Successful treatment of pulmonary invasive fungal infection by Penicillium non‐marneffei in lymphoblastic lymphoma: case report and literature review

Key Clinical Message Penicillium non‐marneffei species rarely cause disease in humans and are encountered most commonly in the clinical laboratory as culture contaminants; however, recently they have emerged as opportunistic pathogens in immunocompromised hosts; therefore, it should not be routinely disregarded without a thorough investigation, especially if normally sterile sites are involved.


Introduction
Members of the genus Penicillium are ubiquitous in the environment, encountered in air, soil, and decaying material, and also they have been reported as the second most dominant fungal contaminant in the International Space Station. Penicillium spp. have also been described as components of human oral and gut microbiota [1]. The conidia become easily aerosolized; this explains their role as environmental contaminants [2]. Historically, non-marneffei species rarely cause disease in humans and are encountered most commonly in the clinical laboratory as culture contaminants; however, recently they have emerged as opportunistic pathogens in immunocompromised hosts. Proof of clinical infection requires histological demonstration of tissue invasion.
Herein, we report a case in which a Penicillium nonmarneffei species pulmonary invasive fungal infection (IFI) was diagnosed synchronously to lymphoblastic lymphoma.

Case Report
A 16-year-old male was admitted with 4 weeks of systemic symptoms, bilateral pleural and pericardial effusions. A chest X-ray and CT scan revealed pulmonary nodules and a mediastinal mass ( Fig. 1A-C). The patient underwent bilateral thoracoscopy. A lung mass biopsy was performed, and lymphoblastic lymphoma was diagnosed with pathology and flow cytometry. Direct smears of the pulmonary nodules were negative, but on fungal culture (lung tissue was inoculated on Sabouraud dextrose agar at 30°C), a colony of rapidly growing mold was isolated, identified as Penicillium spp. (Fig. 1D and E). Multiple attempts to identify species were unsuccessful, including molecular techniques, analysis of sequences of recombinant DNA internal transcribed spacer (ITS 1 and ITS2), and MALDI-TOF MS. The infection was confirmed with histological sections that stained with hematoxylin and eosin showing necrotic lung and pleural tissue with hemorrhagic areas, and Grocott methenamine silver showed numerous branching hyphae surrounded by suppurative granulomatous inflammation and vascular invasion ( Fig. 1F and G). Antifungal susceptibility testing was determined according to CLSI criteria. In vitro susceptibility at 48 h reported minimal inhibitory concentrations for the following antifungals as follows: amphotericin B 1.0 lg/mL and voriconazole 2 lg/mL. The patient completed an 8-week course of amphotericin B deoxycholate, followed by amphotericin B-suppressive treatment during each of the remaining seven cycles of chemotherapy, without recurrence of infection. Nine months later, the patient underwent haploidentic hematopoietic stem cell transplantation. Suppressive treatment with amphotericin B was prescribed again, without recurrence of the infection. Up until the present date, 12 months after HSCT, the patient is in remission, without immunosuppression, and without relapse of the infection. We postulate this case represents infection by Penicillium non-marneffei species as the sample was obtained from a surgical specimen in a sterile site, there was no other isolation, and the galactomannan was negative.
The clinical presentation of Penicilliosis caused by nonmarneffei species reported to date is different to that reported for T. marneffei species; although pulmonary involvement is common in both, non-marneffei species, particularly in patients with hematological malignancies, usually cause disseminated infection in the setting of invasive pulmonary fungal infection, with high fatality rates.
In patients with hematological malignancies, they have emerged as potential opportunistic agents, mainly described in acute lymphoblastic and myeloid leukemia.  Pulmonary invasive fungal infection involving P. citrinum, P. purpurogenus, P. notatum, P. brevicompactum, P. oxalicum, and P. commune has been reported [13][14][15][16][17]; including pericardial involvement [13], coinfection with P. jirovecii [17], pulmonary and hepatosplenic involvement [18], and disseminated disease [19]. In patients undergoing transplantation with hematopoietic progenitors, P. brevicompactum has been described causing invasive pulmonary infection [20], and, after autologous bone marrow transplantation for multiple myeloma, pulmonary infection due to T. purpurogenus has been described [21]. In many cases, the diagnosis was made postmortem with an attributed mortality of 62% [13,14,19]. In summary, according to the review, the most prevalent species causing infection in immunosuppressed patients is P. chrysogenum, which causes systemic and disseminated disease with invasive pulmonary infection. The susceptibility profile is not predictable and may vary according to isolates; some have high MICs to azoles and amphotericin B.
Confirming a diagnosis of Penicilliosis can be difficult by conventional phenotypic methods. Identification at the species level remains challenging, and the high number of species in these genera makes this task even more difficult [1]. The use of molecular methods, however, provides a rapid and relatively simple method for the identification of Penicillium species. Matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) has limitations for identification of dimorphic fungi but may identify species of Penicillium and differentiate T. marneffei from non-marneffei species as P. brevicompactum and P. chrysogenum; however, the scores were below the cutoff required for species identification. This can be explained by the limited number of spectra for the two species so that expansion of MALDI-TOF MS databases is needed [24].
Standard treatment for non-marneffei species has not yet been established; antifungal susceptibility data for clinically available antifungal agents and treatment options for infections caused by Penicillium species are also poorly understood, aside from data published for T. marneffei. In a recent study of 118 clinical isolates, (mainly from the respiratory tract/human bronchoalveolar lavage) terbinafine (TRB) and the echinocandins showed the best in vitro activity against Penicillium species with MIC <0.03 lg/mL for TRB, 0.06 lg/mL for caspofungin and anidulafungin, and 0.125 for micafungin; amphotericin B showed intermediate activity with MIC of 2 lg/mL, and azoles revealed variable activity with MIC ranges of 0.5 lg/mL for posaconazole and 2 lg/mL for voriconazole and itraconazole [25].
In conclusion, Penicillium spp. isolates in immunosuppressed patients should not be routinely disregarded without a thorough investigation, especially if normally sterile sites are involved.