Background

The incidence of surgical site infections (SSIs) caused by fungi varies across settings and regions and is generally rare [1, 2]. Aspergillus welwitschiae is a species belonging to Aspergillus section Nigri, which includes species that can cause diseases in humans and animals and is similar to A. niger [3]. The presence of low antifungal drug susceptibility has been documented in species of Aspergillus section Nigri, as well as in other Aspergillus species [4, 5]. SSIs caused by Aspergillus spp., including A. fumigatus, have been previously reported [6], but no reports on SSIs caused by A. welwitschiae have been published. Here, we report the case of a superficial abdominal SSI caused by A. welwitschiae, which was initially identified as A. niger according to morphological findings and later confirmed as A. welwitschiae through calmodulin gene analysis.

Case presentation

An 86-year-old woman with a history of hypertension, chronic kidney disease, and atrial fibrillation who was taking 6 mg of prednisolone daily for rheumatoid arthritis was admitted to our hospital because of sudden abdominal pain (Day 1). She was diagnosed with sigmoid colon perforation and underwent an open Hartmann operation on the day of admission. Because of the patient’s obesity, the wound was not sutured epidermally after surgery, and she remained on a ventilator in the intensive care unit (ICU). On Day 8, black lesions appeared in the subcutaneous tissue of the wound, and therefore, a wound biopsy culture was performed (Fig. 1). On Day 9, the culture results included black colonies on potato dextrose agar and Sabouraud agar (Fig. 2a). No bacterial growth was observed. A. niger was suspected on the basis of the typical morphology of the black colonies and the microscopic features of the conidial heads (Fig. 2b). Therefore, voriconazole (VRCZ), at a dosage of 400 mg twice on the first day followed by 200 mg twice daily, was administered through a nasogastric tube, and the subcutaneous tissue of the lesion was surgically removed on Day 9. No gross recurrence was observed after the operation. On Day 12, the serum Aspergillus antigen concentration cut-off index (COI) value was 0.6 (Platelia Bio-Rad, normal range: < 0.5 COI). On Days 1 and 17, the serum beta-D-glucan levels (Fungitec G test MK II, Nissui; cut-off value, 20 pg/mL) were 25 and 42 pg/ml, respectively (Table 1). Repeated sputum cultures did not reveal any Aspergillus spp. On Days 26 and 38, the serum Aspergillus antigen levels were 0.5 and 0.4 COI, respectively. VRCZ was continued until Day 40, with serum trough levels of 4.00, 5.66, and 2.56 µg/ml recorded on Days 15, 22, and 29, respectively. No obvious adverse effects of VRCZ, such as photosensitivity or increased transaminase levels, were observed. The patient continued treatment in the ICU because of unstable circulation and respiration, mainly due to ventilator-associated pneumonia and heart failure, and she died on Day 41. After the patient died, the fungal isolate was identified as A. welwitschiae on the basis of the sequence of the calmodulin gene (GenBank/EMBL/DDBJ accession no. LC800242). The neighbour-joining tree generated from the type and authentic strains is shown in Fig. 3. The isolate was deposited at the Medical Mycology Research Centre, Chiba University, as IFM 68,753. The minimum effective concentrations (MECs) and minimum inhibitory concentrations (MICs), determined with the CLSI M38 broth microdilution method [7], were as follows: micafungin (MCFG), 0.03 mg/L; caspofungin (CPFG), 0.25 mg/L; amphotericin-B (AMB), 1 mg/L; itraconazole (ITCZ), 1 mg/L; and VRCZ, 2 mg/L (epidemiologic cut-off value [ECV] proposed for A. niger: 2 mg/L [8]).

Fig. 1
figure 1

Appearance of abdominal subcutaneous tissue on Day 8 after surgery. Black pigmented subcutaneous lesions, 2–5 mm in diameter, were observed

Full size image
Fig. 2
figure 2

(a) Black colony of a sample from the black subcutaneous lesions cultured on Sabouraud agar. (b) Microscopic findings of mycelia from black colonies on Sabouraud agar. This strain was initially suspected to be A. niger on the basis of microscopic findings. The strain was subsequently identified as A. welwitschiae via calmodulin gene analysis

Full size image
Table 1 Timeline of disease progression
Full size table
Fig. 3
figure 3

The phylogenetic position of the isolate IFM 68,753 among Aspergillus section Nigri was determined according to the calmodulin gene sequence. Evolutionary distances were computed via the Kimura 2-parameter method, with a total alignment of 826 bps for 29 type strains. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown to the left of each node (only values > 70% are shown)

Full size image

Discussion and conclusions

This report highlights two important clinical issues. First, we report a rare case of abdominal SSI caused by nosocomial Aspergillus infection. Second, the identification and drug susceptibility testing of Aspergillus section Nigri as well as other Aspergillus species are important because of the reported low susceptibility to azoles in these species.

To our knowledge, this report is the first recorded case of noninvasive postsurgical aspergillosis caused by A. welwitschiae. Abdominal SSIs caused by A. welwitschiae have not been previously reported, partly because A. welwitschia may have been misidentified as A. niger on the basis of microscopic findings alone. A similar case of abdominal SSI caused by A. niger has previously been documented; the fungus was identified via matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) [9]. The species in Aspergillus section Nigri have been reported to be indistinguishable microscopically and not completely distinguishable even by MALDI-TOF MS [10]. Because calmodulin is known to be the optimal gene for species-level classification and has been used previously for the molecular identification of species in Aspergillus section Nigri [11], we performed a phylogenetic analysis of the LC800242 calmodulin gene sequence. However, we did not analyse the β-tubulin gene, as this evaluation provides insufficient information for identifying species in Aspergillus section Nigri [11].

The incidence of SSIs caused by Aspergillus spp. is unknown. The most common organisms causing SSIs are the patient’s endogenous flora, such as Staphylococcus aureus, coagulase-negative Staphylococcus, and Escherichia coli, which account for approximately 70–95% of all SSIs [12, 13]. Fungal SSIs are rare, with a reported incidence of 4/1000 admissions in intensive care unit (ICU) settings [2]. Fungal causative agents include endogenous fungi such as Candida spp. and environmental moulds such as Aspergillus flavus [2]. Abdominal SSIs caused by Aspergillus spp. are uncommon, representing approximately 2% of all cases of postoperative aspergillosis [6]. Invasive mould infections are known to occur in patients in the burn ICU, and Aspergillus has been reported to be one of the most common causative organisms [14]. Notably, A. niger accounted for 43.7% of Aspergillus strains detected in the environment of a university hospital in Iran [15], suggesting that nosocomial transmission of Aspergillus section Nigri is theoretically not rare.

The subcutaneous A. welwitschiae infection in the present case was recognized on Day 8, suggesting a nosocomial infection. Although we did not obtain environmental culture samples from the ICU in this study, there have been no documented cases of black colony-forming Aspergillus in our ICU patients over the past three years. Although A. fumigatus has been reported to cause postsurgical invasive aspergillosis [16], in this case, A. welwitschiae was thought to have caused only a superficial SSI and was not considered related to the patient’s death. Indeed, several factors have been shown to be associated with SSIs [17]. Although the risk factors for postoperative aspergillosis are unknown, the risk factors in this case were believed to be age, diabetes, steroid use, obesity, and potential skin barrier disruption given similar reports of invasive fungal diseases in the ICU [14].

The Aspergillus section Nigri organisms isolated from clinical samples in the UK were found to comprise 17.8% A. tubingensis, 55.6% A. welwitschiae, 13.3% A. niger, 6.7% A. acidus, and 6.7% unknown species on the basis of beta-tubulin/calmodulin sequences [18]. These cryptic species are commonly found to have low susceptibility to ITCZ, and the prevalence of low susceptibility to azoles in A. fumigatus, which is considered the most pathogenic species of the Aspergillus genus for humans, has been documented globally [19]. Additionally, low susceptibility to azoles has been reported in these species in Japan [4, 5]. According to the previously proposed epidemiologic cut-off values for A. niger, 30–55% of A. niger, 79.5–89.7% of A. tubingensis, and 6.8–18.6% of A. welwitschiae strains have been found to have low susceptibility to ITCZ or VRCZ [4]. Despite the low susceptibility of A. welwitschiae to VRCZ in our patient, surgical removal of the infected tissue and continued VRCZ administration were considered effective. Although A. welwitschiae infection was not considered a direct cause of this patient’s death, low susceptibility to azoles may influence the choice of antifungal agents, especially in patients with a poor clinical course. Therefore, we believe that identification and drug susceptibility testing of Aspergillus section Nigri as well as other Aspergillus species is important. AMB is considered a potential treatment option because of its MIC values. The MIC of AMB for A. welwitschiae was 0.5 mg/L, which was within the general MIC range for Aspergillus niger (0.03–0.5 mg/L) [20]. With an MIC of 1 mg/L, ITCZ was an alternative option given the ECV of 4 mg/L provided by the Clinical & Laboratory Standards Institute (CLSI) for A. niger [20].

Our study has several limitations. On Day 9, there was no indication of a rare fungal infection, and we failed to collect a tissue biopsy to confirm the presence of a subcutaneous fungal infection, as the lesion did not reappear following its removal. Although no histological examination was performed and false-positive results of the serum Aspergillus antigen and beta-D-glucan tests, which suggested that the subcutaneous tissue may have been contaminated only by A. welwitschiae, are possible, we nevertheless believe that the presence of rare black subcutaneous lesions in this patient and a positive biopsy culture result indicated an A. welwitschiae subcutaneous infection.

Here, we report the first case of a superficial abdominal SSI caused by A. welwitschiae identified via calmodulin gene analysis. In the past, cryptic species of A. niger might have been mistakenly identified as A. niger on the basis of microscopic findings or MALDI‒TOF MS. Identification and drug susceptibility testing of Aspergillus section Nigri as well as other Aspergillus species are important because these results affect antifungal drug selection.