Background Corticosteroids and immunosuppressant are now widely used as a treatment for collagen diseases, and complications of pneumocystis pneumonia (PCP) have become a problem.
Objectives We study prognostic factors associated with death from PCP complicated by collagen diseases.
Methods We conducted a retrospective analysis on the basis of the medical records of 38 patients with PCP complicated by collagen diseases. For background factors, age, gender, baseline diseases, medications at the time of hospital admission (dose of steroids, immunosuppressant), the presence of complications (respiratory diseases, diabetes, chronic kidney disease), smoking history, trimethoprim-sulfamethoxazole (TMP-SMX) prophylaxis were studied. For laboratory data, CBC (including lymphocyte count), serum LDH, albumin, creatinine, CRP, KL-6, SP-D, IgG level, IgA, IgM, (1,3)-β-D-glucan, arterial blood gas, and sputum PCP-DNA (tested by PCR) at the time of hospital admission were investigated. The eGFR was calculated using creatinine value, age, and gender. We also studied treatment methods after hospitalization (TMP-SMX use, change from TMP-SMX to pentamidine, pulsed-steroid therapy, intratracheal intubation, managing in ICU) and hospitalization period.
Results In laboratory data, the group of non-survivors had a higher creatinine (0.88±0.38 mg/dl vs. 1.16±0.38, p=0.045), a lower albumin (3.13±0.50 g/dl vs. 2.71±0.38, p=0.031), a higher CRP (7.8±7.0 mg/dl vs. 18.9±8.8, p=0.014), a lower IgA (274±108 mg/dl vs. 174±54, p=0.013), a higher serum SP-D (173±106 vs. 285±285, p=0.047), and a lower PaO2/FiO2 (P/F) ratio (268±99 mmHg vs. 147±98, p=0.025). For sputum PCP (by PCR), all patients were positive in the group of non-survivors, but only 18 out of 31 survivors (58%) were positive. All patients in the group of non-survivors underwent pulsed-steroid therapy, but it was assumed that the therapy was performed in order to treat severe respiratory failure because the patients who resulted in death had a low P/F ratio and severely poor respiratory status. On the other hand, as background factors, there was no significant difference.
Conclusions We newly identified prognostic factors associated with death from PCP, which are a low serum albumin and cholinesterase at the time of PCP diagnosis, decreased pulmonary diffusing capacity, and significantly high rate of intratracheal intubation and managing in ICU; in this study, male, older age, a high serum creatinine, a high CRP, a low IgA, a high SP-D, a low P/F ratio, and sputum PCP positive (by PCR) were newly identified as the poor prognostic factors.
Disclosure of Interest None declared