Never Say Never: Unexpected Pulmonary Pathogens Found on Autopsy in Hematopoietic Cell Transplantation Recipients

Ashrit Multani ¹, Libby S. Allard ², Joanna K. Nelson ^{1, *}

1. Division of Infectious Diseases and Geographic Medicine, Department of Medicine, Stanford University School of Medicine, Stanford, California, USA

2. Department of Pathology, Stanford University School of Medicine, Stanford, California, USA

* Correspondence: Joanna K. Nelson

Academic Editor:  Dora Ho

Special Issue: Infectious Complications in Hematopoietic Stem Cell Transplantation

Received: June 03, 2019 | Accepted: July 25, 2019 | Published: July 29, 2019

OBM Transplantation 2019, Volume 3, Issue 3, doi:10.21926/obm.transplant.1903076

Recommended citation: Multani A, Allard LS, Nelson JK. Never Say Never: Unexpected Pulmonary Pathogens Found on Autopsy in Hematopoietic Cell Transplantation Recipients. OBM Transplantation 2019; 3(3): 076; doi:10.21926/obm.transplant.1903076.

© 2019 by the authors. This is an open access article distributed under the conditions of the Creative Commons by Attribution License, which permits unrestricted use, distribution, and reproduction in any medium or format, provided the original work is correctly cited.

1. Introduction

Hematopoietic cell transplantation (HCT) recipients frequently experience infectious disease-related complications, of which pneumonia is one of the most common [1,2,3,4,5]. In this patient population, infectious pneumonia may be caused by a wide variety of pathogens, including bacteria, viruses, fungi, and parasites. However, when isolated from respiratory tract specimens, Candida and Enterococcus species are virtually always disregarded as airway colonizers rather than true pulmonary pathogens [6,7,8,9,10,11]. Some clinicians even go so far as to falsely proclaim with conviction that these organisms never cause pneumonia. Here, we present a case series of HCT recipients with autopsy-proven Candida pneumonia (CP) or enterococcal pneumonia (EP). Because of the paucity of publications about these clinical entities in HCT recipients, we will review the pertinent literature in all patient populations.

2. Methods

This is a single-center, retrospective, autopsy-driven case series of HCT recipients at Stanford University Medical Center (Stanford, CA). Between January 1, 2000 and December 31, 2017, 185 HCT recipients underwent autopsy. The autopsy reports of these patients were reviewed. A diagnosis of CP and/or EP listed on the autopsy record prompted an in-depth review of the patient's clinical record. Data collected included patient demographics (including age and sex), underlying disease, donor and HCT type, graft source, preparative regimen, development of graft-versus-host disease (GVHD), date of death, premortem clinical information, and postmortem autopsy diagnoses. Patients were excluded if they were less than 18 years of age, their autopsy report was either missing or incomplete, or their medical record had insufficient clinical data available for review. This study was exempt from review and approval by the institutional review board of Stanford University since chart review of deceased subjects did not meet the definition of human subjects research.  

3. Part I - Candida Pneumonia

3.1 Illustrative Case I

A 28-year-old man (Table 1, patient #1) with acute lymphoblastic leukemia underwent myeloablative matched unrelated donor bone marrow allogeneic HCT (day 0). He was on acyclovir, ciprofloxacin, and fluconazole prophylaxis per protocol. On day +1, he developed severe headache, photophobia, and nausea without focal deficits, fever, chills, or sweats. On day +2, computed tomography (CT) of the head did not reveal intracranial pathology. Lumbar puncture could not be performed safely due to thrombocytopenia. Tacrolimus was held due to concern for calcineurin-associated neurotoxicity. Cefepime and vancomycin were added. On day +4, total parenteral nutrition was started. On day +13, he developed hyperbilirubinemia for which fluconazole was changed to micafungin. CT of the chest showed bibasilar patchy consolidation (Figure 1). On day +16, transthoracic echocardiogram was negative for vegetations or other evidence of infection. On day +17, he was transferred to the intensive care unit after developing progressive multiorgan dysfunction including respiratory failure and altered mentation. On day +20, blood, respiratory, urine, and stool cultures were obtained. He died on day +21 after which four out of four blood culture bottles grew Candida guilliermondii. Antifungal susceptibility testing was not performed. Postmortem histopathological findings revealed disseminated candidiasis involving the lungs, myocardium, gastrointestinal tract, liver, kidneys, muscle, and central nervous system (occipital cortex, midbrain, and medulla) without corresponding inflammation (Figure 2). CT of the chest was not repeated closer to the patient's death. Therefore, the discrepancy between premortem imaging findings and postmortem autopsy findings is most likely explained by the elapsed time between the two.

Additional cases of autopsy-proven CP are presented in Table 1.

3.2 Epidemiology

CP, formerly termed pulmonary moniliasis, was first described in 1905 by Castellani [12]. The first report of CP to link clinical manifestations with autopsy findings was published in 1933 by Seaborn J. Lewis [13]. CP has since been reported in multiple publications but remains exceedingly rare [6,7,8,14,15,16,17,18,19,20,21]. Because definitive diagnosis of CP requires histopathological evidence of fungal invasion in lung tissue (which is usually obtained only at autopsy) and autopsy rates have been declining over time, its frequency is likely underestimated [8,14,17,20,22,23]. Immunocompromised individuals, such as those with hematologic malignancy or transplant recipients, appear to be at higher risk compared to the general population [8,14,17,19,20,21,24]. The incidence of CP in HCT recipients is not known. Candida airway colonization and CP are associated with poorer outcomes and increased mortality but it is unclear if these are causative or surrogate markers of illness severity [25].

3.3 Risk Factors

Potential risk factors for CP are summarily listed in Table 2. CP primarily occurs via one of two mechanisms: hematogenous spread to the lung(s) or aspiration of organisms from the upper airway into the lung(s) [14,17,18,20]. Risk factors for CP due to hematogenous dissemination are similar to those for candidemia and include diabetes mellitus, critical illness, multifocal colonization, total parenteral nutrition, central venous catheters, renal replacement therapy, surgery (especially if involving the gastrointestinal tract), trauma, corticosteroids, and broad-spectrum antimicrobials. [17,26,27,28,29]. CP resulting from aspiration is mainly a consequence of neurological disorders (including neuromuscular diseases), altered mentation, and antecedent vomiting  [14,17,18,29,30,31]. It is not clear why aspiration, while being such a common phenomena, results in CP only rarely. Multiple contributory factors are likely implicated, including the organism burden in aspirated material as well as local and systemic immune impairment. Immunocompromised state and neutropenia have been cited as risk factors but their absence does not exclude the possibility of CP [8,13,14,16,17,18,20,24,31,32]. Additional factors that have been identified to increase risk for invasive candidiasis in HCT recipients include increased age, increased duration of neutropenia, acute GVHD, and use of total body irradiation as part of the preparative regimen [33]. All of the patients in our cohort had multiple risk factors for CP, including three who underwent total body irradiation-containing preparative regimens (Table 1).

Table 1 Case series of hematopoietic cell transplantation recipients with autopsy-proven Candida pneumonia.

Figure 1 Computed tomography of the chest of Illustrative Case I demonstrating patchy consolidation in both lung bases correlating with autopsy-proven Candida pneumonia.

Figure 2 Hematoxylin and Eosin stain (10X magnification) of postmortem lung tissue of Illustrative Case I. Sections demonstrate a nodular pattern of budding yeast-like forms proliferating in terminal airways and alveolar ducts (black arrows) without host inflammatory response. Inset (40X magnification of area outlined in red) highlights numerous spherical budding yeast forms exhibiting variable size (2-6.5 µm) without pseudohyphae. Postmortem cultures isolated Candida guilliermondii.

Table 2 Potential risk factors for Candida pneumonia and enterococcal pneumonia.

3.4 Clinical Manifestations

The clinical manifestations of CP are nonspecific. Multiple authors have characterized three main clinical forms of CP: mild, moderate, and severe [12,30]. The severe form may either be acute (manifesting as bronchopneumonia) or chronic (mimicking pulmonary tuberculosis) [12,30]. Similar to the patients in our cohort (Table 1), patients may present with fever, night sweats, weakness, fatigue, cough (productive or non-productive), hemoptysis, chest pain, dyspnea, hypoxemia, tachypnea, wheezing, rales, rhonchi, and/or altered mentation (which may be a cause or result of CP) [12,13,14,17,20,21,30,31,34,35]. Almost all of our patients had gastrointestinal symptoms (Table 1). The illness may subsequently progress to acute respiratory distress syndrome or respiratory failure requiring mechanical ventilation, multiple organ dysfunction syndrome, septic shock, and death [8,14,21].

3.5 Microbiologic Findings

According to the 2016 Infectious Diseases Society of America clinical practice guideline for the management of candidiasis, isolation of Candida from respiratory tract specimens usually indicates colonization and rarely requires targeted antifungal therapy [25]. Multiple studies have shown corroborative evidence that the incidence of CP is extremely low (approaching but not consistently reaching zero) in patients who have Candida isolated from respiratory tract specimens [6,7,16,32]. In previous reports of CP, Candida was recovered as either monomicrobial growth or as part of mixed growth [12,13,21,29,30,31,34,35,36]. Highlighting the underappreciation of the ability of Candida to cause pleuropulmonary disease, two of our patients had yeast isolated from a premortem respiratory tract specimen as monomicrobial growth (neither speciated) and two had yeast recovered as part of mixed culture. All five of our patients with autopsy-proven CP had Candida recovered from postmortem culture despite receiving an antifungal at the time of death. Although the vast majority of articles evaluating CP did not provide data regarding speciation, species-related outcomes of CP should not differ from those of extrapulmonary invasive candidiasis [25,37]. Blood cultures may be useful in the setting of hematogenous dissemination, such as in our illustrative case. However, in evaluating for tissue-invasive candidiasis including CP, blood cultures are notoriously insensitive [25,38]. A prior systematic review reported a pooled blood culture sensitivity of only 38% [39]. Therefore, non-culture based methods have been increasingly developed and implemented. Molecular diagnostics employing polymerase chain reaction (PCR)-based assays and miniaturized magnetic resonance technology have demonstrated markedly improved performance characteristics and turnaround times for the detection of candidemia [39,40]. Serum beta-D-glucan (BDG) assays have demonstrated potential in some studies, particularly when measured serially [28,41,42,43,44]. Using higher cutoff values, BDG levels in both endotracheal aspirate and bronchoalveolar lavage specimens (but not serum) have demonstrated good diagnostic value for CP, particularly in the absence of candidemia [41]. However, BDG results need to be interpreted with caution since they are prone to numerable false-positives (e.g., albumin administration, intravenous immunoglobulin administration, blood product transfusion, multiple antimicrobials, hemodialysis, gauze exposure) [28,41,45,46,47].

3.6 Radiologic Findings

Radiologic findings are varied and not specific for CP (Table 3). However, certain patterns may be seen more frequently. The two main radiologic patterns of CP may correlate with the two main mechanisms of its development, with nodules reflecting hematogenous spread and bronchopneumonia resulting from aspiration [28,48]. Chest radiographic findings commonly reported with CP include bronchopneumonia or airspace disease with segmental, lobar, or bilateral diffuse distribution and predilection for basilar involvement [19,29,49,50]. Pagani and Libshitz described CP presenting often as a miliary-nodular pattern which would likely correlate with hematologic dissemination based on pathophysiology [49]. However, this was challenged by studies by Buff et al. and Dubois et al. [19,50]. Thin-section CT findings in HCT recipients are also nonspecific [48,51]. The most frequently reported CT findings were multiple bilateral nodules ranging from 3 to 30mm in diameter [48,51]. The nodules were either well-defined or associated with patchy airspace consolidation, tree-in-bud pattern, and/or ground glass opacity [48,51]. Some cases demonstrated nodules with surrounding ground glass opacity (CT halo sign) but cavitation was infrequently seen [48,51]. Because masslike consolidation, cavitation, miliary-nodular pattern, and pleural effusions were rarely found in their case series, Buff et al. suggested that their presence could be useful in excluding CP [50]. In our cohort, patchy consolidations and ground glass opacities predominated while nodules, masslike consolidation, cavitation, and pleural effusions were infrequently seen (Table 1). However, because some of our patients had additional causes of concomitant pulmonary pathology, it is difficult to determine to what degree the radiologic findings could be attributed to CP in these cases.

Table 3 Radiologic findings of Candida pneumonia and enterococcal pneumonia.

3.7 Histopathological Findings

The definitive diagnosis of CP requires histopathological evidence of fungal invasion in lung tissue [8,14,17,20]. All of our patients had such confirmatory findings on autopsy, which anatomically seemed to correspond to premortem radiologic findings (Table 1). Three cases met criteria for proven CP with evidence of tissue invasion and two cases met criteria for probable CP [52]. In Lewis’s seminal report of CP, he found diffusely-diseased lung parenchyma with several types of lesions, including pseudo-tubercles, fibrosis, granulation tissue, and chronic inflammation [13]. Dubois et al. classified their findings by the potential route of parenchymal infection [19]. Endobronchial CP was characterized (1) by an asymmetrical distribution of macroscopic lesions primarily in the lower lobes, (2) by frequent upper airway mucosal candidiasis, (3) by proliferation of Candida organisms within bronchial lumens occasionally associated with aspirated foreign debris, by parenchymal involvement only in adjacent alveoli, and (4) by the absence of Candida in extrapulmonary sites [19]. Hematogenous CP was characterized (1) by a symmetrical distribution of macroscopic nodules randomly throughout both lungs occasionally with tiny subpleural nodules, (2) by usually absent upper airway candidiasis, (3) by the absence of proliferation of Candida organisms within bronchial lumens, and (4) by the universal presence of Candida in extrapulmonary sites (mainly liver, spleen, and kidneys) [19]. In our cases series, four patients had endobronchial CP and one had hematogenous CP. Inflammatory responses are variably present and largely depend upon the nature and degree of the patient's immunosuppression [14]. When inflammation is seen, polymorphonuclear leukocytic infiltrates and histiocytic or granulomatous lesions predominate [14,20,31]. Microabscesses, granulomas, bronchopneumonia, and intracavitary exudates have also been described [14,18,20]. A recent case report described a patient with CP whose surgical lung biopsies showed extensive suppurative granulomatous inflammation involving >50% of the lung parenchyma [34]. Only one of our patients had granulomas (Table 1). Notably, extensive pulmonary hemorrhage has been described but it is unknown if this is a cause or a consequence of CP [14,31]. Two of our patients had evidence of diffuse alveolar hemorrhage and/or diffuse alveolar damage on autopsy (Table 1).

4. Part II - Enterococcal Pneumonia

4.1 Illustrative Case II

A 49-year-old man (Table 4, patient #6) with non-Hodgkin lymphoma underwent myeloablative matched related donor peripheral blood allogeneic HCT. His posttransplant course was complicated by hepatic GVHD in addition to Nocardia nova pneumonia and bacteremia. Eight months posttransplant while on trimethoprim-sulfamethoxazole and imipenem (which does not have reliable anti-enterococcal activity) for N. nova, he developed chills and a mild non-productive cough. CT of the chest demonstrated new diffuse ground glass opacities with extensive nodular opacities, and improvement in the left lower lobe cavitating nodule corresponding to the known N. nova infection (Figure 3). The following day, he developed epistaxis and acute hypoxemic respiratory failure that was refractory to diuresis. Three days after symptom onset, he suffered a cardiopulmonary arrest and died. Premortem respiratory culture was not obtained. Postmortem histopathological findings demonstrated acute pneumonia with Gram-positive cocci throughout both lungs and cultures grew 4+ enterococci (Figure 4). Additional findings included diffuse alveolar damage bilaterally and a left lower lobe necrotic abscess (correlating with N. nova).

Additional cases of autopsy-proven EP are presented in Table 4.

Table 4 Case series of hematopoietic cell transplantation recipients with autopsy-proven enterococcal pneumonia.

Figure 3 Computed tomography of the chest of Illustrative Case II demonstrating extensive nodular opacities, diffuse ground glass opacities, and a left lower lobe nodule with central cavitation correlating with autopsy-proven enterococcal pneumonia, diffuse alveolar damage, and Nocardia nova abscess, respectively.

Figure 4 Gram stain (20X magnification) of a bacterial abscess within postmortem lung tissue of Illustrative Case II. Sections demonstrate a dense cluster of Gram-positive cocci with accompanying acute neutrophilic inflammation, hemosiderin-laden macrophages, fibrin deposits, and surrounding necrotic tissue. Inset (60X magnification of area outlined in red) highlights Gram-positive cocci in pairs and short chains within the inflammatory milieu. Postmortem cultures isolated 4+ enterococci.

4.2 Epidemiology

EP is even rarer than CP. EP with lung abscess formation was reported as early as 1974 [53]. Additional cases reported within the next decade included an infant's perinatal death due to necrotizing EP after aspiration of infected amniotic fluid [54]. A prospective, observational study of 110 patients with serious enterococcal infections found that 4% had pleuropulmonary involvement [55]. Pleural involvement was even rarer in another study where 1.5% of invasive enterococcal infections were isolated from pleura [56]. Similarly, only a small number of all empyemas are due to enterococci (2-2.5%) [55,57]. A National Healthcare Safety Network report of healthcare-associated infections noted that enterococci accounted for only 0.8% of ventilator-associated pneumonia cases [57]. Apart from this, the medical literature regarding EP is largely limited to a handful of case reports [58,59,60].

4.3 Risk Factors

Potential risk factors for EP are summarily listed in Table 2. Tobacco and alcohol abuse may increase risk for community-acquired EP, presumably as a consequence of impaired local and systemic immunity [10,53,59]. Advanced age, hypertension, and vascular disease have been suggested to increase susceptibility to nosocomial EP but this may represent an observed association rather than causation [53,59]. On a related note, healthcare workers have been mentioned as potential vectors of nosocomial enterococcal transmission [11]. Similar to CP, factors that lead to aspiration, such as stroke and dysphagia, may increase risk [59,61,62]. Since enterococci are commensals of the gastrointestinal flora, it is not surprising that intra-abdominal pathology (e.g., gastrointestinal malignancy, spontaneous bacterial empyema in patients with cirrhosis) has been frequently linked [63,64,65,66,67,68]. Based on pathophysiology, GVHD of the gastrointestinal tract likely increases risk but a correlation has not yet been shown; only one of our patients had intra-abdominal pathology due to gastrointestinal GVHD (Table 4). Direct manipulation of the gastrointestinal tract, either via enteral hyperalimentation or selective gut decontamination with topical antimicrobials to prevent ventilator-associated pneumonia, have also been cited as risk factors [61,69]. In these cases, it is plausible that antimicrobial usage led to dysbiosis favoring increased enterococcal colonization and that the development of EP was ultimately due to aspiration of oropharyngeal or gastrointestinal contents [61,69]. EP may be a postsurgical complication, having been reported in a postpneumonectomy patient with an esophagopleural fistula and in another patient after pneumonectomy and cholecystectomy [68,70]. EP may scarcely be a complication of enterococcal endocarditis [60,71]. It would seem intuitive that immunocompromised patients would be at increased risk for EP.  Previous cases that support this link describe EP in the setting of human immunodeficiency virus infection and acquired immunodeficiency syndrome, acquired asplenia after splenectomy, acute leukemia, and kidney transplantation [58,59,64,72]. A striking case of EP in a patient with chronic lymphocytic leukemia and Strongyloides stercoralis hyperinfection syndrome has also been published [73]. However, these may not be significant risk factors by themselves due to the paucity of published literature suggesting a possible association and because these patients had other risk factors (e.g., older age, tobacco abuse, alcohol abuse, intra-abdominal pathology) [58,59,64]. All of our patients had intra-abdominal pathology, including gastrointestinal and/or hepatic GVHD (Table 4). In an effort to elicit additional commonalities that may predispose to EP, all of our patients were critically ill, had a central venous catheter in-place, and were receiving corticosteroids and broad-spectrum antimicrobials (Table 4).

4.4 Clinical Manifestations

Akin to CP, the clinical manifestations are EP are nonspecific. Patients may have acute or insidious presentations. Reported symptoms and signs include fever, drenching night sweats, malaise, altered mentation (which may be a cause or result of EP), anorexia, weight loss, dyspnea, chest pain (which may be pleuritic), cough (productive or non-productive), palpitations, hypoxemia, tachypnea, decreased respiratory mobility, dullness to percussion, decreased breath sounds, rales, rhonchi, wheezing, vocal fremitus, and other findings related to intra-abdominal pathology (e.g., stigmata of cirrhosis, peritonitis, etc.) [10,53,54,58,59,60,61,62,63,68,73,74,75]. The illness may progress to respiratory failure requiring mechanical ventilation, multiple organ dysfunction syndrome, septic shock, and death [54,58,60,61,62,73]. These manifestations are similar to those seen in our cohort (Table 4). Interestingly, despite the purported low virulence of enterococci, the majority of EP cases have been associated with empyema and/or lung abscess formation [53,55,58,59,60,62,63,65,66,67,68,71,74,76,77]. Many of the aforementioned physical exam findings are most likely a consequence of empyema development. None of the patients in our cohort had evidence of enterococcal empyema or lung abscess (Table 4).

4.5 Microbiologic Findings

Unsurprisingly, recovery of enterococci from clinical samples is relatively easy. Diagnostic assays (e.g., PCR-based methods) beyond routine culture techniques are rarely required. Cultures of respiratory tract or pleural fluid specimens may yield monomicrobial or polymicrobial (usually with other enteric organisms) growth. The bulk of published cases of EP reported large quantities of monomicrobial growth from at least one clinical specimen. Interestingly, none of the patients in our case series had enterococci recovered from premortem respiratory cultures while it was grown from postmortem samples in all cases (Table 4). We hypothesize that this may be explained by sampling bias since autopsy has the inherent advantage of allowing examination of the entire lung and targeted biopsy of clearly diseased tissue. Alse of note, four out of five patients were on an agent with some activity against Enterococcus at the time of death which could have contributed to the lack of growth on premortem cultures. In our cohort, Enterococcus was not further speciated. From review of the literature, the relative frequency of enterococcal species causing EP are similar to other invasive enterococcal infections. The vast majority of cases are due to Enterococcus faecalis [10,62,65,67,68,69,70,71,72,74,76]. A not insignificant number were caused by E. faecium [58,59,60,63,64,65,73]. A single case of EP and empyema involving E. cecorum has been reported [66]. More recent articles have illustrated the growing concern of antimicrobial resistance. Cases of EP due to vancomycin-resistant isolates have been reported [58,59,73]. In such situations, the oxazolidinones (such as linezolid and tedizolid) would generally be recommended since daptomycin is ineffective for treatment of pulmonary infections. Unfortunately, oxazolidinone resistance, albeit quite rare, is an alarming development [68,72]. 

4.6 Radiologic Findings

Radiologic findings of EP are also nonspecific (Table 3). However, EP tends to be more frequently associated with empyema and/or lung abscess formation [10,58,59,64]. Other reported radiographic findings include airspace disease or consolidation [58,59,72]. CT with contrast is the most sensitive and specific radiologic modality to diagnose and distinguish between transudative pleural effusion, empyema, and lung abscess [78,79,80]. The distinction between these three clinical entities is relevant because it carries important therapeutic implications (e.g., pleural drainage in addition to antimicrobial therapy for empyema) [79,80,81,82]. Lung abscess appears as a round cavity containing a gas-fluid level, and occurs most often in the right lower lobe since it is usually caused by aspiration [80]. It is important to mention that a peripheral lung abscess may mimic empyema, but the split pleura sign seen with empyema is considered the most reliable CT sign to distinguish between the two [78,79]. All of our patients had consolidation and/or ground glass opacities but most had another possible explanation (e.g., BCNU-associated pulmonary toxicity) for at least some of these findings (Table 4).

4.7 Histopathological Findings

The only method of attaining a definitive diagnosis of EP is by histopathology [59]. All of our patients had evidence of EP on postmortem examination, more specifically abundant growth of enterococci in alveoli with the addition of adjacent inflammation or tissue destruction in three out of five cases (Table 4). Absence of inflammation does not rule out infection in severely immunocompromised hosts who may be unable to mount an immune response. Local colonization by enterococci with or without inflammation from a secondary process is an alternative explanation for these findings and difficult to exclude. There is no absolute growth threshold to distinguish disease from colonization since it is possible for low organism burdens to cause disease in immunocompromised hosts. Lack of growth on premortem cultures makes diffuse airway colonization unlikely. Because enterococci are easily recovered from clinical samples, more invasive procedures (e.g., transthoracic or transbronchial biopsies of pulmonary tissue) are rarely performed. Accordingly, there is a dearth of published data regarding histopathological findings of EP. As recounted by Grupper et al., Abkarovich and Akimchenkov described an animal model of EP in 1971 that characterized the pleura as confluent and hemorrhagic, and the lung parenchyma as serohemorrhagic with fragmentation, thickening, and defibrillation of elastic fibers of the tissue and blood vessels [10]. The histopathological findings in a case presented by Vanschooneveld et al. included inflammation, thrombosis, microabscesses, and tissue necrosis [59]. In our cohort, enterococci were universally recovered from culture whereas inflammatory changes were variably present (Table 4).

5. Conclusions

Candida and enterococci are rare but bona fide pulmonary pathogens. When isolated from respiratory tract or pleural fluid specimens, these organisms should not be universally disregarded as airway colonizers since they may be pathogenic in susceptible patients. While clinical manifestations are nonspecific, large quantities of monomicrobial growth (especially if recovered from more than one clinical specimen) and compatible radiologic patterns may be suggestive of invasive disease. However, because lower organism burdens may cause disease in immunocompromised patients, a microbiologic threshold should not be used in isolation to distinguish between colonization and disease. We propose that clinicians should consider host risk factors, clinical presentation, and radiographic findings when interpreting microbiologic data. Targeted treatment should be considered in patients who do not demonstrate clinical improvement while receiving antimicrobial therapy lacking Candida and/or enterococcal activity. The higher yield of cultures from postmortem as compared to premortem sampling, especially in the setting of EP, highlights the importance of direct sampling from lung tissue for diagnosis. Clinicians should never say that Candida and enterococci never cause pneumonia.

Acknowledgments

We extend our deepest gratitude to Donald Regula MD for his expertise in reviewing and critically revising the final version of the manuscript.

Author Contributions

All authors conceived and designed the work; AM and LSA collected the data; AM and JKN analyzed the data; AM drafted the manuscript; LSA and JKN critically reviewed and revised the manuscript; all authors approved the final version of the manuscript to be published.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Competing Interests

The authors have declared that no competing interests exist.