MADRID – Several novel risk associations with psoriasis progression were found to differ by sex, and collectively appeared to implicate infections and the “stress response” as a trigger of psoriatic arthritis.
Mitchel L. Zoler/MDedge News
Dr. Alexis Ogdie
The findings come from a risk factor analysis of a U.S. claims database of more than 200,000 adults with psoriasis including more than 4,000 patients who progressed to psoriatic arthritis during nearly 6 years of follow-up.
The new analysis confirmed several previously described risk associations linked with progression to psoriatic arthritis (PsA) that have roughly equal impact on both women and men: fatigue, obesity, and depression, Alexis Ogdie, MD, said at the European Congress of Rheumatology. The new findings also showed several novel, sex-specific associations. In women, these associations included salmonella infection, sepsis, and uveitis; in men, they included gangrene, encephalitis, and hidradenitis suppurativa.
The links with various infections were generally rare; they showed strong nominal associations in multivariate analyses but with wide confidence limits. The findings suggest that events that induce major stress responses, such as infections, often preceded the progression of psoriasis to a diagnosis of PsA, said Dr. Ogdie, director of the psoriatic arthritis clinic at the University of Pennsylvania in Philadelphia. Other, noninfectious clinical features that significantly linked with PsA development but at a lower magnitude included anemia and diabetes in women, and irritable bowel syndrome and venous thromboembolism in men.
Dr. Ogdie cautioned that the findings were preliminary and need confirmation in different data sets, as well as in additional subgroup analyses of the data used in the current analysis, taken from the electronic medical records of 215,386 U.S. residents diagnosed with psoriasis in the Optum medical-claims database for 2006-2017.
The analysis focused on patients who received a second diagnostic code in their EMR for psoriasis during the 12 months after the index psoriasis entry. The identified group averaged 50 years old; 55% of the psoriasis patients were women, and 86% were white.
During the year after their first diagnostic-code entry for psoriasis, 4.6% of the patients received a biological drug and 4.2% received an oral drug for their psoriasis. During 5.6 years of follow-up, 4,288 patients (2%) developed PsA, a rate of 3.5 cases/1,000 patient-years. Dr. Ogdie noted that prior studies have documented the challenge of diagnosing PsA in patients with psoriasis, so this may be a conservative estimate of the progression rate.
The researchers assessed possible linkage with PsA progression for more than 250 different entries in the EMR, but the analysis was limited by the absence of measures of rheumatoid susceptibility, such as immunologic markers, which were not included in the EMR. In multivariate analysis of the full cohort, fatigue at baseline was linked with a 77% higher rate of progression to PsA, obesity was linked with a 48% higher rate, and depression with a 29% higher rate of progression when compared with psoriasis patients without each of these factors. All three differences were statistically significant. Dr. Ogdie cited an article she recently coauthored that detailed the background to this approach in studying the etiology of PsA (Nat Rev Rheumatol. 2019 March;15:153-66).
This is the first study to report sex-linked differences in clinical measures that link with progression to PsA, Dr. Ogdie noted. In women, salmonella infection linked with a 9-fold higher rate of PsA development compared with women with psoriasis without salmonella infection, women with uveitis had a 2.9-fold higher rate of PsA development, and those with sepsis had a 2.4-fold increased rate of PsA. Among men, those with gangrene, encephalitis, or hidradenitis suppurativa each had a greater than 4-fold higher rate of developing PsA, and men with osteomyelitis had a 2.7-fold increase.
All these between-group differences were statistically significant. But because each of these was a relatively rare event, the confidence intervals around these point estimates were wide. For example, in women with salmonella infection from a statistical standpoint the possible range of increased risk could be anywhere from 1.3 to 66. The analysis identified among women and men several additional sex-specific risk associations that were statistically significant but with smaller point estimates.
SOURCE: Ogdie A et al. Ann Rheum Dis. Jun 2019;78(Suppl 2):131-2. Abstract OP0115. doi: 10.1136/annrheumdis-2019-eular.4390.