Although this patient had none of these common risk factors associated with Candida empyema, she did have other factors predisposing her to fungal colonization and subsequent infection.
Candida glabrata empyema
HISTORY OF PRESENT ILLNESS
A 65-year-old woman presented complaining of 2 weeks of sudden-onset dyspnea, cough, and left-sided pleuritic chest pain. Review of systems was notable for increased thirst, polydipsia, polyuria, and nocturia. She denied preceding gastrointestinal or upper respiratory symptoms, although she complained of some nausea and heartburn. She described “white fluffy stuff” on her tongue, accompanied by xerostomia, and diminished taste over the same time period. She denied weight loss or sick contacts. During the 10 days prior to admission, her shortness of breath and chest pain worsened, and she presented to the hospital.
PAST MEDICAL HISTORY
She had a 16-year history of steroid-dependent sarcoidosis and no known history of diabetes, prior surgeries, or malignancies.
She was on a course of 20 mg of prednisone daily for 3 consecutive months for pulmonary sarcoidosis, which was tapered to 10 mg daily 3 weeks prior to admission.
The patient was born in the US. She had no history of intravenous drug use, was a current smoker with a 16 pack-year history, and drank 1 to 6 units of alcohol weekly. She was up-to-date with pneumococcal vaccination, had not traveled in 6 months, and had 1 house cat with no other animal exposures.
On admission she was afebrile (98.3 °F), tachycardic (110 beats per minute), hypotensive (92/57 mm Hg), and tachypneic (28 breaths per minute). Her oxygen saturation was 95% on room air. She was fully oriented but drowsy and complaining of pain in her left chest. She had oral thrush. She had no jugular venous distention. Her heart sounds were regular rate and rhythm without murmurs, rubs, or gallops. The lung examination revealed diminished breath sounds bilaterally, with marked diminishment to the axilla on the left side. She had left upper flank tenderness to palpation. Her abdomen was soft without tenderness or distention. She had no axillary, supraclavicular, or cervical lymphadenopathy. She had no rashes, no joint tenderness or swelling, and no lower limb edema.
Laboratory tests revealed a leukocytosis with neutrophilic predominance: 23.8 k/mm3 wbc (normal, 4.5-11 k/mm3), 93% neutrophils, and 0.9% bands. Her blood glucose level was markedly elevated at 973 mg/dL (normal, 70-140 mg/dL), and her electrolytes and blood gas were consistent with diabetic ketoacidosis with respiratory compensation. Her creatinine level was 1.89 mg/dL (baseline, 1.1 mg/dL), and her liver function tests were within normal limits. An electrocardiogram revealed no ischemic changes, and serially repeated troponins were negative. The HIV test, rapid respiratory viral panel, COVID-19 polymerase chain reaction test, methicillin-resistant Staphylococcus aureus nasal swab, and Legionella urine antigen were negative. Urinalysis was positive for ketones and negative for nitrites and leukocyte esterase and only 1 to 2 white blood cells.
A chest x-ray and computed tomography of the chest with contrast (Figures 1, 2, and 3) showed a large, loculated left-sided pleural effusion, no evidence of esophageal rupture, stable hilar adenopathy, no mediastinitis, and no mass.
She was admitted to the hospital with diabetic ketoacidosis secondary to sepsis, likely from a complicated pneumonia. She received intravenous fluids, electrolyte repletion, and an insulin drip, and she was started on azithromycin and cefepime.
Interventional radiology performed fluoroscopy-guided diagnostic thoracentesis for fluid analysis. The pleural fluid studies revealed an exudative effusion. Laboratory values were as follows: lactate dehydrogenase, 1352 U/L (normal, < 140U/L); total protein, 4.6 g/dL (normal, < 1-2 g/dL); glucose, 165mg/dL; white blood cells, 1381/μL (normal, < 1000) (64% neutrophils); red blood cells, 502 x 106/μL (normal, < 50). The pleural fluid Gram stain was negative, and both routine and fungal cultures grew multiple colonies of Candida glabrata at 72 hours. Two chest tubes were placed for source control; repeat fluid culture was also positive for C glabrata, which revealed susceptibility to micafungin (minimum inhibitory concentration [MIC], 0.015 μg/ml), dose-dependent susceptibility to fluconazole (MIC, 16), and susceptibility to voriconazole (MIC, 0.5 μg/ml). Anaerobic cultures were negative. Blood cultures were negative. Fluid cytology did not yield malignant cells. Ophthalmology found no evidence of ocular involvement on the dilated exam. To further search for an etiology of the fungal empyema, we performed a barium study and esophagogastroduodenoscopy, which showed no evidence of esophageal perforation or other pathology.
TREATMENT AND FOLLOW-UP
Two cycles of chest tube drainage with fibrinolysis achieved moderate improvement. She was treated with caspofungin for 28 days. Her chest tubes were removed 2 weeks prior to completion of antifungal therapy. At 3-month follow-up, she remained asymptomatic without recurrence of pleural effusion.
Definitive diagnosis of fungal empyema should be made based on isolation of fungal species from exudative pleural fluid studies obtained via thoracentesis, clinical presentation consistent with pleural infection (fever, leukocytosis, dyspnea), and isolation of the same organism from pleural fluid on another occasion or from another specimen (including blood culture).1-3 Confirmed Candida empyema thoracis, especially in patients without evidence of candidemia, most commonly occurs in the setting of esophageal disease, especially rupture or recent chest surgery.4-7
Although this patient had none of these common risk factors associated with Candida empyema, she did have other factors predisposing her to fungal colonization and subsequent infection. These risks included a prolonged course of prednisone for sarcoidosis along with uncontrolled hyperglycemia and a new diagnosis of diabetes.8 After excluding the more common etiologies of fungal empyema, we considered other possible mechanisms of spread, such as a bronchopleural fistula, which was not observed on imaging, or hematogenous spread following a transient episode of candidemia that eluded blood culture detection. We further speculated that her prolonged steroid use and uncontrolled hyperglycemia could have predisposed her to Candida overgrowth and facilitated translocation from her respiratory or gastrointestinal mucosa eventually into pleural space. Although Candida species are often thought to be colonizers in sputum and possibly contaminants in pleural fluid, this was not the case in our patient, as multiple colonies of the same species grew on cultures obtained from different samples taken at different times confirming infection.
Drainage and prompt antimicrobial therapy are essential for the treatment of fungal empyema, and choice of empiric antifungal therapy must often be made based on the likelihood of less common pathogens prior to susceptibility results. Candida albicans is the most common causal organism for fungal empyema1; however, other species such as C glabrata as well as fungal pathogens such as Aspergillus and Cryptococcus must be considered.1,9,10 C glabrata becomes more likely in patients with diabetes and those with an abdominal source of infection, particularly when candidemic.11 Broad spectrum antifungals, such as echinocandins, voriconazole, and amphotericin, should be initiated prior to final culture results in these clinical scenarios.
Targeted therapy based on fungal susceptibilities is ultimately required, especially in C glabrata isolates due to high rates of dose-dependent resistance to fluconazole mediated by efflux pumps. Although some C glabrata isolates exhibit a dose-dependent susceptibility to voriconazole, the in vitro susceptibility patterns do not correlate well with in vivo activity and require trough levels that are not clinically feasible.12 Therefore, it is unlikely that voriconazole would be a useful treatment for invasive infection with C glabrata that is resistant to fluconazole. Echinocandins, which target the β-1,3-D-glucan component of fungal cell walls, are appropriate to treat invasive C glabrata infections with an MIC below 0.020 and penetrate the pleural space in sufficient concentration to treat a Candida infection with an MIC of 0.015.13,14
Community-acquired Candida empyema thoracis is rare. Diabetes mellitus and prolonged corticosteroid use both predispose to fungal overgrowth. In patients with these risk factors, transient fungemia or translocation may seed the pleural space. A diagnosis of Candida empyema thoracis should be considered despite a lack of commonly associated risk factors (such as recent surgical intervention, esophageal malignancy, or perforation) when isolated on multiple cultures obtained on different occasions. This infection is associated with a high mortality and must be managed with prompt drainage and targeted antifungal therapy.