A 42-year-old man with X-linked chronic granulomatous disease (CGD) sought care at a hospital in Paris, France, for a 2-week history of cough and night sweats. He had been receiving long-term prophylaxis with itraconazole (400 mg/d) and had normal trough levels (1,240 μg/L) 1 month before his hospital visit.

At admission, blood counts showed mild leukocytosis (leukocytes 9.6 × 10⁹ cells/L, reference range 4–10 × 10⁹ cells/L), with neutrophils at 6.1 × 10⁹ cells/L (reference range 1.5–7 × 10⁹ cells/L) and eosinophils at 2 × 10⁹ cells/L (reference <0.5 × 10⁹ cells/L). Computed tomography (CT) revealed an upper left lobe consolidation (Appendix Figure). We administered broad-spectrum antimicrobial drugs (2 g meropenem 3×/d and 20 mg/kg/d amikacin). Results of bacterial and mycological cultures from sputum were negative, as was serum galactomannan.

The patient’s condition did not improve, so we administered liposomal amphotericin B (5 mg/kg/d) and caspofungin (70 mg/d loading dose followed by 50 mg/d). Bronchoalveolar lavage demonstrated hypercellularity (1.22 × 10⁶ cells/mL); manual differential showed 12% macrophages and 76% eosinophils. Results of bacterial, mycological, and mycobacterial cultures were negative. Pathology studies from a transbronchial biopsy revealed numerous eosinophilic granulomas alongside Charcot-Leyden crystals (Appendix Figure). Grocott methenamine silver staining revealed rare septated filamentous hyphae, but results of mycological cultures were negative. The patient had elevated total serum IgE (1,210 IU/mL, reference <114 IU/mL), elevated serum A. fumigatus IgE (7 IU/mL, reference <0.1 IU/mL) and A. fumigatus IgG (54 IU/mL, reference <5 IU/mL), and precipitating antibodies to A. fumigatus (2 arcs of precipitation in immunoelectrophoresis). Results of parasitologic examination of fecal samples and serologic testing for alternative causes of eosinophilia were negative.

Eosinophilia persisted (1.8–2 × 10⁹ cells/L) despite antiparasitic treatment with ivermectin (5 mg/kg/d at days 1 and 7) and albendazole (400 mg/d for 7 d). Pathology findings from a transthoracic percutaneous biopsy revealed granulomas with Grocott-positive septated hyphae. Result of an Aspergillus section Fumigati PCR on a biopsy specimen were positive, and mycological cultures yielded a mold morphologically identified as Aspergillus. After 5 weeks of liposomal amphotericin B therapy (including 2 weeks of combination therapy with caspofungin), we switched treatment to oral voriconazole (loading dose of 400 mg 2×/d, followed by 200 mg 2×/d). Normalization of eosinophilia occurred at 6 weeks.

We sent mycological cultures from the biopsy specimens to the French National Center for Invasive Mycoses and Antifungals (Paris). Molecular identification based on the partial sequence of the internal transcribed spacer 2, 5.8S ribosomal RNA gene, and internal transcribed spacer 2 (525/526 bp; 99% similarity to the type strain, CBS 130245; GenBank accession no. KF558318.1) and the β-tubulin target gene enabled the identification of Aspergillus felis (109/109 bp; 100% similarity to the type strain, CBS DTO_131-E3 β-tubulin [benA] gene, partial cds; GenBank accession no. KY808576.1). The European Committee for Antimicrobial Susceptibility Testing (EUCAST) MICs with broth microdilution methods (1) were 4 μg/L for voriconazole, 4 μg/L for itraconazole, 0.25 μg/L for posaconazole, 2 μg/L for caspofungin, and 4 μg/L for amphotericin B. Based on EUCAST MIC breakpoints for A. fumigatus (2), we switched treatment to oral posaconazole (loading dose of 300 mg 2×/d, followed by 300 mg/d). Chest CT performed 12 months after treatment initiation showed noticeable improvement of pulmonary lesions.

Invasive pulmonary aspergillosis (IPA) remains a leading cause of death during CGD and IPA typically manifests as subacute pneumonia, with little or no angioinvasion (3). This patient had pulmonary infection caused by A. felis with overlapping features of IPA and allergic bronchopulmonary aspergillosis (ABPA) (4). Sensitization to Aspergillus spp. in patients with CGD (5) and tissue eosinophilia in lung pathology studies during invasive fungal infections (6) have been reported but do not seem to be common features of IPA in patients with CGD (3,7). There was some uncertainty about whether A. felis was responsible for this overlapping phenotype between IPA and ABPA (Table).

A. felis is a member of the A. viridinutans complex, a group of cryptic species belonging to Aspergillus section Fumigati (8). Such fumigati-mimetic molds are increasingly being recognized as sporadic causes of IPA (9). A. felis has been reported as a cause of sino-orbital aspergillosis in cats, but less frequently in humans (8). In one such case of IPA, and in the few reported cases in patients with CGD of IPA caused by the closely related A. pseudoviridinutans and A. udagawae, the course of infection was more protracted than for A. fumigatus infections, and dissemination occurred in a contiguous manner (10). Nonfumigatus Aspergillus spp. exhibit decreased in vitro susceptibility to commonly used antifungal drugs. Most previously reported antifungal susceptibilities from A. felis isolates showed high MICs for voriconazole and itraconazole but lower MICs for posaconazole (8).

Because isolates may be misidentified as A. fumigatus, culture-based morphological identification of invasive fungal infections in CGD may sometimes be insufficient. In cases of breakthrough fungal infections, or when faced with an atypical or refractory course of infection, identification of the fungus at a species level by molecular methods appears to be critical to guiding proper patient management.

Top