This case-patient was identified during routine patient care. Thus, the need for ethics approval was exempted; verbal consent was obtained from the patient.

Figure

Figure. Computed tomography scans of patient with Bordetella hinziipneumonia and severe respiratory syndrome coronavirus 2 infection. A, B) Scan at admission showing bilateral evidence of extensive areas of mainly...

A 63-year-old man with no notable medical history was admitted for cough, asthenia, and shortness of breath starting 3 days before admission. The patient had a positive result for a SARS-CoV-2 rapid antigen autotest. At the emergency department, COVID-19 diagnosis was confirmed by a positive SARS-CoV-2 RT-PCR result for a nasopharyngeal swab sample and chest computed tomography, which showed bilateral ground-glass opacities (50% involvement) and beginning of consolidation in the lower lobes of the lungs (Figure).

He received dexamethasone (6 mg/d for 10 d), subcutaneous, low molecular weight heparin (2 × 6,000 IU/d during the entire hospitalization), ceftriaxone (2 g/d), and spiramycin (1.5 × 10⁶ IU 3×/d). On day 2 of hospitalization, he was transferred to the intensive care unit, antimicrobial drug treatment was stopped, and awake prone positioning was combined with high-flow nasal oxygen therapy.

On day 9, mechanical ventilation was applied because of acute respiratory distress syndrome, worsening hypoxemia, and gas exchange deterioration. There was no documented bacterial superinfection, and after 48 hours, the patient’s oxygenation level had improved.

On day 13, respiratory function worsened; purulent aspiration and fever developed, and inflammatory markers increased (C-reactive protein 254 mg/L [reference <10 mg/L] and procalcitonin 0.35 ng/mL [reference <0.1 ng/mL]). Four-day intravenous piperacillin/tazobactam treatment (4 g/0.5 g 4×/d) was initiated, and an endotracheal aspirate (EA) showed oropharyngeal flora (10⁷ CFU/mL) and 5 × 10⁵ CFU/mL methicillin-susceptible Staphylococcus aureus, 5 × 10³ CFU/mL B. hinzii, 5 × 10² CFU/mL amoxicillin-susceptible Escherichia coli, and 5 × 10² CFU/mL Candida tropicalis.

On day 17, another EA showed oropharyngeal flora (10⁷ CFU/mL), decreased methicillin-susceptible S. aureus (5 × 10³ CFU/mL), increased B. hinzii (10⁶ CFU/mL), amoxicillin-susceptible E. coli (10⁷ CFU/mL), and Aspergillus fumigatus (10² CFU/mL). A. fumigatus was considered as an infection because of worsening of respiratory failure despite piperacillin/tazobactam treatment, ventilatory support for severe acute respiratory distress syndrome, an A. fumigatus‒positive culture on EA (absent on previous EAs), and a computed tomography scan showing cavitating infiltrates (Figure). Voriconazole treatment (400 mg on the first day, followed by 200 mg/12 h) was given for 21 days, in combination with intravenous co-amoxiclav (1 g 3×/d).

EA was repeated on day 25 because of persistence of fever, progressive clinical deterioration, and worsening of radiologic findings and showed 10⁶ CFU/mL B. hinzii, 10⁵ CFU/mL amoxicillin-susceptible E. coli, and 10³ CFU/mL C. tropicalis, which was considered as colonization. We switched treatment to piperacillin/tazobactam (4 g/0.5 g 4×/d for 8 d), which resulted in negative results on subsequent EA samples. Testing of rectal swab samples, blood, and urine cultures remained negative throughout hospitalization. The patient was extubated on day 46 and discharged uneventfully from the hospital.

B. hinzii grew on horse blood agar (bioMérieux, https://www.biomerieux.com) at 37°C after incubation for 24 hours as smooth, gray colonies. We identified B. hinzii by using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (Biotyper; Bruker, https://www.bruker.com) and confirmed it by using whole-genome sequencing (Illumina, https://www.illumina.com) as described (11).

We initially performed antimicrobial susceptibility testing by using Etest (bioMérieux), confirmed results by using disk diffusion and broth microdilution (Thermo Fisher Sensititer System; Thermo Fisher, https://www.thermofisher.com) (Table), and interpreted results by using the 2021 European Committee on Antimicrobial Susceptibility Testing pharmacokinetic/pharmacodynamic (nonspecies related) breakpoints (12). B. hinzii CHAR-1 showed resistance to amoxicillin, cefotaxime, aminoglycosides, and ciprofloxacin; intermediate resistance to amoxicillin/clavulanic acid and ceftazidime; and susceptibility to piperacillin/tazobactam, meropenem, and imipenem. New molecules were also tested and remained susceptible, except for ceftolozane/tazobactam (Table).

We identified β-lactamase activity by using B. hinzii CHAR-1 crude protein extracts as described (13). In silico analysis showed a β-lactamase, Hinzii Bordetella lactamase (HBL-1), which had all canonical boxes of a functional broad-spectrum class A penicillinase (13) (Appendix Figure). HBL-1 had 62.7% amino acid identity with BOR-1 β-lactamase from B. parapertussis (13). Highly related sequences (99.7%‒100% amino acid identity) were identified in genome sequences of B. hinzii available in public databases, suggesting that HBL-1–like enzymes might be native to that species (Appendix Figure). MICs of aminopenicillins and carboxypenicillins might be explained by expression of HBL-1, but, as suggested for B. parapertussis, additional nonenzymatic β-lactam resistance mechanisms, such as impermeability, efflux, or penicillin-binding protein affinities, might be associated for B. hinzii. The complete genome and HBL-1 sequences have been deposited in DDBJ/EMBL (accession no. JAJTJL000000000) and GenBank (accession nos. OM212391).

The lung microbiota of deceased patients who had COVID-19 showed complex bacterial and fungal colonization by opportunistic pathogens (14). SARS‐CoV‐2 infection, antimicrobial drug pressure, alveolar damage, persistent lymphocytic depletion, mechanical ventilation, corticosteroid therapy, and prolonged hospital stays might predispose critically ill COVID-19 patients to opportunistic bacterial or fungal superinfection (2,14). Critically ill COVID-19 patients have the highest percentage of secondary pulmonary infections (34.5%) compared with percentages for severely ill (8.3%) and moderately ill (3.9%) COVID-19 patients (15). COVID-19‒associated pulmonary aspergillosis is a worrisome complication in critically ill patients who show increased illnesses and deaths (2). A. fumigatus co-infections are frequent among critically ill COVID-19 patients (2,3). Rothe et al. showed that in a group of 50 critically ill COVID-19 patients admitted to the intensive care unit, 34% were co-infected with Enterobacterales and 18% with A. fumigatus (15).

Human cases of B. hinzii infection are rare and associated mostly with immunosuppression and anterior poultry exposure (7–10). Our patient reported recent exposure to geese and ducks, which probably led to latent or chronic colonization (digestive or respiratory tract) before infection at a time when he was most vulnerable (e.g., COVID-19 and aspergillosis superinfection). Reports of patients who have B. hinzii infections seem to be increasing in recent years, which might reflect emergence of this pathogen or availability of better identification methods, such as matrix-assisted laser desorption/ionization time-of-flight mass spectrometry, 16S rRNA gene sequencing, and whole-genome sequencing (8). Among the few B. hinzii infections described, none reported Aspergillus infections (9).

Reported B. hinzii isolates were frequently multidrug resistant, including resistance to cephalosporins, aminoglycosides, and quinolones, but remained susceptible to piperacillin/tazobactam, ceftazidime, tigecycline, and meropenem (9,10). Interpretation of antimicrobial susceptibility testing is not established, and the choice of antimicrobial drugs and treatment duration are not standardized. Cases with documented pneumonia were successfully treated with piperacillin/tazobactam or cefmetazole (9). Our patient was successfully treated with piperacillin/tazobactam, but treatment with amoxicillin/clavulanic acid failed, probably because of intermediate susceptibility of B. hinzii to this antimicrobial drug. Our study suggests that B. hinzii needs to be taken into account when initiating antimicrobial drug therapy.

Increasing reports of invasive B. hinzii might indicate its emergence as a pathogen in immunocompromised patients. We describe a B. hinzii and A. fumigatus co-infection in a SARS-CoV-2‒infected immunocompetent patient who had no underlying conditions but had probable transient immunosuppression caused by dexamethasone treatment and SARS-CoV-2 infection. Our study highlights the role of opportunistic infections (by fungal or rare bacterial species) in COVID-19 patients and the need to serially monitor the bacteria/fungi in the lower respiratory tract for timely personalized treatment.