive fungal infection (IFI) during HSCT; IA accounted for 59% of IFI followed by invasive candidiasis (25%), zygomycosis (7%) and IFI due to other molds (7%) [2,3]. Since 2002, the availability and increasing use of voriconazole, an antifungal triazole, may have contributed to a reduced incidence of IA, while zygomycosis and other mold infections are encountered more commonly. Encouragingly, the survival probability in patients diagnosed with IA between 2002 and 2004 was significantly higher than in those diagnosed between 1990 and 2001 [4]. Nevertheless, IA continues to play a dominant role in HSCT recipients with graft-versus-host disease (GVHD), and poses significant diagnostic and therapeutic challenges.

Making an early diagnosis of IFI, as with serious bacterial infections, portends a better outcome [5–7]. Likewise, a delay in the diagnosis of IA leads to a worse outcome [8]. It is well known that clinical symptoms and signs of IFI during HSCT are mostly nonspecific, and are thus not helpful for early diagnosis. Establishing a definitive diagnosis of IA early in the course of infection, based on culture and histopathology of infected tissue, is not an easy task. Frequently, invasive diagnostic procedures (e.g., lung biopsy) are contraindicated in HSCT recipients due to the presence of thrombocytopenia or severe illness. Even when tissue specimens are obtained, accurate etiologic diagnosis is not always established – the micoscopic appearance of aspergillus resembles that of many other pathogenic molds; moreover, cultures do not always yield the infecting pathogen.

Surrogate markers, namely high-resolution thoracic CT scan and serological tests (β-D-glucan and aspergillus galactomannan [GM]) are useful for diagnosis, but are not confirmatory. Typical radiological findings of IA on CT scan, that is, macronodules (>1 cm), halo and air crescent signs, are not always present and are not pathognomonic of IA; other features (e.g., ground glass appearance or tree-in-bud appearance) are nonspecific [9]. Serum (β-D-glucan is nonspecific and has not been carefully evaluated during IA. Serum aspergillus GM, although more specific, has widely variable sensitivity, is associated with false positivity, and a negative GM result, particularly in the setting of antimold prophylaxis, does not exclude IA [10]. GM in the bronchoalveolar lavage fluid appears to have better sensitivity than in the serum [11]; however, more data are needed. Given the foregoing limitations in diagnostic techniques, early identification of this serious infection remains elusive. As such, clinicians do not hesitate to initiate aspergillus-specific therapy, justifiably so, during ‘suspected’ IA. In clinical practice, microbiologically confirmed cases of IA in HSCT constitute only a small percentage of all cases presumptively treated for IA.

Limited ability in early diagnosis of IA has led to the development of different treatment strategies for an improved outcome. These strategies are far from perfect, and include prophylaxis, pre-emptive and empiric approaches.

Chemoprophylaxis

Prophylaxis is administered to high-risk patients (e.g., HSCT patients with GVHD) in the absence of infection. Posaconazole is currently recommended for use in those with GVHD based on prospective data [12]. The drug is available only as oral formulation and has to be taken three-times daily along with food. Food/drug intake is often a challenge for HSCT patients with GVHD and gastrointestinal symptoms of vomiting and diarrhea. Low serum levels of posaconazole have been associated with breakthrough aspergillosis. Voriconazole, the drug of choice for treatment of IA, was shown to decrease the frequency of IA but did not improve fungal infection-free survival and overall survival compared with fluconazole (no activity against aspergillus) among allogeneic recipients [13]. Both triazoles, now liberally used for prophylaxis during allogeneic HSCT, are not without potentially serious adverse effects. They cause abnormalities in liver function tests and have considerable interaction with many transplant-associated drugs (e.g., cyclosporine, tacrolimus and sirolimus) metabolized via the hepatic cytochrome (primarily CYP3A4) system. As with any antimicrobial class of drugs, widespread and prolonged use of triazoles may lead to the emergence of drug-resistant organisms [14]. Several reports have linked voriconazole use to a rising incidence of zygomycosis [15,16]. Moreover, unlike with fluconazole, many clinicians routinely obtain serum levels of these azoles in order to ensure efficacy and minimize toxicity. Available data do suggest that monitoring the new triazole levels are helpful [17]; however, the practice has caused considerable confusion in routine management; assays to measure serum levels are not available in most institutions and add to the cost of care. Echinocandins and polyenes, the other two antifungal drug classes with anti-aspergillus activity, are parenteral drugs not suited for long-term prophylaxis. Intravenous polyenes (amphotericin B and its lipid forms) are potentially nephrotoxic; aerosol formulations, although promising, have not been adequately examined. No good data are available for the efficacy of echinocandins and polyenes to prevent IA in HSCT recipients.

Empiric therapy

Drug therapy is empirically initiated based on minimal, nonspecific evidence. For example, in the neutropenic patient with leukemia and persistent, antibiotic-refractory fever, antifungal therapy is begun with the assumption that the presence of fever is indicative of an ‘occult’ IFI. Fever may be due to many etiologies other than IFI in such settings, and thus this strategy has led to excessive and inappropriate antifungal drug use; yet, for the past 25 years, this practice has been the standard of care worldwide in the management of all patients, including HSCT recipients, with refractory fever and neutropenia. It needs to be emphasized that the empiric antifungal approach has not been examined during the pre-engraftment neutropenic phase or in graft failure with prolonged aplasia in HSCT recipients.

Pre-emptive therapy

In an attempt to reduce excessive antifungal drug use, this strategy provides early antifungal therapy only to those high-risk patients with ‘some’ evidence of IA (probable diagnosis), based on high-resolution thoracic CT scan and serum GM assay. Microbiologic or histopathologic confirmation is not available before therapy is provided. This strategy has been evaluated in patients with hematologic malignancies undergoing intensive chemotherapy, but not in allogeneic HSCT recipients. A pilot trial showed a substantial decrease in the use of antifungal drugs compared with the empiric strategy [18]. More recently, a randomized, prospective trial comparing pre-emptive and empiric strategies, mostly in patients with acute leukemia and neutropenia, showed that the former strategy led to increased diagnosis of IFI, no increased mortality, and decreased antifungal drug cost [19]. Again, as with empiric therapy, data on pre-emptive strategy in HSCT recipients, although much needed, are not available.

The drug of choice for proven/probable IA is voriconazole. With this alternative drug therapy to previously used polyene compounds for this infection, there has been a great reduction in morbidity/mortality of patients with IA. Despite voriconazole use, the favorable response rate in IA among HSCT recipients is reported to be approximately 30% [20]. Lipid forms of amphotercin B have similar suboptimal efficacy. Whether a combination of antifungal drugs may be superior to voriconazole monotherapy is a current question. Based on in vitro, animal and retrospective clinical data, a clinical trial comparing the efficacies of voriconazole plus anidulafungin versus voriconazole in the treatment of IA is well underway. It is hoped that the study will be adequately powered to address the value of combination therapy in HSCT population. Until such data are available, it is reasonable to use a combination of antifungal drugs (voriconazole with an echinocandin) in severely ill patients with IA (e.g., profoundly neutropenic patients and those with cerebral aspergillosis). The duration of antifungal therapy remains unclear; clinical trials have used a 12-week treatment period.

Management guidelines are absent in patients who are failing primary therapy with voriconazole. Drugs from other drug classes are added during salvage therapy; uncontrolled data support a clinical response of approximately 40% under such circumstances. No evidence-based data are available for the management of those developing breakthrough IA while receiving antimold prophylaxis or for those with a prior history of IA requiring secondary prophylaxis for subsequent immunosuppression. Clinical trials are not feasible or extremely difficult to address such issues, thus an individualized approach is the current option. Adjunct drugs to enhance immunity such as the colony-stimulating factors and IFN-γ are of anecdotal benefit.

Doubtless many advances have been made in the past decade, vastly improving the management of IA. Better identification of high-risk patients, tools for early diagnosis and better-tolerated antifungal drugs are the key elements of progress. What is anticipated in the near future? Long overdue is the availability of molecular technique (i.e., PCR) for the early diagnosis of IA. Just as progress has been witnessed in the management of cytomegalovirus infection with PCR testing, the same can be anticipated in the management of IA. In addition, accurate species identification of aspergillus with polyphasic taxonomy is likely to contribute to a better understanding of the epidemiology and management of the infection. Above all, novel strategies for normalization of the immune status of the HSCT recipient are paramount for the prevention and treatment of opportunistic infections. For the present, the challenges in the management of IA are to be met with excellent clinical judgment, including a high index of suspicion, prompt intervention and judicious use of laboratory data and pharmacotherapy.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.