Molecular epidemiology, genotype distribution and drug resistance of Pneumocystis jirovecii in a Turkish tertiary care center

Background Pneumocystis jirovecii pneumonia (PJP) is a life-threatening fungal infection that disproportionately affects non-HIV immunocompromised patients, for whom epidemiological and outcome data remain scarce, particularly from the Eastern Mediterranean region. This study aimed to evaluate diagnostic methods for PJP, describe the molecular epidemiology of P. jirovecii , investigate trimethoprim-sulfamethoxazole resistance-associated mutations, and explore clinical correlates of disease severity and mortality in a Turkish tertiary care center. Methods A total of 102 specimens from 73 patients were evaluated; 21 patients positive by real-time quantitative PCR (Rt-qPCR) were involved in the study prospectively. Diagnostic performance was assessed using a composite reference standard (≥ 2 methods positive), with inter-method agreement evaluated using Cohen’s kappa. Genotypes were determined by MLST, and DHPS resistance mutations were analysed by RFLP. Clinical data and 30-day mortality were retrieved from hospital records. Results Against the composite reference, Giemsa staining showed 66.7% sensitivity, and CW staining 40.0% sensitivity, and Rt-qPCR was positive in 6 composite-negative patients with negative microscopy, highlighting the diagnostic challenge of distinguishing active infection from colonization. Inter-method agreement was moderate-to-fair (κ = 0.276–0.533). Exploratory ROC analysis in non-HIV patients suggested a fungal load threshold of 41.6 copies/mL for predicting IFM positivity (AUC 0.750), however, specimen type heterogeneity and the inability of fungal load alone to distinguish infection from colonization underline the need for further validation. The predominant genotypes were mtLSU genotype 7, CYB1, and SOD1; two patients with MLG X and two with MLG Y had overlapping hospital stays, raising the possibility of an epidemiological link. No DHPS mutations were identified. The 30-day mortality rate was 47.6%, LDH was significantly higher in non-survivors, while platelet count and albumin were significantly lower ( p = 0.02, p = 0.02, and p = 0.02, respectively). Conclusions Rt-qPCR was the most sensitive diagnostic method, but our findings suggest fungal load should not be used alone to distinguish active infection from colonization. Quantitative PCR results should always be interpreted together with clinical and radiological findings. Genotype distribution appeared to differ from those reported in European studies, and mortality was high, with LDH, platelet count, and albumin identified as key prognostic markers. Larger multicenter studies are needed to confirm these findings.