Infectious Disease · Medical Literature
Accuracy and safety of C-TST for tuberculosis infection: A clinical trial in Brazil.
Abstract
Tuberculosis infection (TBI) diagnosis in BCG-vaccinated, high-burden settings remains challenging. We aimed to assess the sensitivity, specificity, and safety of the C-tuberculin skin test (C-TST), compared with PPD RT-23, for tuberculosis infection in BCG-vaccinated adults. Randomised, double-blind clinical trial in Brazil. Adults with pulmonary tuberculosis (PTB) and asymptomatic controls without known TB exposure or prior TB/TPT underwent C-TST, PPD RT-23, and QuantiFERON-TB Gold Plus (QFT-Plus). Skin tests were read by two strategies: (a) induration only or (b) induration or erythema, at a ≥5-mm cut-off. Sensitivity was assessed in PTB; specificity for C-TST and PPD RT-23 was estimated against QFT-Plus. Adverse events were monitored. Among 446 participants (141 PTB; 305 controls), at the 5-mm cut-off, C-TST showed lower sensitivity but higher specificity than PPD RT-23. With strategy (a) and (b), C-TST sensitivity was 0.67 and 0.68, specificity was 0.92 and 0.9, compared to 0.74 sensitivity and 0.78 specificity for PPD RT-23. Adverse events were more frequent in PTB: 12.1% (C-TST) and 6.4% (PPD RT-23); none were serious. In BCG‑vaccinated adults, C-TST showed a trade-off of higher specificity and lower sensitivity than PPD RT-23 at the 5-mm cut-off; and a favourable safety profile. Adding erythema to readings did not improve accuracy, suggesting that C-TST may be a useful alternative for programmatic TBI screening in high-burden settings. Trial registry number RBR-7tn2ysw (https://ensaiosclinicos.gov.br/ registered on 25 January 2021).
Clinical relevance
In BCG-vaccinated adults in Brazil, does the C-tuberculin skin test (C-TST, an ESAT-6/CFP-10-based antigen-specific skin test) demonstrate acceptable sensitivity and specificity for tuberculosis infection diagnosis compared with PPD RT-23 tuberculin skin test, using QuantiFERON-TB Gold Plus (QFT-Plus) as the reference standard for specificity? __SNAPSHOT__ {"studyDesign":"Randomized, double-blind clinical trial (diagnostic accuracy study)","setting":"Multi-center (eight health centers across different regions of Brazil); conducted July 2023 – September 2024","sampleSize":"n=446 total (TB group: n=141 microbiologically confirmed pulmonary TB; control group: n=305 asymptomatic adults without known TB exposure or prior TB/TPT)","population":"BCG-vaccinated adults in Brazil; TB group comprised microbiologically confirmed pulmonary tuberculosis patients (used to estimate sensitivity); control group comprised healthy, asymptomatic adults without known TB exposure, prior TB, or tuberculosis preventive therapy (used to estimate specificity against QFT-Plus reference standard)","intervention":"C-TST (ESAT-6/CFP-10 recombinant fusion protein antigen-based skin test, Anhui Zhifei Longcom), administered intradermally in one arm; read at 48–96 h using two pre-specified strategies: (a) induration only, or (b) induration or erythema, at a ≥5 mm cut-off","comparator":"PPD RT-23 tuberculin skin test (standard TST), administered intradermally in the opposite arm (randomized right–left allocation); QFT-Plus served as reference standard for specificity estimation in controls","primaryOutcome":"Sensitivity (in PTB group) and specificity (in control group, referenced against QFT-Plus) of C-TST versus PPD RT-23 at a ≥5 mm cut-off, assessed by two reading strategies (induration only; induration or erythema); safety assessed by active adverse event monitoring","followUp":"Skin test reading at 48–96 hours post-application; adverse event monitoring duration not reported in the available abstract","fundingSource":"Not reported in the available abstract; search results indicate CNPq (Brazilian national research council) and Stop TB Partnership funding for related Brazilian TB diagnostics work; specific funding for this trial not confirmed from abstract alone"}
Key findings
At the prespecified ≥5 mm cut-off, C-TST demonstrated lower sensitivity but higher specificity than PPD RT-23. Using strategy (a) (induration only), C-TST sensitivity was 0.67 and specificity was 0.92; using strategy (b) (induration or erythema), sensitivity was 0.68 and specificity was 0.90. By contrast, PPD RT-23 sensitivity was 0.74 and specificity was 0.78 (strategy not explicitly differentiated for PPD RT-23 in the abstract). The absolute difference in sensitivity favored PPD RT-23 by approximately 6–7 percentage points; the absolute difference in specificity favored C-TST by approximately 12–14 percentage points. Confidence intervals and p-values for these diagnostic performance estimates were not reported in the available abstract. NNT and NNH are not calculable from the reported data. Adverse events (AEs) were more frequent in the PTB group than in controls. In the PTB group, AE rates were 12.1% for C-TST and 6.4% for PPD RT-23; no serious adverse events were reported in either arm. AE data for the control group were not reported in the available abstract. Calculated: the absolute difference in AE rates between C-TST and PPD RT-23 in the PTB group was 5.7 percentage points (calculated).
Practice implications
This trial provides evidence that C-TST achieves higher specificity than PPD RT-23 in BCG-vaccinated Brazilian adults—a clinically meaningful finding in a setting with near-universal neonatal BCG vaccination, where PPD-based TST is prone to false positives. The specificity gain (~12–14 percentage points) is potentially important for reducing unnecessary tuberculosis preventive therapy in programmatic settings. However, C-TST's lower sensitivity (~6–7 percentage points below PPD RT-23) raises concern about missed TBI diagnoses, particularly in high-burden contexts where under-treatment carries substantial public health risk. Sensitivity was estimated using microbiologically confirmed active pulmonary TB as a proxy—a standard but imperfect approach, since TBI (latent infection) cannot be microbiologically confirmed and active TB patients may not be representative of those with latent infection. The clinical importance of the sensitivity gap depends on intended use: for screening asymptomatic high-risk contacts where specificity matters most, C-TST's profile may be advantageous; for ruling out TBI in immunocompromised individuals where sensitivity is paramount, the lower sensitivity is a more significant limitation. Patients with HIV co-infection, children, and immunocompromised individuals were not adequately characterized in the abstract and may not be well represented. Safety appeared acceptable, with no serious AEs; the modestly higher local AE rate with C-TST (12.1% vs. 6.4% in PTB group) appears clinically manageable. CRITICAL LIMITATION: Only the truncated abstract was available for this Pulmonology citation (PMID 42467493); confidence intervals, p-values, subgroup analyses (e.g., by HIV status, age), and complete AE data from the control arm are not reported here and are essential for full interpretation. Findings should not be generalized beyond BCG-vaccinated adult populations in high-burden settings similar to Brazil. __LIMITATIONS__ Only a truncated abstract was available; confidence intervals, p-values, and complete safety data (especially for the control group) are not reported, limiting precise assessment of statistical significance and precision of estimates.,Sensitivity was estimated using microbiologically confirmed active pulmonary TB as a proxy for TB infection—a methodological necessity but one that may overestimate sensitivity for latent TBI detection, as immune responses may differ between active and latent disease.,Specificity was estimated against QFT-Plus as reference standard, not a microbiological gold standard; QFT-Plus itself has imperfect sensitivity and specificity, introducing reference standard bias.,Asymptomatic controls were defined by absence of known TB exposure and negative QFT-Plus, but subclinical or unrecognized TB infection cannot be fully excluded, potentially inflating apparent specificity.,The study population was restricted to BCG-vaccinated adults; generalizability to non-BCG-vaccinated populations, children, people living with HIV, and immunocompromised patients is not established from this abstract.,Adverse event follow-up duration was not reported; delayed or systemic reactions beyond the skin test reading window may not have been captured.,The two reading strategies (induration only vs. induration or erythema) produced only marginally different results for C-TST (sensitivity 0.67 vs. 0.68), suggesting limited added value of erythema assessment, but the clinical and operational implications of including erythema in programmatic reading require further study.,Industry sponsorship status not reported; potential for sponsorship bias cannot be assessed.,Multi-center design across diverse Brazilian regions strengthens external validity within Brazil but may introduce inter-rater variability in skin test reading, the extent of which is not reported in the abstract.,No comparison with QFT-Plus sensitivity in the same TB group was reported, preventing a full three-way diagnostic accuracy comparison. __BOTTOM_LINE_CATEGORY__ potentially_practice_informing __BOTTOM_LINE__ In BCG-vaccinated Brazilian adults, C-TST offers meaningfully higher specificity than PPD RT-23 (~0.90–0.92 vs. ~0.78) at a ≥5 mm cut-off, with a modest but clinically relevant reduction in sensitivity (~0.67–0.68 vs. ~0.74), and an acceptable safety profile with no serious adverse events. These findings support C-TST as a potentially superior alternative to PPD RT-23 for TBI screening in high-burden, BCG-vaccinated populations where reducing false positives and unnecessary preventive therapy is a priority. Full publication with confidence intervals, subgroup analyses, and complete safety data is needed before clinical adoption decisions are made.