Abstract
Background
Pneumocystis jirovecii is a major cause of morbidity and mortality in HIV. It also poses a significant risk to non-HIV immunocompromised patients, especially those with cancer or recent chemotherapy. We evaluated risk factors, diagnostic markers—including lactate dehydrogenase (LDH)—and mortality predictors in PCR-confirmed pneumocystis pneumonia (PJP).
Methods
We retrospectively analyzed patients with hypoxemia, pulmonary involvement and/or fever who underwent PJP PCR testing at Koç University Hospital between January 2020 and November 2023. Cases were defined as patients with a positive PCR result who fulfilled the EORTC/MSGERC criteria for PJP, whereas the non-PJP group included PCR-negative patients and PCR-positive patients without clinically compatible disease. Clinical, laboratory, and radiologic data—including LDH, fungal load, co-infections, and 4-week mortality—were collected.
Results
Of 235 patients, 59 had PCR-confirmed PJP. PJP patients were older (median 70 vs. 64.5, p < 0.001), more often had solid organ malignancies (67.8% vs. 41.5%, p < 0.001), recent chemotherapy (66.1% vs. 41.5%, p = 0.001), and higher fungal loads (median 10,892 vs. 0, p < 0.001). Bilateral lung lesions (81% vs. 66.5%, p = 0.037) and ground-glass opacities (82.8% vs. 67.7%, p = 0.041) were more frequent. Serum LDH increased by 81.6% in PJP vs. 20.7% in controls (p = 0.001), whereas no significant difference was observed among patients with elevated baseline LDH (29.7% vs. 16.3%, p = 0.397). Multivariate analysis identified older age (OR 1.042, p = 0.017), solid organ malignancy (OR 2.304, p = 0.048), and CMV co-infection (OR 3.786, p = 0.006) as independent factors associated with PJP, whereas ICU admission at diagnosis was inversely associated with PJP (OR 0.239, p = 0.005). The 4-week mortality was 37%, with bacterial co-infection as the strongest predictor (OR 5.854, p = 0.009).
Conclusion
Older age, solid organ malignancy, and CMV co-infection increased PJP risk, while bacterial co-infection predicted mortality. LDH changes supported diagnosis, but not when baseline levels were already elevated, in this predominantly oncology cohort.
Keywords: PJP, Oncology, Lactate dehydrogenase, Mortality, Risk factors, PCR diagnosis
Introduction
Pneumocystis jirovecii pneumonia (PJP) is a significant cause of morbidity and mortality not only in HIV-infected patients but also in non-HIV immunocompromised hosts [1, 2]. Patients with cancer are particularly vulnerable, especially in centers with active transplant programs [1, 2]. In these patients, especially those who develop hypoxia, PJP often enters the differential diagnosis, and delayed test results frequently lead to empirical treatment [3]. Diagnosis is challenging due to overlapping clinical and imaging findings and the limitations of conventional microscopy [4]. Real-time polymerase chain reaction (PCR) has improved detection of PJP, yet distinguishing true infection from colonization remains challenging [5].
Lactate dehydrogenase (LDH), a commonly requested test, can support clinical suspicion [6, 7]. However, its interpretation is complicated in patients with already elevated baseline levels, a situation frequently encountered in oncology [6, 7]. Confounders such as tumor burden, hemolysis, organ failure, or concurrent infections reflecting nonspecific cellular injury and inflammation limit its specificity [7–9]. Clinical practice relies mainly on single LDH measurements alongside imaging, and most studies have evaluated LDH as a prognostic rather than diagnostic marker for PJP [7, 10, 11]. Although levels above 220–250 IU/L (and up to 500 IU/L) are associated with diagnosis, severity, and mortality, the significance of incremental rises remains unclear [12–15]. Nevertheless, it is logical to expect that PJP can lead to measurable increases even when baseline levels are already elevated [16–18].
In our study, we included non-HIV immunocompromised patients mainly presenting with hypoxia and/or lung involvement who underwent PCR testing for PJP. Within this cohort, we aimed to analyse clinical and laboratory findings—including LDH in those with elevated baseline levels—and factors associated with mortality in PCR-confirmed cases.
Material and methods
This study was conducted at Koç University Hospital (KUH) which is a tertiary private university hospital with 426 beds, including 55 intensive care unit (ICU) beds. Solid organ and bone marrow transplantations are currently being performed at KUH. The departments of Internal Medicine (including oncology and hematology) and Pulmonology frequently encounter patient groups at high risk for PJP.
All patients for whom a PJP PCR test was requested between 1 Jan 2020 and 27 Nov 2023 were recorded. PJP PCR testing was performed in patients presenting with at least one of the following: hypoxemia, unilateral or bilateral lung involvement, or symptoms such as fever and cough. According to the EORTC/MSGERC (European Organization for Research and Treatment of Cancer / Mycoses Study Group Education and Research Consortium) criteria, probable PJP requires host factors, compatible clinical and radiological findings, and a positive PCR result from a respiratory specimen [19]; accordingly, all PJP cases in our study were classified as probable PJP.
Among patients with positive PJP PCR results, those without compatible clinical and radiological findings for PJP were considered colonized rather than infected. PCR values below 2000 copies/mL were not used as an absolute cutoff but were considered supportive evidence of possible colonization. However, in patients with values above 2000 copies/mL, no interpretation of colonization was made. Alternative explanations for their presentation included bacterial pneumonia, COVID-19, lung cancer with unilateral lung consolidation, or fever without significant ground-glass opacities. The non-PJP group therefore included patients with negative PJP PCR results as well as PCR-positive patients without compatible clinical findings. Ultimately, all patients were classified as either PJP or non-PJP, with PCR values reported for both groups.
Patients’ medical records were reviewed to collect data on age, sex, underlying conditions (hematologic malignancies, solid organ malignancies, bone marrow transplantation, chronic lung disease, etc.), laboratory parameters, presence of simultaneous cytomegalovirus (CMV) and bacterial co-infections, radiological findings, clinical signs and symptoms, and medications used—particularly corticosteroids and temozolomide. For patients with a positive PJP PCR result, quantitative values were recorded. Four-week mortality data were also collected. CMV infection was defined as a plasma CMV PCR viral load > 1000 IU/mL regardless of whether anti-CMV therapy was administered. Bacterial co-infection was defined as infection diagnosed and treated by the attending clinician at the time the PJP test was requested, based on clinical, microbiological, or laboratory findings. LDH levels were recorded for each group at three time points: before PJP diagnosis (within the 4-week period prior to PCR sampling), at the time of diagnosis (PCR sampling date), and after diagnosis (within 10 days following the PCR sampling). Increases in LDH rates were analyzed as percentages. In addition, patients who already had elevated LDH levels before the PJP diagnosis were identified and reanalyzed separately.
Microbiological diagnosis
Pneumocystis PCR was performed on respiratory specimens, including induced sputum, endotracheal aspirates, and bronchoalveolar lavage samples, using the RealStar Pneumocystis jirovecii PCR assay according to the manufacturer’s instructions. Quantitative results were reported as copies/mL and interpreted in conjunction with clinical and radiological findings, as values may vary by assay platform and specimen type. Low PCR values (e.g., < 2000 copies/mL) were not used as an absolute cutoff but considered supportive evidence of possible colonization when PJP was clinically unlikely.
Reactions (30 µL) included 20 µL reaction mix and 10 µL template DNA. Thermocycling conditions were 95 °C for 2 min, then 45 cycles of 95 °C for 15 s, 58 °C for 45 s, and 72 °C for 15 s on a CFX96 Real-Time PCR system (Bio-Rad, Hercules, CA, USA). Pneumocystis jirovecii probes were FAM-labeled; the internal control used JOE fluorophore. Each run included positive and no-template controls. Amplifications with well-defined exponential curves were considered positive; tests lacking internal control amplification were invalid.
Statistical analysis
The data obtained in this study were analyzed using the IBM SPSS Statistics software (version 26). Descriptive statistics were presented as frequencies and percentages (%) for categorical variables. The normality of data distribution was assessed using the Kolmogorov-Smirnov test, histogram evaluation, and the Shapiro-Wilk test. Continuous variables that did not meet the assumption of normality were expressed as median (interquartile range, IQR). For comparing categorical variables, Pearson’s chi-square test or Fisher’s exact test were applied. For comparisons of continuous variables, the Mann-Whitney U test was used. We conducted multivariable logistic regression analyses to evaluate predictors of PJP diagnosis and potential risk factors for all-cause 4-week mortality. Odds ratios (ORs) with 95% confidence intervals (CIs) and corresponding p-values were reported.
Variables with p < 0.05 in univariate analyses and those considered clinically relevant based on previous literature were evaluated for inclusion in multivariable logistic regression models. To reduce multicollinearity and overfitting, variables with substantial clinical overlap or strong correlations were carefully assessed before final model construction. Variables directly related to PJP diagnosis or sampling characteristics, such as fungal load and respiratory specimen type, were not included. Among radiological variables with overlapping significance, only ground-glass opacity was retained in the final model. Given the limited number of PJP events, the number of variables included was restricted to preserve model stability.
Results
A total of 235 patients underwent PJP PCR testing on respiratory specimens (Fig. 1). Of these, 64 patients were PCR positive. Among them, 59 were diagnosed with probable PJP according to EORTC/MSGERC criteria, with PJP PCR values ranging from 11 to 5,314,492 copies/mL. The remaining five patients (7.8%) were considered colonized rather than having active PJP because they lacked compatible clinical and radiological findings. Their PCR values ranged from 10 to 229 copies/mL. These cases included two oncology patients with fever but no significant ground-glass opacities who recovered without trimethoprim-sulfamethoxazole (TMP-SMX), one oncology patient with bacterial pneumonia who improved with antibiotic treatment, one hematology patient with COVID-19 who recovered with COVID-19 treatment alone, and one oncology patient with lung cancer and unilateral consolidation who was considered colonized despite receiving a short course of TMP-SMX. Together with the 171 PCR-negative patients, these five colonized patients were included in the control group (n = 176).
Fig. 1.
Flow diagram showing how patients were classified based on PCR results, clinician review, and inclusion in either the PJP or non-PJP groups. *PJP PCR: Pneumocystis jirovecii polymerase chain reaction. ϕEORTC/MSGERC: European Organization for Research and Treatment of Cancer / Mycoses Study Group Education and Research Consortium criteria
Since our study population consisted of patients with hypoxemia and/or mostly ground-glass opacities, the findings were as follows: hypoxemia was observed in 44.3% of the non-PJP group and 52.5% of PJP patients, with no significant difference (p = 0.273). In contrast, ground-glass opacities on computed tomography (CT) were more frequent in the PJP group (82.8% vs. 67.7%, p = 0.041).
Patients with PJP were older (median age 70 vs. 64.5 years, p < 0.001) and more frequently had solid organ malignancies (67.8% vs. 41.5%, p < 0.001) and recent chemotherapy (66.1% vs. 41.5%, p = 0.001), whereas hematologic malignancies were more common in the non-PJP group (26.7% vs. 13.6%, p = 0.039). CMV co-infection was significantly higher in PJP patients (29.6% vs. 11%, p = 0.001) (Table 1). Among the 34 patients with CMV infection (viremia with or without anti-CMV therapy), viral loads ranged from 1,100 to 3,400,000 IU/mL.
Table 1.
Clinical and laboratory characteristics of patients with pneumocystis pneumonia (PJP) compared to those without PJP
| Control (n = 176) | Cases (n = 59) | Total (n = 235) | P | |
|---|---|---|---|---|
| Demographics | ||||
| Sex (Male) | 101 (57.4) | 42 (71.2) | 143 (60.9) | 0.06 |
| Age (median, IQRϕ) | 64.5 (53.5–71.5) | 70 (65.5–77) | 67 (56–72) | 0.000 |
| BMIϕ (median, IQR) | 25.4 (22.4–28.4) | 24.4 (21.9–26.4) | 24.9 (22.3–27.9) | 0.122 |
| Wards at diagnosis | 0.000 | |||
| Hematology ward | 31 (17.6) | 3 (5.1) | 34 (14.5) | |
| Oncology ward | 36 (20.5) | 28 (47.5) | 64 (27.2) | |
| ICUϕ | 23 (13.1) | 5 (8.5) | 28 (11.9) | |
| Others | 86 (48.9) | 23 (39) | 109 (46.4) | |
| Microbiology | ||||
| Sample source | 0.013 | |||
| BALϕ | 78 (44.3) | 16 (27.1) | 94 (40) | |
| ETAϕ | 32 (18.2) | 8 (13.6) | 40 (17) | |
| Sputum | 66 (37.5) | 35 (59.3) | 101 (43) | |
| Fungal load (copies) median IQR | 0 (0–0) | 10,892 (1123–71614) | 0 (0–31.5) | 0.000 |
| Time to result (days) median IQR | 9 (5–16) | 8 (3-14.5) | 9 (5–16) | 0.076 |
| Underlying conditions | ||||
| Smoking | 59 (43.1) | 26 (51) | 85 (45.2) | 0.332 |
| Diabetes mellitus | 40 (23.3) | 16 (28.6) | 56 (24.6) | 0.422 |
| COPDϕ | 13 (7.8) | 4 (7) | 17 (7.6) | 1 |
| Solid organ malignancy | 73 (41.5) | 40 (67.8) | 113 (48.1) | 0.000 |
| Hematologic malignancy | 47 (26.7) | 8 (13.6) | 55 (23.4) | 0.039 |
| History of chemotherapy | 73 (41.5) | 39 (66.1) | 112 (47.7) | 0.001 |
| Chronic kidney disease | 17 (9.7) | 4 (6.8) | 21 (8.9) | 0.606 |
| Solid organ transplant | 9 (5.1) | 2 (3.4) | 11 (4.7) | 0.735 |
| Bone marrow transplant | 15 (8.5) | 1 (1.7) | 16 (6.8) | 0.080 |
| Autoimmune disease | 23 (13.1) | 6 (10.2) | 29 (12.3) | 0.558 |
| Co-infections | ||||
| Bacterial co-infection | 80 (45.7) | 22 (37.3) | 102 (43.6) | 0.259 |
| Gram-negative pathogens | 47 (28.8) | 10 (21.3) | 57 (27.1) | 0.305 |
| Gram-positive pathogens | 5 (3.1) | 3 (6.4) | 8 (3.8) | 0.382 |
| Invasive Candida infection | 38 (21.7) | 14 (23.7) | 52 (22.2) | 0.748 |
| CMVϕ co-infection | 18 (11) | 16 (29.6) | 34 (15.7) | 0.001 |
| COVID-19 (last 3 months) | 40 (23) | 7 (11.9) | 47 (20.2) | 0.066 |
| Various | ||||
| CRPϕ median IQR | 105 (44–187) | 108 (59–153) | 105 (44.5–178.5) | 0.893 |
| Hypoxemia | 78 (44.3) | 31 (52.5) | 109 (46.4) | 0.273 |
| Bilateral lung lesion | 107 (66.5) | 47 (81) | 154 (70.3) | 0.037 |
| CTϕ ground-glass appearance | 109 (67.7) | 48 (82.8) | 157 (71.7) | 0.041 |
| Mechanical ventilation at diagnosis | 63 (35.8) | 9 (15.3) | 72 (30.6) | 0.003 |
| ICU admission at diagnosis | 63 (35.8) | 11 (18.6) | 74 (31.5) | 0.014 |
| Steroid (> 14 days) | 46 (26.3) | 17 (28.8) | 63 (26.9) | 0.705 |
| 4-week mortality | 72 (40.9) | 22 (37.3) | 94 (40) | 0.623 |
| Anti-PJP prophylaxis | 19 (10.8) | 4 (6.8) | 23 (9.8) | 0.456 |
All patients underwent PJP PCR testing on respiratory samples for various clinical indications
ϕBAL: Bronchoalveolar lavage; BMI: Body mass index; CMV: Cytomegalovirus; COPD: Chronic obstructive pulmonary disease; CRP: C-reactive protein; CT: Computed tomography; ETA: Endotracheal aspirate; ICU: Intensive care unit; IQR: Interquartile Range; PJP: Pneumocystis pneumonia
Fungal load was markedly elevated in PJP patients (median 10,892 copies vs. 0, p < 0.001). Imaging findings such as bilateral lung lesions (81% vs. 66.5%, p = 0.037) and ground-glass appearance on chest CT (82.8% vs. 67.7%, p = 0.041) were more frequent in PJP cases. Interestingly, mechanical ventilation (MV) and intensive care unit (ICU) admission at diagnosis were more common in the non-PJP group (MV 35.8% vs. 15.3%, p = 0.003; ICU 35.8% vs. 18.6%, p = 0.014).
Among PJP patients, 39% (23/59) had elevated LDH levels prior to diagnosis, compared with 45% (80/176) in the non-PJP group. The percentage change in LDH from pre- to post-diagnosis was 20.7% (IQR − 1.23–66.6) in the non-PJP group and 81.6% (IQR 23.3–170) in PJP patients (p = 0.001). However, in patients with already elevated baseline LDH, the percentage change was 16.3% in the non-PJP group and 29.7% in the PJP group, a difference that was not statistically significant (p = 0.397) (Table 2).
Table 2.
Serum lactate dehydrogenase increments in patients with pneumocystis pneumonia (PJP) compared to those without PJP
| Control (n = 176) | Cases (n = 59) | Total (n = 235) | P | |
|---|---|---|---|---|
| LDH (pre-diagnosis), median (IQR) | 242 (192-406) | 217 (180-315) | 236 (189-369) | 0.163 |
| LDH (at diagnosis) median (IQR) | 319 (225–431) | 354 (277–497) | 327 (233–446) | 0.122 |
| LDH (post-diagnosis) median (IQR) | 305 (217–523) | 413 (329–562) | 353 (235–532) | 0.026 |
| LDH change % (pre- vs. post-diagnosis) median (IQR) | 20.7 (-1.23–66.6) | 81.6 (23.3–170) | 31.8 (2.35–95) | 0.001 |
| LDH change % in patients with elevated baseline median (IQR) | 16.3 (-4.9–57.6) | 29.7 (1.7–81.5) | 19.1 (-2.3 -61.9) | 0.397 |
LDH: Lactate dehydrogenase
Other factors, including body mass index, diabetes, chronic obstructive pulmonary disease, chronic kidney disease, steroid use, smoking, bacterial co-infections, and 4-week mortality, did not differ significantly between groups. Sample source also differed: PJP patients were more often tested via sputum (59.3% vs. 37.5%, p = 0.013) and non-PJP patients via bronchoalveolar lavage (44.3% vs. 27.1%).
In multivariate analysis comparing PJP and non-PJP cases, increasing age (OR 1.042, 95% CI 1.007–1.077, p = 0.017), solid organ malignancy (OR 2.304, 95% CI 1.007–5.270, p = 0.048), and CMV co-infection (OR 3.786, 95% CI 1.456–9.844, p = 0.006) were independently associated with PJP (Table 3). Ground-glass opacities showed a non-significant trend toward association (OR 2.298, 95% CI 0.896–5.895, p = 0.084), whereas the percentage change in LDH (pre- vs. post-diagnosis) was not significantly associated with PJP (OR 1.002, 95% CI 1.000–1.005, p = 0.103). ICU admission at diagnosis was inversely associated with PJP (OR 0.239, 95% CI 0.088–0.653, p = 0.005), suggesting that patients requiring ICU care at presentation were less likely to have PJP.
Table 3.
Multivariate analysis of diagnostic factors associated with Pneumocystis pneumonia (PJP) versus non-PJP patients
| OR | 95% CI | p | |
|---|---|---|---|
| Age | 1.042 | 1.007–1.077 | 0.017 |
| Solid organ malignancy | 2.304 | 1.007–5.270 | 0.048 |
| Cytomegalovirus co-infection | 3.786 | 1.456–9.844 | 0.006 |
| Hematologic malignancy | 0.594 | 0.212–1.665 | 0.322 |
| LDH change % (pre- vs. post-diagnosis) | 1.002 | 1.000–1.005 | 0.103 |
| Presence of ground-glass opacity | 2.298 | 0.896–5.895 | 0.084 |
| ICU admission at diagnosis | 0.239 | 0.088–0.653 | 0.005 |
All individuals underwent PJP PCR testing on respiratory samples due to various clinical indications, including hypoxemia, ground-glass opacities, and fever etc.
LDH: Lactate dehydrogenase; ICU: Intensive care unit
Variables were selected based on univariate significance and clinical relevance
Among the 59 patients with PJP, 37 survived and 22 did not (Table 4). In univariate analysis, non-survivors more frequently had bacterial co-infections (63.6% vs. 21.6%, p = 0.001), CMV co-infections (47.4% vs. 20%, p = 0.035), ICU admission at diagnosis (36.4% vs. 8.1%, p = 0.013), and hypoxemia (72.7% vs. 40.5%, p = 0.017). They also had higher median CRP levels (134 vs. 97 mg/L, p = 0.046) and a greater, though not statistically significant, median percentage change in LDH from pre- to post-diagnosis (95.8% vs. 62.6%). In multivariate analysis, bacterial co-infection remained independently associated with mortality (OR 5.854, 95% CI 1.544–22.189, p = 0.009), whereas CMV co-infection and hypoxemia did not.
Table 4.
Multivariate analysis of factors associated with 4-week mortality in patients with pneumocystis pneumonia (n = 59)
| Survivors (n = 37) | Non-survivors (n = 22) |
Total (n = 59) |
p
(Univariate) |
OR | 95% CI |
p
(multivariate) |
|
|---|---|---|---|---|---|---|---|
| Age median (IQR) | 69 (62-77) | 72 (67-77) | 70 (65.5-77) | 0.994 | - | - | - |
| Male gender | 23 (62.2) | 19 (86.4) | 42 (71.2) | 0.074 | - | - | - |
| Bacterial co-infection | 8 (21.6) | 14 (63.6) | 22 (37.3) | 0.001 | 5.854 | 1.544–22.189 | 0.009 |
| Cytomegalovirus co-infection | 7 (20) | 9 (47.4) | 16 (29.6) | 0.035 | 2.215 | 0.553–8.871 | 0.261 |
| ICU admission at diagnosis | 3 (8.1) | 8 (36.4) | 11 (18.6) | 0.013 | - | - | - |
| Hypoxemia | 15 (40.5) | 16 (72.7) | 31 (52.5) | 0.017 | 3.018 | 0.803–11.349 | 0.102 |
|
LDH change % median (IQR) |
62.6 (20.9–164) |
95.8 (63–170.9) |
81.6 (23.3–170) |
0.276 | |||
|
CRP median (IQR) |
97 (27–142) |
134 (93–183) |
105 (59–153) |
0.046 | |||
|
Pneumocystis jirovecii fungal load (copies/ml) median (IQR) |
11,017 (1385–99161) |
4147 (959-60998) |
10,892 (1123–71614) |
0.541 | - | - | - |
LDH: Lactate dehydrogenase; ICU: Intensive care unit
Discussion
We observed that patients with PJP were significantly older, and more frequently presented with solid organ malignancies, consistent with previous cohorts [1, 2]. Age was an independent risk factor (OR 1.042; p = 0.017), in line with studies showing increased susceptibility among cancer and non-HIV immunocompromised patients [1, 20]. Solid organ malignancy was also independently associated with PJP (OR 2.304; p = 0.048), supporting multicenter findings identifying this condition as a key predisposing factor outside the HIV setting [1, 2]. Notably, cytomegalovirus co-infection emerged as a significant independent risk factor (OR 3.786; p = 0.006), underscoring its potential role in the pathogenesis and clinical course of PJP in immunocompromised hosts.
A definitive diagnosis of PJP requires visualization of cystic or trophic forms in respiratory specimens [21]. Because clinical and radiological findings are non-specific, PCR positivity alone may reflect colonization rather than active infection, particularly in immunocompromised hosts [21–25]. Therefore, we applied EORTC/MSGERC criteria to improve diagnostic specificity and classified PCR-positive patients without compatible clinical and radiological findings as colonized. Approximately 8% of PCR-positive patients were included in the control group on this basis. Furthermore, the non-PJP group intentionally consisted of patients in whom PJP was clinically suspected but ultimately excluded according to PCR results and EORTC/MSGERC criteria. These patients frequently had alternative diagnoses, including pulmonary edema, interstitial lung disease exacerbations, immune-related pneumonitis, hypersensitivity pneumonitis, or other pulmonary infections. We believe that this design reflects routine clinical practice and enhances the applicability of our findings.
Radiologically, bilateral lung involvement and ground-glass opacities were significantly more frequent among patients with PJP, consistent with previous studies [22, 26]. Nevertheless, these findings are not specific and may also occur in other infectious or non-infectious pulmonary diseases in immunocompromised hosts [21]. The inclusion of a control group drawn from the same clinically relevant population strengthens the applicability of our results to routine practice.
Elevated LDH likely reflects lung injury and metabolic stress in PJP [1, 27]. Because baseline LDH levels varied substantially among patients and were frequently elevated due to underlying conditions, we focused on percentage changes rather than absolute LDH values to better assess the dynamic impact of PJP. Accordingly, LDH increased significantly more in PJP patients than in controls (81.6% vs. 20.7%, p = 0.001). However, among patients with elevated baseline LDH, the difference was attenuated and not statistically significant (29.7% vs. 16.3%, p = 0.397), suggesting limited discriminatory value in this subgroup. These findings indicate that serial LDH measurements and relative changes may be more informative than absolute values, particularly when baseline LDH levels are within the normal range. This interpretation is supported by previous studies in non-HIV immunocompromised populations reporting higher LDH levels in PJP compared with non-PJP patients [16–18]. Given that 39% of our PJP cases already had elevated baseline LDH, both baseline values and serial changes may provide complementary clinical information in patients with hypoxia and pulmonary involvement. However, the precise threshold for a clinically meaningful LDH increase remains undefined, and larger studies are needed to better establish the clinical utility of LDH kinetics.
CMV co-infection was independently associated with PJP and is recognized as an important complicating factor in immunocompromised lung disease [26, 28]. While CMV co-infection correlated with higher mortality in univariate analysis, it was not significant in multivariate models. This finding suggests that although CMV may reflect greater immunosuppression, other factors—particularly bacterial superinfection—drive adverse outcomes [21]. Bacterial co-infection was the strongest predictor of mortality among probable PJP cases (OR 5.854, p = 0.009), aligning with previous studies showing that secondary bacterial infections, often nosocomial or ventilator-associated, substantially increase mortality [1, 29].
ICU admission and mechanical ventilation were more frequent in the control group, and ICU admission was inversely associated with PJP (OR 0.239, p = 0.005). This finding should be interpreted with caution because of potential selection bias. Only patients who underwent PJP PCR testing were included in the study. In clinical practice, critically ill patients are more likely to undergo extensive diagnostic evaluation, including respiratory sampling for PJP PCR, regardless of the underlying diagnosis. As a result, many ICU patients with negative PCR results were included in the control group. Therefore, the higher rates of ICU admission and mechanical ventilation in controls may reflect testing practices rather than true differences between patients with and without PJP. This phenomenon was also reported in other retrospective studies of immunocompromised patients [22, 24].
In our study, mortality among PJP patients was 37%, consistent with reported rates for non-HIV immunocompromised populations (20–60%), with elevated risk in those presenting with co-infections or severe hypoxemia, compared to 7–20% in HIV-positive patients [29, 30]. PJP remains highly lethal, with worse outcomes in HIV-negative patients due to delayed diagnosis from lower fungal burden and more severe inflammatory lung injury. In contrast, HIV-positive patients have higher organism loads and reduced inflammation, which improve diagnostic yield and mitigate lung damage [5, 21, 25]. Thus, HIV status is critical in interpreting risk of mortality and diagnostic testing.
Non-survivors in our cohort were more likely to have bacterial or CMV co-infections, ICU admission, hypoxemia, and elevated CRP levels in univariate analysis. This pattern mirrors mortality risk patterns reported for non-HIV PJP in both cancer and solid organ transplant recipients [2, 30]. Importantly, neither age, sex, LDH change, nor fungal burden significantly predicted death, consistent with recent meta-analyses indicating that comorbid infections, rather than absolute fungal load, are the primary contributors to mortality in this population [4]. In multivariate analysis, bacterial co-infection remained independently associated with mortality (OR 5.854, 95% CI 1.544–22.189, p = 0.009), whereas CMV co-infection and hypoxemia did not. Recent studies similarly highlight mixed infections, hypoxemia, and systemic inflammation as major mortality drivers [2, 5, 23].
The overall use of PJP prophylaxis in our cohort was relatively low, with only 23 of 235 patients (9.8%) receiving prophylactic treatment. Although prophylaxis was more frequent in the control group (10.8%) than in patients who developed PJP (6.8%), this difference was not statistically significant. Universal PJP prophylaxis is not currently recommended for all patients with solid tumors or hematologic malignancies, and prophylactic strategies are generally guided by the underlying disease, treatment regimen, and estimated level of immunosuppression. Because detailed data on specific chemotherapy, immunotherapy, and targeted therapy regimens were not systematically evaluated in our study, we were unable to determine which treatment-related factors may have influenced prophylaxis decisions or PJP risk. Nevertheless, the occurrence of PJP in this heterogeneous oncology population suggests that risk assessment for PJP and the potential role of prophylaxis may warrant greater attention in selected high-risk patients.
Our study provides insight into the clinical characteristics, risk factors, diagnostic features, and mortality predictors of patients with probable PJP using EORTC/MSGERC criteria and PCR-based diagnosis. A major strength is the assessment of LDH dynamics through measurements obtained both before and after diagnosis, including separate analyses of patients with elevated baseline LDH levels.
Several limitations should be acknowledged. The retrospective, single-center design and relatively small sample size may limit generalizability and introduce bias. Variability in the timing of LDH measurements and respiratory specimen types may have affected diagnostic accuracy and fungal load interpretation. Because histopathology or direct fluorescent antibody testing was not routinely available, a definitive microbiological gold standard could not be established; consequently, all cases were classified as probable PJP, and an optimal quantitative PCR cutoff for distinguishing infection from colonization could not be validated. Detailed immunological status and specific chemotherapy, immunotherapy, and targeted therapy regimens were not systematically assessed. Mortality was evaluated only at 4 weeks, precluding assessment of long-term outcomes. Although EORTC/MSGERC criteria improved diagnostic specificity, atypical or subclinical cases may have been missed. The inclusion of a small number of likely colonized patients in the control group may have resulted in limited misclassification. Additional biomarkers, including (1, 3)-β-D-glucan and procalcitonin, were not consistently available. Finally, because only patients undergoing PCR testing were included, selection bias is possible, and critically ill PCR-negative patients may have been overrepresented in the control group, potentially influencing comparisons of ICU admission and mechanical ventilation rates.
Conclusions
We found that older age, solid organ malginancy, and CMV co-infection were associated with an increased risk of PJP. LDH showed a greater relative increase in PJP patients than in controls, suggesting its potential diagnostic value as an adjunctive marker, although this distinction was less clear in patients with elevated baseline LDH. Bacterial co-infection was the main determinant of mortality, underscoring the importance of early recognition and management. The role of extended prophylaxis in this high-risk population warrants further evaluation in prospective studies.
Acknowledgements
We will always remember Bilge Emine Dikenelli, MD, who died from cancer on March 7, 2024. We dedicate this writing to her memory. May you rest in peace, Bilge. We thank Burak Uslu for providing the figure and conducting the visual inspection. We also thank Zeliha Genç for her valuable support for the data extraction.
Author contributions
All authors meet the ICMJE authorship criteria. Author contributions are as described below.Conceptualization: Mahir Kapmaz, Berk Mızrak, Şevval Nur Bektaş, Pelin İrkören, Süda Tekin.Data curation: Berk Mızrak, Şevval Nur Bektaş, Ayşe Okan, Pelin İrkören, Mahir Kapmaz, Meyha Şahin.Formal analysis: Berk Mızrak, Şevval Nur Bektaş, Pelin İrkören, Süda Tekin, Mahir Kapmaz, Meyha Şahin.Supervision: Mahir Kapmaz, Süda Tekin, Önder Ergönül.All authors contributed to the writing of the final manuscript.
Funding
No financial support was provided relevant to this article.
Data availability
The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request and subject to approval by the relevant institutional authorities.
Declarations
Ethics approval
This study was approved by the Koç University Committee on Human Research in Istanbul with a registry number of 2022.451.IRB1.177.
Consent for publication
Not applicable.
Consent to participate
This was a retrospective observational study and informed consent to participate was waived in accordance with the ethics approval.
Competing interests
The authors declare no competing interests.
Footnotes
Note: A part of this study, titled as “Impact of lactate dehydrogenase, bacterial and cytomegalovirus co-infections on mortality in non-HIV-infected patients with Pneumocystis jirovecii pneumonia”, authored by Berk Mizrak, Sevval Nur Bektas, Mahir Kapmaz, Ayse Okan, and Suda Tekin, was presented as poster no. 2761 at ESCMID Global 2024 in Barcelona, Spain, on 29 April 2024.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Chatterji S, Franchi V, Konnakottu AJ, Das P, Mukherjee S, Bhattacharya S, Chatterjee A. Pneumocystis jirovecii Pneumonia in Cancer Patients, a Lethal Yet Fully Preventable Disease: Insights From a Tertiary Cancer Center in East India. Asia Pac J Clin Oncol. 2025;21(3):319–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Benedict K, Baggs J, Gold JAW, Epidemiology. Clinical Characteristics, and Diagnostic Testing Practices for Pneumocystis Pneumonia-Associated Hospitalizations, United States, 2019–2022. Open Forum Infect Dis. 2024;11(2):ofae054. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Fujioka H, Matsui H, Homma Y, Nagai T, Otsuki A, Ito H, Ohmura S, Miyamoto T, Shichi D, Tomohisa W, Otsuka Y, Nakashima K. Association of time-to-treatment with prognosis in pneumocystis pneumonia among immunocompromised patients without HIV infection: a multi-center, retrospective observational cohort study. BMC Infect Dis. 2025;25(1):531. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Perret T, Kritikos A, Hauser PM, Guiver M, Coste AT, Jaton K, Lamoth F. Ability of quantitative PCR to discriminate Pneumocystis jirovecii pneumonia from colonization. J Med Microbiol. 2020;69(5):705–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Guegan H, Robert-Gangneux F. Molecular diagnosis of Pneumocystis pneumonia in immunocompromised patients. Curr Opin Infect Dis. 2019;32(4):314–21. [DOI] [PubMed] [Google Scholar]
- 6.Mappin-Kasirer B, Del Corpo O, Gingras MA, Hass A, Hsu JM, Costiniuk CT, Ezer N, Fraser RS, Lee TC, McDonald EG. Development of a clinical risk score for the prediction of Pneumocystis jirovecii pneumonia in hospitalised patients. BMC Infect Dis. 2024;24(1):1032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Chai D, Yang T, Zhang L, Hui Y, Feng J, Wang W. Prognostic value of the lactate dehydrogenase to albumin ratio in cancer patients. Front Nutr. 2025;12:1610487. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kanbay M, Mizrak B, Alper EN, Copur S, Ortiz A. Acute kidney injury following CAR-T cell therapy: a nephrologist’s perspective. Clin Kidney J. 2024;18(1):sfae359. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Shanmugavel Geetha H, Prabhu S, Sekar A, Gogtay M, Singh Y, Mishra AK, Abraham GM, Martin S. Use of inflammatory markers as predictor for mechanical ventilation in COVID-19 patients with stages IIIb-V chronic kidney disease? World J Virol. 2023;12(5):286–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Tini P, Cinelli E, Yavorska M, Donnini F, Marampon F, Pastina P, Rubino G, Chibbaro S, Cerase A, Mazzei MA, Di Giacomo AM, Minniti G. Serum lactate dehydrogenase as a prognostic marker for treatment response in IDH wild-type glioblastoma patients undergoing stupp protocol. J Neurooncol. 2025;171(2):413–21. [DOI] [PubMed] [Google Scholar]
- 11.Galli E, Guarneri A, Sorà F, Viscovo M, Pansini I, Maiolo E, Alma E, Annunziata S, Sica S, Leccisotti L, Hohaus S. Baseline Tumor Burden Assessed With AI-Guided PET/CT Total Metabolic Tumor Volume (TMTV) and LDH Levels Predict Efficacy of CAR-T in Aggressive B-Cell Lymphoma. Hematol Oncol. 2025;43(1):e70029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Miyake K, Senoo S, Shiiba R, Itano J, Kimura G, Kawahara T, Tamura T, Kudo K, Kawamura T, Nakahara Y, Higo H, Himeji D, Takigawa N, Miyahara N, Okayama Respiratory Disease Study Group (ORDSG). Pneumocystis jirovecii pneumonia mortality risk associated with preceding long-term steroid use for the underlying disease: A multicenter, retrospective cohort study. PLoS ONE. 2024;19(2):e0292507. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Zhao M, Yue R, Wu X, Gao Z, He M, Pan L. The diagnostic value of metagenomic next-generation sequencing for identifying Pneumocystis jirovecii infection in non-HIV immunocompromised patients. Front Cell Infect Microbiol. 2022;12:1026739. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Grønseth S, Rogne T, Hannula R, Åsvold BO, Afset JE, Damås JK. Semiquantitative Real-Time PCR to Distinguish Pneumocystis Pneumonia from Colonization in a Heterogeneous Population of HIV-Negative Immunocompromised Patients. Microbiol Spectr. 2021;9(1):e0002621. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.McDonald EG, Afshar A, Assiri B, Boyles T, Hsu JM, Khuong N, Prosty C, So M, Sohani ZN, Butler-Laporte G, Lee TC. Pneumocystis jirovecii pneumonia in people living with HIV: a review. Clin Microbiol Rev. 2024;37(1):e0010122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Cao Y, Chen J, Dong L. Supplementary Role of Immunological Indicators in the Diagnosis and Prognosis of Pneumocystis Pneumonia in Non-HIV Immunocompromised Patients. Infect Drug Resist. 2022;15:4675–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Long Y, Zhang C, Su L, Que C. Pneumocystis jirovecii dihydropteroate synthase gene mutations in a group of HIV-negative immunocompromised patients with Pneumocystis pneumonia. Exp Ther Med. 2014;8(6):1825–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Sun R, Lv D, Xiao M, Zhang L, Xu J, Yu X, Zhu H, Yang J. Diagnostic accuracy of the 1,3-beta-D-glucan test and lactate dehydrogenase for pneumocystis pneumonia in non-HIV patients. Sci Rep. 2021;11(1):9226. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Donnelly JP, Chen SC, Kauffman CA, Steinbach WJ, Baddley JW, Verweij PE, Clancy CJ, Wingard JR, Lockhart SR, Groll AH, Sorrell TC, Bassetti M, Akan H, Alexander BD, Andes D, Azoulay E, Bialek R, Bradsher RW, Bretagne S, Calandra T, Caliendo AM, Castagnola E, Cruciani M, Cuenca-Estrella M, Decker CF, Desai SR, Fisher B, Harrison T, Heussel CP, Jensen HE, Kibbler CC, Kontoyiannis DP, Kullberg BJ, Lagrou K, Lamoth F, Lehrnbecher T, Loeffler J, Lortholary O, Maertens J, Marchetti O, Marr KA, Masur H, Meis JF, Morrisey CO, Nucci M, Ostrosky-Zeichner L, Pagano L, Patterson TF, Perfect JR, Racil Z, Roilides E, Ruhnke M, Prokop CS, Shoham S, Slavin MA, Stevens DA, Thompson GR, Vazquez JA, Viscoli C, Walsh TJ, Warris A, Wheat LJ, White PL, Zaoutis TE, Pappas PG. Revision and Update of the Consensus Definitions of Invasive Fungal Disease From the European Organization for Research and Treatment of Cancer and the Mycoses Study Group Education and Research Consortium. Clin Infect Dis. 2020;71(6):1367–76. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Pulsipher AM, Vikram HR, Gotway MB, Cartin-Ceba R, Henry KJ, Johnson AM, Gassner RR, Zhou K, Green DB, Thompson ER, Limper AH, Sen A, Lee AS, Ham K. An Observational Analysis of a Large Cohort of Pneumocystis jirovecii Pneumonia. Open Forum Infect Dis. 2025;12(7):ofaf390. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Brown L, Rautemaa-Richardson R, Mengoli C, Alanio A, Barnes RA, Bretagne S, Chen SCA, Cordonnier C, Donnelly JP, Heinz WJ, Jones B, Klingspor L, Loeffler J, Rogers TR, Rowbotham E, White PL, Cruciani M. Polymerase Chain Reaction on Respiratory Tract Specimens of Immunocompromised Patients to Diagnose Pneumocystis Pneumonia: A Systematic Review and Meta-analysis. Clin Infect Dis. 2024;79(1):161–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Abramovich T, Korem M, Kuint R, Michael-Gayego A, Moran-Gilad J, Olshtain-Pops K. The Role of Polymerase Chain Reaction (PCR) and Quantification Cycle Values in the Diagnosis of Pneumocystis jirovecii Pneumonia. J Fungi (Basel). 2025;11(8):557. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Yang SL, Wen YH, Wu YS, Wang MC, Chang PY, Yang S, Lu JJ. Diagnosis of Pneumocystis pneumonia by real-time PCR in patients with various underlying diseases. J Microbiol Immunol Infect. 2020;53(5):785–90. [DOI] [PubMed] [Google Scholar]
- 24.Doyle L, Vogel S, Procop GW, Pneumocystis PCR. It Is Time to Make PCR the Test of Choice. Open Forum Infect Dis. 2017;4:ofx193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Bateman M, Oladele R, Kolls JK. Diagnosing Pneumocystis jirovecii pneumonia: A review of current methods and novel approaches. Med Mycol. 2020;58(8):1015–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Korkmaz Ekren P, Töreyin ZN, Nahid P, Doskaya M, Caner A, Turgay N, Zeytinoglu A, Toz S, Bacakoglu F, Guruz Y, Erensoy S. The association between Cytomegalovirus co-infection with Pneumocystis pneumonia and mortality in immunocompromised non-HIV patients. Clin Respir J. 2018;12(11):2590–7. [DOI] [PubMed] [Google Scholar]
- 27.Sarasombath PT, Thongpiya J, Chulanetra M, Wijit S, Chinabut P, Ongrotchanakun J, Jitmuang A, Wanachiwanawin D. Quantitative PCR to Discriminate Between Pneumocystis Pneumonia and Colonization in HIV and Non-HIV Immunocompromised Patients. Front Microbiol. 2021;12:729193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Li J, Mu X, Li H, Liu X. Clinical course and prognostic factors of Pneumocystis pneumonia with respiratory failure in non-HIV patients. Front Cell Infect Microbiol. 2024;14:1380494. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Kao TW, Ruan SY, Huang YT, Liu WD, Liu CJ, Chen YY, Hsueh PR, Yu CJ, Chien JY. TACTICS (TAiwan CollaboraTive Intensive Care Study) Group. Evolving risk factors and predisposing conditions of Pneumocystis pneumonia in non-HIV patients: A seven-year multicenter study. J Infect. 2025;91(3):106592. [DOI] [PubMed] [Google Scholar]
- 30.Rhoads S, Maloney J, Mantha A, Van Hook R, Henao-Martínez AF. Pneumocystis jirovecii Pneumonia in HIV-Negative, Non-transplant Patients: Epidemiology, Clinical Manifestations, Diagnosis, Treatment, and Prevention. Curr Fungal Infect Rep. 2024;18(2):125–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request and subject to approval by the relevant institutional authorities.