Nucleic Acid Amplification Testing for Tuberculosis Diagnosis

Specimen Selection and Testing

The American Thoracic Society, Infectious Diseases Society of America, and Centers for Disease Control and Prevention recommend performing nucleic acid amplification testing (NAAT) on respiratory specimens, not tissue biopsies, as the standard diagnostic approach for tuberculosis diagnosis, with established sensitivity of 96% in AFB smear-positive cases and 66% in smear-negative cases 1, 2
The Xpert MTB/RIF assay demonstrates 85% sensitivity and 98% specificity when performed on respiratory specimens, according to the Clinical Infectious Diseases journal 2, 3
For respiratory specimens, NAAT sensitivity is 96-97% and specificity is 85-95% in AFB smear-positive specimens, and 66-79% sensitivity and 80-98% specificity in AFB smear-negative specimens 1, 4

When AFB smear is positive and NAAT is positive, presume tuberculosis and initiate treatment immediately, with a positive predictive value of >95% 4, 5
When AFB smear is negative and NAAT is positive, use clinical judgment and consider testing additional specimens to confirm, with two or more positive NAAT results allowing presumptive tuberculosis diagnosis 4, 5

Culture remains the gold standard and must always be performed regardless of NAAT results, with both liquid and solid cultures recommended for all specimens 1, 4, 6
Sufficient specimen volume must be reserved for culture before NAAT testing, and NAAT results should be available within 48 hours of specimen collection 4, 6

A single negative NAAT cannot definitively exclude tuberculosis, especially with moderate-to-high clinical suspicion, and NAAT results must be interpreted within the clinical context, not in isolation 4, 6
Positive NAAT results should be treated as critical values and reported immediately to clinicians and public health authorities, with a turnaround time of 48 hours or less 4, 6

The presence of inhibitors can cause false-negative results; if AFB smear is positive but Xpert is negative, evaluate for inhibitors and consider collecting a new specimen, as recommended by the Clinical Infectious Diseases guideline 7

Xpert MTB/RIF provides rifampicin resistance results simultaneously with TB detection, although the American Thoracic Society and other guideline societies recommend that all molecular resistance results must be confirmed by phenotypic culture-based drug susceptibility testing, with Xpert MTB/RIF showing 96% sensitivity and 98% specificity for resistance detection 7

In patients with suspected miliary tuberculosis, CB‑NAAT (GeneXpert MTB/RIF) yields a positive result in only ≈ 50‑60 % of cases, markedly lower than the ≈ 85‑96 % positivity observed in AFB smear‑positive pulmonary tuberculosis. 8
A negative CB‑NAAT result cannot exclude miliary tuberculosis; the false‑negative rate ranges from 40 % to 50 % in this population. [8][9]
For extrapulmonary specimens relevant to disseminated disease (e.g., pleural fluid, cerebrospinal fluid), CB‑NAAT sensitivity is 56‑62 % with a high specificity of ≈ 98 %. [8][9]
A positive CB‑NAAT from any site provides > 95 % specificity, offering strong diagnostic confirmation and justifying immediate treatment initiation. [8][9]
Mycobacterial culture remains essential: sputum culture detects ≈ 76 % of miliary TB cases, outperforming CB‑NAAT on respiratory specimens. 8

Positive CB‑NAAT (any specimen) → presume miliary tuberculosis and start anti‑TB therapy promptly. 9
All CB‑NAAT negative but high clinical suspicion → do not withhold therapy; commence empiric treatment while awaiting culture results. [8][9]
Negative CB‑NAAT results are common (38‑44 % false‑negative) and must not delay treatment when clinical suspicion is strong. [8][9]

Specimen strategy: obtain multiple specimen types (e.g., ≥ 3 sputum samples, gastric aspirate, urine, blood, and relevant extrapulmonary fluids) and submit each for both CB‑NAAT and mycobacterial culture; reliance on CB‑NAAT alone is insufficient. 8
Culture priority: despite CB‑NAAT testing, culture remains the gold standard and should be performed on all collected specimens, recognizing a turnaround of 2‑6 weeks.
Treatment decision rule: a single positive CB‑NAAT result warrants immediate initiation of anti‑TB therapy; a completely negative CB‑NAAT panel does not preclude treatment if clinical assessment indicates probable miliary disease.

These recommendations are derived from clinical evidence reported in Clinical Infectious Diseases (2017).