乌鸦传媒

Treating inflammatory bowel disease (IBD) with tumor necrosis factor (TNF) inhibitors was paradoxically associated with an increased risk for other immune-mediated inflammatory diseases (IMIDs) in a large national study.

The study included large cohorts of patients with IBD in Denmark and France -- more than 100,000 individuals -- and found that use of TNF inhibitors was associated with a 76% increase in risk for rheumatoid arthritis, psoriasis, and hidradenitis suppurativa (HR 1.76, 95% CI 1.63-1.91), Daniel Ward, PhD, of Aalborg University in Copenhagen, Denmark, and colleagues reported.

As shown in their study in , the risk estimates were similar when the researchers separately examined the Danish (HR 1.66, 95% CI 1.34-2.07) and French (HR 1.78, 95% CI 1.63-1.94) cohorts, which suggested that the findings were robust, the team said. The analyses were adjusted for multiple potential confounders, including sex, IBD subtype and severity, IBD-related procedures, and various comorbidities and medications.

The researchers further assessed the strength of their findings by conducting an active comparator analysis of TNF inhibitor monotherapy versus azathioprine monotherapy. TNF inhibitor use was associated with a significantly increased risk for IMIDs when compared with use of azathioprine (HR 2.94, 95% CI 2.33-3.70).

"Our findings are unexpected and warrant further research into the mechanisms behind the paradoxical anti-TNF-caused occurrence of diseases that are normally indications for anti-TNF," Ward and colleagues wrote. "If correct, the paradoxical effect of anti-TNF could have serious clinical implications."

A spokesperson for the American Gastroenterological Association, Joseph Feuerstein, MD, of Beth Israel Deaconess Medical Center and Harvard Medical School in Boston, urged caution, however, in interpreting the results.

"This study does not show a true causal relationship between anti-TNF and the IMIDs but rather an association between taking them and developing the IMIDs," Feuerstein, who was not involved with the study, told 乌鸦传媒 via email. "Patients should not stop their anti-TNF thinking they are at risk of IMIDs as a result of the anti-TNF. More studies are needed to understand this finding and its meaning in our disease management."

"The explanation for why anti-TNFs would 'bring out' these IMIDs is unclear," he continued. "Patients with IBD are at risk of other IMIDs, but in these cohorts, when compared to other drug classes (thiopurines – azathioprine), they did not see the same risk with that exposure."

Most IMIDs are "two hit" processes, Feuerstein explained. They require the underlying genetics, which many patients with IBD already have, and then a second change to bring these disease processes out. "It is conceivable that altering the immune system may be bringing some of these disease processes out 'earlier' than they might have," he said.

Another possibility, Ward's group suggested, is that TNF inhibitors might dysregulate the immune system through altered paracrine signaling, despite the anti-inflammatory action. In addition, a reported a link between anti-TNF exposure and demyelinating diseases of the central nervous system in patients with IBD, further suggesting that TNF inhibitors may alter regulation of the immune system in susceptible individuals.

Ward and colleagues analyzed data from Danish national health registries and the French National Health Insurance database from 2005 through 2018. The main analyses included 18,258 patients with IBD from Denmark and 88,786 from France, half of whom had received TNF inhibitors and half of whom had not.

In the Danish cohort, the median age at entry was 38 years, 47% were male, 50% had Crohn's disease (CD), and 50% had ulcerative colitis (UC). In the French cohort, the median age was 36 years, 47% were male, 68% had CD, and 32% had UC.

The researchers used Cox models to estimate hazard ratios for the association between TNF inhibitor exposure and IMIDs, and then pooled the estimates from the two cohorts. The primary outcome was a composite of rheumatoid arthritis, psoriasis, and hidradenitis suppurativa. Secondary outcomes were each of these diseases analyzed separately, with associations seen for each (pooled data shown here):

• Rheumatoid arthritis: HR 1.47 (95% CI 1.25-1.73)
• Psoriasis: HR 1.83 (95% CI 1.66-2.03)
• Hidradenitis suppurativa: HR 2.12 (95% CI 1.65-2.72)

A chief limitation of the study was a lack of information on smoking, which is an important risk factor for psoriasis in patients with IBD, Ward and co-authors said. "Our e-value estimate suggests that unmeasured confounding of moderate strength could nullify the association observed. This should encourage careful interpretation of study findings with particular regard to the differential effect of smoking on CD and UC."

The team noted: "The paradoxical effect of anti-TNF suggested by the present study might also occur in the setting of other inflammatory diseases, and further cohort studies of patients with rheumatoid arthritis or psoriasis exposed to anti-TNF are needed."

Comment