Baseline characteristics of patients (Table 1)
A total of 29 patients with intestinal BD were included in this study. The baseline patient characteristics are shown in Table 1. Among the 29 patients, 14 (48.3%) were men. The average age at initiation of anti-TNFα therapy was 42 years (range, 17-74 years), and the average disease duration before anti-TNFα therapy was 6 years (range, 0-41 years). Regarding the type of anti-TNFα agent, 16 (55%) patients were treated with IFX and 13 (45%) were treated with ADA. Ocular lesions, skin lesions, oral ulcers, and genital ulcers were present in 9, 12, 28, and 10 patients, respectively. In terms of HLA serological typing (A/B), 21 patients (21/29, 72.4%) were investigated for HLA serological typing (A/B). We identified 5 (5/21, 23.8%) patients positive for HLA-B51 and three (3/21, 14.3%) positive for HLA-A26. Concomitant medications administered with anti-TNFα agents included systemic corticosteroids (20/29, 69.0%), 5-ASA (11/29, 37.9%), cyclosporine A (2/29, 6.9%), methotrexate (3/29, 10.3%), AZA/6-mercaptopurine (18/29, 62.1%), and colchicine (16/29, 55.2%). Four patients (13.8%) had a history of major abdominal surgery. The mean baseline global GI score was 1.69, while the mean baseline C-reactive protein (CRP) level was 1.81 mg/dL.
(Table 1)
Table 1 Baseline characteristics of patients
|
n=29
|
Male sex (%)
|
14 (48.3)
|
Age at disease onset (mean±SD, years old)
|
36±14
|
Body weight, kg
|
|
Mean±SD
|
51.7±11.1
|
Range
|
32.0-78.0
|
Tobacco, nonsmoker (%)
|
22 (75.9)
|
Alcohol, nondrinker (%)
|
17 (58.7)
|
Age at initiation of anti-TNFα agent (mean±SD, years old)
|
42±14
|
Disease duration before anti-TNFα agent
(mean±SD, years)
|
6±9
|
Type of anti-TNFα agent (IFX/ADA) administered
|
16/13
|
Major symptoms
|
|
Ocular lesions (%)
|
9 (31.0)
|
Skin lesions (%)
|
12 (41.4)
|
Oral ulcers (%)
|
28 (96.6)
|
Genital ulcers (%)
|
10 (34.5)
|
Minor symptoms
|
|
Arthritis (%)
|
8 (27.6)
|
Vascular involvement (%)
|
1 (3.4)
|
HLA-B51 positivity (n=21, %)
|
5/21 (23.8)
|
HLA-A26 positivity (n=21, %)
|
3/21 (14.3)
|
Concomitant medication administered with the anti-TNFα agent
|
-
|
Systemic corticosteroids (%)
|
20 (69.0)
|
≧20 mg corticosteroids (%)
|
9 (31.0)
|
5-aminosalicylic acid (%)
|
11 (37.9)
|
CyA (%)
|
2 (6.9)
|
MTX (%)
|
3 (10.3)
|
Azathioprine/6-mercaptopurine (%)
|
18 (62.1)
|
Colchicine (%)
|
16 (55.2)
|
Previous major abdominal surgery (%)
|
4 (13.8)
|
baseline Global GI symptoms score
(mean±SD); Score 0-4
|
1.69±1.4
|
baseline CRP level (mean±SD, mg/dL)
|
1.81±3.8
|
Abbreviations: ADA, adalimumab; CRP, C-reactive protein; CyA, cyclosporine A; HLA, human leukocyte antigen; IFX, infliximab; MTX, methotrexate; SD, standard deviation; TNFα, tumor necrosis factor α
During the observation period, AEs occurred in 14 patients (48.3%). AEs included nasopharyngitis in 7 patients (24.1%), hepatic events in 2 patients (6.9%), dizziness in 2 patients (6.9%), urinary tract infection in 1 patient (3.4%), pulmonary infection in 1 patient (3.4%), influenza in 1 patient (3.4%), and headache in 1 patient (3.4%). Among these, two serious events were observed. One patient died of lung cancer, which developed one year after the initial IFX administration. The other patient experienced ileum perforation due to exacerbation of intestinal BD.
Relation between failure of anti-TNFα agent and CRP levels (Table 2)
Next, we evaluated the continuation rate of anti-TNFα therapy. As shown in Table 2, 25 (86.2%) and 22 (75.9%) patients continued anti-TNFα therapy at weeks 8 and 48, respectively. We compared CRP levels at baseline in the CG and DG groups. At week 8, the mean CRP levels at baseline in the CG and DG were 1.16±2.09 mg/dL and 5.93±8.60 mg/dL, respectively (p=0.017). At week 48, the mean CRP levels at baseline in the CG and DG were 0.84±1.41 mg/dl and 4.86±6.80 mg/dl, respectively (p=0.006).
Moreover, 9 (36.0%) patients in the CG had baseline CRP levels of 1 mg/dL or higher, whereas two (50.0%) patients in the DG showed similar levels. At week 48, 6 (27.3%) patients in the CG had baseline CRP levels of 1 mg/dL or higher when compared with 5 (71.4%) patients in the DG.
Table 2 Relationship between the continuation of anti-TNFα agent and CRP levels
|
At week 8
|
At week 48
|
|
continuation
|
discontinuation
|
p-value
|
continuation
|
discontinuation
|
p-value
|
N (%)
|
25 (86.2)
|
4 (13.8)
|
|
22 (75.9)
|
7 (24.1)
|
|
CRP level at baseline (mean±SD, mg/dl)
|
1.16±2.09
|
5.93±8.60
|
0.017
|
0.84±1.41
|
4.86±6.80
|
0.006
|
CRP≥1 mg/dL in the group (%)
|
9 (36.0)
|
2 (50.0)
|
|
6 (27.3)
|
5 (71.4)
|
|
Abbreviations: CRP, C-reactive protein; SD, standard deviation.
Continuation rates and efficacy of anti-TNFα agents (Figure 2, Figure 3)
We investigated the overall clinical course of patients with intestinal BD who received first-line anti-TNFα therapy (Figure 2). The cumulative continuation rate of anti-TNFα agents was calculated using the Kaplan-Meier curve. Kaplan-Meier curves demonstrated that the cumulative continuation rates of anti-TNFα agents at weeks 8, 48, and 96 were 86.2%, 72.0%, and 60.4%, respectively. The median duration of continuation was 63 months.
Furthermore, we compared the cumulative continuation rates calculated by the Kaplan-Meier curve according to the type of anti-TNFα agent (IFX or ADA) and the difference in baseline CRP levels (CRP ≥ 1 mg/dL vs. CRP less than 1 mg/dL) using the log-rank test (Figure 2). No significant difference in cumulative continuation rates was observed between the IFX and ADA groups (p=0.5). In contrast, baseline CRP levels of 1 mg/dL or higher were associated with discontinuation of anti-TNFα agents (p=0.04).
Next, we investigated the response rate of anti-TNFα agents in patients with intestinal BD at the indicated time points (Figure 3). At week 48, 14 (48.3%), 11 (37.9%), and 14 (48.3%) patients achieved MI, CR, and MH, respectively, whereas 11 (37.9%) patients achieved MI, CR, and MH at week 96.
Predictors of discontinuation of anti-TNFα agents (Table 3)
We investigated factors associated with the discontinuation of anti-TNFα agents in patients with intestinal BD. To identify predictors of discontinuation of anti-TNFα therapy, we conducted a logistic regression analysis stratified by demographic variables such as sex (male vs. female), age at disease onset (≥40 vs. <40 years), disease duration before anti-TNFα agent initiation (≥5 vs. <5 years), type of anti-TNFα agent administered (IFX vs. ADA), CRP level at baseline (≥1 vs. <1 mg/dL), and concomitant use of systemic corticosteroids at a dose of ≥20 mg (yes vs. no). The results are presented in Table 3. Univariate analysis showed no significant differences between the two groups in terms of sex, age at disease onset, disease duration before anti-TNFα agent initiation, type of anti-TNFα agent administered, and concomitant use of corticosteroids (≥ 20 mg). Notably, only the baseline CRP level was significantly associated with the discontinuation of anti-TNFα agents at week 48 (p=0.019). In addition, multivariate analysis revealed that only a baseline CRP level of 1 mg/dL or higher was statistically associated with the discontinuation of anti-TNFα therapy at week 48 (p=0.033).
Table 3 Univariate and multivariate logistic regression analyses of predictors for discontinuation of anti-TNFα therapy at week 48
|
Univariate analysis
|
Multivariate analysis
|
|
OR
|
95% CI
|
p-value
|
OR
|
95% CI
|
p-value
|
Sex (male vs. female)
|
4.9
|
0.79-30.3
|
0.09
|
2.7
|
0.31-22.7
|
0.37
|
Age at disease onset
(≥40 vs. <40 years old)
|
2.0
|
0.38-10.5
|
0.41
|
1.9
|
0.20-17.5
|
0.59
|
Disease duration before anti-TNFα agent initiation (≥5 vs. <5 years)
|
0.4
|
0.07-2.7
|
0.38
|
0.4
|
0.04-4.96
|
0.49
|
Type of anti-TNFα agent
administered (IFX vs. ADA)
|
0.8
|
0.15-3.8
|
0.73
|
|
|
|
CRP level at baseline
(≥1 vs. <1 mg/dL)
|
9.6
|
1.45-63.5
|
0.019
|
10.5
|
1.21-90.2
|
0.033
|
Concomitant use of systemic corticosteroid (≥20 mg) (yes vs. no)
|
1.5
|
0.27-8.3
|
0.64
|
|
|
|
Abbreviations: ADA, adalimumab; CI, confidence interval; CRP, C-reactive protein; IFX, infliximab; OR, odds ratio; TNFα, tumor necrosis factor α.