COVID-19 and Pneumocystis jirovecii Pulmonary Coinfection—The First Case Confirmed through Autopsy

Background: Establishing the diagnosis of COVID-19 and Pneumocystisjirovecii pulmonary coinfection is difficult due to clinical and radiological similarities that exist between the two disorders. For the moment, fungal coinfections are underestimated in COVID-19 patients. Case presentation: We report the case of a 52-year-old male patient, who presented to the emergency department for severe dyspnea and died 17 h later. The RT-PCR test performed at his admission was negative for SARS-CoV-2. Retesting of lung fragments collected during autopsy revealed a positive result for SARS-CoV-2. Histopathological examination showed preexisting lesions, due to comorbidities, as well as recent lesions: massive lung thromboses, alveolar exudate rich in foam cells, suprapleural and intra-alveolar Pneumocystis jirovecii cystic forms, and bilateral adrenal hemorrhage. Conclusion: COVID-19 and P. jirovecii coinfection should be considered, particularly in critically ill patients, and we recommend the systematic search for P. jirovecii in respiratory samples.


Introduction
Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2), commonly referred to as coronavirus disease 2019 (COVID- 19), has become a worldwide public health emergency. Although SARS-CoV-2 has caused millions of infections and hundreds of thousands of deaths in the world, pathophysiological mechanisms of the disease are not yet elucidated. The analysis of clinical cases information is useful, helping to improve medical practice and to prevent complications and coinfections [1].
Approximately 5% of patients with COVID-19 and 20% of inpatients develop acute respiratory distress syndrome (ARDS), septic shock and/or multiple organ failure, requiring intensive care treatments [2]. The mortality of patients admitted to the intensive care units reaches 40% [3]. The knowledge of coinfections associated with COVID-19 is relevant in reducing morbidity and mortality [4,5].
Pneumocystis jirovecii (previously called Pneumocystis carinii f. sp. hominis) is an atypical unicellular fungus and an opportunistic pathogen [6]. In immunosuppressed patients (due to HIV, malignancies, immunosuppressive therapy, organ transplantation, or congenital immunodeficiencies), Pneumocystis can cause severe pneumonia-known as Pneumocystis pneumonia or pneumocystosis [7]. Pneumocystis pneumonia is rarely found in patients with active and competent immune systems: they can be infected with Pneumocystis, but usually they are asymptomatic or the disease manifests as a mild respiratory infection. The first infection occurs usually in early life [8].
The occurrence of COVID-19 and Pneumocystis coinfection cases has previously been reported based on laboratory diagnosis [4,[9][10][11] and seems to be underestimated in clinical practice [11].
The diagnosis of COVID-19 and Pneumocystis coinfections is challenging due to overlapping clinical and radiological features, as well as limitations in laboratory diagnostics.
In this paper, we report a case of COVID-19 and Pneumocystis coinfection confirmed through autopsy, with massive thrombosis in the pulmonary circulation and bilateral adrenal hemorrhage, deceased 17 h after presentation to the emergency department. The main conclusion is that the COVID- 19 and Pneumocystis coinfections should be investigated, systematically, in all severe cases of COVID (patients admitted in the intensive care units), even in those without risk factors for immunosuppression.

Case Presentation
A 52-year-old male patient, a chronic smoker and drinker, a welder (exposed to occupational respiratory hazards for about 20 years), presented to the emergency department for severe dyspnea with sudden onset on the previous day, cough and right posterior chest pain, chills, and high fever (39 • C). The patient had a series of comorbidities: hepatic steatosis, chronic alcohol liver disease, hypertensive and ischemic heart disease, and essential hypertension.
Chest X-ray evidenced a pulmonary condensation area in the posterior inferior third of the right lung ( Figure 1A,B).
The diagnosis of right basal pneumonia, acute respiratory failure, acute chronic liver dysfunction, and acute kidney injury was established. The patient was transferred to the intensive care unit. Anti-inflammatory steroid, antibiotic, antihypertensive, oxygen therapy was initiated. The patient died 17 h after presentation. Laboratory tests revealed leukopenia with lymphopenia (white blood cells 1110/μL, lymphocytes 190/μL), thrombocytopenia (127.000/μL), prolonged prothrombin time (PT) (15.6 s), international normalized ratio (INR) 1.15, hepatocytolysis (aspartate aminotransferase 88.8 units/L, alanine aminotransferase 82.8 units/L), total bilirubin 2.88 mg/dL, reduced creatinine clearance (63.8 mL/min), nitrogen retention (creatinine 1.82 mg/dL), inflammatory syndrome (C-reactive protein 450 mg/L), antibodies against HIV-1 and HIV-2 were not detectable, nonreactive anti-SARS CoV-2 antibodies. Combined throat/nasal sampling (with collection device produced by Sanimed International Impex, Bucharest, Romania) was performed for real-time PCR (RT-PCR) SARS CoV-2 (QuantStudio v5 analysis method with TaqPath COVID-19 CE-IVD RT-PCR Kit, The QuantStudioTM Design & Analysis Software, Thermo Fisher Scientific, Pleasanton, CA, USA). The result was negative. Respiratory secretions were not collected for microbiological diagnosis. The necropsy was performed 12 h after death. Necropsy exam revealed the following: about 1700 mL of serous fluid in the right pleural cavity, with gray deposits, which were also present focally on the surface of the right pleura, and about 500 mL of serous citrine fluid in the left pleural cavity. The middle lower lobe of the right lung had a well-demarcated red-gray area, with small violaceous areas of increased consistency, having a dry, granular cross section surface. The adrenal glands showed bilateral medullary hemorrhage. No signs of autolysis were identified.
Given the persistence of the suspicion of SARS-CoV-2 infection, five lung fragments (0.5/0.3 cm) were collected by cutting with scalpel from the red-gray area, middle lower lobe (with collection device produced by Sanimed International Impex, Bucharest, Romania). RT-PCR testing was performed on these lung fragments and the result was positive. Total RNA isolation was performed with NucleoSpin RNA for tissue (Macherey-Nagel, Dueren, Germany) according to the manufacturer's instructions. The E gene assay (Envelope gene) was used as the first line screening tool (Ct 23) then followed by confirmatory testing with an RdRp gene assay (Ct 25.3) (RNA-dependent RNA polymerase gene-inside the Orf1ab polyprotein gene) and N gene assay (Ct 26.5) according to the Charite/Berlin protocol [15].
The histopathological examination of the case evidenced preexisting lesions in various organs, which could be assigned to preexisting morbidities, favoring the severe evolution of COVID-19 infection. The histopathological examination also detected recent lesions occurring in the context of COVID-19 infection.
In the lung tissue, a background of lung fibrosis (Figure 2A), with emphysematous areas ( Figure 2B) and thickened septa with myofibroblasts ( Figure 2C,D) was seen. On this fibrosis background, the large and small arterial lung vessels show many thromboses ( Figure 3). The thromboses are accompanied by subpleural hemorrhagic foci (Tardieu spots, resulting from acute hypoxia), as well as by acute pulmonary edema foci in the rare alveoli unoccupied by inflammatory exudate. In the thrombosed vessels, lymphocytes adherent to the endothelium, with a significance of endotheliitis ( Figure 4) but with low intensity, can be seen. In the dilated perivascular lymphatics, there is a cell-rich On this fibrosis background, the large and small arterial lung vessels show many thromboses ( Figure 3). The thromboses are accompanied by subpleural hemorrhagic foci (Tardieu spots, resulting from acute hypoxia), as well as by acute pulmonary edema foci in the rare alveoli unoccupied by inflammatory exudate. In the thrombosed vessels, lymphocytes adherent to the endothelium, with a significance of endotheliitis ( Figure 4) but with low intensity, can be seen. In the dilated perivascular lymphatics, there is a cell-rich lymph fluid, with a cellularity similar to that of alveolar exudate ( Figure 4A). On this fibrosis background, the large and small arterial lung vessels show many thromboses ( Figure 3). The thromboses are accompanied by subpleural hemorrhagic foci (Tardieu spots, resulting from acute hypoxia), as well as by acute pulmonary edema foci in the rare alveoli unoccupied by inflammatory exudate. In the thrombosed vessels, lymphocytes adherent to the endothelium, with a significance of endotheliitis ( Figure 4) but with low intensity, can be seen. In the dilated perivascular lymphatics, there is a cell-rich lymph fluid, with a cellularity similar to that of alveolar exudate ( Figure 4A).   Eosinophilic ( Figure 5A) and basophilic microthrombi ( Figure 5B) can also be observed, which represent aspects of disseminated intravascular coagulation (DIC). Eosinophilic ( Figure 5A) and basophilic microthrombi ( Figure 5B) can also be observed, which represent aspects of disseminated intravascular coagulation (DIC). Eosinophilic ( Figure 5A) and basophilic microthrombi ( Figure 5B) can also be observed, which represent aspects of disseminated intravascular coagulation (DIC). The alveolar exudate is rich in foam cells (Figures 6 and 7). Alveolar exudate immunolabeling is positive for the leukocyte common antigen (CD45, a commonly used marker of hematopoietic cells except for erythrocytes and platelets) ( Figure 8A), but negative for CD3 (a common antibody for identifying T cells) ( Figure 8B). Immunolabeling of the innate immune system cells for induced nitric oxide (iNOS) is extremely weak (Figure 9). The alveolar exudate is rich in foam cells (Figures 6 and 7). Alveolar exudate immunolabeling is positive for the leukocyte common antigen (CD45, a commonly used marker of hematopoietic cells except for erythrocytes and platelets) ( Figure 8A), but negative for CD3 (a common antibody for identifying T cells) ( Figure 8B). Immunolabeling of the innate immune system cells for induced nitric oxide (iNOS) is extremely weak (Figure 9).        In the alveoli filled with macrophages, cystic forms with multinuclear content were also present ( Figure 6B). These proved to be positive by Giemsa stain (Figure 10A,B), Grocott-Gomori's methenamine silver stain ( Figure 10C), and PAS stain ( Figure 10D). In the alveoli filled with macrophages, cystic forms with multinuclear content were also present ( Figure 6B). These proved to be positive by Giemsa stain (Figure 10A,B), Grocott-Gomori's methenamine silver stain ( Figure 10C), and PAS stain ( Figure 10D).
Giemsa stain also evidenced the fact that focally, some alveoli were invaded by a population of filamentous bacteria, in those areas the alveolar exudate having a bronchopneumonic, bacterial superinfection appearance ( Figure 11).
In the pleura, the whitish exudate described macroscopically was a mixture of fibrin and basophilic granular material, rich in round cells with vesicular nuclei. These cells were positive by Giemsa and Grocott staining, supporting the diagnosis of pneumocystosis ( Figure 12). Starting from these pleural images suggestive of Pneumocystis, it was inferred that the foamy alveolar material phagocyted by the activated macrophages was probably identical to this basophilic granular material, and the cystic forms described in Figure 10 were Pneumocystis cystic forms.  In the alveoli filled with macrophages, cystic forms with multinuclear content were also present ( Figure 6B). These proved to be positive by Giemsa stain (Figure 10A,B), Grocott-Gomori's methenamine silver stain ( Figure 10C), and PAS stain ( Figure 10D). Giemsa stain also evidenced the fact that focally, some alveoli were invaded by a population of filamentous bacteria, in those areas the alveolar exudate having a bronchopneumonic, bacterial superinfection appearance ( Figure 11). In the pleura, the whitish exudate described macroscopically was a mixture of fibrin and basophilic granular material, rich in round cells with vesicular nuclei. These cells were positive by Giemsa and Grocott staining, supporting the diagnosis of pneumocystosis ( Figure 12). Starting from these pleural images suggestive of Pneumocystis, it was inferred that the foamy alveolar material phagocyted by the activated macrophages was probably identical to this basophilic granular material, and the cystic forms described in Figure 10 were Pneumocystis cystic forms.  The myocardial muscle fibers are hypertrophied and fragmented, and in the medium intramyocardial venous vessels, rare microthrombi are present ( Figure 13A). In the liver, alcoholic steatosis is seen. In the medium and large branches of the suprahepatic veins, there is inflammatory infiltrate which allows the diagnosis of phlebitis ( Figure 13B). The kidney presents diffuse tubular necrosis (shock kidney) ( Figure 13C), and in the adrenal glands, massive medullary hemorrhage can be observed ( Figure 13D). The kidneys (mainly located within the interstitium) and liver (portal spaces and sinusoids) are CD45 positive. The myocardial muscle fibers are hypertrophied and fragmented, and in the medium intramyocardial venous vessels, rare microthrombi are present ( Figure 13A). In the liver, alcoholic steatosis is seen. In the medium and large branches of the suprahepatic veins, there is inflammatory infiltrate which allows the diagnosis of phlebitis ( Figure 13B). The kidney presents diffuse tubular necrosis (shock kidney) ( Figure 13C), and in the adrenal glands, massive medullary hemorrhage can be observed ( Figure 13D). The kidneys (mainly located within the interstitium) and liver (portal spaces and sinusoids) are CD45 positive.

Discussions
Respiratory viral infections predispose to the development of secondary bacterial and fungal infections [16]. Approximately 30% of COVID-19 patients could develop secondary pneumonia without an identified etiology [17].
Immunological imbalances produced in COVID-19 increase the risk of serious fungal infections (invasive pulmonary aspergillosis, invasive candidiasis or Pneumocystis pneumonia). Reported cases of invasive fungal coinfections were rare, probably due to the small number of bronchoscopies and necropsies performed so far [18]. Therefore, for the moment, fungal coinfections are underestimated. A multicenter cohort study of invasive fungal infections in intensive care COVID-19 patients show an incidence of 26.7% [19].
Pathogenesis and immune response. In the life cycle of Pneumocystis there are two morphological forms: the cystic form, or ascus, and the trophic form (formerly called trophozoite-from Greek, "animal that feeds")-pleomorphic [6].
Cystic form (Figure 10) is the infective form, with a thick protective cell wall, rich in β-glucans, enabling the parasite to survive in the outside environment-the dormant stage. After respiratory inhalation, the cystic form is deposited in the alveoli and release spores, from which trophic forms develop-the active stage, in the host, associated with pathogenesis [6,20].
Pneumocystis is a pathogen commonly found in patients with defects in T cell immunity, especially CD4+ lymphopenia [12,21,22]. In patients with severe SARS-CoV-2 infection, lymphopenia is a common feature, with drastically reduced numbers of CD4+ [2].
In the most part of individuals, Pneumocystis is dormant and sparsely dispersed in the lung, without an apparent response from the host (latent infection) [13]. Considering

Discussions
Respiratory viral infections predispose to the development of secondary bacterial and fungal infections [16]. Approximately 30% of COVID-19 patients could develop secondary pneumonia without an identified etiology [17].
Immunological imbalances produced in COVID-19 increase the risk of serious fungal infections (invasive pulmonary aspergillosis, invasive candidiasis or Pneumocystis pneumonia). Reported cases of invasive fungal coinfections were rare, probably due to the small number of bronchoscopies and necropsies performed so far [18]. Therefore, for the moment, fungal coinfections are underestimated. A multicenter cohort study of invasive fungal infections in intensive care COVID-19 patients show an incidence of 26.7% [19].
Pathogenesis and immune response. In the life cycle of Pneumocystis there are two morphological forms: the cystic form, or ascus, and the trophic form (formerly called trophozoite-from Greek, "animal that feeds")-pleomorphic [6].
Cystic form (Figure 10) is the infective form, with a thick protective cell wall, rich in β-glucans, enabling the parasite to survive in the outside environment-the dormant stage. After respiratory inhalation, the cystic form is deposited in the alveoli and release spores, from which trophic forms develop-the active stage, in the host, associated with pathogenesis [6,20].
Pneumocystis is a pathogen commonly found in patients with defects in T cell immunity, especially CD4+ lymphopenia [12,21,22]. In patients with severe SARS-CoV-2 infection, lymphopenia is a common feature, with drastically reduced numbers of CD4+ [2].
In the most part of individuals, Pneumocystis is dormant and sparsely dispersed in the lung, without an apparent response from the host (latent infection) [13]. Considering Menon [9], we believe that SARS-CoV-2 infection led to a state of functional immune suppression related to CD4+ lymphopenia, which predisposes to the activation and proliferation of the Pneumocystis, with the breakout of Pneumocystis pneumonia. Also, COVID-19 patients could develop acute respiratory distress syndrome (ARDS), requiring steroids and immunomodulatory therapies, well-known susceptibility factors for developing Pneumocystis pneumonia. In our patient, no risk factors for Pneumocystis pneumonia were identified.
Patients with severe COVID-19 pneumonia also present lower CD3, monocyte, CD8, and CD45 counts. CD45 is necessary for T cell activation [1]. In this case, alveolar exudate immunolabeling was positive for CD45, but negative for CD3. This fact indicates a severe lack of immune response from T lymphocytes, which results in a functional deficit translating into very weak immunolabeling of the innate immune system cells-macrophages, neutrophils, and natural killer cells-for iNOS [23], see Figure 9.
The functional deficit due to the absence of T lymphocytes translates into a compensatory activation of the function of macrophages (type 2 pneumocytes and blood macrophages). Macrophages become foamy ( Figure 6B) due to the fact they have intensely phagocyted a preexisting foamy alveolar material, releasing in this process a wide variety of proinflammatory cytokines (cytokine storm) [23]. Again, Pneumocystis elicits a response of the alveolar macrophages and phagocytosis [13]. Thus, both pathogens cause a massive activation of macrophages, which can be observed in our case as well: most of the alveoli have a cell content with rare lymphocytes, but rich in foamy macrophages. Foamy macrophages were identified in COVID-19 infection also by other authors [24].
Due to the lack of identification of another prothrombogenic factor, in this case, we consider COVID-19 infection as a determinant of the breakout of massive pulmonary thrombosis. Coagulopathy associated with COVID-19 is a combination of low grade DIC and localized pulmonary thrombotic microangiopathy [25].
In our patient, pulmonary fibrosis and emphysematous areas were interpreted to be lesions preceding SARS-CoV-2 infection. In addition, the patient, a welder, has long been exposed to occupational respiratory hazards. Respiratory damage may have contributed, over time, to the development of pulmonary fibrosis on a toxic basis. However, three other COVID-19 positive patients, with various comorbidities, who died in the same hospital and were examined, presented pulmonary fibrosis with different degrees of severity, and in one of them we could notice the presence of Hamman-Rich syndrome (also known as acute interstitial pneumonia [26]), in which pulmonary fibrosis was developed very rapidly, on the background of COVID-19 infection.
COVID-19 infection is associated with bilateral adrenal hemorrhage [27,28], which can also be identified in our case ( Figure 13D).
Imaging and Laboratory. Multifocal ground-glass opacities are observed both in pneumonia with Pneumocystis and SARS-CoV-2 infection, making radiographic differential diagnosis difficult [29][30][31][32]. In our case, the radiological lung appearance showed a relatively well-demarcated lobar infiltration.
Sensitivity of RT-PCR from throat swabs to detect SARS-CoV-2 RNA has been reported to be up to 95% [15]. However, the sensitivity of RT-PCR depends on the type of specimen and the assay used [33]. It is unclear why in our patient the SARS-CoV-2 PCR testing results were negative.
Standard laboratory diagnosis involves the histological detection of Pneumocystis in respiratory specimens (bronchoalveolar lavage fluid, induced sputum, or lung biopsy samples), through chemical staining (Giemsa, crystal violet, and Diff-Quick stains detect both the cystic and trophic forms; toluidine blue or Grocott-Gomori's methenamine silver stains color the walls [34]) or immunofluorescence staining with anti-Pneumocystis monoclonal antibodies [35].
The diagnosis of Pneumocystis pneumonia can be made by molecular PCR tests [36]. The significance of detected Pneumocystis DNA using PCR alone remains uncertain, and can be due to colonization of the respiratory tract [37]. Real-time quantitative (RTqPCR) assays are the only PCR method recommended for diagnosis [38]. A study conducted in 108 critical patients with COVID-19 from the intensive care unit showed that 9.3% had positive Pneumocystis RTqPCR tests from bronchoalveolar lavage/tracheal aspirate/sputum samples [11].
Serological tests for detection of anti-Pneumocystis antibodies have not proven clinically useful for establishing a diagnosis of pneumonia [35,39]. β-1,3-glucan is sensitive, but not specific [38,40]. The identification of specific biomarkers would be particularly useful, especially in patients where manipulation of the airway is limited and therefore the collection of respiratory samples is difficult.
In this patient, the diagnosis of Pneumocystis pneumonia was established through histological detection in lung fragments. The lack of confirmation of the presence of Pneumocystis in lung fragments and by molecular diagnosis PCR is a limitation of our study. However, the identification of Pneumocystis cystic forms in four distinct histological stains, from four different paraffin blocks, greatly reduces the possibility that they were artifacts.
Clinical Management. The firstline drug for the treatment of Pneumocystis pneumonia is trimethoprim-sulfamethoxazole administered orally or intravenously. For patients who cannot tolerate this treatment, or if it is inefficient, the alternative is clindamycinprimaquine, dapsone-trimethoprim, intravenous pentamidine, and atovaquone. Prophylaxis with trimethoprim-sulfamethoxazole, or cotrimoxazole, dapsone, atovaquone, and aerosol pentamidine is effective in preventing Pneumocystis pneumonia in patients at risk [41][42][43].
For COVID-19 and Pneumocystis coinfection, high-dose corticosteroid therapy is controversial [10] but can be a therapeutic necessity. For COVID-19 patients with Pneumocystis positive respiratory tests who require steroids, prophylaxis was administered to avoid the risk of pneumocystosis [11]. Table 1 presents a comparison of COVID-19 and Pneumocystis coinfection cases reported so far.

Conclusions
This paper is in line with those raising a red flag about COVID-19 and Pneumocystis coinfection, which should be seriously taken into consideration, particularly, in critically ill patients, even if patients do not have classical risk factors for Pneumocystis pneumonia. Hence, we recommend a systematic search for Pneumocystis in respiratory samples for COVID-19 critically ill patients. Tests combining molecular techniques and biomarkers could optimize the diagnosis of Pneumocystis pneumonia in patients with diagnostic suspicion. Data Availability Statement: The study did not report any data, apart from those reported already in the article.