Pulmonary cryptococcosis beyond human immunodeficiency virus: a five-year cohort study of diagnostic challenges, radiologic mimics, and clinical predictors

Abstract

Background

Cryptococcosis is an opportunistic fungal infection that increasingly affects non-human immunodeficiency virus (HIV), immunocompetent or mildly immunocompromised individuals. Pulmonary cryptococcosis, in particular, poses diagnostic challenges due to its radiological resemblance to malignant lung lesions and often subtle or absent clinical symptoms. The aim of this study is to characterize the demographic features, clinical presentation, diagnostic pathways, radiological and pathological findings, and treatment patterns among hospitalized patients with histopathologically or molecularly confirmed pulmonary cryptococcosis.

Methods

This retrospective descriptive study included 61 non-HIV patients hospitalized with confirmed pulmonary cryptococcal infection over a 5-year period. Data were extracted from medical records, including demographics, comorbidities, presenting symptoms, imaging features, diagnostic procedures, histopathological and molecular confirmation, laboratory parameters, and treatment details. Pearson correlation and linear regression analyses were used to explore predictors of hospital length of stay.

Results

The cohort comprised predominantly male patients (72.1%) with a mean age of 51.6 years. Most cases (63.3%) were incidentally detected on imaging, while the remainder presented with non-specific respiratory symptoms. Radiological findings frequently included multiple pulmonary nodules (56.4%) and ground-glass opacities (16.4%). Surgical resection was performed in over half of the cases, primarily due to initial suspicion of malignancy. Diagnosis was confirmed through periodic acid-Schiff (PAS) (72.1%) or silver staining (57.4%), and supported by molecular testing in selected cases. Three patients (4.9%) had central nervous system (CNS) involvement confirmed by lumbar puncture and imaging. All patients received fluconazole as first-line antifungal therapy, while those with CNS involvement received a combination of amphotericin B and fluconazole followed by fluconazole maintenance therapy. In multivariate analysis, bilateral pulmonary nodules were associated with shorter hospital stays (β =−5.36, P=0.041), while the presence of cavitary lesions predicted longer hospitalization (β =5.04, P=0.03).

Conclusions

Pulmonary cryptococcosis may mimic malignancy and frequently presents without overt symptoms, contributing to diagnostic delays and surgical interventions. Recognition of its radiological patterns and clinical spectrum in non-HIV patients is essential for timely diagnosis and appropriate antifungal treatment. This study highlights the importance of considering cryptococcosis in the differential diagnosis of pulmonary nodules, even in asymptomatic, immunocompetent individuals.

Highlight box.

Key findings

• This retrospective cohort study included 61 non-human immunodeficiency virus (HIV) hospitalized patients with histopathologically or molecularly confirmed pulmonary cryptococcosis. Most were male (72.1%), mean age 51.6 years; 63.3% were asymptomatic, diagnosed incidentally via imaging. Dominant radiological findings were multiple pulmonary nodules (56.4%) and ground-glass opacities (16.4%), 20% with bilateral involvement. Over half had invasive surgical resection (e.g., lobectomy) due to suspected malignancy; diagnosis was confirmed by periodic acid-Schiff staining (72.1%) and molecular methods like next-generation sequencing (NGS). Central nervous system involvement was rare (4.9%). All received fluconazole. Bilateral nodules predicted shorter hospital stays (β =−5.36, P=0.041), cavitary lesions longer (β =5.04, P=0.03).

What is known and what is new?

• Cryptococcosis, linked to advanced immunosuppression (e.g., HIV), is increasingly seen in non-HIV, even immunocompetent individuals. Pulmonary forms mimic malignancy, causing diagnostic delays and unnecessary surgeries; inflammatory markers like C-reactive protein and erythrocyte sedimentation rate are unreliable.

• This study adds a 5-year cohort analysis of non-HIV patients, showing high incidental diagnosis (63.3%) and novel predictors of hospital stay (e.g., bilateral nodules reduce stays by 5.36 days). It emphasizes cryptococcosis can occur without overt immunosuppression and the role of molecular diagnostics (e.g., NGS) in reducing invasive procedures.

What is the implication, and what should change now?

• These findings imply misdiagnosis as cancer in non-HIV patients, leading to avoidable surgeries. Clinicians should include cryptococcosis in differential diagnosis of pulmonary nodules in asymptomatic, immunocompetent individuals and prioritize non-invasive molecular testing (e.g., NGS). Changes should include reducing routine surgical interventions, reviewing imaging for cryptococcal patterns, and initiating early antifungal therapy.

Introduction

Cryptococcosis is a potentially life-threatening opportunistic infection caused primarily by Cryptococcus neoformans and Cryptococcus gattii, two encapsulated yeast species with a predilection for the pulmonary and central nervous systems (CNS) (1). While the infection is classically associated with advanced immunosuppression (2,3)—particularly among individuals with human immunodeficiency virus/acquired immune deficiency syndrome (HIV/AIDS)—it has increasingly been reported in non-HIV populations (4-6), including those with diabetes, malignancy, chronic pulmonary disease, and other forms of immune dysregulation. This shift in host profiles presents diagnostic and therapeutic challenges, especially in settings where the infection manifests atypically or mimics malignancy on imaging studies.

Pulmonary cryptococcosis is the most common initial presentation, often discovered incidentally as solitary or multiple pulmonary nodules during routine radiological evaluation (7). These nodules may be indistinguishable from primary lung neoplasms or metastatic lesions, frequently prompting invasive procedures such as biopsy or lobectomy before a definitive diagnosis is reached. In some cases, dissemination to the CNS or other extrapulmonary sites may occur, particularly in patients with impaired cellular immunity (8). Delays in diagnosis are common, owing to the variable clinical presentation and the low index of suspicion in immunocompetent or asymptomatic patients.

Histopathological staining [e.g., periodic acid-Schiff (PAS), silver methenamine] and antigen-based assays remain diagnostic cornerstones, yet recent advances in molecular diagnostics—such as next-generation sequencing (NGS), metagenomic NGS (mNGS), and targeted NGS (tNGS)—have enhanced pathogen detection sensitivity, especially in challenging or smear-negative cases (9). Despite these developments, data remain limited regarding the real-world diagnostic pathways, clinical features, and outcomes of cryptococcal infection across diverse patient populations.

The present study aims to address this gap by characterizing the demographic, clinical, radiological, and laboratory profiles of hospitalized patients with histopathologically or molecularly confirmed cryptococcosis. We further examine diagnostic strategies, treatment patterns, and length of hospital stay (LOS), with a particular focus on potential predictors of prolonged hospitalization. By synthesizing insights from a 5-year institutional cohort, this study seeks to inform the early recognition and management of cryptococcosis in both immunocompromised and non-immunocompromised patients. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-914/rc).

Methods

Study design and setting

This was a retrospective descriptive study conducted at a tertiary academic medical center in China (Beijing Chest Hospital of Capital Medical University). The study included patients who were hospitalized for pulmonary cryptococcal infections between December 24^th, 2019 and November 21^st, 2024. All diagnoses were confirmed through histopathological and/or molecular diagnostic methods. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Beijing Chest Hospital of Capital Medical University and the requirement for individual informed consent was waived due to the retrospective design and full de-identification of data prior to analysis.

It is noteworthy that hospitalization in this cohort was primarily due to diagnostic evaluation or surgical intervention, as pulmonary malignancy was the initial consideration in most cases. Only a minority of patients were admitted because of symptomatic cryptococcosis.

Case identification and inclusion criteria

Eligible patients were identified through the hospital’s electronic medical records system, using keyword and International Classification of Diseases (ICD)-based queries across pathology reports, discharge diagnoses, and microbiology records. Inclusion criteria required hospitalization for a pulmonary cryptococcal infection confirmed by either positive histopathological staining (PAS and/or hexamine silver stain) or molecular techniques, including NGS, mNGS, or tNGS. All patients included in the cohort were HIV-negative. Patients without confirmed diagnosis or with incomplete clinical documentation were excluded.

Although some patients underwent lumbar puncture and brain imaging to assess for dissemination, the diagnosis in all included cases was primarily pulmonary in origin. Only three patients were diagnosed with cryptococcal meningitis. The remainder had no evidence of CNS involvement.

Data collection and variables

Data were collected through manual chart review by two independent investigators using a standardized case report form. Extracted information included demographic characteristics such as age, sex, height, weight, and comorbidities including diabetes mellitus, hypertension, autoimmune diseases, malignancies, and history of environmental exposure such as pigeon breeding.

Clinical presentation was assessed based on initial hospital admission notes, including symptoms such as cough, fever, shortness of breath, chest pain, and the duration of symptoms prior to admission. Hospitalization data included LOS and initial clinical impression upon presentation. Radiologic findings were recorded from chest computed tomography (CT) scans, including the number, size, and location of nodular lesions, as well as presence of cavities, consolidation, and ground-glass opacities. Laterality and lobar distribution were also documented.

Pathological confirmation included histological evaluation for granulomatous inflammation, necrosis, and positivity for PAS and hexamine silver stains. Molecular diagnostic confirmation was based on the identification of Cryptococcus species in bronchoalveolar lavage or resected tissue using NGS, mNGS, or tNGS platforms. Microbiological testing included fungal culture, India ink staining, and cryptococcal capsular antigen testing, particularly in the cerebrospinal fluid (CSF) samples from patients who underwent lumbar puncture.

Laboratory parameters at admission were documented, including white blood cell (WBC) counts, neutrophil counts, C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), procalcitonin, calcitonin, and immunological markers such as CD4 and CD8 T-cell levels. Tumor marker panels, obtained in selected patients to rule out malignancy, included carcinoembryonic antigen (CEA), cancer antigen 125 (CA125), neuron-specific enolase (NSE), squamous cell carcinoma antigen (SCC), cytokeratin 19 fragment (CYFRA21-1), and pro-gastrin-releasing peptide (ProGRP). Normal reference ranges were as follows: WBC (4.0–10.0)×10⁹/L, neutrophils (2.0–7.5)×10⁹/L, CRP 0–8 mg/L, ESR 0–20 mm/h, procalcitonin <0.05 ng/mL, calcitonin <10 pg/mL, CD4 410–1,590 cells/µL, CD8 190–1,140 cells/µL, CEA <5 ng/mL, and ProGRP <50 pg/mL.

While most patients underwent surgical intervention due to suspicion of malignancy based on imaging, no antifungal therapy was administered immediately after resection. Following histopathologic confirmation, all patients received medical treatment with oral fluconazole as first-line antifungal therapy. Surgical notes were reviewed to confirm that resection was the initial management approach, and inpatient medical records were used to verify subsequent pharmacologic treatment.

According to hospital protocol and Infectious Diseases Society of America (IDSA) guidelines, patients with isolated pulmonary disease received oral fluconazole, whereas those with confirmed CNS involvement were treated with a combination of amphotericin B (liposomal formulation) and fluconazole during hospitalization, followed by fluconazole maintenance therapy after discharge.

Statistical analysis

Descriptive statistics were used to summarize demographic, clinical, radiologic, and laboratory variables. Continuous variables were reported as means with standard deviations or medians with ranges, depending on distribution. Categorical variables were summarized as frequencies and percentages.

Pearson correlation coefficients were calculated to examine relationships between selected continuous variables, particularly to explore associations with LOS. Correlation was assessed between CRP, ESR, WBC, neutrophil count, nodule size, and symptom duration. Univariate linear regression models were performed to identify candidate predictors of LOS. Variables with a P value less than 0.10 were subsequently entered into a multivariate linear regression model using stepwise backward elimination. A two-sided P value <0.05 was considered statistically significant. All analyses were conducted using STATA version 18.0 (StataCorp LLC, College Station, TX, USA).

Results

Patient demographics and baseline characteristics

A total of 61 patients were hospitalized with histopathologically or molecularly confirmed cryptococcal infections over the study period. The majority were male (72.1%), with a mean age of 51.6 (standard deviation: 14.5) years and a median of 55 (range, 15–79) years. Mean height and weight were 167.9 cm and 71.6 kg, respectively. Most patients (63.3%) were asymptomatic and diagnosed incidentally through radiological detection of pulmonary shadows or nodules. Among symptomatic cases, cough (21.7%), sputum production (13.3%), intermittent chest pain (11.7%), and fever (8.3%) were the most frequently reported complaints. The median duration of symptoms prior to hospitalization was 45 (range, 4–730) days, and the median LOS was 12 (range, 2–42) days (Table 1). In most cases, hospitalization occurred for diagnostic or surgical purposes [e.g., video-assisted thoracoscopic surgery (VATS) lobectomy or biopsy] rather than for the management of cryptococcal symptoms, which were often minimal or absent.

Table 1. Demographic and baseline information of included patients with cryptococcal hospitalization in the past 5 years.

Variable	Number/mean ± SD	%/median (range)
Gender (n=61)
Female	17	27.87
Male	44	72.13
Age (years) (n=61)	51.57±14.49	55 (15–79)
Height (cm) (n=61)	167.93±8.22	168 (150–186)
Weight (kg) (n=61)	71.61±13.21	68 (50–105)
Clinical presentation (n=60)
Incidental finding on radiological examination (shadows and/or nodules)	38	63.33
Fever	5	8.33
Cough	13	21.67
Sputum	8	13.33
Chest tightness	3	5
Intermittent chest pain	7	11.67
Shortness of breath	2	3.33
Headache	1	1.67
Nausea	1	1.67
Symptom duration (days) (n=52)	111.69±166.38	45 (4–730)
Length of hospital stay (days) (n=60)	12.25±6.91	12 (2–42)

n, number of patients with analyzable data per each variable; SD, standard deviation.

Comorbidities

Hypertension (37.7%) and diabetes mellitus (18.0%) were the most prevalent comorbidities. Other conditions included cerebral infarction (8.2%), fatty liver (4.9%), coronary artery disease (4.9%), and prior tuberculosis (4.9%). Less common immunological or oncologic comorbidities included systemic lupus erythematosus, Sjögren’s syndrome, thyroid cancer, and lymphoma. Notably, 19.7% of patients were without any documented comorbidities (Table 2).

Table 2. Comorbidities and medical history of patients with cryptococcal hospitalization in the past 5 years.

Variable	Number	Percentage (%)
Diabetes	11	18.03
Sjögren syndrome	1	1.64
Liver damage	1	1.64
Brain tumor	1	1.64
Premature ventricular contractions	1	1.64
Reflux esophagitis	3	4.92
Hyperlipidemia	1	1.64
Asthma	1	1.64
Systemic lupus erythematosus	1	1.64
History of tuberculosis	3	4.92
History of pulmonary embolism	1	1.64
Coronary artery disease	3	4.92
Tuberculous pleurisy	1	1.64
Emphysema	1	1.64
Fatty liver	3	4.92
Hypothyroidism	2	3.28
Thyroid cancer	2	3.28
Hypertension	23	37.70
Smoking	2	3.28
Cerebral infarction	5	8.20
Hepatitis B virus	2	3.28
Lymphoma	1	1.64
Liver cirrhosis	1	1.64
Splenomegaly	1	1.64
Asymptomatic	12	19.67

Radiological patterns and distribution

Among the 61 patients, nodular lesions were the dominant radiological finding. Multiple nodules were observed in 56.4% of cases, while 43.6% exhibited solitary nodules. The right lung was most frequently involved (45%), followed by the left (35%) and bilateral presentations (20%). The lower lobes were affected in 61.1% of cases, with a minority involving multiple lobes. Nodule sizes ranged from 0.6 to 4.1 cm (median: 1.2 cm). Other radiographic findings included small cavitary lesions (21.3%), ground-glass opacities (16.4%), and patchy consolidations (8.2%) (Table 3).

Table 3. Radiographic findings in patients with cryptococcal hospitalization.

Variable	Number/mean ± SD	%/median (range)
Radiographic features (n=61)
Patchy consolidation	5	8.20
Small cavitary lesions	13	21.31
Calcifications	2	3.28
Consolidation	7	11.48
Mixed-density shadow	3	4.92
High-density shadow	2	3.28
GGO	10	16.39
Nodular characteristics
Nodule number (n=55)
Single nodule	24	43.64
Multiple nodules	31	56.36
Nodule shadow	3	4.92
Nodule laterality (n=60)
Bilateral	12	20
Left	21	35
Right	27	45
Affected lobe (n=54)
Lower	33	61.11
Middle	2	3.70
Upper	16	29.63
Upper + lower	2	3.70
Upper + middle + lower	1	1.85
Nodule size (n=20) (cm)	1.59±0.88	1.2 (0.6–4.1)

GGO, ground glass opacity; n, number of patients with analyzable data per variable; SD, standard deviation.

Diagnostic approaches and coinfections

Invasive diagnostic procedures were performed in 40 patients. Single-port VATS lobectomy was the most common intervention (57.5%), followed by CT-guided biopsy (35%) and wedge resection (5%). Bronchoscopic biopsy was rare (2.5%). Among 16 patients who underwent lavage-based molecular testing, various strains of Cryptococcus were identified using NGS, mNGS, and tNGS platforms. Coinfections were documented in four patients and included H. parahaemolyticus, K. pneumoniae, M. avium complex, and S. pneumoniae (Table 4).

Table 4. The diagnostic summary of included patients with cryptococcal hospitalization.

Variable	Number	Percentage (%)
Diagnostic intervention (n=40)
CT-guided puncture biopsy	14	35
Single-port VATS lobectomy	23	57.50
TBLB	1	2.50
Wedge resection	2	5
Lavage fluid (n=16)
Not specified spp.	1	6.25
NGS Cryptococcus 106	1	6.25
NGS Cryptococcus 16	1	6.25
NGS Cryptococcus 189	1	6.25
NGS Cryptococcus 45	1	6.25
NGS Cryptococcus 32	1	6.25
mNGS Cryptococcus 10	1	6.25
mNGS Cryptococcus 1909	1	6.25
mNGS Cryptococcus 49	1	6.25
mNGS Cryptococcus 60	1	6.25
mNGS Cryptococcus 628	1	6.25
tNGS Cryptococcus 45	1	6.25
tNGS Cryptococcus 50	2	12.50
tNGS Cryptococcus 628	1	6.25
Coinfections (n=4)
H. parahaemolyticus	1	25
K. pneumoniae	1	25
M. avium complex 242	1	25
S. pneumoniae	1	25

CT, computed tomography; mNGS, metagenomic next-generation sequencing; NGS, next-generation sequencing; spp., species; TBLB, transbronchial lung biopsy; tNGS, targeted next-generation sequencing; VATS, video-assisted thoracoscopic surgery.

Pathological and laboratory confirmation

PAS staining was positive in 72.1% of specimens, and silver staining in 57.4%. Chronic granulomatous inflammation was observed in 67.2%, and necrotic lung tissue in 16.4% of patients. Brain magnetic resonance imaging (MRI) was performed in 51 patients; 7.8% showed lesions suggestive of CNS involvement. These involved white matter degeneration (two cases), old infract lesion on one side (one case), ischemic and demyelinated changes in white matter (one case), and spot-like abnormalities in the subcortical white matter and corona radiata of the three frontal lobes (one case). Lumbar puncture was conducted in 22.4% of patients; of those, 23.1% had a positive India ink stain. Cryptococcal capsular antigen testing was positive in 57.1% of tested cases.

Laboratory data revealed elevated procalcitonin in 97% of tested patients, mostly in the mildly elevated range (0.05–0.5 ng/mL). CRP was elevated in 37.5% of cases (median: 5.03 mg/L), while ESR was elevated in 31.6% (median: 15.5 mm/h). CD4, CD8, and calcitonin levels were within normal limits for all tested patients. Neutrophil and WBC counts varied, with some patients presenting elevated values (Table 5).

Table 5. Pathological, histopathology, and laboratory findings in patients with cryptococcal hospitalization.

Variable	Number/mean ± SD	%/median (range)
Pathological & histological findings (n=61)
Hexamine silver stain (+ve)	35	57.38
PAS stain (+ve)	44	72.13
Necrotic lung tissue	10	16.39
Granulomatous inflammation (chronic)	41	67.21
CNS involvement
Brain MRI (n=51)
Negative	47	92.16
Brain lesions	4	7.84
Lumbar puncture (n=58)
Not performed	45	77.59
Performed (with negative results)	13	22.41
CSF analysis (India ink stain) (n=13)
Negative	10	76.92
Positive	3	23.08
Immune & laboratory status
Cryptococcus caps. antigen (n=14)
Negative	6	42.86
Positive	8	57.14
CD4 (n=4)
Normal range	4	100
CD8 (n=4)
Normal range	4	100
Calcitonin (n=6)
Normal range	6	100
Procalcitonin (n=34)
Normal range	1	2.94
Mildly increased (0.05–0.5 ng/mL)	30	88.24
Moderately increased (0.5–2 ng/mL)	3	8.82
Tumor markers (n=6)
Negative	5	83.33
Positive (for cytokeratin 19 fragment)	1	16.67
ESR (n=38) (mm/h)	23.24±23.63	15.5 (2–109)
CRP (n=40)	19.06±27.05	5.03 (0.26–114.2)
Normal range	25	62.50
Abnormally high	15	37.50
Neutrophil count (n=49) (×109/L)	5.46±3.36	4.6 (1.43–19.05)
WBC count (n=56) (×109/L)	7.66±3.27	6.93 (2.61–21.97)
Lung function test (n=8)
Normal	8	100
Blood group (n=52)
A	9	17.31
B	14	26.92
AB	8	15.38
O	21	40.38

+ve, positive; caps., capsulatus; CD4/CD8, cluster of differentiation 4/8; CNS, central nervous system; CRP, C-reactive protein; CSF, cerebrospinal fluid; ESR, erythrocyte sedimentation rate; MRI, magnetic resonance imaging; n, number of patients with analyzable data per variable; PAS, periodic acid-Schiff; SD, standard deviation; WBC, white blood cell.

Correlations and predictors of hospital stay

Correlation analysis (Table 6) showed a strong positive association between CRP and ESR (r=0.81, P<0.001), and moderate correlations between neutrophils and WBCs (r=0.97, P<0.001). However, no significant correlations were found between LOS and symptom duration, inflammatory markers, or nodule size.

Table 6. Correlation between LOS and other measured covariates in patients with cryptococcal hospitalization.

Variable	CRP	ESR	Neutrophil	WBC	Nodule size	Symptom duration	LOS
CRP
r	1	–	–	–	–	–	–
P	–	–	–	–	–	–	–
ESR
r	0.8089	1	–	–	–	–	–
P	<0.001***	–	–	–	–	–	–
Neutrophil
r	0.4087	0.4996	1	–	–	–	–
P	0.02*	0.002**	–	–	–	–	–
WBC
r	0.4771	0.5039	0.9693	1	–	–	–
P	0.003**	0.001**	<0.001***	–	–	–	–
Nodule size
r	−0.1418	−0.1235	−0.0861	−0.145	1	–	–
P	0.64	0.72	0.73	0.54	–	–	–
Symptom duration
r	0.0911	0.1973	−0.0007	−0.0097	0.2569	1	–
P	0.63	0.27	>0.99	0.95	0.36	–	–
LOS
r	0.0352	−0.1807	−0.0372	0.0037	0.282	−0.1657	1
P	0.84	0.28	0.80	0.98	0.23	0.25	–

r: correlation coefficient. *, P<0.05; **, P<0.01; ***, P<0.001. CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; LOS, length of hospital stay; WBC, white blood cell.

Univariate regression (Table 7) suggested that bilateral nodular involvement (β =−7.12, P=0.004) and multiple nodules (β =−4.14, P=0.03) were associated with reduced LOS, whereas the presence of cavitary lesions showed a trend toward longer hospitalization (β =3.71, P=0.09). In the adjusted model, only two predictors remained significant: bilateral nodules (β =−5.36, P=0.041) and cavitary lesions (β =5.04, P=0.03). Other variables such as age, gender, hypertension, or CRP levels were not independently associated with LOS.

Table 7. Univariate and multivariate linear regression model of the predictors of LOS (per day).

Variable	Unadjusted linear regression model						Adjusted linear regression model
	Coefficient	SE	Low CI	High CI	P value		Coefficient	SE	Low CI	High CI	P value
Male (vs. female)	3.222	1.952	−0.686	7.129	0.10		–	–	–	–	–
Age	0.010	0.063	−0.115	0.135	0.88		–	–	–	–	–
Weight	0.049	0.068	−0.087	0.186	0.47		–	–	–	–	–
Height	−0.005	0.110	−0.224	0.215	0.97		–	–	–	–	–
Incidental shadow/nodule found during examination	2.591	1.879	−1.172	6.355	0.17		–	–	–	–	–
Cough (vs. none)	−1.303	2.266	−5.841	3.234	0.57		–	–	–	–	–
Chest pain (vs. none)	−3.407	2.793	−8.999	2.186	0.23		–	–	–	–	–
Symptom duration	−0.007	0.006	−0.019	0.005	0.25		–	–	–	–	–
Diabetes (vs. none)	−2.644	2.300	−7.248	1.960	0.26		–	–	–	–	–
Hypertension (vs. none)	−1.816	1.836	−5.490	1.859	0.33		–	–	–	–	–
Asymptomatic presentation (vs. symptomatic)	2.083	2.233	−2.387	6.554	0.36		–	–	–	–	–
Cavitary lesion on imaging (vs. none)	3.707	2.130	−0.556	7.970	0.09		5.038	2.216	0.586	9.49	0.03*
GGO (vs. none)	0.359	2.520	−4.685	5.404	0.89		–	–	–	–	–
Multiple nodules (vs. single nodule)	−4.137	1.881	−7.909	−0.365	0.03*		−3.999	2.268	−8.557	0.558	0.08
Laterality (reference: left)
Right	−0.654	1.920	−4.500	3.193	0.74		−0.851	2.161	−5.195	3.492	0.70
Bilateral	−7.117	2.377	−11.878	−2.356	0.004**		−5.359	2.551	−10.485	−0.234	0.041
Affected lobe (reference: lower)
Middle lobe	1.076	4.640	−8.259	10.411	0.82		–	–	–	–	–
Upper lobe	−2.558	1.984	−6.549	1.434	0.20		–	–	–	–	–
Nodule size	2.608	2.091	−1.786	7.002	0.23		–	–	–	–	–
ESR	−0.045	0.041	−0.129	0.038	0.28		–	–	–	–	–
CRP	0.010	0.050	−0.091	0.112	0.84		–	–	–	–	–
Neutrophil	−0.062	0.245	−0.554	0.430	0.80		–	–	–	–	–
WBC	0.008	0.291	−0.576	0.591	0.98		–	–	–	–	–
Blood group (reference: A)
B	−0.556	3.025	−6.637	5.526	0.86		–	–	–	–	–
AB	2.694	3.440	−4.222	9.611	0.44		–	–	–	–	–
O	0.683	2.820	−4.988	6.353	0.81		–	–	–	–	–

*, P<0.05; **, P<0.01. CI, confidence interval; CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; GGO, ground-glass opacity; LOS, length of hospital stay; SE, standard error; WBC, white blood cell.

Discussion

This study presents one of the largest retrospective analyses of histopathologically and/or molecularly confirmed pulmonary cryptococcosis in non-HIV hospitalized patients to date. Our findings highlight the diagnostic complexity, frequent misclassification, and under-recognition of pulmonary cryptococcosis, particularly in immunocompetent individuals, and add depth to the growing literature on the presentation and management of this disease in non-AIDS populations (7,10-12).

The predominance of asymptomatic or minimally symptomatic presentations in our cohort is consistent with prior reports by Deng et al. (13), who noted that more than half of patients had no complications or systemic illness at diagnosis. Similarly, Fox et al. (14) emphasized that CT findings in immunocompetent individuals were often indistinguishable from neoplastic lesions, frequently prompting unnecessary surgical resection. In our series, over half of the patients underwent surgical interventions due to radiologic suspicion of malignancy, underscoring the diagnostic dilemma and the need for heightened clinical suspicion.

Radiologically, multiple nodules and solitary lesions were the most frequent patterns observed—aligning with patterns reported by Min et al. (15), who also noted a strong association between radiographic extent and serum cryptococcal antigen positivity in immunocompetent patients. Ground-glass opacities and cavitary changes, although less common, were present in a subset of cases. Notably, our multivariate analysis revealed that bilateral involvement was associated with shorter hospitalization, whereas cavitary lesions predicted longer stays. Importantly, these findings should be interpreted within the context that hospitalization in most patients was diagnostic or perioperative rather than therapeutic (10,11). Thus, the longer stay among those with cavitary lesions likely reflects more complex diagnostic evaluation or postoperative care, rather than disease severity. Patients with mild or incidental lesions would typically be managed as outpatients once the diagnosis is established.

Our laboratory data support the known limitations of inflammatory markers in cryptococcosis. As in the study by Yu et al. (16), elevations in CRP and ESR were inconsistent, and leukocytosis was uncommon, particularly in the absence of systemic dissemination. The limited diagnostic yield of serum biomarkers and the often-normal immune panels, including CD4/CD8 counts and procalcitonin, emphasize the importance of tissue diagnosis. Serum cryptococcal antigen, although valuable in disseminated disease, had suboptimal sensitivity in isolated pulmonary cases, as similarly observed by Singh et al. (17). In our cohort, serum cryptococcal antigen (CrAg) testing was performed in only a small subset (n=14), and while broader radiographic involvement appeared more common among CrAg-positive cases, no formal statistical correlation was conducted due to limited sample size. Ground-glass opacities and cavitary changes, although less common, were present in a subset of cases.

Our cohort was entirely composed of HIV-negative patients, many of whom had no overt immunosuppressive risk factors. Nevertheless, consistent with previous studies by Baddley et al. (18) and by Yu et al. (16), diabetes and environmental exposures such as pigeon handling were present in a subset, reinforcing that so-called immunocompetent patients may still harbor subtle predispositions. Moreover, the absence of CNS involvement in all but three cases suggests that routine lumbar puncture may not be necessary in the absence of neurological signs, a position supported by recent data questioning the universal application of this practice.

Surgical intervention was primarily diagnostic in our series, with resected tissue providing the first and often only means of diagnosis. However, our findings align with recommendations by Baddley et al. (18) who caution against surgery as first-line therapy in favor of less invasive techniques when fungal infection is suspected. All patients in our cohort ultimately received fluconazole, consistent with IDSA guidelines for mild-to-moderate pulmonary cryptococcosis in non-immunocompromised hosts. None experienced relapse or mortality within the observation period, supporting the favorable prognosis of localized disease in HIV-negative individuals when appropriately treated.

Of particular interest is our correlation analysis, which demonstrated a strong relationship between CRP and ESR but no meaningful associations with symptom duration or nodule size—metrics often presumed to reflect disease burden. Our regression modeling adds new insight to the field by identifying cavitary morphology as a practical predictor of longer hospitalization, potentially due to its confounding with bacterial abscess or malignancy.

The current study builds upon prior retrospective cohorts (13,15,18-23) and reinforces key messages: pulmonary cryptococcosis is frequently indolent, radiologically nonspecific, and easily mistaken for malignancy. The role of cryptococcal antigen testing remains nuanced, with its utility maximized in diffuse pulmonary or extrapulmonary disease rather than early or isolated nodular forms.

This study has several limitations. As a single-center retrospective analysis, it is subject to selection and information biases. Despite a relatively large sample size, the number of cases with meningitis was small, precluding meaningful subgroup analysis. Also, long-term outcomes post-discharge were not uniformly available for all patients. Nevertheless, our strict inclusion criteria, comprehensive data extraction, and multi-dimensional analysis strengthen the validity of the findings.

Conclusions

Pulmonary cryptococcosis in non-HIV patients often masquerades as malignancy and presents with minimal or absent symptoms. Diagnosis requires a high index of suspicion, especially in patients with incidental radiologic findings. Tissue biopsy remains essential in many cases, particularly when molecular testing is unavailable. Our study supports a nuanced approach to lumbar puncture and highlights the importance of incorporating non-invasive screening tests, particularly serum CrAg, as an initial step in suspected cases—especially in areas where pulmonary cryptococcosis is relatively prevalent. While CrAg sensitivity may be limited in isolated pulmonary disease, early testing could help guide clinical suspicion and reduce unnecessary surgical procedures. Tissue biopsy should be pursued when CrAg results are negative or inconclusive but radiologic suspicion remains high. Cavitary lesions in our cohort were associated with longer hospitalization, likely reflecting diagnostic or procedural factors rather than intrinsic disease severity. These findings may inform both diagnostic pathways and clinical decision-making in patients with suspected pulmonary cryptococcosis.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Beijing Chest Hospital of Capital Medical University and the requirement for individual informed consent was waived due to the retrospective design and full de-identification of data prior to analysis.

Data Sharing Statement

Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-914/dss

jtd-18-02-101-dss.pdf ^{(67KB, pdf)}

DOI: 10.21037/jtd-2025-914