Differential Diagnoses for Miliary Pattern on Chest X-Ray
The most common causes of a miliary pattern on chest radiography are tuberculosis, fungal infections (particularly coccidioidomycosis in endemic areas), metastatic malignancy, sarcoidosis, and pneumoconiosis, with the specific etiology heavily dependent on geographic location, immunization status, and immune competence. 1, 2
Primary Diagnostic Considerations by Clinical Context
Infectious Etiologies
Tuberculosis remains the classic cause of miliary pattern globally, particularly in patients with travel to or residence in TB-endemic regions, immunocompromised states (HIV, immunosuppressive therapy), or lack of BCG vaccination 3, 1, 2
Coccidioidomycosis is the leading cause (53.7%) in Coccidioides-endemic regions (southwestern United States, parts of Mexico and Central/South America), especially when serum eosinophilia >500 cells/μL is present (100% predictive in one cohort) 1
Other fungal infections including disseminated histoplasmosis (endemic to Ohio and Mississippi River valleys) and Candida albicans (typically in severely immunocompromised patients) 1, 4
Atypical mycobacteria such as Mycobacterium simiae can rarely present with miliary patterns 1
Malignant Etiologies
Metastatic adenocarcinoma accounts for approximately 17% of miliary patterns, with primary sources including thyroid, kidney, lung (particularly mucinous adenocarcinoma), and trophoblastic tumors 1, 2, 5
Lymphoma can rarely present with miliary nodules 1
Sarcomas may metastasize in a miliary distribution 2
Non-Infectious, Non-Malignant Etiologies
Sarcoidosis presents with miliary or pseudo-miliary patterns in rare cases (true miliary pattern without fissural nodules in only 34.6% of reported cases), typically in older patients (>40 years) with more comorbidities and symptomatic presentation 2, 6, 4
Pneumoconiosis from occupational dust exposures (silicosis, coal worker's pneumoconiosis) 2
Hypersensitivity pneumonitis in the setting of specific environmental or occupational exposures 2
High-Resolution CT Pattern Recognition
HRCT is essential for characterizing nodule distribution, which narrows the differential diagnosis significantly 2, 4:
Random distribution (no relationship to secondary lobular structures): Most consistent with hematogenous spread from tuberculosis, fungal infections, or metastatic disease 2, 4
Perilymphatic distribution (along bronchovascular bundles, interlobular septa, and pleural surfaces): Strongly suggests sarcoidosis, though this represents "pseudo-miliary" rather than true miliary pattern 2, 6, 4
Centrilobular distribution: Indicates bronchogenic spread, seen in hypersensitivity pneumonitis and some infections 2
Critical Distinguishing Features
Favoring Sarcoidosis
- Peripheral/subpleural predominance of nodules 4
- Thickened interlobular septa and interlobar fissures 4
- Upper/middle lung zone predominance 4
- Presence of fissural nodules (excludes true miliary pattern) 6
Favoring Tuberculosis or Metastatic Disease
- Multiple cyst-like lesions (12 mm diameter) 4
- Random nodule distribution without perilymphatic features 2, 4
- Upper/middle lung zone predominance (tuberculosis) 4
Favoring Coccidioidomycosis
- Serum eosinophilia >500 cells/μL 1
- Elevated β-D-glucan levels 1
- Geographic exposure to endemic areas 1
Diagnostic Algorithm
Obtain detailed exposure history: Travel to TB-endemic or fungal-endemic regions, occupational dust exposures, immunization status (BCG), HIV status, immunosuppressive medications 3, 7
Order HRCT chest with IV contrast using thin-section reconstruction (≤1.5 mm slices) with three-plane imaging to characterize nodule distribution pattern 8, 2
Laboratory evaluation: Complete blood count with differential (eosinophil count), β-D-glucan, tuberculin skin test or interferon-gamma release assay, fungal serologies based on endemic exposure 1
Tissue diagnosis: Bronchoscopy with transbronchial biopsy is typically required when imaging and clinical context are not definitive, as cytology alone is insufficient 3, 5
Common Pitfalls
Assuming tuberculosis without considering endemic fungal infections in appropriate geographic regions leads to delayed antifungal therapy 1
Misinterpreting sarcoidosis with subtle perilymphatic features as true miliary pattern when fissural nodules are present 6
Overlooking metastatic adenocarcinoma in younger patients (as young as 35 years) who present with respiratory symptoms mimicking infection 5
Failing to recognize the association between tuberculosis and subsequent sarcoidosis, which may occur 52 weeks (median) after TB treatment 6