GeneXpert MTB/RIF on Stool Specimens for Tuberculosis Diagnosis
Stool GeneXpert (Xpert MTB/RIF Ultra preferred) should be used as a diagnostic tool for pediatric tuberculosis, particularly in children under 15 years who cannot produce respiratory specimens, with collection of approximately 0.6-1 gram of stool processed using centrifuge-free methods like Simple-One-Step (SOS) or Optimized Sucrose Flotation (OSF).
When to Use Stool GeneXpert
Primary Indication
- Pediatric patients (<15 years) with presumptive pulmonary tuberculosis who cannot produce sputum, gastric aspirates, or other respiratory specimens 1, 2
- Children under 5 years show equivalent performance between stool and respiratory specimens (36.4% positivity rate for both) 3
Clinical Context
While existing guidelines 4, 5 focus on respiratory specimens for NAA testing, they do not address stool specimens. However, recent high-quality evidence demonstrates stool's utility in pediatric populations where respiratory sampling is challenging.
Collection and Processing
Specimen Collection
- Collect 0.6-1.0 gram of stool in a sterile container 6, 2
- 0.6g samples performed better than larger volumes (1.2g) in validation studies 6
- Stool swabs in PrimeStore MTM Molecular Transport Medium are equivalent to direct stool processing and eliminate cold chain requirements 7
Processing Methods (Centrifuge-Free Preferred)
Recommended approaches:
Simple-One-Step (SOS) method - Easiest method, requires minimal manipulation and no additional reagents, preferred by 6 of 7 laboratory operators 2
Optimized Sucrose Flotation (OSF) method - WHO-endorsed alternative 2
PrimeStore MTM with stool swabs - Inactivates MTB, stabilizes DNA without cold chain, increases access in underserved areas 7
Critical advantage: Centrifuge-free methods (SOS, OSF) improve sensitivity to 77% compared to 61% for centrifuge-requiring methods, while maintaining >96% specificity 1, 2. This is crucial for LMIC implementation where centrifuges may be unavailable.
Diagnostic Performance
Sensitivity and Specificity
- Xpert Ultra on stool: 73% sensitivity (95% CI: 63-81%) against bacteriological confirmation from respiratory samples 1
- Specificity: Consistently >93-100% across studies 1, 8, 2
- Additionality: Stool Xpert Ultra increases bacteriological confirmation by 38.6% overall, even when respiratory specimens are also tested 1
Performance by Age
- Children <5 years: Equivalent performance to respiratory specimens 3
- Analytical limit of detection: 1,000 CFU/gram of stool 6
Interpretation
Positive Result
- Treat as confirmed tuberculosis - Specificity >93% means false positives are rare
- Check for rifampicin resistance detection (probe mutations, particularly probe E covering codons 529-533, most common at 40.64%) 9
- HIV-positive patients and previously treated patients have higher risk of rifampicin resistance 9
Negative Result
- Does not exclude tuberculosis - Sensitivity is only 73% with Xpert Ultra
- Consider clinical diagnosis based on symptoms, imaging, and tuberculosis scoring systems
- Attempt respiratory specimen collection if feasible
- In high-risk patients (HIV-positive, previous TB treatment, disseminated disease), strongly consider empiric treatment despite negative stool test 3, 9
Critical Caveats
Limited utility for isolated intestinal TB - Stool Xpert showed poor performance for intestinal-only tuberculosis 3
Not a standalone test - Use as supplementary tool alongside clinical assessment, particularly given 73% sensitivity 3
Age considerations - Performance equivalent to respiratory specimens only in children <5 years; older children may benefit more from respiratory sampling if obtainable 3
HIV co-infection - HIV-positive patients have nearly twice the odds of rifampicin resistance (AOR 1.91); positive stool results should prompt immediate drug susceptibility testing 9
Processing quality matters - Peripheral laboratory performance (63% sensitivity) was significantly lower than central reference laboratory (84% sensitivity), emphasizing need for proper training and quality control 8