Characteristics of culture-positive invasive pulmonary aspergillosis in patients with hematologic diseases: Comparison between Aspergillus fumigatus and non-fumigatus Aspergillus species.

While the epidemiology and clinical differences of various Candida spp. has been relatively well-identified, data regarding invasive aspergillosis (IA) caused by different Aspergillus spp. are insufficient.We aimed to determine the epidemiology of culture-positive invasive pulmonary aspergillosis (IPA) and to compare the characteristics and outcomes of Aspergillus fumigatus IPA with those of non-fumigatus IPA in patients with hematologic diseases. All consecutive cases of IPA from 2011 to 2015 were reviewed retrospectively.There were 430 proven/probable IPA and 76 culture-positive proven/probable IPA. Excluding cases of multiple species of fungi or cases having difficulties in species-level identification, 41 A fumigatus and 22 non-fumigatus IPA (Aspergillus flavus [n = 11], Aspergillus niger [n = 6], and Aspergillus terreus [n = 5]) were compared. There were no significant differences in baseline characteristics between the 2 groups. However, disseminated IA was more common in non-fumigatus IPA (2.4% vs 18.2%; P = .046). Paranasal sinus (PNS) involvement was more common in non-fumigatus IPA. There was a trend towards higher peak serum galactomannan values in non-fumigatus IPA than in A fumigatus IPA group (median 1.33 [interquartile 0.98-3.29] vs 0.97 [0.66-1.97]; P = .084). Clinical response and mortality did not differ between groups.The culture-positive rate of proven/probable IPA was 17.7%, of which non-fumigatus Aspergillus accounted for about one-third. Disseminated IA, especially involving the PNS, was more frequent in non-fumigatus IPA than in A fumigatus IPA.

Introduction

Aspergillus fumigatus is the most common pathogen causing invasive aspergillosis (IA) in patients with hematologic malignancies. However, several studies have reported the emergence of IA caused by non-fumigatus Aspergillus spp. such as Aspergillus flavus or Aspergillus terreus.[ The clinical importance of the non-fumigatus Aspergillus spp. is based on the potential differences in in vitro susceptibilities, clinical courses, and finally the treatment outcomes of patients. In addition, there are recent reports regarding “cryptic species” among “Aspergillus complex” denominating a group of species that are very closely related and almost indistinguishable by morphologic methods, which might have decreased or affected susceptibility to amphotericin B deoxycholate or voriconazole.[ However, species-level identification of Aspergillus is still limited in current routine microbiology work in many clinical settings.

While the epidemiology and clinical differences of various Candida spp. has been relatively well-identified during recent decades, data regarding IA caused by different Aspergillus spp. are insufficient. This may be attributable to the low culture-positive rate in IA, which shows only 10% to 30% of patients with IA at any time.[ In addition, fungus culture may be falsely positive or negative,[ which might be difficult to be interpreted and differentiate true infections from colonization and/or contaminations in some cases. Nevertheless, there is a need to investigate the causative fungal organisms of IA, considering the development of antimold active prophylaxis in this decade.

The aim of this study was to identify the epidemiology, characteristics, and outcomes of culture-positive proven or probable invasive pulmonary aspergillosis (IPA) cases, and to compare the clinical characteristics of A fumigatus IPA and those of non-fumigatus IPA in patients with hematologic diseases.

Materials and methods

Patients and clinical setting

We retrospectively reviewed all consecutive cases of invasive fungal diseases (IFDs) from January 2011 to December 2015 at the Catholic Blood and Marrow Transplantation (BMT) Centre, Seoul St. Mary's Hospital. This is a 1300-bed, university-affiliated, tertiary hospital that performs over 500 stem cell transplantations (SCTs) annually. Serum galactomannan assay was routinely screened twice per week and examined daily for sequential 3 to 4 days if IFDs were suspected. Computed tomography (CT) was performed when (1) patients had symptoms (pleuritic chest pain, blood tinged sputum, or hemoptysis) or signs that suggest newly developed pneumonia, tenderness, or swelling around the paranasal sinus (PNS) or orbital area, ulcerating lesions or eschar in the nose; (2) neutropenic fever did not resolve within 3 to 5 days of initial empirical antibacterial agents, regardless of symptom; (3) serum galactomannan test was positive. There were no changes in the diagnostic strategies during the whole study periods.

Study design

Only culture-positive proven/probable IPA cases in adult (≥18 years of age) patients with hematologic diseases were included in this study. IA of deep-seated organs without lung involvement, IPA cases involving mixed growth of 2 or more fungal organisms, or cases where the Aspergillus isolate could not be identified at the species level were excluded from the analysis. Respiratory specimens were considered as appropriate when they were obtained from the lower respiratory tract and demonstrated clinical significance: sputum (group 4, 5, or 6; grade of sputum quality satisfying epithelial cells ≤25/low power field [LPF] and white blood cells [WBC] >25/LPF), bronchial washing fluid, bronchoalveolar lavage fluid, or bronchial brush according to the revised definition of IFD from the European Organization for the Research and Treatment of Cancer/Mycoses Study Group (EORTC/MSG).[ Identification of mold relied mainly on the observation of morphological characteristics by clinical microbiologists in this institute during the study period. Data on baseline characteristics, results regarding IFDs (clinical, microbiologic, laboratory, and radiologic results), and outcomes of patients were collected. The Institutional Review Board of Seoul St. Mary's Hospital approved the research protocol and waived the need for informed consent due to the anonymous and retrospective design of the study (KC16RISI0623).

Definitions

Neutropenia was defined as an absolute neutrophil count (ANC) <500/mm3 or ANC <1000/mm3 with a predicted reduction to <500/mm3 within 2 to 3 days. Severe neutropenia was defined as an ANC <100/mm3.[ IFDs were categorized as proven or probable according to the revised definition from EORTC/MSG. Disseminated IFD was defined as involvement of ≥2 noncontiguous organs.[ Responses were assessed at 6 weeks according to the previous studies with some modifications as follows.[ Complete response (CR) was defined as resolution of all clinical signs and symptoms and >90% reduction in radiologic lesions attributable to IA. Partial response (PR) was defined as clinical improvement plus >50% reduction in radiologic lesions attributable to IA. Stable response (SR) was defined as no or minimal improvement. Progression was defined as worsening of IA disease. In cases of IA involving ≥2 nonadjacent organs, the response was graded on the basis of the organ with the worse condition. Success was defined as CR or PR. Unsuccessful outcomes were defined as SR or failure. Mortality was assessed at 6 weeks, 12 weeks, and 1 year.

Statistical analysis

Categorical and continuous variables are presented as n (%) and median (interquartile range [IQR]). The chi-square test or Fisher exact test was used to compare categorical variables, and the Student t test or Mann–Whitney test was used to compare continuous variables. The Kaplan–Meier method was used to compare trends of survival. Cox proportional-hazard model was used to identify independent risk factors for 12-week mortality. A 2-tailed P value <.05 was considered significant. Statistical analyses were performed using SPSS software version 24.0 (SPSS Korea, Seoul, Korea).

Results

Categories of IPA and isolated fungal organisms

During the 5 years of study period, among 521 proven/probable IFDs, 430 cases of proven/probable IPA were identified at the Catholic BMT Center. Aspergillus spp. were isolated from clinically significant respiratory specimens in 4 of 30 proven IPA cases and 72 of 400 probable IPA cases. The culture-positive rate of proven/probable IPA was 17.7% (76/430) (Fig. 1). Eighty-three Aspergillus clinical isolates were identified from 76 proven/probable IPA cases. A fumigatus was the most common species and accounted for 56.6% (47/83) of isolates, followed by A flavus (18.1%, 15/83), Aspergillus niger (12.0%, 10/83), A terreus (9.6%, 8/83), and Aspergillus spp. (3.6%, 3/83) without species-level identification by morphology.

Figure 1

Categories of invasive pulmonary aspergillosis. ∗Reported as Aspergillus spp. without species-level identification by morphology. IPA = invasive pulmonary aspergillosis.

To compare IPA caused by A fumigatus and non-fumigatus Aspergillus spp., cases with both A fumigatus and non-fumigatus Aspergillus spp. (n = 5), IPA with combined IFD other than aspergillosis (n = 5), or reported only as Aspergillus spp. (n = 3) were excluded from the analysis. Mixed growth of non-Aspergillus spp. included Penicillium scopulariopsis (n = 1), Penicillium spp. (n = 2), Trichosporon asahii (n = 1), and Fusarium spp. (n = 1). After these exclusions, in all, 41 patients with proven/probable IPA caused by A fumigatus (A fumigatus group) and 22 patients with proven/probable IPA by non-fumigatus Aspergillus spp. (non-fumigatus group) were compared.

Patient characteristics

As shown in Table 1, there were no significant differences in baseline patient characteristics, including age, sex, underlying diseases, treatment for hematologic diseases, severity of neutropenia, and graft-versus-host disease (GVHD) between the A fumigatus and non-fumigatus groups. The most common underlying disease was acute myeloid leukemia (49.2%, 31/63), followed by acute lymphoblastic leukemia (19.0%, 12/63), lymphoma (11.1%, 7/63), multiple myeloma (11.1%, 7/63), myelodysplastic syndrome (3.2%, 2/63), aplastic anemia (3.2%, 2/63), and chronic lymphocytic leukemia (1.6%, 1/63). SCT was the most common treatment (54.0%, 34/63) for underlying hematologic diseases before the diagnosis of IPA, followed by intensive chemotherapy (22.2%, 14/63). Of the SCT recipients, 73.5% (25/34) of patients received immunosuppressive therapy for acute (grade ≥II) or chronic GVHD as the main treatment at the time of IPA diagnosis.

Table 1

Patient characteristics.

Comparison of A fumigatus IPA and non-fumigatus Aspergillus IPA

The clinical course and outcomes of patients are shown in Table 2. Serum galactomannan positivity was confirmed in about half of the cases, which did not differ between the A fumigatus and non-fumigatus groups. However, there was a trend towards higher peak serum galactomannan levels in the non-fumigatus group (median 1.33, IQR 0.98–3.29) compared with the A fumigatus group (median 0.97, IQR 0.66–1.97) (P = .084).

Table 2

Clinical course and outcomes of patients.

As shown in Table 2, dissemination rate was significantly higher in the non-fumigatus group (18.2% [4/22]) than A fumigatus group (2.4% [1/41]) (P = .046). Among the non-fumigatus Aspergillus species, A terreus showed significantly higher dissemination rate (40% [2/5]) when compared with other non-terreus Aspergillus species (5.2% [3/58]) (P = .046). Five disseminated aspergillosis cases developed in patients who had risk factors of disseminated fungal infection, such as neutropenia, chronic GVHD, and steroid use, whereas underlying diseases were varied. Organs to which dissemination was observed included PNS (n = 2), bone (n = 1), kidney (n = 1), and external auditory canal (n = 1).

The initial choice of antifungal agent for the treatment of IPA was amphotericin B deoxycholate or liposomal amphotericin B (52.4%) and voriconazole (44.4%), followed by voriconazole maintenance therapy. The distribution of antifungal agent usage showed no difference between groups (P = .829). Treatment success at 6 weeks (43.9% vs 50.0%; P = .914), recurrence of IPA (13.2% vs 15.0%; P = .569), and all-cause mortality at 6 weeks (26.8% vs 27.3%; P = .970), 12 weeks (41.5% vs 40.9%; P = .966), and 1 year (63.2% vs 61.9%; P = .924) were not statistically different between groups (Table 2).

Discussion

We reported herein the culture-positive proven/probable IPA cases in patients with hematologic diseases. In addition, as there are relatively few data regarding Aspergillus spp. from IPA patients due to the lower culture-positive rate, we compared the characteristics and outcome between IPA caused by A fumigatus and non-fumigatus Aspergillus spp. During the study period, the culture-positive rate of proven/probable IPA was 17.7% at this institute. Interestingly, however, we found that the culture-positive rate has doubled to 38% since January 2016, accompanied by clinicians’ active diagnostic efforts—increase of fungus culture prescriptions and performance rates—to identify causative Aspergillus spp. from patients with suspected IPA. About 50% of our culture-positive cases showed negative results on the serum galactomannan assay, despite compatible clinical manifestations of IPA. In such cases, the diagnostic evidence was strengthened from possible to probable categories based on the fungus culture of appropriate respiratory specimens according to the EORTC/MSG criteria.

In this study, A fumigatus was the dominant species (65%), whereas in a previous report non-fumigatus spp. comprised 70% of IA cases.[ This supports the idea that distribution of Aspergillus spp. causing IA can show institution-specific differences. Furthermore, the increase of non-fumigatus Aspergillus spp. can be related to the previous exposure to azole or amphotericin B due to the antifungal selection pressure, and azole resistance can also be detected in non-fumigatus Aspergillus in such cases.[ Antimold active prophylaxis became available to significant number of GVHD patients since July 2015 based on reimbursement practices in Korea, which could affect the future mold epidemiology.

Non-fumigatus Aspergillus spp., which comprised one-third of culture-positive IPA, presented with multiple organ involvement more frequently, especially PNS, compared with A fumigatus IPA. Although there was a relatively small number of disseminated cases in culture-positive IPA patients, this finding suggests that upon identification of non-fumigatus Aspergillus spp., there is a need to identify the foci of IA other than the lung. The characteristics and differences in the pathogenicity of each Aspergillus spp. depends on the various factors such as conidial size, virulence, germination rate, adhesion, or phagocytosis. For example, A flavus, the second most common pathogenic species of IA, produces conidia larger than A fumigatus. Such characteristics of A flavus is related to the sinus aspergillosis, whereas A fumigatus is known as the main cause of IPA since A fumigatus conidia can reach pulmonary alveoli easier.[ On the contrary, A terreus is known to have better sporulation capacity in tissue and blood, which can lead to a higher rate of dissemination compared with A fumigatus.[ In the present study, the dissemination rate was 40% in A terreus IPA and 2.4% in A fumigatus IPA, which is consistent with an Austrian study where proven IA caused by A terreus is common in leukemia patients with 60% of dissemination rate.[ However, there are intraspecies variations in virulence or pathogenicity of Aspergillus spp., and human data are still insufficient. In addition, host factor of patients can also significantly affect the dissemination rate and clinical course. Therefore, long-term human clinical data accumulation is necessary to elucidate the relationship between IA and different Aspergillus spp.

Currently, the diagnosis of IPA primarily depends on indirect biomarkers such as serum galactomannan or 1, 3, β-D-glucan assays.[ While specific antibodies and molecular methods have been developed, there still remain unmet needs for diagnosis of aspergillosis.[ Since culture result from an appropriate specimen is the gold standard for diagnosis of infection, efforts to identify causative Aspergillus spp. should be emphasized to improve the outcome based on a detailed pathogen diagnosis. Sputum culture with an adequate collection procedure could be helpful when performed in a timely manner for diagnosis and managing IPA. The strength of this study is that it was based on the infectious diseases specialist's clinical judgment of sputum culture results.

The limitation of this study is that it is based on data collected before establishing an in vitro susceptibility test with cryptic species-level identification of Aspergillus from clinically significant fungal pathogen of IFD patients. At our institute, identification of fungal pathogens is now performed by using both morphological characteristics and sequence-based molecular methods. Further investigations by our study group are ongoing. Differences between non-fumigatus Aspergillus and A fumigatus IPA, and furthermore, the characteristics of each species, should be studied by means of an in vitro susceptibility test of Aspergillus isolates from a large number of patients.

Conclusions

In conclusion, non-fumigatus comprises one-third of culture-positive IPA cases in this institute. Multiple organ involvement should be checked in non-fumigatus IPA compared with A fumigatus IPA.