Background/Purpose: Inflammatory bowel disease (IBD) is a matter of interest in patients with juvenile idiopathic arthritis (JIA) treated with biologics. Methods: Baseline demografics, clinical characteristics and medication have been documented in the German BIKER JIA registry between 2001 and 2013. Exposure time of treatment was used to calculate IBD incidence. Published data of IBD incidence in the general pediatric population were used as reference.

Results:

3070 patients were followed in the BIKER registry with a total observation time of 8389 patient years (py). 13 events of IBD-flares in 12 patients (33% male), 8 with Crohn’s disease (CD, 67%) and 4 with ulcerative colitis (UC, 33%). were documented.

4 patients had extended oligoarthritis (eoJIA, 33%), 3 seronegative polyarthritis (RF-JIA, 25%), 3 enthesitis-related arthritis (ERA, 25%) and 2 psoriatic JIA (PsA, 17%). HLA-B27 was positive in 3 (25%) and anti-nuclear antibodies were positive in 7 IBD patients (58%).

Thus patient with IBD as adverse event more commonly had ERA, eoJIA and PsA than expected in the total BIKER cohort, but not statistitically significant. No IBD occurred in systemic JIA, RF positive or pers-oligo JIA. IBD incidence in patients with pers-oligo JIA was significantly decreased.

IBD incidence in JIA patients was 143/100,000 py and significantly higher than IBD incidence in the general pediatric population (5.2/100,000 py, p<0.001; OR 27.50 [15.54-48.67]). Mean disease duration until IBD was 7.8 ± 4.0 years (1.8-16.8).

The total exposure time was 4575PY for NSAIDs, 1981PY for corticosteroids, 195PY for sulfasalazine (SUL), 5455PY for methotrexate (MTX), 212PY for leflunomide (LEF), 3557PY for ETA and 369PY for adalimumab (ADA). Incidence of IBD was significantly higher in patients treated with etanercept (p<0.05; OR 4.53 [1.25-16.47]), leflunomide (p<0.05; OR 7.03 [1.55-31.91]) and also sulfasalazine (p<0.001; OR [12.63 (3.45-46.24]). In patients treated with methotrexate the IBD incidence decreased significantly (p<0.05; OR 0.16 [0.04-0.59]).

At admission to the registry, JIA patients later developing IBD had been treated with MTX (100%), steroids (67%), SUL (33%), azathioprine (25%), LEF (17%) and hydroxychloroquine (8%). At time of IBD diagnosis 83% were on  ETA, 33% on steroids, 25% on MTX, 25% on SUL, 17% on LEF and 1 on ADA. Thus, therapy with non-biologic agents has markedly been reduced. 

 

Results			IBD events no.	IBD-rate/1000PY	p-value	odds ratio (95% CI)
Patient cohort	JIA registry		12	1,430	<0.001	27.50 (15.54-48.67)
	ref. population*		739	0,052
Treatment  group	NSAIDs	+	6	1.31	0.5442	0.71 (0.24-2.13)
		–	7	1.84
	Steroids	+	4	2.02	0.5437	1.44 (0.44-4.68)
		–	9	1.40
	Sulfasalazine	+	3	15.41	<0.001	12.63 (3.45-46.24)
		–	10	1.22
	MTX	+	3	0.55	0.0015	0.16 (0.04-0.59)
		–	10	3.41
	Leflunomide	+	2	9.45	0.0032	7.03 (1.55-31.91)
		–	11	1.35
	Etanercept	+	10	2.81	0.0118	4.53 (1.25-16.47)
		–	3	0.62
	Adalimumab	+	1	2.71	0.5638	1.81 (0.24-13.95)
		–	12	1.50

* Reference population children aged <16 years (see Sawczenko et al.)

Conclusion: Incidence of IBD in JIA patients is higher than expected. Especially patients with ERA, PsA and eoJIA  are at risk. IBD seems to be associated with treatment of ETA, LEF and SUL, while MTX turned out to be protective. The impact of discontinuation of MTX, AZA and corticosteroids in patients later on developing IBD has to be discussed. Unfortunately case numbers so far are too small for further statistical analyses including linear regression models.

---

« Back to 2013 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/inflammatory-bowel-disease-in-juvenile-idiopathic-arthritis-patients-upon-biologics/