Browse the corpus

BCG Revaccination for the Prevention of Mycobacterium tuberculosis Infection. BACKGROUND: In a previous phase 2 trial, bacille Calmette-Guérin (BCG) revaccination was not shown to provide protection from primary Mycobacterium tuberculosis infection but prevented sustained M. tuberculosis infection, defined by an initial conversion on a QuantiFERON-TB (QFT) test (an interferon-γ release assay) from negative to positive, followed by two additional positive QFT tests at 3 and 6 months after the initial conversion (a secondary end point). A vaccine efficacy of 45% (95% confidence interval [CI], 6 to 68) was observed. METHODS: We performed a phase 2b, double-blind, randomized, placebo-controlled trial to evaluate the efficacy of BCG revaccination, as compared with placebo, for the prevention of sustained QFT test conversion (primary end point) in QFT test-negative, human immunodeficiency virus (HIV)-negative adolescents. Adverse events were assessed in a secondary analysis, and immunogenicity was assessed in an exploratory analysis. Vaccine efficacy was evaluated in the modified intention-to-treat population, which included all the participants who had undergone randomization, received the BCG vaccine or placebo, and had a negative QFT test 10 weeks after receipt of BCG vaccine or placebo; the last criterion was added to exclude participants with M. tuberculosis infection around the time that the vaccine or placebo was administered. Hazard ratios and 95% confidence intervals were estimated from a stratified Cox proportional-hazards model. RESULTS: A total of 1836 participants underwent randomization; 918 received the BCG vaccine, and 917 received placebo. After a median 30 months of follow-up, a sustained QFT test conversion was observed in 62 of 871 participants in the BCG-vaccine group and 59 of 849 participants in the placebo group. The hazard ratio for a sustained QFT test conversion (BCG vaccine vs. placebo) was 1.04 (95% CI, 0.73 to 1.48), for a vaccine efficacy point estimate of -3.8% (95% CI, -48.3 to 27.4). Adverse events occurred more frequently in the BCG-vaccine group than in the placebo group, and most were due to injection-site reactions (pain, redness, swelling, and ulceration). BCG revaccination induced cytokine-positive type 1 helper CD4 T cells. CONCLUSIONS: BCG revaccination in QFT-test negative, HIV-negative adolescents did not provide protection from sustained M. tuberculosis infection. (Funded by the Gates Foundation; ClinicalTrials.gov number NCT04152161.).

This randomized, placebo-controlled, observer-blind, phase 2b trial enrolled participants at 5 sites in South Africa: Worcester, Cape Town, and Mbekweni in Western Cape; Durban in KwaZulu-Natal; and Johannesburg in Gauteng Province. The primary objective was to demonstrate the efficacy of BCG revaccination against sustained Mtb infection versus placebo in QuantiFERON®-TB-Gold-Plus assay (QFT)-negative, healthy adolescents. Secondary objectives were to evaluate the durability of efficacy, and safety and reactogenicity of BCG revaccination. Exploratory objectives included the immunogenicity of BCG revaccination, VE based on alternative definitions of QFT conversion, and incident cases of laboratory-confirmed TB. Participants were randomized 1:1 to receive BCG (BCG Danish 1331 vaccine, AJ Vaccines, Copenhagen, Denmark) or placebo intradermally (Figure 1). Since BCG does not include antigens used in the QFT assay, revaccination is not associated with QFT conversion and negative QFT result at Day (D) 71 could be used to define the mITT population for VE assessment. Randomization was stratified by age, sex, trial site, and school cluster (Worcester site only). BCG was administered as per product package insert for South Africa (0.1mL intradermally). Randomization to BCG or placebo was assigned using a validated Interactive Voice/Web Response System. The trial was observer-blinded until the primary endpoint analyses were completed. Inadvertent unblinding occurred due to recognizable lesion formed at the BCG injection site.

rt for South Africa (0.1mL intradermally). Randomization to BCG or placebo was assigned using a validated Interactive Voice/Web Response System. The trial was observer-blinded until the primary endpoint analyses were completed. Inadvertent unblinding occurred due to recognizable lesion formed at the BCG injection site. Eligible participants were ≥10 and ≤18 years old and tested negative for HIV and QFT. Detailed eligibility criteria are described in the Supplement and protocol at nejm.org. The primary endpoint was sustained QFT conversion based on positive QFT results using the manufacturer’s (Qiagen®, Germany) assay threshold of 0.35 IU/mL IFNγ. Sustained Mtb infection was defined as sustained QFT conversion from a negative to a positive test, with initial conversion at any time after a first negative QFT post randomization, and remaining QFT-positive at 3- and 6-months post conversion. Figure 1A shows the sampling schedule. Secondary endpoints of safety and reactogenicity of BCG revaccination were evaluated based on the adverse events (AEs) reported as described in the supplement. Exploratory efficacy endpoints included initial QFT conversion using various IFN-γ thresholds ranging from 0.36 IU/mL to 10 IU/mL.

The primary endpoint was sustained QFT conversion based on positive QFT results using the manufacturer’s (Qiagen®, Germany) assay threshold of 0.35 IU/mL IFNγ. Sustained Mtb infection was defined as sustained QFT conversion from a negative to a positive test, with initial conversion at any time after a first negative QFT post randomization, and remaining QFT-positive at 3- and 6-months post conversion. Figure 1A shows the sampling schedule. Secondary endpoints of safety and reactogenicity of BCG revaccination were evaluated based on the adverse events (AEs) reported as described in the supplement. Exploratory efficacy endpoints included initial QFT conversion using various IFN-γ thresholds ranging from 0.36 IU/mL to 10 IU/mL. The first 80 10-12-year-old participants randomized at the Worcester site were enrolled in the immunogenicity sub-cohort since this younger age group had not previously been evaluated for immunogenicity. Immunogenicity was evaluated as the frequency of BCG-specific CD4 or CD8 T cells expressing one or more of the following cytokines: IFN-γ, tumor necrosis factor (TNF), interleukin (IL)-2, IL-17 and/or IL-22, by whole blood intracellular cytokine staining (WB-ICS) assay.

not previously been evaluated for immunogenicity. Immunogenicity was evaluated as the frequency of BCG-specific CD4 or CD8 T cells expressing one or more of the following cytokines: IFN-γ, tumor necrosis factor (TNF), interleukin (IL)-2, IL-17 and/or IL-22, by whole blood intracellular cytokine staining (WB-ICS) assay. An independent data monitoring committee reviewed unblinded safety data every three months during enrollment and every six months thereafter, as well as the outcomes of the primary analyses. The trial was conducted in accordance with the International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use (ICH) Good Clinical Practice (GCP) Guidelines, SAHPRA (South African Health Products Regulatory Authority) Regulations, the IRB/IEC, and other applicable country and local requirements. All authors contributed to data collection, data analysis, reviewed the manuscript, and vouch for the accuracy and completeness of the data presented. The trial was designed to provide 90% power with a 1-sided alpha of 2.5% for the primary endpoint analysis. Assuming a true VE of 45%, at least 118 sustained QFT conversion events were required to demonstrate VE with a lower bound of zero, where VE is calculated as 1 – HR(BCG/Placebo).

An independent data monitoring committee reviewed unblinded safety data every three months during enrollment and every six months thereafter, as well as the outcomes of the primary analyses. The trial was conducted in accordance with the International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use (ICH) Good Clinical Practice (GCP) Guidelines, SAHPRA (South African Health Products Regulatory Authority) Regulations, the IRB/IEC, and other applicable country and local requirements. All authors contributed to data collection, data analysis, reviewed the manuscript, and vouch for the accuracy and completeness of the data presented. The trial was designed to provide 90% power with a 1-sided alpha of 2.5% for the primary endpoint analysis. Assuming a true VE of 45%, at least 118 sustained QFT conversion events were required to demonstrate VE with a lower bound of zero, where VE is calculated as 1 – HR(BCG/Placebo). The modified intention-to-treat (mITT) population included all participants who received the trial treatment and were QFT negative at the D71 visit, or at the first trial visit post D71 for which a QFT result was available. The safety population included all participants who received the trial treatment. Frequency (n) and percentages (%) were used to summarize categorical variables; mean, standard deviation (SD), median, minimum, and maximum were used to summarize continuous variables.

visit post D71 for which a QFT result was available. The safety population included all participants who received the trial treatment. Frequency (n) and percentages (%) were used to summarize categorical variables; mean, standard deviation (SD), median, minimum, and maximum were used to summarize continuous variables. The primary efficacy endpoint was analyzed using a log-rank test, stratified by sex and age group (10-11 years old, 12-14 years old, and >14 years old) to evaluate differences in distributions of event times between the BCG revaccination and placebo groups. AEs in different categories were summarized by treatment groups and the 95% CI were calculated for a single proportion using the mid-p binomial option. AEs were coded using Medical Dictionary for Regulatory Activities (MedDRA) version 26.0 or higher. AEs were graded using the Division of AIDS Table for Grading for Severity of Adult and Pediatric Adverse Events Version 2.1, July 20179. The percentage of missing data was low, with a discontinuation rate of 5.1% and 3.9% for the BCG and placebo groups, respectively (including 1.0% and 0.2% lost to follow up in the BCG and placebo groups and 2.5% participant withdrawal in each group). Given that the missingness was minimal and did not appear to be dependent on known factors, we assume the data to be missing completely at random and therefore, only the observed data were used in analyses.

An independent data monitoring committee reviewed unblinded safety data every three months during enrollment and every six months thereafter, as well as the outcomes of the primary analyses. The trial was conducted in accordance with the International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use (ICH) Good Clinical Practice (GCP) Guidelines, SAHPRA (South African Health Products Regulatory Authority) Regulations, the IRB/IEC, and other applicable country and local requirements. All authors contributed to data collection, data analysis, reviewed the manuscript, and vouch for the accuracy and completeness of the data presented.

The trial was designed to provide 90% power with a 1-sided alpha of 2.5% for the primary endpoint analysis. Assuming a true VE of 45%, at least 118 sustained QFT conversion events were required to demonstrate VE with a lower bound of zero, where VE is calculated as 1 – HR(BCG/Placebo). The modified intention-to-treat (mITT) population included all participants who received the trial treatment and were QFT negative at the D71 visit, or at the first trial visit post D71 for which a QFT result was available. The safety population included all participants who received the trial treatment. Frequency (n) and percentages (%) were used to summarize categorical variables; mean, standard deviation (SD), median, minimum, and maximum were used to summarize continuous variables. The primary efficacy endpoint was analyzed using a log-rank test, stratified by sex and age group (10-11 years old, 12-14 years old, and >14 years old) to evaluate differences in distributions of event times between the BCG revaccination and placebo groups. AEs in different categories were summarized by treatment groups and the 95% CI were calculated for a single proportion using the mid-p binomial option. AEs were coded using Medical Dictionary for Regulatory Activities (MedDRA) version 26.0 or higher. AEs were graded using the Division of AIDS Table for Grading for Severity of Adult and Pediatric Adverse Events Version 2.1, July 20179.

the 95% CI were calculated for a single proportion using the mid-p binomial option. AEs were coded using Medical Dictionary for Regulatory Activities (MedDRA) version 26.0 or higher. AEs were graded using the Division of AIDS Table for Grading for Severity of Adult and Pediatric Adverse Events Version 2.1, July 20179. The percentage of missing data was low, with a discontinuation rate of 5.1% and 3.9% for the BCG and placebo groups, respectively (including 1.0% and 0.2% lost to follow up in the BCG and placebo groups and 2.5% participant withdrawal in each group). Given that the missingness was minimal and did not appear to be dependent on known factors, we assume the data to be missing completely at random and therefore, only the observed data were used in analyses.

From 16 Oct 2019 to 22 July 2021, 3653 participants were screened and 1836 were randomized 1:1 to BCG and placebo groups (Figure 1B). Table S1 shows the representativeness of trial participants. At primary analysis (data cutoff: June 30, 2023), 1752 participants were in trial follow-up and 83 (4.5%) discontinued. Among the 3653 screened participants, 307 had no blood drawn, 12 had indeterminate QFT results, and 1 had invalid results, and of the remaining 3333 participants with screening QFT results available, 1219 (36.6%) were QFT positive (Table S2). QFT positivity rates at screening were higher at sites in the Western Cape (37.8%-39.6%) than in KwaZulu-Natal (29.1%) or Gauteng (23.6%) provinces. Demographics and baseline characteristics were comparable between the two groups (Table S3). The median age was 13 years (range, 10-18) and most participants (1464 [79.8%]) self-identified as Black African.

ere higher at sites in the Western Cape (37.8%-39.6%) than in KwaZulu-Natal (29.1%) or Gauteng (23.6%) provinces. Demographics and baseline characteristics were comparable between the two groups (Table S3). The median age was 13 years (range, 10-18) and most participants (1464 [79.8%]) self-identified as Black African. The mITT population included 871 and 849 participants in the BCG and placebo groups, respectively (Figure 1B). Thirty-eight (4.1%) and 58 (6.3%) participants were excluded from the mITT population due to positive QFT at D71; 9 (1.0%) and 8 (0.9%) were excluded due to missing D71 and no negative result at the first visit after D71; and 0 and 2 (0.2%) were excluded due to indeterminate results at D71. At primary analysis after a median of 30 months (interquartile range, 25 to 33) of follow up, sustained QFT conversion was observed in 62 participants in the BCG group and 59 in the placebo group (one-sided P=0.58; Table 1, Figure 2A & 2B). The sustained QFT conversion rate (95% CI) was 7.1% (5.4 to 8.8) in the BCG group and 7.0% (5.2 to 8.7) in the placebo group. The overall sustained conversion incidence rate and the incidence rate over time were similar in both groups (Table 1, Figure 2A & 2B). The VE point estimate for the primary endpoint was −0.038 (95% CI, −0.483 to 0.274) (Table 1). Site specific VE point estimates are listed in Table S4.

.2 to 8.7) in the placebo group. The overall sustained conversion incidence rate and the incidence rate over time were similar in both groups (Table 1, Figure 2A & 2B). The VE point estimate for the primary endpoint was −0.038 (95% CI, −0.483 to 0.274) (Table 1). Site specific VE point estimates are listed in Table S4. Overall, there were 135 (15.5%) initial QFT conversions (after D71) in the BCG group and 125 (14.7%) conversions in the placebo group (Table 1, Figure 2C). QFT reversion rates were similar in both groups (Table 1, Table S5). Amongst initial QFT converters with 3 consecutive positive or negative QFT results available, 62 of 111 BCG recipients (56%) and 59 of 101 placebo recipients (58%) developed sustained QFT conversion, while 49 (44%) and 42 (42%) developed QFT reversion at D84 or Month 6 post conversion. The initial QFT conversion rates based on a range of increasingly stringent thresholds were similar between the BCG and placebo groups, regardless of threshold (Figure 2D, Table S6). For example, initial QFT conversions based on a QFT threshold of 4.0 IU/mL were observed in 60 (6.9%) participants in the BCG group and 64 (7.5%) in the placebo group (Figure 2D, Table S6, Figure S1). Six participants in the safety population (3 in each of the BCG and placebo groups) developed laboratory-confirmed TB.

The initial QFT conversion rates based on a range of increasingly stringent thresholds were similar between the BCG and placebo groups, regardless of threshold (Figure 2D, Table S6). For example, initial QFT conversions based on a QFT threshold of 4.0 IU/mL were observed in 60 (6.9%) participants in the BCG group and 64 (7.5%) in the placebo group (Figure 2D, Table S6, Figure S1). Six participants in the safety population (3 in each of the BCG and placebo groups) developed laboratory-confirmed TB. Solicited local AEs were reported by 77.8% participants in the BCG group and 38.1% in the placebo group; 40.5% and 34.3% reported any solicited systemic AEs, respectively (Table 2). Swelling was the most common solicited local AE, and tiredness was the most common solicited systemic AE. Unsolicited non-serious AEs were experienced by 185 (20.2%) participants in the BCG group and 117 (12.8%) in the placebo group (Table 2). The most common unsolicited non-serious AEs with BCG were headache (3.9%) and injection-site pain (2.3%). Most AEs were mild to moderate in severity. Three participants (0.3%) in each group experienced serious AEs, which were assessed by the site investigator as not related to trial intervention. No SAEs with the outcome of death, serious ADRs or AEs leading to premature trial discontinuation were reported. In the BCG group, 32 participants developed injection site scars >10mm. Of them, 9 were >15mm and 2 were >20mm. One larger keloid (2.5 cm max. diameter) was surgically removed at the participant’s request.

. No SAEs with the outcome of death, serious ADRs or AEs leading to premature trial discontinuation were reported. In the BCG group, 32 participants developed injection site scars >10mm. Of them, 9 were >15mm and 2 were >20mm. One larger keloid (2.5 cm max. diameter) was surgically removed at the participant’s request. BCG revaccination increased the frequencies of antigen-specific CD4 T cells expressing any combination of IFN-γ, TNF, IL-2, IL-17, and IL-22, compared to placebo, which remained higher than baseline, 6 months after vaccination (Figures 3, S2 and S3). Th1 responses (CD4 T cells expressing IL-2, IFN-γ, and/or TNF) predominantly accounted for this observation, rather than IL-17 or IL-22 producing CD4 T cells (Figure S3). A transient increase of CD8 responses to BCG was observed at D29 post-vaccination (Figure S4).

The mITT population included 871 and 849 participants in the BCG and placebo groups, respectively (Figure 1B). Thirty-eight (4.1%) and 58 (6.3%) participants were excluded from the mITT population due to positive QFT at D71; 9 (1.0%) and 8 (0.9%) were excluded due to missing D71 and no negative result at the first visit after D71; and 0 and 2 (0.2%) were excluded due to indeterminate results at D71. At primary analysis after a median of 30 months (interquartile range, 25 to 33) of follow up, sustained QFT conversion was observed in 62 participants in the BCG group and 59 in the placebo group (one-sided P=0.58; Table 1, Figure 2A & 2B). The sustained QFT conversion rate (95% CI) was 7.1% (5.4 to 8.8) in the BCG group and 7.0% (5.2 to 8.7) in the placebo group. The overall sustained conversion incidence rate and the incidence rate over time were similar in both groups (Table 1, Figure 2A & 2B). The VE point estimate for the primary endpoint was −0.038 (95% CI, −0.483 to 0.274) (Table 1). Site specific VE point estimates are listed in Table S4. Overall, there were 135 (15.5%) initial QFT conversions (after D71) in the BCG group and 125 (14.7%) conversions in the placebo group (Table 1, Figure 2C). QFT reversion rates were similar in both groups (Table 1, Table S5). Amongst initial QFT converters with 3 consecutive positive or negative QFT results available, 62 of 111 BCG recipients (56%) and 59 of 101 placebo recipients (58%) developed sustained QFT conversion, while 49 (44%) and 42 (42%) developed QFT reversion at D84 or Month 6 post conversion.

both groups (Table 1, Table S5). Amongst initial QFT converters with 3 consecutive positive or negative QFT results available, 62 of 111 BCG recipients (56%) and 59 of 101 placebo recipients (58%) developed sustained QFT conversion, while 49 (44%) and 42 (42%) developed QFT reversion at D84 or Month 6 post conversion. The initial QFT conversion rates based on a range of increasingly stringent thresholds were similar between the BCG and placebo groups, regardless of threshold (Figure 2D, Table S6). For example, initial QFT conversions based on a QFT threshold of 4.0 IU/mL were observed in 60 (6.9%) participants in the BCG group and 64 (7.5%) in the placebo group (Figure 2D, Table S6, Figure S1). Six participants in the safety population (3 in each of the BCG and placebo groups) developed laboratory-confirmed TB.

Solicited local AEs were reported by 77.8% participants in the BCG group and 38.1% in the placebo group; 40.5% and 34.3% reported any solicited systemic AEs, respectively (Table 2). Swelling was the most common solicited local AE, and tiredness was the most common solicited systemic AE. Unsolicited non-serious AEs were experienced by 185 (20.2%) participants in the BCG group and 117 (12.8%) in the placebo group (Table 2). The most common unsolicited non-serious AEs with BCG were headache (3.9%) and injection-site pain (2.3%). Most AEs were mild to moderate in severity. Three participants (0.3%) in each group experienced serious AEs, which were assessed by the site investigator as not related to trial intervention. No SAEs with the outcome of death, serious ADRs or AEs leading to premature trial discontinuation were reported. In the BCG group, 32 participants developed injection site scars >10mm. Of them, 9 were >15mm and 2 were >20mm. One larger keloid (2.5 cm max. diameter) was surgically removed at the participant’s request.

BCG revaccination increased the frequencies of antigen-specific CD4 T cells expressing any combination of IFN-γ, TNF, IL-2, IL-17, and IL-22, compared to placebo, which remained higher than baseline, 6 months after vaccination (Figures 3, S2 and S3). Th1 responses (CD4 T cells expressing IL-2, IFN-γ, and/or TNF) predominantly accounted for this observation, rather than IL-17 or IL-22 producing CD4 T cells (Figure S3). A transient increase of CD8 responses to BCG was observed at D29 post-vaccination (Figure S4).

This trial was conducted to assess the findings of an earlier trial that reported a significant reduction in sustained QFT conversion rates following BCG revaccination of QFT-negative, HIV-negative adolescents3, assuming that a confirmatory prevention of sustained Mtb infection result and supportive correlates of protection data could de-risk and motivate a future POD trial. While a prevention of sustained Mtb infection VE point estimate of 45.4% (95% CI, 6.4 to 68.1) was observed in the earlier trial, no efficacy was observed in the trial reported here (VE point estimate −3.8% [95% CI, −48.3 to 27.4]). The differences in observed efficacy are unlikely due to differences in BCG strain, case definition, biospecimen processing, or assay performance, which were intentionally harmonized. This trial differed from the previous trial in that it was powered for sustained QFT conversion as the primary endpoint using QFT TB Gold-Plus rather than the QFT TB Gold In-tube test; the sample size per treatment arm was approximately 3 times larger; the age range increased to include 10- and 11-year-olds; and the geographical footprint expanded to include additional communities in the Western Cape, Gauteng, and KwaZulu Natal provinces with a more diverse population.

n the QFT TB Gold In-tube test; the sample size per treatment arm was approximately 3 times larger; the age range increased to include 10- and 11-year-olds; and the geographical footprint expanded to include additional communities in the Western Cape, Gauteng, and KwaZulu Natal provinces with a more diverse population. However, geographical or ethnic differences between trials are unlikely to explain the observed differences in vaccine efficacy, since the observed site-specific VE in this trial at the Worcester site where 92.7% of the earlier trial was enrolled, was similar to the overall VE observed in this trial. The overall initial QFT conversion rate in this trial (6.5-6.9% per person-year) was lower than the 9.9% per person-year reported for the earlier trial. This could be due, in part, to the COVID-19-related lockdown with extended school-closures in 2020 and 202110, and inclusion of trial sites with a lower force of infection: two sites outside the Western Cape province enrolled 20% of participants but only contributed 5.0% [6/121] of primary endpoint cases. One observation that differed between this trial and the earlier trial and may have contributed to the difference in observed VE is that the QFT reversion rate in the placebo arm was higher in our trial (42%) than the earlier trial (25%) while the reversion rates in the BCG group were comparable (44% vs 46%). Whether differences in baseline characteristics (inclusion of younger children, more diverse population) or lower intensity of Mtb exposure, e.g., related to COVID-19 lock-down restrictions, contributed to the observed difference in reversion rates remains unknown.

n rates in the BCG group were comparable (44% vs 46%). Whether differences in baseline characteristics (inclusion of younger children, more diverse population) or lower intensity of Mtb exposure, e.g., related to COVID-19 lock-down restrictions, contributed to the observed difference in reversion rates remains unknown. The safety and reactogenicity data reported here agrees with the known safety profile of the marketed BCG vaccine. This vaccine is well characterized. Larger scars (>10mm) at the site of BCG administration were observed in approximately 3% of BCG vaccinees, consistent with previous reports 11. A BCG revaccination trial for prevention of sustained Mtb infection in 2,000 adult health-care workers in Brazil recently reported the absence of VE against either initial or sustained QFT conversion12. Given that two randomized controlled trials in settings with different force of infection, geography, climate, exposure to environmental mycobacteria, ancestry and age, observed a lack of VE, the evidence supporting BCG revaccination for the prevention of Mtb infection appears weakened. However, our data do not allow conclusions to be drawn on the potential efficacy of BCG revaccination for POD, for which supporting evidence is limited.13,14 A BCG revaccination trial for POD is underway in India and may provide informative data on that endpoint15.

for the prevention of Mtb infection appears weakened. However, our data do not allow conclusions to be drawn on the potential efficacy of BCG revaccination for POD, for which supporting evidence is limited.13,14 A BCG revaccination trial for POD is underway in India and may provide informative data on that endpoint15. Significant attention has been given to the predictive power of the QFT assay for TB disease, and higher IFN-γ conversion values were previously reported to be associated with a higher risk of progression to TB13,14. However, for POI, we observed no changes in VE with increasing IFN-γ thresholds, arguing against the possibility that false-positive converters contributed to the observed lack of VE. The strengths of this dataset include that this trial employed stringent statistical criteria to assess VE, and that the primary endpoint data are unambiguous. BCG revaccination did not provide any protection from Mtb infection as assessed by QFT, whether defined as sustained conversion, initial conversion, or conversion with higher IFN-γ thresholds. Treatment groups were well balanced, and the trial was well executed, with stringent monitoring and oversight. Participant retention and community engagement were excellent throughout.

infection as assessed by QFT, whether defined as sustained conversion, initial conversion, or conversion with higher IFN-γ thresholds. Treatment groups were well balanced, and the trial was well executed, with stringent monitoring and oversight. Participant retention and community engagement were excellent throughout. While this trial does not allow us to draw firm conclusions on the efficacy of BCG revaccination for POD, the absence of VE for POI probably decreases the likelihood of BCG revaccination conferring POD, unless by prevention of progression from infection to disease. Other limitations include that enrollment was paused for four months due to COVID-19 pandemic restrictions, and schools were closed for several additional months10, which may have contributed to a lower incidence rate.