Tuberculosis (TB) - Diagnosis & Investigations

1. Clinical Suspicion - When to Investigate

• Cough - persistent, productive, often with mucus or blood (note: 10-25% of bacteriologically confirmed TB patients do not report cough)
• Systemic symptoms - fever, night sweats, weight loss
• Hemoptysis - seen with more extensive disease
• Lymphadenopathy - especially in HIV-positive patients

• Murray & Nadel's Textbook of Respiratory Medicine, p. 1187-1188

2. Chest Radiograph

• Middle or lower lung zone consolidation with ipsilateral hilar/paratracheal lymphadenopathy
• Atelectasis from airway compression
• May show pleural effusion

• Upper lobe involvement (apical and posterior segments of right upper lobe; apical-posterior segment of left upper lobe)
• Cavitation is common
• Fibrotic scarring with lung shrinkage and calcification on healing
• Bronchogenic spread can cause lower lobe opacities
• Miliary pattern if hematogenous dissemination occurs

• Early HIV - typical upper lobe pattern
• Advanced HIV (low CD4) - atypical pattern: lower zone / diffuse opacities, intrathoracic adenopathy, less cavitation

• Murray & Nadel's Textbook of Respiratory Medicine, p. 1188-1189

3. Microbiological Tests

A. AFB Smear Microscopy (Ziehl-Neelsen / fluorescence stain)

• Collect at least 2 sputum specimens (US: 3 specimens are routine)
• Specimens include: early morning sputum, spot sputum, induced sputum, BAL (if unable to produce)
• Sensitivity: up to 90% with 3 specimens; however, often lower in paucibacillary disease
• AFB smear cannot distinguish M. tuberculosis from non-tuberculous mycobacteria (NTM)
• Remains the most widely available test globally

B. Mycobacterial Culture

• More sensitive than smear (~100% when adequate sputum is provided)
• Results take 2-8 weeks (liquid culture faster than solid media)
• Definitive identification of M. tuberculosis complex
• Also allows drug susceptibility testing (DST)
• Requires specialized lab infrastructure; often unavailable in resource-limited settings
• Textbook of Family Medicine, 9e, p. 763

C. Nucleic Acid Amplification Testing (NAAT) - Xpert MTB/RIF

• Xpert MTB/RIF (GeneXpert): Self-contained, automated, results in ~90 minutes
• Simultaneously detects M. tuberculosis complex AND rifampicin resistance (as a proxy for MDR-TB)
• Distinguishes M. tuberculosis from NTM (unlike smear)
• Interpretation:
  • AFB smear-positive + NAAT-positive = rapid confirmation of TB
  • AFB smear-negative + NAAT-positive = presumptive evidence in intermediate/high-probability cases
  • NAAT-negative does NOT exclude active TB (insufficient sensitivity to replace culture)
• WHO recommendation: Preferred initial test for all adults/children with suspected TB where feasible, especially for HIV-positive patients and those at risk of MDR-TB
• US recommendation: Use as adjunct to smear + culture
• Murray & Nadel's Textbook of Respiratory Medicine, p. 1189-1193

D. Line Probe Assays (LPAs)

• Molecular test detecting resistance to isoniazid, rifampin, fluoroquinolones, and injectables
• Useful for rapid MDR-TB and XDR-TB screening
• Applied to culture isolates or directly to smear-positive specimens

4. Tests for TB Infection (Latent TB)

A. Tuberculin Skin Test (TST / Mantoux)

• Intradermal injection of PPD (purified protein derivative)
• Read at 48-72 hours; induration measured in mm
• Cutoffs:
  • ≥5 mm: HIV-infected, close contacts, immunosuppressed, CXR with fibrotic changes
  • ≥10 mm: Recent immigrants, IV drug users, high-risk occupational groups, children <4 yrs
  • ≥15 mm: No known risk factors
• Limitations:
  • False-negative: immunosuppression, overwhelming TB, very early infection
  • False-positive: prior BCG vaccination, NTM infection
  • Low specificity due to cross-reactivity with conserved mycobacterial proteins
  • Operator-dependent; batch variations

B. Interferon-Gamma Release Assays (IGRAs)

• Measure T-cell IFN-γ release in response to TB-specific antigens ESAT-6 and CFP-10 (RD1-encoded, absent from BCG and most NTMs)
• Types:
  • QuantiFERON-TB Gold Plus (QFT-Plus): Whole-blood ELISA; measures both CD4+ and CD8+ T-cell responses
  • T-SPOT.TB (Oxford Immunotec): ELISPOT assay
• Advantages over TST: Higher specificity; not affected by BCG vaccination; no booster effect; single visit
• Limitations: False-negatives in immunosuppressed; higher cost; need lab processing

C. New Antigen-Based Skin Tests (TBSTs)

• Use ESAT-6 and CFP-10 antigens (same as IGRAs) but in skin-test format
• WHO assessment: accuracy similar to IGRAs and greater than TST
• Useful also for HIV-positive individuals, children, and BCG-vaccinated persons
• Harrison's Principles of Internal Medicine 22E (2025), p. 1437

5. Non-Sputum-Based Tests

Urine LAM (Lipoarabinomannan) Antigen Test

• Detects LAM - a mycobacterial cell-wall component - in urine
• Particularly useful when sputum cannot be produced
• Best performance in: HIV-positive patients with low CD4 counts (<200 cells/µL)
  • Pooled sensitivity 42-62% (higher in hospitalized, symptomatic, low-CD4 patients)
  • Pooled specificity ~91%
• WHO recommendation: Recommended for all HIV-positive inpatients with TB symptoms, regardless of CD4 count; and for outpatients with CD4 <100 cells/µL
• Not currently available in the US
• Murray & Nadel's Textbook of Respiratory Medicine, p. 1193

6. Additional Investigations

7. Summary Algorithm

Clinical suspicion (symptoms + risk factors)
        ↓
Chest X-ray
        ↓
Sputum × 2-3 for: AFB smear + mycobacterial culture + NAAT (Xpert MTB/RIF)
        ↓
[If smear-negative + high suspicion] → CT chest, bronchoscopy/BAL, induced sputum
        ↓
[Infection screening (no active disease suspected)] → TST or IGRA
        ↓
[HIV-positive / cannot produce sputum] → Urine LAM + Xpert
        ↓
Definitive diagnosis = positive culture OR positive NAAT (with clinical correlation)

• Smear alone cannot confirm TB species - NAAT distinguishes M. tuberculosis from NTM
• A single negative NAAT does not rule out active TB - culture remains the gold standard
• TST and IGRAs detect infection, not active disease
• Radiographs are suggestive but never diagnostic; microbiologic confirmation is always needed
• In HIV, expect atypical presentations and use urine LAM + Xpert as initial tests