Table 4.
Predictors of lower respiratory tract infections in GEMINI 1 and 2.a
| Variable | All patients | Patients with AE | Adjusted results | |
|---|---|---|---|---|
| HR [95% CI] | p-Value | |||
| UC population [GEMINI 1] | n = 769 | n = 38 | ||
| Age, mean [SD] years | 40.3 [13.0] | 42.5 [13.3] | 1.00 [0.98–1.03] | 0.770 |
| Female sex, n [%] | 313 [40.7] | 22 [57.9] | 2.11 [1.07–4.14] | 0.030 |
| Disease duration ≥7 years, n [%] | 267 [34.7] | 18 [47.4] | 1.49 [0.77–2.88] | 0.242 |
| Previous TNF antagonist therapy, n [%] | 384 [49.9] | 25 [65.8] | 2.20 [1.10–4.41] | 0.027 |
| Baseline disease activity, mean [SD] complete Mayo score | 6.1 [1.6] | 5.8 [1.6] | 0.97 [0.79–1.18] | 0.733 |
| Concomitant narcotic use, n [%] | 148 [19.2] | 8.0 [21.1] | 0.82 [0.37–1.83] | 0.628 |
| Concomitant corticosteroid use, n [%] | 409 [53.2] | 19 [50.0] | 0.76 [0.40–1.46] | 0.415 |
| Concomitant immunomodulator use, n [%] | 257 [33.4] | 16 [42.1] | 1.51 [0.79–2.91] | 0.217 |
| Current smoker, n [%]b | 47 [6.1] | 1.0 [2.6] | 0.61 [0.08–4.63] | 0.635 |
| Former smoker, n [%]b | 254 [33.0] | 18 [47.4] | 2.01 [0.99–4.07] | 0.053 |
| In-study surgery, n [%] | 64 [8.3] | 4.0 [10.5] | 1.11 [0.25–4.92] | 0.896 |
| Vedolizumab treatment, n [%] | 620 [80.6] | 31 [81.6] | 0.72 [0.31–1.69] | 0.452 |
| CD population [GEMINI 2] | n = 962 | n = 55 | ||
| Age, mean [SD] years | 35.9 [12.14] | 37.1 [12.19] | 0.99 [0.97–1.02] | 0.603 |
| Female sex, n [%] | 514 [53.4] | 37 [67.3] | 1.64 [0.92–2.92] | 0.093 |
| Disease duration ≥7 years, n [%] | 483 [50.2] | 34 [61.8] | 1.35 [0.75–2.43] | 0.324 |
| Previous TNF antagonist therapy, n [%] | 607 [63.1] | 40 [72.7] | 1.41 [0.73–2.72] | 0.303 |
| Baseline disease activity, mean [SD] score | 11.2 [3.8] | 12.0 [3.7] | 1.05 [0.98–1.13] | 0.136 |
| Concomitant narcotic use, n [%] | 315 [32.7] | 20 [36.4] | 0.85 [0.47–1.54] | 0.597 |
| Concomitant corticosteroid use, n [%] | 488 [50.7] | 24 [43.6] | 0.83 [0.49–1.42] | 0.493 |
| Concomitant immunomodulator use, n [%] | 321 [33.4] | 13 [23.6] | 0.67 [0.36–1.27] | 0.225 |
| Current smoker, n [%]b | 250 [26.0] | 24 [43.6] | 2.37 [1.26–4.45] | 0.008 |
| Former smoker, n [%]b | 219 [22.8] | 13 [23.6] | 1.51 [0.73–3.15] | 0.271 |
| In-study surgery, n [%] | 139 [14.4] | 10 [18.2] | 2.03 [0.83–4.95] | 0.119 |
| Vedolizumab treatment, n [%] | 814 [84.6] | 48 [87.3] | 0.94 [0.42–2.10] | 0.872 |
| Combined UC and CD population [GEMINI 1 and 2] | n = 1731 | n = 93 | ||
| Age, mean [SD] years | 37.9 [12.7] | 39.3 [12.8] | 1.00 [0.98–1.02] | 0.994 |
| Female sex, n [%] | 827 [47.8] | 59 [63.4] | 1.84 [1.19–2.83] | 0.006 |
| Disease duration ≥7 years, n [%] | 750 [43.3] | 52 [55.9] | 1.35 [0.88–2.09] | 0.173 |
| Previous TNF antagonist therapy, n [%] | 991 [57.3] | 65 [69.9] | 1.70 [1.06–2.73] | 0.028 |
| Baseline disease activity, mean [SD] scorec | 5.8 [1.7] | 5.9 [1.7] | 1.07 [0.95–1.22] | 0.278 |
| Concomitant narcotic use, n [%] | 463 [26.7] | 28 [30.1] | 0.87 [0.55–1.39] | 0.565 |
| Concomitant corticosteroid use, n [%] | 897 [51.8] | 43 [46.2] | 0.83 [0.55–1.25] | 0.382 |
| Concomitant immunomodulator use, n [%] | 578 [33.4] | 29 [31.2] | 0.96 [0.62–1.50] | 0.862 |
| Current smoker, n [%]b | 297 [17.2] | 25 [26.9] | 1.97 [1.17–3.31] | 0.011 |
| Former smoker, n [%]b | 473 [27.3] | 31 [33.3] | 1.74 [1.06–2.87] | 0.029 |
| In-study surgery, n [%] | 203 [11.7] | 14 [15.1] | 1.65 [0.78–3.49] | 0.195 |
| Vedolizumab treatment, n [%] | 1434 [82.8] | 79 [84.9] | 0.85 [0.48–1.52] | 0.585 |
LRTI is defined according to the MedDRA high-level term ‘lower respiratory tract and lung infection’.
AE, adverse event; CD, Crohn’s disease; CI, confidence interval; HBI, Harvey–Bradshaw Index; HR, hazard ratio, SD, standard deviation; TNF, tumour necrosis factor; UC, ulcerative colitis.
aAnalysis based on Cox proportional hazards model.
bCurrent and former smoker estimates are relative to non-smokers.
cBaseline disease activity scores based on partial Mayo Score for GEMINI 1 [UC], HBI score for GEMINI 2 [CD], and common index for pooled GEMINI 1 and 2. Common index ranged from 0 to 9 to allow the combination of baseline partial Mayo and HBI scores in the pooled analysis.