1609: Comorbidity Clusters in Ankylosing Spondylitis and Their Association with Disease Activity and Functional Impairment: Data from the PSOAS Cohort

Paras Karmacharya¹, Cynthia Crowson², Ryan Lennon³, Dilli Poudel⁴, John Davis³, Alexis Ogdie⁵, Jean Liew⁶, Michael Ward⁷, Mariko Ishimori⁸, Michael Weisman⁹, Matthew Brown¹⁰, Mohammad Rahbar¹¹, Mark Hwang¹², John Reveille¹³ and Lianne Gensler¹⁴, ¹Vanderbilt University, Nashville, TN, ²Mayo Clinic, Eyota, MN, ³Mayo Clinic, Rochester, MN, ⁴Indiana Regional Medical Center, Indiana, PA, ⁵Department of Medicine, Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA, ⁶Boston University, Boston, MA, ⁷National Institutes of Health, Rockville, MD, ⁸Cedars-Sinai Health System, Los Angeles, CA, ⁹Adjunct Professor of Medicine, Stanford University; Distinguished Professor of Medicine Emeritus, David Geffen School of Medicine at UCLA, Los Angeles, CA, ¹⁰Genomics England, London, United Kingdom, ¹¹University of Texas Health Science Center at Houston, Houston, TX, ¹²McGovern Medical School at the University of Texas Health Science Center at Houston, Houston, TX, ¹³University of Texas McGovern Medical School, Houston, TX, ¹⁴Department of Medicine, Division of Rheumatology, University of California San Francisco, San Francisco, CA

Background/Purpose: Comorbidities occur more frequently in patients with ankylosing spondylitis (AS) than in the general population and are associated with higher morbidity and mortality. Some comorbidities may be associated, making one more likely in the presence of another, and different combinations of comorbidities may have differential considerations for AS management and outcomes. The objective of our study was to examine the association of clusters of comorbidities at baseline with disease activity and functional status over time in AS.

Methods: We performed a longitudinal study from the Prospective Study Of Ankylosing Spondylitis (PSOAS) cohort, a multicenter, prospective cohort from five centers (4 in the US, 1 in Australia). AS patients ≥ 18 years fulfilling modified New York criteria for AS (study period 2002-20) were included. Patient-reported AS comorbidities (N=28) and extra-musculoskeletal manifestations (EMMs) within 3 years of enrollment (pre-specified on the baseline case-report form) which occurred in ≥1%, were included. Undocumented comorbidities were assumed to be absent if missing in < 15% of patients, and those missing in >50% of patients were excluded. Comorbidity clusters were identified using K-median clustering. The optimal number of clusters was determined using a scree plot of the sum of squared errors. Baseline characteristics of the clusters were compared. Comorbidities and EMMs were collected at baseline and every 2 years, and the disease activity (ASDAS-CRP) and functional status (BASFI) measures were collected every 6 months. Generalized estimating equation models with an auto-regressive correlation structure (to account for multiple observations per patient) were used to examine the associations between comorbidity clusters and measures of disease activity and functional status over time.

Results: We included 1,270 AS patients (9,885 visits) with a mean age of 44.6 ±14.3 years, 74.4% male, and 80.1% white. Mean AS symptom duration was 21.6±14 years, 83% were HLA-B27 positive, and CRP was elevated in 42% of patients at baseline. The median follow-up was 2.9 years (IQ range: 1.0-6.8 years). Depression was the most prevalent comorbidity (33%) followed by hypertension (28%); uveitis was the most common EMM (34%). We identified five comorbidity clusters: depression (n=345, 27%), no comorbidities (n=281, 22%), hypertension (n=266, 21%), uveitis (n=249, 20%), and miscellaneous comorbidities (n=129, 10%) (Figure). The cluster with no comorbidities was significantly younger, with a lower symptom duration (p< 0.001). Depression and uveitis clusters had a significantly higher number of females (33% and 34% respectively) compared to the cluster with no comorbidities (20%). Patients in the depression cluster tended to have more comorbidities and worse disease activity and functional status compared to those with no comorbidities (Table).

Conclusion: Distinct comorbidity clusters were identified in AS patients in the PSOAS cohort. In addition to the number of comorbidities, the type of comorbidity seems to be important to longitudinal outcomes in AS. The depression cluster was associated with worse disease activity and function.
Figure. Baseline comorbidity clusters in ankylosing spondylitis from the Prospective Study Of Ankylosing Spondylitis (PSOAS) cohort.

Table. Age, sex, race, time from baseline adjusted associations between comorbidity clusters, compared to cluster 2 (no comorbidities), and disease activity/ functional status measures over time in ankylosing spondylitis based on generalized linear models.
Disclosures: P. Karmacharya, None; C. Crowson, None; R. Lennon, None; D. Poudel, None; J. Davis, Pfizer; A. Ogdie, AbbVie, Amgen, Novartis, Pfizer Inc, Bristol-Myers Squibb, Celgene, Janssen, CorEvitas, Gilead Sciences, Eli Lilly, GlaxoSmithKline, Happify Health, UCB; J. Liew, None; M. Ward, None; M. Ishimori, None; M. Weisman, None; M. Brown, UCB Pharma, Pfizer, Clementia, Ipsen, Incyte, Regeneron, Grey Wolf Therapeutics, Xinthera, Novartis; M. Rahbar, None; M. Hwang, Novartis; J. Reveille, Eli Lilly; L. Gensler, Novartis, Pfizer Inc, UCB Pharma, AbbVie, Eli Lilly, Janssen, Gilead, Moonlake.