Study participant characteristics
A total of 936 adolescents (age: 13-15 years) were screened for evidence of TB infection: 164 were IGRA-positive or borderline at baseline and were excluded from immunization; 667 were IGRA-negative and eligible for the immunization phase of the trial. A total of 48 participants converted to IGRA-positive at day 60, 420, or 720, including 20 early converters (initial positive at day 60, 11 females/9 males) and 29 late converters (initial positive at day 420 or 720, 17 females/12 males). One female early converter was excluded because a paired CBC sample was not collected at the time of IGRA conversion. All other participants who converted to IGRA-positive at day 60, 420, or 720 are included in the present study. Among the IGRA-positive participants, 22 were persistent positives and 25 were transient positives. Twenty-one received the DAR-901 vaccine (12 persistent/9 transient), while 26 received saline placebo (10 persistent/16 transient). Ninety-eight randomly selected participants who were IGRA-negative at baseline and remained IGRA-negative throughout the entire trial period (56 female/42 male) served as negative controls. Of these, 47 received the DAR-901 vaccine, and 51 received saline placebo. The IGRA converters and IGRA-negative controls were similar with respect to both gender (p=1.00) and treatment assignment (i.e., placebo vs vaccine, p=0.73).
Baseline CBC values of subsequent IGRA converters and IGRA-negative participants
We first evaluated whether baseline CBC values (obtained at screening) differed between subsequent IGRA converters (n=47) and IGRA-negative controls (n=98). At baseline, subsequent IGRA converters had modest but statistically significant reductions in red blood cell (RBC) count (4.6 vs 4.8 × 106/mcL, p=0.008), hemoglobin (12.3 vs 12.6 g/dL, p=0.01), and hematocrit (37.8 vs 38.8 %, p=0.004), compared with IGRA-negative controls (Table 1). None of the WBC-associated measures differed between IGRA converters and controls. Among IGRA converter sub-groups, persistent converters had the greatest number of CBC components that were statistically different from IGRA-negative controls (n=4: absolute neutrophil count, RBC count, hemoglobin, and hematocrit), followed by late converters (n=3: RBC count, hemoglobin, and hematocrit), transient converters (n=1: hematocrit), and early converters (n=1: hematocrit).
Table 1. Baseline CBC results for subsequent IGRA converters and IGRA-negative participants.
IGRA- All Early Late Transient Persistent
Negative Converters Converters Converters Converters Converters
(n=98) (n=47) (n=18) (n=29) (n=25) (n=22)
Laboratory Component Mean Mean p-val Mean p-val Mean p-val Mean p-val Mean p-val
(S.D.) (S.D.) (S.D.) (S.D.) (S.D.) (S.D.)
White blood cells (WBCs)
WBC count (× 103/mcL) 6.14 5.81 0.20 5.79 0.46 5.82 0.24 5.79 0.36 5.83 0.29
(1.36) (1.36) (1.22) (1.47) (1.43) (1.32)
1Neutrophil count (× 103/mcL) 2.66 2.26 0.06 2.34 0.29 2.22 0.08 2.37 0.36 2.15 0.05
(0.96) (0.93) (0.94) (0.94) (1.00) (0.85)
Neutrophil % 42.5 38.2 0.11 39.6 0.41 37.4 0.12 39.4 0.48 36.8 0.07
(8.6) (11.8) (11.2) (12.2) (11.1) (12.6)
Lymphocyte count (× 103/mcL) 2.68 2.73 0.91 2.66 0.94 2.78 0.92 2.66 0.96 2.82 0.90
(0.56) (0.74) (0.63) (0.81) (0.61) (0.87)
Lymphocyte % 44.5 47.8 0.12 46.5 0.51 48.6 0.11 47.2 0.39 48.5 0.11
(7.6) (10.0) (9.2) (10.5) (9.6) (10.5)
Monocyte count (× 103/mcL) 0.484 0.459 0.40 0.467 0.99 0.454 0.26 0.425 0.10 0.498 0.68
(0.151) (0.175) (0.153) (0.189) (0.147) (0.197)
Monocyte % 7.9 7.9 0.61 8.0 0.92 7.8 0.44 7.3 0.18 8.5 0.52
(2.0) (2.5) (2.2) (2.7) (1.6) (3.1)
Eosinophil count (× 103/mcL) 0.220 0.253 0.26 0.230 0.88 0.268 0.16 0.249 0.17 0.258 0.75
(0.259) (0.229) (0.222) (0.236) (0.197) (0.266)
Eosinophil % 3.5 4.4 0.17 4.2 0.80 4.5 0.09 4.5 0.13 4.3 0.58
(3.6) (3.9) (4.4) (3.6) (4.0) (3.8)
Basophil count (× 103/mcL) 0.100 0.097 0.98 0.093 0.66 0.100 0.77 0.084 0.90 0.112 0.93
(0.075) (0.068) (0.069) (0.069) (0.036) (0.091)
Basophil % 1.6 1.7 0.60 1.7 0.99 1.7 0.46 1.5 0.69 1.9 0.68
(1.1) (1.2) (1.4) (1.2) (0.8) (1.6)
Red blood cells (RBCs)
1,2RBC count (× 106/mcL) 4.8 4.6 0.008 4.7 0.20 4.5 0.008 4.7 0.08 4.5 0.02
(0.4) (0.5) (0.8) (0.4) (0.6) (0.4)
1,2Hemoglobin (g/dL) 12.6 12.3 0.01 12.5 0.08 12.1 0.03 12.4 0.09 12.1 0.02
(1.0) (1.6) (2.4) (0.9) (2.0) (1.0)
1,2,3,4Hematocrit (%) 38.8 37.8 0.005 38.2 0.03 37.5 0.03 38.1 0.04 37.4 0.02
(3.0) (4.5) (6.5) (2.7) (5.6) (3.0)
MCV (fL) 81.9 82.2 0.67 81.4 0.90 82.7 0.50 81.4 0.68 83.1 0.26
(6.2) (5.7) (6.7) (5.1) (5.2) (6.3)
MCH (pg) 26.5 26.8 0.50 26.7 0.58 26.8 0.62 26.6 0.96 27.0 0.30
(2.2) (2.1) (2.5) (1.9) (2.1) (2.2)
MCHC (g/dL) 32.4 32.5 0.36 32.8 0.12 32.4 0.94 32.6 0.31 32.5 0.71
(0.9) (0.9) (0.8) (1.9) (0.9) (1.0)
RDW (%) 16.1 16.7 0.33 16.9 0.38 16.6 0.51 17.2 0.15 16.1 0.99
(1.5) (2.4) (2.9) (2.2) (3.0) (1.5)
Platelets
Platelet count (× 103/mcL) 312 300 0.20 306 0.24 297 0.38 288 0.19 315 0.51
(103) (118) (141) (104) (107) (131)
S.D.: standard deviation; MCV: mean corpuscular volume; MCH: mean corpuscular hemoglobin, MCHC: mean corpuscular hemoglobin concentration; RDW: red cell distribution width.
Baseline CBC-derived ratios of subsequent IGRA converters and IGRA-negative participants
Next, we sought to determine whether any of the baseline CBC-derived ratios (MLR, NLR, and PLR) differed between subsequent IGRA converters and IGRA-negative controls. There were no significant differences between all IGRA converters and controls with respect to MLR (0.17 vs 0.18, p=0.10), NLR (0.88 vs 1.02, p=0.08), or PLR (115 vs 120, p=0.28) (Table 2). With respect to sub-group analysis, there was a modest but statistically significant decrease in MLR among transient converters relative to controls (0.16 vs 0.18, p=0.03). No differences were observed in sub-group analyses for either NLR or PLR.
The performance of baseline MLR, NLR, and PLR with respect to identifying subsequent TB infection was evaluated using ROC curves, with the AUROC used to quantify test performance, and Youden’s index used to define optimal cutoff values. With respect to all converters, MLR yielded an AUROC of 0.58 (95% CI: 0.48-0.69), NLR an AUROC of 0.59 (95% CI: 0.49-0.70), and PLR an AUROC of 0.56 (95% CI: 0.44-0.67), with none of these ratios performing statistically better than random (Figure 1). Of note, RBC count (AUROC: 0.64, 95% CI: 0.54-0.73), hemoglobin (AUROC: 0.63, 95% CI: 0.53-0.73), and hematocrit (AUROC: 0.65, 95% CI: 0.55-0.75) did perform statistically better than random for differentiating between overall IGRA converters and controls. With respect to analysis of IGRA converter sub-groups, MLR performed better than random among transient converters only (AUROC: 0.64, 95% CI: 0.52-0.76), and neither NLR nor PLR performed better than random among IGRA converter sub-groups.
Table 2. Baseline MLR, NLR, and PLR for subsequent IGRA converters and IGRA-negative participants.
Group n Mean S.D. Q1 Q3 p-value
MLR
IGRA Negatives 98 0.18 0.05 0.15 0.21 Reference
All Converters 47 0.17 0.06 0.13 0.20 0.10
Early Converters 18 0.18 0.07 0.13 0.22 0.56
Late Converters 29 0.16 0.06 0.13 0.18 0.07
Transient Converters 25 0.16 0.04 0.13 0.18 0.03
Persistent Converters 22 0.18 0.08 0.14 0.22 0.73
NLR
IGRA Negatives 98 1.02 0.40 0.74 1.22 Reference
All Converters 47 0.88 0.42 0.56 1.21 0.08
Early Converters 18 0.93 0.42 0.62 1.25 0.35
Late Converters 29 0.85 0.42 0.50 1.20 0.09
Transient Converters 25 0.91 0.39 0.68 1.69 0.38
Persistent Converters 22 0.85 0.45 0.48 1.10 0.06
PLR
IGRA Negatives 98 120 46 96 135 Reference
All Converters 47 115 50 82 152 0.28
Early Converters 18 119 58 89 136 0.41
Late Converters 29 113 45 80 158 0.40
Transient Converters 25 112 43 79 155 0.22
Persistent Converters 22 119 57 87 142 0.69
S.D.: standard deviation; Q1: first quartile (25th percentile); Q3: third quartile (75th percentile).
Concurrent CBC values at the time of IGRA conversion
We next sought to determine whether concurrent CBC values differed between IGRA converters at the time of IGRA conversion (n=18 early converters and 25 late converters) and time-matched IGRA-negative controls (n=93 at day 60 and n=95 at day 720). Follow-up CBCs were not available for a subset of IGRA-negative controls (n=5 at day 60 and n=3 at day 720) and were excluded from these analyses accordingly. Although included in the baseline analysis, participants who had converted to IGRA-positive at day 420 but had reverted to IGRA-negative by day 720 are excluded from the concurrent analysis since a CBC was not obtained at the time of IGRA conversion. Comparison groups were similar with respect to treatment assignment (i.e., vaccine vs placebo) and gender for early converters versus time-matched negative controls (p=0.57 for treatment, p=0.44 for gender), late converters versus time-matched controls (p=0.63 for treatment, p=1.00 for gender), and early converters versus late converters (p=0.76 for treatment, p=0.76 for gender). There were no significant differences noted with respect to any CBC components or CBC-derived ratios (MLR, NLR, or PLR) for either early converters or late converters relative to time-matched controls. In the comparison of early versus late converters, there were significant differences observed with respect to both hemoglobin (12.1 vs 13.0 g/dL, p=0.04) and hematocrit (37.7 vs 40.7 %, p=0.05), although these differences were also noted in the comparison of day 60 and day 720 IGRA-negative controls (Table 3).
We also determined whether CBC components and/or CBC-derived ratios differed between persistent and transient IGRA converter sub-groups. The persistent (n=22) and transient (n=21) converter groups were similar with respect to treatment (p=0.12), gender (p=0.55), and early versus late converter status (p=0.76). With respect to CBC components, only absolute monocyte count was significantly different between persistent and transient converters (2.06 vs 2.81 × 103/mcL, respectively, p=0.04). Furthermore, none of the CBC-derived ratios differed significantly between the two groups.
Table 3. Concurrent CBC results and CBC-derived ratios for early and late converters at first study visit where IGRA conversion was detected versus time-matched controls.
Early Negatives Late Negatives
Converters at Day 60 Converters at Day 720 p-value p-value p-value
Laboratory Mean Mean Mean Mean (EC vs (LC vs (EC vs
Component (S.D.) (S.D.) (S.D.) (S.D.) N60) N720) LC)
White blood cells (WBCs)
WBC count1 5.62 (1.18) 5.52 (1.44) 5.86 (1.48) 5.43 (1.42) 0.53 0.13 0.48
Neutrophil count1 2.16 (0.85) 2.27 (1.12) 2.62 (1.24) 2.28 (1.02) 0.92 0.24 0.31
Neutrophil % 37.5 (11.1) 39.3 (10.7) 43.2 (12.6) 41.2 (12.0) 0.57 0.75 0.24
Lymphocyte count1 2.69 (0.54) 2.53 (0.63) 2.53 (0.71) 2.41 (0.71) 0.33 0.45 0.40
Lymphocyte % 48.8 (7.6) 47.0 (9.4) 44.4 (11.5) 45.0 (9.6) 0.50 0.81 0.15
Monocyte count1 0.405 (0.147) 0.428 (0.204) 0.374 (0.158) 0.448 (0.365) 0.77 0.33 0.42
Monocyte % 7.2 (2.1) 7.9 (3.9) 6.5 (2.5) 8.4 (7.0) 0.64 0.09 0.30
Eosinophil count1 0.185 (0.205) 0.177 (0.209) 0.269 (0.281) 0.163 (0.224) 0.84 0.13 0.28
Eosinophil % 3.3 (3.7) 3.1 (3.3) 4.6 (4.8) 3.0 (3.5) 0.66 0.17 0.29
Basophil count1 0.181 (0.292) 0.123 (0.116) 0.071 (0.040) 0.143 (0.201) 0.62 0.46 0.42
Basophil % 3.2 (4.7) 2.4 (2.6) 1.3 (0.9) 2.6 (3.4) 0.55 0.07 0.24
Red blood cells (RBCs)
RBC count2 4.6 (0.6) 4.7 (0.5) 4.7 (0.5) 4.8 (0.5) 0.53 0.68 0.44
Hemoglobin (g/dL) 12.1 (1.4) 12.3 (1.0) 13.0 (1.4) 12.7 (1.3) 0.29 0.42 0.04
Hematocrit (%) 37.7 (4.5) 38.5 (3.3) 40.8 (4.8) 40.1 (4.1) 0.22 0.50 0.05
MCV (fL) 82.3 (6.1) 82.7 (6.7) 86.0 (4.7) 84.0 (7.0) 0.90 0.27 0.04
MCH (pg) 26.5 (2.2) 26.6 (2.8) 27.4 (1.8) 26.7 (2.5) 0.98 0.17 0.14
MCHC (g/dL) 32.2 (1.0) 32.1 (2.1) 31.9 (1.5) 31.8 (1.4) 0.78 0.35 0.68
RDW (%) 17.1 (2.1) 16.6 (1.5) 16.4 (1.6) 16.8 (2.1) 0.57 0.64 0.38
Platelets
Platelet count1 276 (81) 311 (168) 272 (95) 289 (79) 0.32 0.37 0.93
CBC-derived Ratios
MLR 0.15 (0.05) 0.18 (0.10) 0.16 (0.07) 0.20 (0.25) 0.54 0.20 0.88
NLR 0.82 (0.38) 0.94 (0.57) 1.13 (0.69) 1.01 (0.54) 0.45 0.71 0.15
PLR 105 (32) 138 (132) 114 (46) 128 (43) 0.11 0.21 0.53
EC: early converters; N60; IGRA-negative at day 60 (time-matched controls for early converters); LC: late converters; N720: IGRA-negative at day 720 (time-matched controls for late converters); MCV: mean corpuscular volume; MCH: mean corpuscular hemoglobin, MCHC: mean corpuscular hemoglobin concentration; RDW: red cell distribution width; 1: units are × 103/mcL; 2: units are × 106/mcL.
Effect of vaccination on CBC components and CBC-derived ratios
Finally, we sought to evaluate whether vaccination with DAR-901 had an effect on CBC components or CBC-derived ratios relative to treatment with saline placebo. When including all participants irrespective of disease status, there were no significant differences with respect to CBC components or CBC-derived ratios between participants who received DAR-901 versus saline placebo at day 60 and/or 720. When groups were stratified according to disease status (IGRA-positive versus IGRA-negative), there was a modest but significant difference in RDW between late converters who received vaccine versus those who received placebo (15.6 vs 17.0 %, respectively, p=0.03).