Disease Burden and Patient-Reported Outcomes Among Ulcerative Colitis Patients According to Therapy at Enrollment Into CorEvitas’ Inflammatory Bowel Disease Registry

Abstract Background To evaluate disease burden and patient-reported outcomes (PROs) of ulcerative colitis (UC) patients at enrollment into CorEvitas’ Inflammatory Bowel Disease Registry by therapy class. Methods Between May 3, 2017 and September 3, 2019, 773 UC registry patients were categorized by therapy class at enrollment: patients on 5-aminosalicylic acids (5-ASAs) only (n = 290), and patients on biologics/Janus kinase inhibitors (JAKi) alone or in combination with 5-ASAs or immunosuppressant therapies (BIO/JAKi) (n = 315). To quantify between group differences, the mean/proportional differences and corresponding 95% CIs were calculated. Results Among 605 UC patients at enrollment, BIO/JAKi patients were younger (44.1 vs. 50.9 years) more were female (58.0% vs. 49.7%), had lower remission (45.4% vs. 60.0%), had more moderate/severe disease (16.5% vs. 7.1%), experienced less proctitis (10.5% vs. 22.1%), but more pancolitis (54.6% vs. 34.1%), more corticosteroid experience (70.8% vs. 44.5%), previous biologic experience (1 prior: 21.6% vs. 2.4%; 2+ prior: 12.1% vs. 0.3%), and shorter duration of current UC therapy (1.6 vs. 3.5 years) than 5-ASAs patients. BIO/JAKi patients had higher current employment than 5-ASAs patients (70.7% vs. 62.4%) and higher mean Work Productivity and Activity Impairment (WPAI) domains for absenteeism (7.3 vs. 2.8) and activity impairment (22.0 vs. 17.5). Conclusions Among UC patients in a real-world setting, BIO/JAKi patients had less remission, more moderate-to-severe disease, and worse PROs than 5-ASAs patients. These results suggest that despite increased therapeutic options, patients with UC currently being treated with biologics or JAKi may still experience disease burden and continued unmet needs.


Introduction
The clinical characteristics and outcomes of ulcerative colitis (UC) patients on the available treatments are not well understood. Clinical trials data are limited and fail to capture disease burden and the spectrum of clinical response completely. Residual or persistent symptoms have not been adequately explored, particularly with respect to the differences between the newer therapies, including biologics and tofacitinib, and older therapies. Patient characteristics and outcomes across different treatments can be depicted using systematically collected real-world data, thus providing the opportunity to gain insight into the disease burden in UC patients. The examination of real-world data may inform health care providers on ways to improve treatment strategies and to tailor therapy for individual patients to achieve optimal disease control.
To address the disease burden in UC patients in a realworld setting, we describe the characteristics of patients with UC enrolled in CorEvitas' Inflammatory Bowel Disease (IBD) Registry, by therapy class (5-ASAs alone  compared to biologics/Janus kinase inhibitors (JAKi) alone or in combination with immunosuppressant therapies [ISTs] or 5-ASAs [BIO/JAKi]). Variables of interest included sociodemographic 2 Disease Burden and Patient-Reported Outcomes Among Ulcerative Colitis Patients characteristics, history of comorbidities, disease activity, disease characteristics, therapy history, and patient-reported outcomes (PROs).

Study Setting
Launched in May 2017, CorEvitas' IBD Registry is a prospective, noninterventional, US research registry of patients who are at least 18 years of age or older with IBD under the care of a board-certified gastroenterologist. To better focus on the study of safety and effectiveness of biologic and small molecule therapies explicitly used in the treatment of UC and Crohn's disease (CD), in January 2019, the Registry changed the protocol to only enroll patients initiating biologic/JAKi therapy. Newly enrolling patients were required to have initiated or switched to an approved biologic or JAKi to treat UC or CD within the previous 12 months.
Eligible medications for enrollment include FDA-approved biologic treatments for IBD (eg, tumor necrosis factor inhibitors (TNFi), an interleukin-12/23 inhibitor (IL-12/23i), an integrin α4β7 inhibitor, and an integrin α4 inhibitor), and JAKi. Enrollment data are collected from the patients and their treating gastroenterologist during routine clinical visits using CorEvitas registry questionnaires. As of December 31, 2019, the CorEvitas' IBD Registry database included 1858 patients and 3384 patient visits, with a mean duration of patient follow-up of 1.0 years (median 1.0 years). Patients were recruited from 57 private (n = 48) and academic (n = 9) practice sites across 20 states in the United States, with 123 participating gastroenterologists.

Assessment of Patient Characteristics and Outcomes
Data were collected for sociodemographic characteristics, history of comorbidities, disease activity, disease characteristics, therapy history, and PROs. Sociodemographic characteristics included age, gender, race (White or non-White), ethnicity (Hispanic or non-Hispanic), body mass index (BMI), type of health insurance plan, and education. History of physicianreported comorbidities included: cardiovascular, autoimmune, gastrointestinal, respiratory, digestive/hepatic and neurologic diseases, cancer, infections, diabetes, osteoporosis, depression, anxiety, and other nonserious medical conditions. The modified Charlson Comorbidity Index (mCCI) was calculated as the sum comorbid conditions available in the registry. 1 Disease activity measures included the Mayo Score (Partial) without physician global assessment and by physician's global assessment (PGA) (remission [0-1], mild disease [2][3][4], moderate disease [5][6], severe disease [7][8][9]). Disease characteristics included location of disease (proctitis, left-sided, pancolitis), history of hospitalization and Emergency Room (ER) use for UC-related issues, and difference in time between symptoms and UC diagnosis (years). History of extraintestinal manifestations (eg, arthritis, skin manifestations, eye involvement) was also collected.
Data included the history of previous drug therapies including the number of prior biologics and JAKi, prior 5-ASAs, prior ISTs, corticosteroid use, antibiotic use, and the duration of current UC therapy. PROs to measure health status and functioning included the Patient Reported Outcomes Measurement Information System (PROMIS) and the Work Productivity and Activity Impairment (WPAI) questionnaires. We included PROMIS measures for fatigue, sleep disturbance, pain interference, depression, and anxiety. PROMIS uses a T-score metric in which 50 is the mean of a relevant reference population (eg, US general population) and 10 is the SD of that population. Higher scores indicate a worse outcome and more severity of the domain being measured. 2 PROMIS scores are grouped as within normal limits (<55), mild (55 to <60), moderate (60 to <70), and severe (≥70) based on domain scores measured in the US general population. 2,3 The WPAI measures absenteeism (work hours missed), presenteeism (impairment at work/reduced on-the-job effectiveness), work productivity loss (overall work impairment/ absenteeism plus presenteeism), and activity impairment (daily activities impaired). The WPAI outcomes are scored as percentages of impairment, with higher numbers indicating more significant impairment and less productivity (ie, worse outcomes). 4

Statistical Analyses
Descriptive statistics characterized patients at enrollment overall and within each cohort separately to highlight differences in patient characteristics across cohorts. Categorical variables were summarized using frequency counts and percentages. Continuous variables were summarized by the number of observations, mean, and SD. To quantify between group differences, the mean or proportional differences and their corresponding 95% CIs were calculated. Compared to P values, CIs aide clinical interpretation by reporting a plausible range of values in the actual units of data measured along with the direction and strength of the effect. 5 Confidence intervals for differences that do not include 0 are considered statistically noteworthy. The outcome comparisons are cross-sectional and were not adjusted for these cohort differences and should not be interpreted as a difference in response to treatment. R version 3.6.2 (The R Foundation for Statistical Computing) was used for analyses.

Ethical Considerations
All participating IBD Registry investigators were required to obtain full board approval for conducting research involving human subjects. Sponsor approval and continuing review were obtained through a central Institutional Review Board (IRB) (IntegReview, Protocol number Corrona-IBD-600).   For academic investigative sites that did not receive a waiver to use the central IRB, approval was obtained from the respective governing IRBs, and documentation of approval was submitted to the Sponsor before initiating any study procedures. All registry subjects were required to provide written informed consent before participating.

Results
To  Table 3 presents PROMIS scores and WPAI domains at enrollment. For both BIO/JAKi and 5-ASAs patients, the median PROMIS scores were within normal limits (<55) based on population norms. When evaluating the amount of impairment at work and impact on daily activities due to UC, BIO/JAKi patients had higher mean impairment score for absenteeism (7.3% vs. 2.8%) and activity impairment (22.0% vs. 17.5%) than 5-ASAs patients, respectively (Figure 2).  Unadjusted mean difference (95% CI) between groups for continuous variables and difference in percentage points (95% CI) between groups for categorical variables.

Discussion
Because real-world systematically collected data provide a clearer picture of the burden of disease in UC, the objective of this study was to describe characteristics of patients with UC, including demographics, history of comorbidities, disease activity, disease characteristics, therapy history, and PROs, by therapy, in a US population-based IBD Registry. Our study found that those UC patients who received advanced therapies (BIO/JAKi) had a history of more severe disease, and worse clinical outcomes including hospitalization and ER use for UC-related issues, work absenteeism, and impaired daily activities compared to patients receiving 5-ASAs. In the clinical setting, patients with mild-to-moderate UC disease severity are treated with 5-ASAs, and biologics/JAKi are used when 5-ASAs do not adequately control disease. Although biologic therapies have been available for IBD for over 20 years, patient outcomes have not generally changed over this time. According to the American Gastroenterological Association (AGA) recommendations for managing UC patients with moderate-to-severe disease activity and/or who are at high risk of colectomy, clinicians should implement biologic agents with or without an immunomodulator early rather than gradual step-up therapy after failure of 5-ASAs. 6 Given the prior therapy history in the BIO/JAKi cohort, it is likely that patients in this group had treatment refractory disease. The differences between BIO/JAKi and 5-ASAs are likely due to differences in disease severity between groups and not necessarily the effects of treatments.
Our findings correlate with the two most recent real-world studies addressing treatment patterns 7 and the association between disease activity and quality of life 8 among moderate-tosevere UC patients in the United States and Europe. In the first study, Armuzzi et al found many patients with moderate-tosevere IBD initiated advanced treatments early in their disease course and that combining ISTs and biologics was common, as conventional therapies were not considered an adequate initial treatment. 7 In the second study, the same researchers    found that patients with moderate-to-severe UC treated with biologics still had significant impairment in health-related quality of life, including work and daily activities. 8 To achieve the best possible long-term outcomes, control of clinical symptoms and achieving mucosal healing is vital. Thus, defining unmet needs in UC patients is paramount in increasing the awareness of the problems associated with current therapeutic management. 9 Due to an increase in the number of pharmacologic agents available to treat moderate-to-severe UC over the last 5 years, more exploration is needed to evaluate some critical knowledge gaps in treating UC patients. These knowledge gaps include: identifying biomarkers predictive of response to individual therapies to facilitate the optimal choice of treatments; developing clinical prediction models to help identify patients who have low vs. high probability of response to treatment; and creating novel combinations of available therapies in patients with moderate-to-severe UC. 10 To further tailor treatment to individual patients, clinicians will need to determine which drugs to use and which molecular pathway to target.
Identifying subpopulations of patients who respond to specific medications based on an increased understanding of pharmacogenomics, biomarkers, and clinical characteristics is vital. 11 Implementing these suggestions while characterizing treatment patterns and sequencing in patients with UC will help researchers identify new and more efficacious treatments.
This study's limitations include the cross-sectional analysis of observational data that did not evaluate improvement over time, specifically about patients who started biologics or advanced therapeutics. Comparisons were made to highlight the differences in patient characteristics across cohorts and should not be construed as differences in response to treatment. As with all registry-based research, selection bias may be introduced during enrollment if certain subgroups of patients are routinely included or excluded from the registry. The patient population might not be generalizable to UC patients outside the United States. Caution is warranted when interpreting these results as this was an exploratory analysis and was not adjusted for potential confounding factors. However, these unadjusted results are  Unadjusted mean difference (95% CI) between groups for continuous variables and difference in percentage points (95% CI) between groups for categorical variables. c a preliminary step to provide data to develop future research which would include adjusted models. We believe the fact that patients with higher disease severity are more likely to be on biologics is intuitive. In contrast to these limitations, we do not believe selection bias occurred in the CorEvitas Registry and thus did not influence our study. CorEvitas observational data consist of community practice and referral tertiary care IBD centers and includes high and low acuity patients. Our study involved real-world systematically collected data that provide a more realistic view of the burden of disease in UC patients and is more reflective of the UC population than those in clinical trials. Moreover, our results further add to the evidence that there continue to be unmet treatment needs in patients with moderate-to-severe UC. 7,9 As with the most recent real-world studies in UC patients in the United States and Europe, 8 many of our patients initiated advanced biologic therapies and experienced a low health-related quality of life.

Conclusions
At enrollment in CorEvitas' IBD Registry among UC patients in a real-world setting, patients treated with biologics and JAKi had less remission, more moderate-to-severe disease, increased absenteeism at work, and added negative impact on daily activities than those on 5-ASAs. These findings may suggest, along with recent population-based studies in UC patients, that despite increased therapeutic options, there is a significant disease burden and continued unmet needs among patients with UC currently being treated with biologics or JAKi. More research needs to be conducted to identify new and more efficacious and safe treatments that benefit patients based on clinical assessments and PROs, how clinicians can optimize treatments, and discover better methods for sequencing UC therapies.

Funding
Support for third-party writing assistance was furnished by Michelle D. Karpman, PhD, of CorEvitas, LLC, and funding was provided by Eli Lilly and Company, Indianapolis, IN, USA, in accordance with Good Publication Practice (GPP3) guidelines (Ann Intern Med 2015;163:461-464). The CorEvitas' Inflammatory Bowel Disease Registry is sponsored by CorEvitas, LLC (formerly known as Corrona, LLC). CorEvitas, LLC has been supported through contracted subscriptions in the last two years by AbbVie, Amgen, Boehringer Ingelheim, Bristol Myers Squibb, Celgene, Eli Lilly and Company, Genentech, Gilead Sciences, Inc., GSK, Janssen, LEO, Novartis, Ortho Dermatologics, Pfizer Inc., Regeneron, Sanofi, Sun Pharma, and UCB.