Abstract
Objectives
This study compared the incidence and time-to-event outcomes of ocular extraintestinal manifestations (O-EIMs) and EIMs among patients with human leukocyte antigen (HLA)-B27-associated diseases receiving different classes of immunotherapy.
Methods
A retrospective cohort study was conducted using aggregated electronic health records from the TriNetX network between January 1, 2014, and December 31, 2024. Patients with HLA-B27-associated diseases were included if they were newly prescribed tumor necrosis factor (TNF), janus kinase (JAK), or interleukin (IL) inhibitors on or after their initial diagnosis. The cumulative incidence of O-EIMs and EIMs was calculated per 100 person-years. Kaplan–Meier estimates were used to graph time-to-event outcomes using R Studio (version 4.4.1).
Results
We identified 22,287 patients [mean age 47.2 (16.8) years; 13,132 (58.9) % women]. The cumulative incidence of anterior uveitis from the index date was 0.11 cases per 100 person-years. According to first medication prescribed, the incidence of anterior uveitis was lowest in patients prescribed IL inhibitors, followed by JAK and TNF inhibitors (0.07%, 0.10%, and 0.20%, cases per 100 person-years respectively). Time-to-event analysis showed a higher likelihood of anterior uveitis with initial TNF inhibitor use, although the difference was not statistically significant. Regarding EIMs, incident sacroiliitis was highest with TNF (0.91%), followed by JAK (0.72%) and IL (0.37%). Psoriatic dermatitis incidence was highest with JAK inhibitors (2.85%), compared to TNF (2.19%) and IL (2.17%).
Conclusion
In this cohort study of patients with HLA-B27-associated diseases, the lack of statistical significance indicates similar effectiveness across immunotherapy classes in preventing O-EIMs and EIMs.
Keywords: Extraintestinal manifestations, HLA-B27-associated diseases, Immunotherapy, Ocular extraintestinal manifestations
Introduction
The human leukocyte antigen B27 (HLA-B27)-associated diseases are a group of immune-mediated inflammatory conditions strongly linked to the presence of the HLA-B27 allele. They fall under the broader category of seronegative spondyloarthropathies, characterized by the absence of elevated serum rheumatoid factor [1, 2]. Among individuals who are HLA-B27 positive, groups of diseases may develop such as ankylosing spondylitis, reactive arthritis, psoriatic arthritis, or inflammatory bowel disease, all of which share common pathophysiologic mechanisms and genetic predisposition [1]. They share clinical features such as axial and peripheral arthritis, enthesitis, dactylitis, and a tendency for extra-articular involvement [2].
HLA-B27 may influence autoimmune disorders by affecting immune responses to certain microbial antigens. It is linked to both intestinal tolerance and loss of ocular immune privilege, but its exact mechanism remains unclear [3]. Dermatologic, cardiovascular, musculoskeletal, oral, hepatopancreatobiliary, metabolic, renal, and ophthalmic systems are established extraintestinal manifestations (EIMs) [4]. Ocular involvement, particularly anterior uveitis, is one of the most common ocular EIMs (O-EIMs) [2], accounting for approximately 90% of presentations [5, 6]. Less typical presentations are also documented such as development of posterior segment inflammation [1, 7], as well as episcleritis and scleritis [2]. Notably, some patients presenting with idiopathic ocular diseases have positive HLA-B27, and ocular inflammation may precede systemic symptoms, serving as a valuable early indicator of underlying disease [8].
Early detection and specific therapy are fundamental to avoiding sight-threatening complications [9, 10]. Apart from sufficient treatment with recommended disease-modifying antirheumatic drugs (DMARDs) [11, 12], several strategies have been employed to use approved agents like tumor necrosis factor (TNF) [13–15] or interleukin (IL) inhibitor [14–18]. Additionally, a newer class of medication, janus kinase (JAK) inhibitors, which first gained food and drug administration approval in 2018 for ulcerative colitis, provide another treatment option for patients with moderate to severe IBD [19], particularly when failure and intolerance of existing therapies remains high [18, 20]. Moreover, patients with psoriatic arthritis refractory to biologic-DMARDs, showed great improvement in their clinical manifestations when JAK inhibitor was used [21]. Nonetheless, their use in ocular inflammatory disorders remain off-label.
These agents have demonstrated a plausible effect in controlling the HLA-B27-associated diseases, as well as preventing the episodes of O-EIMs [13–20]. However, no large-scale studies have evaluated the comparative risk of O-EIMs and EIMs in patients with HLA-B27-associated diseases treated with different immunotherapy agents. Addressing this gap could provide critical insights for optimizing therapeutic strategies and improving patient outcomes. The objective of this study was to compare the risk of incidents and EIMs among patients with HLA-B27-associated disease receiving different class of immunotherapy.
Methods
Study population
In this retrospective cohort study, data of patients were derived from TriNetX, an electronic health records (EHR) research network comprising multiple large health care organizations within the United States and globally. This research network includes deidentified EHR data of more than 150 million patients within 93 health care organizations across 93 healthcare organizations in four countries (United States, Taiwan, Georgia, and Brazil), with 69 of them located in the United States. TriNetX is compliant with the Health Insurance Portability and Accountability Act and is certified to the International Organization for Standardization 27,001:2013 standard. All data from the TriNetX platform are displayed in aggregate form. This study only contained deidentified patient electronic medical records, therefore it was exempted from institutional review board approval and patient consent was not required for this research.
The most updated information for this study was gathered on December 31, 2024, from the TriNetX Health Research Network, which granted access to patients’ demographic information, electronic medical records (EMRs) diagnoses, procedures, medications, laboratory values, and genomic information. At the time of data retrieval, the TriNetX database encompassed a total of 167,367,388 patients (Fig. 1). Our study adheres to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for reporting cohort studies. The research employed a retrospective cohort design, involving the examination of EMRs. The platform utilizes standardized coding systems, including International Classifications of Diseases, Tenth Revision (ICD-10), and Current Procedural Terminology (CPT) codes for diagnoses and procedures.
Fig. 1.
Flow diagram of patient selection from TriNetX
Study design
This study was conducted from January 1 st, 2014, to December 31 st, 2024. Subjects who were newly diagnosed with HLA-B27-associated diseases (ankylosing spondylitis, reactive arthritis, psoriatic arthritis, Crohn’s disease, and ulcerative colitis) and were prescribed a class of immunotherapy were included in the study. Prescribed class of immunotherapies were TNF inhibitor (infliximab, adalimumab, golimumab, or certolizumab), JAK inhibitor (tofacitinib, upadacitinib, or filgotinib), or IL inhibitors (tocilizumab, ustekinumab, risankizumab, tildrakizumab, guselkumab, siltuximab, or sarilumab).
The index date was defined by the first prescription of a class of immunotherapy (TNF, JAK, or IL inhibitors) that occurred on or after the date when a patient received their first HLA-B27-associated disease diagnosis code. Any patients who were prescribed immunotherapy before receiving their first diagnostic code or had less than 1 year of follow-up time before the index date were excluded. Based on ICD-10 coding, patients with any prior history of uveitis (anterior, intermediate, posterior, and panuveitis), scleritis, episcleritis, colon cancer, colectomy, sacroiliitis, psoriatic dermatitis, urethritis within 12 months prior to HLA-B27-associated disease diagnosis were excluded.
Demographic data were extracted from the TriNetX database and included age at the index date, sex (male, female, or unknown, as coded in TriNetX), race (White, Black or African American, Asian, or other/unknown), ethnicity (Hispanic/Latino, non-Hispanic/non-Latino, unknown), risk factors (inflammatory polyarthritis, erythema nodosum, pyoderma gangrenosum, nicotine dependence, disorder of lipid metabolism, diabetes mellitus, and hypertension), DMARDs treatment before the index date (mesalamine and sulfasalazine, methotrexate, and azathioprine) and prescriptions for different immunotherapies as the first and second line of treatment (TNF, JAK, or IL inhibitors). Race, ethnicity, and sex were determined based on the presence of these designations within the EMR.
The primary analysis assessed the incident rate of O-EIMs including uveitis (anterior uveitis, intermediate, posterior and panuveitis), scleritis, and episcleritis and EIMs comprising colon cancer, colectomy, sacroiliitis, psoriatic dermatitis, and urethritis in patients with the most common International Classification Disease (ICD)−10 categories (Supplementary Table 1).
Outcome
The primary outcome was the first occurrence of any diagnostic code for O-EIMs and EIMs starting from 2 weeks after the index date. The secondary outcome was also the first occurrence of any diagnostic code for EIMs starting from 2 weeks after the index date.
Statistical analysis
TriNetX provides real-time live analytics on its platform. Continuous variables were normally distributed and are represented as mean (SD), while categorical variables are reported as counts and percentages. The primary outcome was time to incident O-EIMs and EIMs. The cumulative incidence of uveitis (anterior, intermediate, posterior, and panuveitis), scleritis, episcleritis, colon cancer, colectomy, sacroiliitis, psoriatic dermatitis, and urethritis were calculated per 100 person-years. A Cox proportional hazard model was used to determine hazard ratios (HRs), and Kaplan–Meier estimates were used to graph time-to-event outcome with log-rank test using R Studio (version 4.4.1). Significance tests were 2-sided and paired. P values less than 0.05 were considered significant, and 95% confidence intervals (CIs) were calculated.
Results
Characteristics of the study population
A cohort comprising 167,367,388 patients was identified within the TriNetX Global Collaborative Network. After excluding patients based on the pre-stated exclusion criteria, a total of 22,287 patients with HLA-B27-associated diseases and receiving immunotherapy were identified (Fig. 1). Table 1 demonstrates the baseline characteristics of patients prescribed immunotherapies separated by class. The average age of patients diagnosed with HLA-B27-associated diseases was 47.2 (16.8) years. More than half of the patients in the cohort were female (58.9%), the majority identified as White (78.8%), and non-Hispanic/non-Latino ethnicity (73.6%).
Table 1.
Demographic and clinical characteristics of patients with HLA-B27 associated diseases
| Characteristics | Overall cohort (n = 22,287) | Immunotherapy categories |
||
|---|---|---|---|---|
| TNF inhibitor (n = 5,672) | JAK inhibitor (n = 9,282) | IL inhibitor (n = 7,333) | ||
| Age at diagnosis, mean (SD) | 47.2 (16.8) | 45.1 (16.2) | 48.6 (17.4) | 46.7 (17.0) |
| Sex, n (%) Male |
8,496 (38.1) | 1,786 (31.5) | 3,852 (41.5) | 2,858 (39.0) |
| Female | 13,132 (58.9) | 3,750 (66.1) | 5,242 (56.5) | 4,140 (56.4) |
| Unknown | 659 (3.0) | 136 (2.4) | 188 (2.0) | 335 (4.6) |
| Race/ethnicity, n (%) White |
17,561 (78.8) | 4,358 (76.8) | 7,179 (77.3) | 6,024 (82.2) |
| Black/African American | 1,307 (5.9) | 330 (5.8) | 535 (5.8) | 442 (6.0) |
| Asian | 563 (2.5) | 159 (2.8) | 305 (3.3) | 99 (1.3) |
| Other/unknown race | 2,856 (12.8) | 825 (14.6) | 1,263 (13.6) | 768 (10.5) |
| Hispanic/Latino | 1,105 (5.0) | 247 (4.3) | 598 (6.4) | 260 (3.5) |
| Non-Hispanic/non-Latino | 16,410 (73.6) | 4,249 (75.0) | 6,646 (71.6) | 5,515 (75.2) |
| Unknown ethnicity | 4,772 (21.4) | 1,176 (20.7) | 2,038 (22.0) | 1,558 (21.3) |
| Prior history of medications, n (%) Mesalamine |
6,086 (27.3) | 1,170 (20.6) | 2,918 (31.4) | 1,998 (27.2) |
| Sulfasalazine | 2,742 (12.3) | 754 (13.3) | 1,348 (14.5) | 640 (8.7) |
| Azathioprine | 3,345 (15.0) | 936 (16.5) | 1,124 (12.1) | 1,285 (17.5) |
| Methotrexate | 5,900 (26.5) | 1,579 (28.0) | 2,900 (31.2) | 1,421 (19.4) |
| Risk factors, n (%) Inflammatory polyarthropathies |
7,010 (31.4) | 2,105 (37.1) | 3,410 (36.7) | 1,495 (20.4) |
| Erythema nodosum | 225 (1.0) | 64 (1.1) | 73 (0.8) | 88 (1.2) |
| Pyoderma gangrenosum | 198 (0.8) | 64 (1.1) | 65 (0.7) | 69 (0.9) |
| Nicotine dependence | 3,084 (13.8) | 740 (13.0) | 1,174 (12.6) | 1,170 (16.0) |
| Disorder of lipid metabolism | 6,455 (29.0) | 1,305 (23.0) | 3,225 (34.7) | 1,925 (26.2) |
| Diabetes mellitus | 3,005 (13.5) | 623 (11.0) | 1,477 (16.0) | 905 (12.3) |
| Hypertension | 7,279 (32.6) | 1,588 (28.0) | 3,425 (37.0) | 2,266 (30.9) |
| First line of treatment, n (%) TNF inhibitor |
7,333/22,287 (26.1) | 7,333 (100) | - | - |
| JAK inhibitor | 6,652/22,287 (42.6) | - | 6,652 (100) | - |
| IL inhibitor | 5,013/22,287 (31.2) | - | - | 5,013 (100) |
| Second line of treatment, n (%) * TNF inhibitor |
2,011/11,622 (16.8) | 2,011 (100) | - | - |
| JAK inhibitor | 3,592/11,622 (29.2) | - | 3,592 (100) | - |
| IL inhibitor | 5,988/11,622 (53.8) | - | - | 5,988 (100) |
10,665 (48%) patients do not have a pathway and 31 (0.2%) patients had other pathways
The most common risk factor was presence of inflammatory polyarthropathies (31.4%), followed by hypertension (32.6%), disorder of lipid metabolism (29.0%), nicotine dependence (13.8%), diabetes mellitus (13.5%), erythema nodosum (1.0%), and pyoderma gangrenosum (0.8%). Conventional medication prescribed before the index date were mesalamine (27.3%), methotrexate (26.5%), azathioprine (15.0%), and sulfasalazine (12.3%). The most common immunotherapy prescribed initially after the diagnosis of HLA-B27-associated disease was JAK inhibitor (41.6%), followed by IL inhibitor (33.0%), and TNF inhibitor (25.4%). Among 11,622 patients who switched therapy, the most common second immunotherapy prescribed was IL inhibitor (53.8%) followed by JAK inhibitor (29.2%), and TNF inhibitor (16.8%) (Table 1). The average duration of immunotherapy before switching was 657.6 (581.8) days for IL inhibitors, 551.2 (472.1) days for JAK inhibitors, and 1,362.8 (1,130.0) days for TNF inhibitors.
Incidence of ocular ‑extraintestinal manifestations with different immunotherapies
During the study period, out of 22,287 patients with HLA-B27-associated diseases, 263 patients (0.11 per 100 person-years) developed anterior uveitis, while 36 patients (0.02 per 100 person-years) experienced intermediate, posterior, or panuveitis. The incidence of anterior uveitis per 100 person-years was 0.20 for TNF inhibitors (115 events in 56,720 person-years), 0.10 for JAK inhibitors (92 events in 92,820 person-years), and 0.07 for IL inhibitors (56 events in 73,330 person-years). The incidence of intermediate/posterior/panuveitis was 0.02 cases per 100 person-years for patients prescribed TNF inhibitors (24 events in 56,720 person-years), 0.02 cases per 100 person-years for patients prescribed JAK inhibitors (16 events in 92,820 person-years), and 0.01 cases per 100 person-years for patients prescribed IL inhibitors (10 events in 73,330 person-years) (Table 2).
Table 2.
Cumulative incidence of ocular and extra-intestinal manifestations in patients with HLA-B27-associated diseases
| Cumulative incidence of O-EIM and EIM, per 100 person-years (events/ person-years) | Overall cohort (n = 22,287) | First biologic or JAK inhibitor class prescribed |
||
|---|---|---|---|---|
| TNF inhibitor (n = 5,672) | JAK inhibitor (n = 9,282) | IL inhibitor (n = 7,333) | ||
| Ocular-extraintestinal manifestation Anterior uveitis |
0.11 (263/222,870) | 0.20 (115/56,720) | 0.10 (92/92,820) | 0.07 (56/73,330) |
| Intermediate/posterior/panuveitis | 0.02 (36/222,870) | 0.02 (10/56,720) | 0.02 (16/92,820) | 0.01 (10/73,330) |
| Episcleritis | 0.02 (55/222,870) | 0.04 (24/56,720) | 0.02 (21/92,820) | 0.01 (10/73,330) |
| Scleritis | 0.03 (77/222,870) | 0.06 (36/56,720) | 0.03 (30/92,820) | 0.01 (11/73,330) |
| Extraintestinal manifestation Colectomy |
0.30 (670/222,870) | 0.30 (156/56,720) | 0.21 (195/92,820) | 0.43 (319/73,330) |
| Colon cancer | 0.07 (173/222,870) | 0.07 (41/56,720) | 0.08 (72/92,820) | 0.08 (60/73,330) |
| Sacroiliitis | 0.65 (1,463/222,870) | 0.91 (520/56,720) | 0.72 (668/92,820) | 0.37 (275/73,330) |
| Psoriatic dermatitis | 2.46 (5,483/222,870) | 2.19 (1,242/56,720) | 2.85 (2,646/92,820) | 2.17 (1,595/73,330) |
| Urethritis | 0.04 (99/222,870) | 0.04 (25/56,720) | 0.05 (44/92,820) | 0.04 (30/73,330) |
O-EIM Ocular extra-intestinal manifestation, EIM Extraintestinal manifestation
Additionally, episcleritis was observed in 55 patients (0.02 per 100 person-years). Episcleritis incidence was 0.04 cases per 100 person-years for TNF inhibitors (24 events in 56,720 person-years), 0.02 cases for JAK inhibitors (21 events in 92,820 person-years), and 0.01 cases for IL inhibitors (10 events in 73,330 person-years) (Table 2).
Scleritis was reported in 77 patients (0.03 per 100 person-years) (Table 2). Scleritis incidence was 0.06 cases per 100 person-years for TNF inhibitor users 36 events in 56,720 person-years), 0.03 for JAK inhibitor users (30 events in 92,820), and 0.01 for IL inhibitor users (11 events in 73,330) (Table 2).
Even though the number of incident anterior uveitis was slightly higher in patients prescribed TNF inhibitor, in analysis of time to incident, the probability was not significantly different compared to other immunotherapies. Time to incident anterior uveitis after initiation of immunotherapies is demonstrated in Fig. 2. The incident intermediate/posterior/panuveitis and scleritis were slightly higher in JAK inhibitor with no statistically significant differences among different immunotherapies. Episcleritis and scleritis were more frequently observed in patients on TNF inhibitors, however this difference was not statistically significant as well.
Fig. 2.
Time to incident anterior uveitis among patients with HLA-B27-associated diseases initiating immunotherapies
Incidence extraintestinal manifestations in different immunotherapies
The most EIM reported in descending order were psoriatic dermatitis with an incidence of 2.46 (5,483 events in 222,870 person-years), sacroiliitis at 0.65 (1,463 events in 222,870 person-years), colectomy at 0.30 (670 events in 222,870 person-years), colon cancer at 0.07 (173 events in 222,870 person-years), and urethritis at 0.04 (99 events in 222,870 person-years) (Table 2). Time to psoriatic dermatitis and sacroiliitis are also demonstrated in Figs. 3 and 4, respectively. Even though the rates of EIMs varies in different immunotherapies, the incidence remain statistically insignificant.
Fig. 3.
Time to incident psoriatic dermatitis among patients with HLA-B27-associated diseases initiating immunotherapies
Fig. 4.
Time to incident sacroiliitis among patients with HLA-B27-associated diseases initiating immunotherapies
Discussion
In this large retrospective cohort study using real-world electronic health record data from the TriNetX network, we investigated the incidence of O-EIMs (anterior uveitis, intermediate/posterior/panuveitis, episcleritis, and scleritis) and EIMs (colon cancer, colectomy, sacroiliitis, psoriatic dermatitis, and urethritis) in patients with HLA-B27-associated diseases receiving different immunotherapies. Our findings indicate that the overall occurrence of O-EIMs and EIMs were low across all treatment class. In addition, while anterior uveitis was frequently observed in patients prescribed TNF inhibitors, the time-to-event analysis demonstrated no significant difference in the probability of O-EIMs development among the different immunotherapies.
Our findings demonstrate that anterior uveitis was the most frequently observed O-EIMs, with a slightly higher incidence in patients receiving TNF inhibitors; however, its cumulative incidence remained low across all treatment classes. The relationship between TNF inhibitors and ocular inflammation has been variably reported in the literature. Some studies suggest that TNF inhibitors may delay uveitis flares when compared to placebo, while others have shown a significant reduction in ocular inflammation among patients with refractory noninfectious uveitis treated with TNF inhibitors [22–24]. In contrast, Barberio et al. found no statistically significant difference in uveitis risk between TNF inhibitors and nonbiologic agents in patients with inflammatory bowel disease (IBD) [13]. Consistent with our findings, Roche et al. observed rare anterior uveitis flares with both TNF and IL inhibitor treatments in patients with spondyloarthritis; however, unlike our results, their study demonstrated a protective effect of TNF inhibitors compared to IL inhibitors or placebo [25]. Despite these findings, there remains a scarcity of studies directly comparing the incidence of O-EIMs in patients with HLA-B27-associated diseases across various treatment with conventional DMARDs and immunotherapies.
While biologic agents and JAK inhibitors have demonstrated effectiveness in managing refractory scleritis associated with HLA-B27 [26], comprehensive data on their impact on scleritis and episcleritis are lacking. The results of our study align with previous research demonstrating a low overall incidence of scleritis and episcleritis reinforcing prior observations that with treatment, scleritis and episcleritis remains low [27, 28]. One factor contributing to the lower percentage of O-EIMs in our study compared to existing literature, is our exclusion of patients who had O-EIMs prior to HLA-B27-associated disease diagnosis. For instance, it has been shown that patients who develop O-EIMs are at two-fold increased risk of developing IBD [29–31]. However, in our study, we cannot determine whether pre-existing O-EIMs were directly related to HLA-B27 positivity. To address this potential bias, we implemented a one-year follow-up period before the index date to include only patients with no prior O-EIMs before their HLA-B27-associated disease diagnosis. Additionally, we adjusted for conventional treatment as a proxy for disease severity.
Although our findings showed slight variations in the incidence of EIMs across immunotherapy classes, these differences may reflect treatment selection biases based on underlying HLA-B27-associated disease phenotypes. Our study showed modestly higher incidence of sacroiliitis in patients receiving TNF inhibitor. One possible explanation is that this group had a greater prevalence of inflammatory polyarthropathies, prompting more frequent TNF inhibitor prescriptions. Previous studies have suggested that TNF inhibitors are more effective in managing both articular and ocular EMIs in IBD [32], which may have influenced prescribing patterns and, consequently, higher incidence of sacroiliitis in this class. Annunciato et al., reported that the presence of EIMs significantly influenced biologic selection in spondyloarthritis, with TNF inhibitors preferred for musculoskeletal and ocular symptoms, and JAK and IL inhibitors often chosen for dermatologic or intestinal involvement [33]. Consistent with this, our study also observed a higher incidence of psoriatic dermatitis among patients treated with JAK inhibitors. Although we could not determine the specific HLA-B27-associated diseases treated with JAK inhibitors, JAK inhibitors may have been more frequently prescribed to patients with psoriasis, which may explain the increased incidence of psoriatic dermatitis in this group. We could not determine the specific HLA-B27-associated diagnoses for which JAK inhibitors were prescribed. However, JAK inhibitors were more frequently used in patients with psoriasis, which may explain the higher incidence of psoriatic dermatitis in this group. Overall, while immunotherapy may decrease the risk of EIMs, the observed incidence may be due to individualizing treatment based on the HLA-B27-associated disease rather than the drug efficacy alone.
As O-EIMs and EIMs secondary to HLA-B27 has immune-mediated origin, it is plausible that immunotherapy may reduce risk of O-EIMs and EIMs. Pro-inflammatory cytokines play a fundamental role in disease pathophysiology and its associated complications, including ocular and systemic inflammation [17, 18, 20]. Among these, TNF-α, IL-12, IL-23, and IL-17 have been identified as key mediators of disease progression, and Janus kinase-signal transducer and activator of transcription (JAK-STAT) pathway is a crucial intracellular signaling mechanism involved in cytokine-mediated immune regulation. This pathway is particularly important in mediating the effects of IL-6, IL-12, and IL-23, which drive chronic inflammation in HLA-B27-associated diseases. The IL-12/23 pathway is a critical driver of Th1 and Th17 immune responses, which are implicated in both intestinal and ocular inflammation [17, 20, 34]. Based on our findings, this paper provides evidence that these targeted treatments may be effective in managing both O-EIMs and EIMs. These immunotherapies help control Th1 and Th17 cell activity, limiting cytokine-driven damage to ocular tissues which results in decreased incidence and severity of O-EIM and EIMs in patients with HLA-B27-associated diseases [5, 14, 35–39].
Our study has several limitations, primarily due to its retrospective design and the use of de-identified, aggregated data, which prevented validation through individual patient chart review. Additionally, we were unable to account for potential confounding factors such as socioeconomic status, as our analysis relied on coded electronic health records. The TriNetX platform also does not allow for the calculation of overlapping manifestations, limiting our ability to identify patients with multiple O-EIMs and EIMs. Moreover, the decision to prescribe immunotherapy for HLA-B27-associated disease is likely influenced by several unmeasured confounders, including disease severity and concomitant systemic diseases that may not be adequately captured in electronic health records. One notable limitation of this study is the potential confounding effect of HLA-B27 itself on the incidence of breakthrough uveitis, independent of immunotherapy class. While we attempted to exclude patients with a known history of uveitis prior to the initiation of immunotherapy, the limitations of the TriNetX platform prevented us from fully ensuring that all instances of HLA-B27-associated uveitis were excluded. It is important to note that the TriNetX database does not provide information on disease severity, limiting us to fully elucidate the reasons behind the slightly higher incidence of O-EIMs and EIMs in each treatment class. Another challenge is the under-recognition of O-EIMs due to their often subtle presentation and limited awareness among gastroenterologists, potentially leading to underreporting. As a result, the true prevalence of O-EIMs in our study population may be higher than observed. Additionally, some patients may have sought evaluation and treatment for ocular conditions outside of the included institutions, leading to incomplete documentation within the TriNetX database.
The key strengths of this study include a large and diverse population using TriNetX, which is a live, real-world database with new patient data added daily. This is the first study to evaluate and compare the time-to-event analyses of different classes of immunotherapy, allowing for direct comparisons of their incidence. Utilizing real-world data from routine clinical practice enables observations that may be more reflective of actual patient outcomes compared to clinical trials, which often have strict inclusion criteria. It can also provide a broader insight into the incidence of O-EIMs and EIMs in HLA-B27-associated diseases receiving different class of immunotherapies.
Our findings revealed that anterior uveitis was the most frequently observed O-EIMs across all therapeutic classes, with a slightly higher incidence noted in patients receiving TNF inhibitors. Although these differences did not reach statistical significance in time-to-event analyses. The low incidence of O-EIMs across all treatment groups suggests effective disease control with current immunotherapy classes. Moreover, it indicates that therapeutic decisions may need to be guided by individual patient characteristics and disease phenotype rather than broad class-based efficacy. Our findings also highlighted that EIMs, such as psoriatic dermatitis and sacroiliitis, were the most common non-ocular manifestations observed. Notably, the prolonged time on TNF inhibitors before switching therapies may reflect their sustained clinical efficacy in managing HLA-B27-associated diseases. Meanwhile, the variability in switching patterns among JAK and IL inhibitors suggests possible differences in therapeutic response or tolerability that requires further investigation.
Future studies with head-to-head comparisons of these classes of immunotherapy are necessary to validate our findings and refine treatment algorithms for HLA-B27-associated diseases.
Supplementary Material
Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s10067-025-07670-y.
Key Points.
Large-scale comparative study assessing time-to-incident O-EIMs and EIMs in patients with HLA-B27-associated diseases treated with TNF, JAK, or IL inhibitor therapies using real-world data from the TriNetX network.
Low overall incidence of O-EIMs and EIMs was observed across all immunotherapy classes, with no statistically significant difference in time-to-event outcomes, indicating similar effectiveness in preventing these manifestations.
Funding
This work was supported by the National Eye Institute P30 Vision Research Core Grant (P30-EY026877), Research to Prevent Blindness, and National Eye Institute (1K12EY033745).
Footnotes
Declarations
Ethical Approval This study only contained deidentified patient electronic medical records, therefore it was exempted from institutional review board approval.
Consent to Participate Not applicable.
Consent to Publish Not applicable.
Competing interests AK: Consultant–Immunovant, Genentech, Lassen, AmGen, Acelyrin. Independent Contracted Research Viridian Therapeutics. Sling Therapeutics, Kriya Therapeutics. QDN: Consultant–Genentech, Kriya Therapeutics, Regeneron. Research Support–Belite Bio, Genentech, Kriya Therapeutics, Priovant, Regeneron. PM: Consultant–Alcon, Amgen, Aura Biosciences, Castle Biosciences, Alcon, Genentech, Dutch Ophthalmic, Immunogen; Financial support–Genentech; Lecturer–Alcon. ER: Consultant–Apellis Pharmaceuticals, Iveric Bio, Genentech, Regeneron, Allergan, Zeiss, Google; Speaker– Apellis Pharmaceuticals, Genentech, Regeneron, Allergan. The other authors have disclosed no conflicts of interest.
Data Availability
Not applicable.
References
- 1.Dodds EM, Lowder CY, Meisler DM (1999) Posterior segment inflammation in HLA-B27+ acute anterior uveitis: clinical characteristics. Ocul Immunol Inflamm 7(2):85–92 [DOI] [PubMed] [Google Scholar]
- 2.Waheed NK (n.d.). HLA-B27 associated uveitis. Massachusetts Eye and Ear Infirmary. Unpublished manuscript [Google Scholar]
- 3.Rosenbaum JT, Asquith M (2018) The microbiome and HLA-B27-associated acute anterior uveitis. Nat Rev Rheumatol 14(12):704–713 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Alsoudi AF, Wai KM, Koo E, McConnell RA, Pham NH et al. (2024) Risk of intestinal complications, extraintestinal morbidity, and mortality in patients with Crohn’s disease and ocular involvement. Ocul Immunol Inflamm 32(10):2486–2493 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Chang JH, McCluskey PJ, Wakefield D (2005) Acute anterior uveitis and HLA-B27. Surv Ophthalmol 50(4):364–388 [DOI] [PubMed] [Google Scholar]
- 6.İnanç M, Şimşek M, Çakar Özdal MP (2019) Etiological and clinical characteristics of HLA-B27-associated uveitis in a tertiary referral center. Turk J Ophthalmol 49(1):10–14 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Or C, Lajevardi S, Ghoraba H, Park JH, Onghanseng N et al. (2023) Posterior segment ocular findings in HLA-B27 positive patients with uveitis: a retrospective analysis. Clin Ophthalmol 17:1271–1276 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Al-Ani HH, Sims JL, Niederer RL (2023) Long term complications and vision loss in HLA-B27 uveitis. Eye 37(8):1673–1677 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Troncoso LL, Biancardi AL, de Moraes HV Jr, Zaltman C (2017) Ophthalmic manifestations in patients with inflammatory bowel disease: a review. World J Gastroenterol 23(32):5836. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Migliorisi G, Vella G, Dal Buono A, Gabbiadini R, Busacca A et al. (2024) Ophthalmological manifestations in inflammatory bowel diseases: keep an eye on it. Cells 13(2):142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Bouzid N, Jamilloux Y, Chapurlat R, Pradat P, De Parisot A et al. (2020) Impact of systemic treatments on the course of HLA-B27-associated uveitis: a retrospective study of 101 patients. PLoS One 15(3):e0230560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Zu Hoerste MM, Walscheid K, Tappeiner C, Zurek-Imhoff B, Heinz C et al. (2018) The effect of methotrexate and sulfasalazine on the course of HLA-B27-positive anterior uveitis: results from a retrospective cohort study. Graefes Arch Clin Exp Ophthalmol 256(10):1985–1992 [DOI] [PubMed] [Google Scholar]
- 13.Barberio J, Kim SC, Roh M, Lewis JD, Desai RJ (2020) Risk of uveitis in patients with inflammatory bowel disease on immunosuppressive drug therapy. Crohns Colitis 360 2(3):otaa041. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Gossec L, Kerschbaumer A, Ferreira RJ, Aletaha D, Baraliakos X et al. (2024) EULAR recommendations for the management of psoriatic arthritis with pharmacological therapies: 2023 update. Ann Rheum Dis 83(6):706–719 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Ciccia F, McGonagle D, Thomas R, Marzo-Ortega H, Martin DA et al. (2025) Jak inhibition and axial spondyloarthritis: new steps on the path to understanding pathophysiology. Front Immunol 16:1488357. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.McLean MH, Neurath MF, Durum SK (2014) Targeting interleukins for the treatment of inflammatory bowel disease—what lies beyond anti-TNF therapy? Inflamm Bowel Dis 20(2):389–397 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Bourgonje AR, Ungaro RC, Mehandru S, Colombel J-F (2024) Targeting the IL-23 pathway in inflammatory bowel disease. Gastroenterology [DOI] [PubMed] [Google Scholar]
- 18.Tian Z, Zhao Q, Teng X (2024) Anti-IL23/12 agents and JAK inhibitors for inflammatory bowel disease. Front Immunol 15:1393463. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Olivera PA, Lasa JS, Bonovas S, Danese S, Peyrin-Biroulet L (2020) Safety of Janus kinase inhibitors in patients with inflammatory bowel diseases or other immune-mediated diseases: a systematic review and meta-analysis. Gastroenterology 158(6):1554–1573.e1512 [DOI] [PubMed] [Google Scholar]
- 20.Honap S, Agorogianni A, Colwill MJ, Mehta SK, Donovan F et al. (2024) JAK inhibitors for inflammatory bowel disease: recent advances. Frontline Gastroenterol 15(1):59–69 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Mease PJ, Lertratanakul A, Anderson JK, Papp K, Van den Bosch F et al. (2021) Upadacitinib for psoriatic arthritis refractory to biologics: SELECT-psa 2. Ann Rheum Dis 80(3):312–320 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Jaffe GJ, Dick AD, Brézin AP, Nguyen QD, Thorne JE et al. (2016) Adalimumab in patients with active noninfectious uveitis. N Engl J Med 375(10):932–943 [DOI] [PubMed] [Google Scholar]
- 23.Mackensen F, Heinz C, Jakob E, Grewing V, Lorenz H-M et al. (2018) Randomized controlled study to evaluate the efficacy of adalimumab in patients with different forms of refractory uveitis. Ocul Immunol Inflamm 26(7):1015–1022 [DOI] [PubMed] [Google Scholar]
- 24.Nguyen QD, Merrill PT, Jaffe GJ, Dick AD, Kurup SK et al. (2016) Adalimumab for prevention of uveitic flare in patients with inactive non-infectious uveitis controlled by corticosteroids (VISUAL II): a multicentre, double-masked, randomised, placebo-controlled phase 3 trial. Lancet 388(10050):1183–1192 [DOI] [PubMed] [Google Scholar]
- 25.Roche D, Badard M, Boyer L, Lafforgue P, Pham T (2021) Incidence of anterior uveitis in patients with axial spondyloarthritis treated with anti-TNF or anti-IL17A: a systematic review, a pairwise and network meta-analysis of randomized controlled trials. Arthritis Res Ther 23:1–11 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Gueudry J, Thorne JE, Bansie R, Braun J, Van Hagen PM et al. (2017) Biologic therapy for HLA-B27-associated ocular disorders. Ocul Immunol Inflamm 25(2):169–178 [DOI] [PubMed] [Google Scholar]
- 27.Cury DB, Moss AC (2010) Ocular manifestations in a community-based cohort of patients with inflammatory bowel disease. Inflamm Bowel Dis 16(8):1393–1396 [DOI] [PubMed] [Google Scholar]
- 28.Hedin C, Vavricka S, Stagg A, Schoepfer A, Raine T et al. (2019) The pathogenesis of extraintestinal manifestations: implications for IBD research, diagnosis, and therapy. J Crohns Colitis 13(5):541–554 [DOI] [PubMed] [Google Scholar]
- 29.King D, Chandan JS, Thomas T, Denniston AK, Braithwaite T et al. (2022) Risk of a subsequent diagnosis of inflammatory bowel disease in subjects with ophthalmic disorders associated with inflammatory bowel disease: a retrospective cohort analysis of UK primary care data. BMJ Open 12(5):e052833. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Vavricka SR, Rogler G, Gantenbein C, Spoerri M, Prinz Vavricka M et al. (2015) Chronological order of appearance of extraintestinal manifestations relative to the time of IBD diagnosis in the Swiss inflammatory bowel disease cohort. Inflamm Bowel Dis 21(8):1794–1800 [DOI] [PubMed] [Google Scholar]
- 31.Vavricka SR, Schoepfer A, Scharl M, Lakatos PL, Navarini A et al. (2015) Extraintestinal manifestations of inflammatory bowel disease. Inflamm Bowel Dis 21(8):1982–1992 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Ferretti F, Monico MC, Cannatelli R, Carmagnola S, Lenti MV et al. (2022) The impact of biologic therapies on extra-intestinal manifestations in inflammatory bowel disease: a multicenter study. Front Med 9:933357. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Annunciato DdR, Oliveira TL, Magalhães VO, Pinheiro MdM (2023) Extra-musculoskeletal manifestations driving the therapeutic decision-making in patients with Spondyloarthritis: a 12-month follow-up prospective cohort study. Adv Rheumatology 63:44. [DOI] [PubMed] [Google Scholar]
- 34.Herrera-deGuise C, Serra-Ruiz X, Lastiri E, Borruel N (2023) JAK inhibitors: a new dawn for oral therapies in inflammatory bowel diseases. Front Med 10:1089099. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Fröhlich F, Micheroli R, Hebeisen M, Kissling S, Bürki K et al. (2023) HLA-B27 as a predictor of effectiveness of treatment with TNF inhibitors in axial spondyloarthritis: data from the Swiss Clinical Quality Management Registry. Clin Rheumatol 42(5):1267–1274 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.van der Heijde D, Song IH, Pangan AL, Deodhar A, van den Bosch F et al. (2019) Efficacy and safety of upadacitinib in patients with active ankylosing spondylitis (SELECT-AXIS 1): a multicentre, randomised, double-blind, placebo-controlled, phase 2/3 trial. Lancet 394(10214):2108–2117 [DOI] [PubMed] [Google Scholar]
- 37.Sandborn WJ, Peyrin-Biroulet L, Quirk D, Wang W, Nduaka CI et al. (2022) Efficacy and safety of extended induction with tofacitinib for the treatment of ulcerative colitis. Clin Gastroenterol Hepatol 20(8):1821–1830.e1823 [DOI] [PubMed] [Google Scholar]
- 38.Zhao X, Li M, Zhang X, Tian Y, Li D et al. (2025) Efficacy and Safety of IL-17 and JAK Inhibitors in Ankylosing Spondylitis: a systematic review and network meta-analysis. Int Arch Allergy Immunol [DOI] [PubMed] [Google Scholar]
- 39.Chen S, Huang W, Wan Q, Tang Z, Li X et al. (2024) Investigation of the acute pathogenesis of spondyloarthritis/HLA-B27-associated anterior uveitis based on genome-wide association analysis and single-cell transcriptomics. J Transl Med 22(1):271. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Not applicable.