Advantages and disadvantages of possible end-point definitions for tuberculosis (TB) vaccine efficacy trials

 IGRA (QTF) conversion  (>0.35 IU·mL−1)∼10-fold more frequent than TB disease
Widely used threshold in routine practice
Higher sensitivity for detecting MTB infection compared to higher thresholds
Does not detect all true MTB infections
IGRA reversion is common, but of unclear clinical significance
Significance of protection unclear
Test performance in PLHIV, people with immune-mediated inflammatory diseases taking immunosuppressive treatment, and very young children is unclear
 IGRA (QTF) conversion  (>4.0 IU·mL−1)Higher risk of progression to TB disease than conversion at manufacturer threshold
Higher specificity for detecting MTB infection than manufacturer threshold
Less frequent event than conversion at manufacturer threshold
Lower sensitivity for detecting MTB infection than manufacturer threshold
 Sustained IGRA  conversion (6 months)Might represent sustained (persistent) MTB infection
Might be associated with higher risk of TB disease compared to a single conversion
Lower risk of false-positive result compared to a single test
Less frequent event than initial IGRA conversion
Does not encompass IGRA conversion–reversion–conversion events
Need for TB preventive therapy precludes nested POI in POD efficacy trial design
 MTB liquid culture  (sputum)Gold standard (most sensitive) tool
Allows genotyping of MTB strain
Need for central laboratory
Variable contamination rate (largely laboratory-dependent)
Lower yield in pauci-bacillary TB disease:
- HIV-associated TB
- childhood TB
  - One sampleLogistically simple
Lower sensitivity
Potential false-positive results
  - Two or more separate samples (processed independently)Higher specificity (if both need to be positive)Logistically complex
Less frequent than single positive sample
Decreased sensitivity (if both need to be positive), especially in pauci-bacillary TB disease:
- HIV-associated TB
- childhood TB
  - Before treatment of the TB episode startsNot affected by effect of vaccination on response to TB therapyLogistically complex
Lower sensitivity than before or after treatment starts
 Xpert Ultra (sputum)Does not need central laboratory
Available at district-level hospitals
Rapid turnaround
Does not allow to genotype MTB strain
Lower sensitivity and specificity than culture (higher false-positive rate, especially among trace results)
 MTB liquid culture OR Xpert  Ultra (sputum)
  - With symptomsProtection associated with direct health benefitNo opportunity to assess vaccine efficacy against subclinical disease
  - Without symptomsAllows assessment of vaccine efficacy against subclinical TB diseaseSignificance of clinical protection unclear
Need for TB treatment precludes subsequent assessment of vaccine efficacy against symptomatic TB disease
 Digital chest radiograph  (with or without CAD)High sensitivity for TB, widely available and with high added value in populations with paucibacillary disease, such as:
- HIV-associated TB
- childhood TB
Limited specificity
 Urine LAMHigh specificity in PLHIV with low CD4 countsLimited sensitivity with increasing levels of CD4 counts
 TB symptoms (clinical  diagnosis)High sensitivity, especially in pulmonary TB among HIV-negative individuals, cheap, no laboratory infrastructure needed
Sometimes used to define unconfirmed TB
Limited specificity
Subjective interpretation of symptoms
Prevention of recurrence/therapeutic
M. tuberculosis liquid culture  (sputum)Gold standard
Allows genotyping of MTB strain (true relapse versus reinfection) if collected before treatment
Need for central laboratory
 Xpert Ultra (sputum)Logistically simple at treatment startCannot distinguish viable from non-viable MTB bacilli
No opportunity to genotype MTB strain

CAD: computer-aided detection; IGRA: interferon gamma release assay; LAM: lipoarabinomannan MTB: Mycobacterium tuberculosis; PLHIV: people living with HIV; POD: prevention of disease; POI: prevention of infection; QFT: QuantiFERON-TB.