The Development and Initial Findings of A Study of a Prospective Adult Research Cohort with Inflammatory Bowel Disease (SPARC IBD)

Abstract

Background

Clinical and molecular subcategories of inflammatory bowel disease (IBD) are needed to discover mechanisms of disease and predictors of response and disease relapse. We aimed to develop a study of a prospective adult research cohort with IBD (SPARC IBD) including longitudinal clinical and patient-reported data and biosamples.

Methods

We established a cohort of adults with IBD from a geographically diverse sample of patients across the United States with standardized data and biosample collection methods and sample processing techniques. At enrollment and at time of lower endoscopy, patient-reported outcomes (PRO), clinical data, and endoscopy scoring indices are captured. Patient-reported outcomes are collected quarterly. The quality of clinical data entry after the first year of the study was assessed.

Results

Through January 2020, 3029 patients were enrolled in SPARC, of whom 66.1% have Crohn’s disease (CD), 32.2% have ulcerative colitis (UC), and 1.7% have IBD-unclassified. Among patients enrolled, 990 underwent colonoscopy. Remission rates were 63.9% in the CD group and 80.6% in the UC group. In the quality study of the cohort, there was 96% agreement on year of diagnosis and 97% agreement on IBD subtype. There was 91% overall agreement describing UC extent as left-sided vs extensive or pancolitis. The overall agreement for CD behavior was 83%.

Conclusion

The SPARC IBD is an ongoing large prospective cohort with longitudinal standardized collection of clinical data, biosamples, and PROs representing a unique resource aimed to drive discovery of clinical and molecular markers that will meet the needs of precision medicine in IBD.

Inflammatory bowel disease (IBD), a chronic, relapsing and remitting, inflammatory disorder of the gastrointestinal tract, is estimated to affect 3.1 million Americans.¹ While recent epidemiologic studies suggest the incidence of IBD in western countries is stabilizing, the incidence is rapidly increasing in newly industrialized countries and may approximate those of the western hemisphere in the 20th century, suggesting that the global burden of IBD will continue to grow over the next century. Inflammatory bowel disease comprises 2 main subtypes: Crohn’s disease (CD) and ulcerative colitis (UC)²; the term IBD unclassified (IBD-U) is reserved for those in whom it is not possible to distinguish between CD and UC due to overlapping features.³ Within each disease subtype, there is heterogeneity of disease location, behavior, severity, and responsiveness to therapy, preventing a “one-size-fits-all” approach to treatment.

Significant breakthroughs have led to new and effective treatments for IBD.^4–7 However, despite the growing therapeutic options for IBD, better tools are needed to personalize our treatment approach to determine the most effective treatment for each individual patient based on specific disease characteristics. One approach to achieve this goal is to assemble large-scale databases that combine high-resolution clinical and patient-reported data with molecular data derived from biosamples to discover mechanisms of disease and predictors of response and disease relapse.

Large populations are needed to develop precise clinical and molecular subcategories of IBD. Multi-institutional biobanks are capable of building larger biorepositories but face numerous challenges due to the multitude of stakeholders participating in these endeavors. These challenges include best practices to ensure standardization across participating sites and legal and ethical concerns surrounding sharing of clinical data and biosamples. Herein, we describe the development and implementation of Study of a Prospective Adult Research Cohort with Inflammatory Bowel Disease (SPARC IBD), a prospective cohort of well-phenotyped adult IBD patients with longitudinal clinical and patient-reported data and biosamples, initiated by the Crohn’s & Colitis Foundation (Foundation). Additionally, we describe the characteristics of the study cohort and the results of a phenotyping data quality exercise.

MATERIALS AND METHODS

The data and biosamples collected from SPARC IBD, a longitudinal research cohort, reside in IBD Plexus. IBD Plexus is a robust research information exchange platform designed to accelerate discovery, clinical research development, safety and surveillance, and outcomes research in IBD. The conception and development of IBD Plexus was the result of a partnership of the Foundation and academic and industry researchers after an extensive landscape analysis, which identified an opportunity to integrate the data and biosamples being collected in Foundation-supported IBD research and quality of care programs (IBD Partners, Risk Stratification in Crohn’s Disease [pediatric], and IBD Qorus; Fig. 1).^8–10 This analysis also identified a gap in its existing programs that was the catalyst to create SPARC IBD to include prospectively collected clinical data, patient-reported outcomes, and serially collected biosamples from adult patients with IBD over the course of their disease.

Figure 1.

Foundation-supported IBD research and quality of care programs and IBD plexus.

SPARC IBD

The principal objectives of SPARC IBD are to identify clinical and molecular predictors of response to therapy and relapse of disease to develop precision medicine strategies and new treatment targets and biomarkers. The priorities of this effort are to recruit a geographically diverse sample of patients with IBD across the United States and standardize data and sample collection methods and sample processing techniques to enable rapid sharing of data and knowledge. IBD Plexus provides the infrastructure and platform to house, organize, aggregate, and disseminate data and provides a researcher portal to maximize use of these data and samples.

In 2015, the Foundation issued a request for applications for sites to participate in SPARC IBD. Initially, 7 academic institutions were selected to serve as vanguard sites and assist with the planning of the data and biosample collection process, protocol design, methodology, and development of case report forms. A biobanking task force composed of academic and industry experts was created to advise on the sample protocol of SPARC IBD. In addition, an IBD SmartForm work group was established to create guidelines on IBD SmartForm data collection and ensure data collection standardization and quality across sites. In the second year of the study, the Foundation issued a second request for applications for additional sites, and 10 institutions were selected to join the SPARC IBD network.

During the first 2 years, site principal investigators held biweekly calls to discuss study procedures, including but not limited to recruitment, data collection, biosample collection, exchange of ideas, and best practices. The site principal investigators continue to meet approximately quarterly either through webinars or biannually at the Crohn’s & Colitis Congress and Digestive Diseases Week. The Foundation also organizes biweekly calls and an annual in-person meeting for all site research coordinators to review study procedures, data and biosample collection, and sharing of best practices. All sites undergo an initiation visit and subsequent biannual site visits by a member of the Foundation study team. The SPARC IBD principal investigators and the Foundation’s study team have weekly meetings to monitor progress, discuss logistics, and develop future strategies.

Patient Consent and Authorization

All patients enrolled in the study are consented for participation through a universal consent form authorizing use of their personal information for future IBD studies. These data include information about health and health-related issues, information collected from the medical record, insurance claims data, and biosamples collected as part of the research study.

Clinical Data Collection and Extraction

Participating institutions have agreed to embed an electronic case report form (eCRF), referred to as the IBD SmartForm in their Epic electronic health record. Epic SmartForms are customizable data collection tools. The IBD SmartForm was created by Epic employees in collaboration with clinicians at the University of Pennsylvania and with design input from the SPARC IBD Vanguard sites. The IBD SmartForm allows a clinician or research coordinator to record clinical data pertinent to a patient’s IBD history. The IBD SmartForm is available in the encounter navigator and can be updated at any encounter or between visits. Data captured into the SmartForm include IBD diagnosis, year of IBD symptom onset and diagnosis, summary of disease course, medication history, Montreal classification phenotype, surgical history, cancer-related history, and extraintestinal manifestations, in addition to pertinent laboratory, radiographic, endoscopic, and histologic results. Current symptoms and quality of life measures are collected through Epic’s MyChartor Epic’s Welcome module, which can be implemented on computers or smartphones. A physician global assessment is captured, and disease activity scores (Simple Crohn’s Disease Activity Index or 6-point Mayo score) are calculated based on data entry into the SmartForm. The IBD SmartForm is updated at every study encounter, allowing for longitudinal data collection. Study encounters include enrollment and all subsequent standard of care office visits. Using smart phrases, the data contained within the IBD SmartForm can be imported into a consultation or progress note and serve as the clinical documentation for that visit. The SmartForm was revised based on feedback from the participating sites, and the most recent version is available in Epic 2018 for all Epic users. For the few sites that use other electronic health records, a separate, similar electronic data capture (EDC) tool built in Salesforce, a cloud-based platform, is available.

At enrollment and at the time of colonoscopy or flexible sigmoidoscopy (event), IBD-specific patient-reported outcomes, current medications, and endoscopic scoring indices (Simple Endoscopic Score for Crohn’s Disease [SES-CD] and Modified Mayo Endoscopic Score for ulcerative colitis [MMES]) are collected and captured through the SPARC IBD EDC. Furthermore, electronic surveys generated through the Salesforce platform are delivered to patients every 3 months to capture IBD-related symptoms, current IBD therapies, and hospitalizations. These quarterly surveys allow for tracking of patient-reported disease activity between office visits, confirmation of current medications and dosage, and capture of IBD hospitalizations.

Biosample Collection

Sample banking across multiple sites requires standardization in sample collection, annotation, processing, and storage.¹ A range of biosamples (blood, stool, and intestinal tissue) are collected as part of SPARC IBD (Table 1) and stored in a central biobank. Biosamples are collected at the time of consent (blood and stool) and at time of usual care sigmoidoscopy or colonoscopy (blood, stool, and mucosal biopsies). Blood and stool samples are also obtained approximately 3 months after a change in therapy following a colonoscopy or sigmoidoscopy if the patient has a follow-up office visit during that time. Each site is trained by a member of the Foundation’s SPARC IBD study team and provided a SPARC lab manual for standardized collection, processing, and shipping of biosamples. In addition, quality control rules and monitoring processes have been implemented to ensure completeness, standardization, and quality of the samples. Samples obtained for clinical use are not accessed for research purposes.

Table 1.

SPARC IBD Data & Biosamples Collection Summary

	Study Enrollment (baseline)	Office Visit	Colonoscopies or sigmoidoscopies (event)	3 months after event	Quarterly
Data
Medical record		╳
Ibd smartform	╳	╳
Colonoscopy and sigmoidoscopy indices			╳
Patient-reported Outcomes	╳ a	╳ a	╳ a	╳ a	╳ a
Biosamples
Blood	╳	╳	╳	╳ b
Stool	╳		╳	╳
Intestinal Biopsies			╳

^aHospitalization is only collected when applicable.

^bBlood collected if patient has a change in IBD medications after the colonoscopy/sigmoidoscopy and is scheduled for the clinical visit.

Blood

Peripheral blood samples are allocated into DNA, plasma, RNA, and peripheral blood mononuclear cells (PBMCs). These samples can be used for a variety of research purposes including assessing genotype and gene expression, protein and metabolite concentrations, and ex vivo analysis. These biosamples—except plasma—are shipped unprocessed to Brook Life Sciences, which serves as the central biobank. Plasma is extracted and aliquoted at each site before it is shipped to central biobank.

Stool

Stool samples are collected to explore gut microbiome composition, inflammatory markers (such as calprotectin), various proteins, and metabolites. For patients undergoing colonoscopy, the first stool of the day before initiation of a bowel preparation is collected. Samples are collected by the patient at home and shipped directly to the biobank through a courier service. Participants aliquot a small sample of stool into alcohol (95% ethanol) to preserve the sample for measurement of metabolomics. Additional aliquots from the stool sample are created by the biobank and frozen without preservative. In addition, fecal calprotectin (fCAL) is measured by the central biobank.

Intestinal Tissue

In designing SPARC IBD, we assumed that most patients would have an endoscopic procedure before a change in treatment regimen. It is at the discretion of the endoscopist to assess whether collection of biopsies for research purposes poses an unacceptable risk to a subject. In the absence of a contraindication to obtaining biopsies, research biopsies are obtained even if no biopsies are obtained for clinical purposes. Intestinal biopsies are collected to aid efforts to better understand the biology of disease at the mucosal level and identify potential biomarkers that predict response to therapy, adverse reactions related to therapy, or disease relapse. Biopsies are processed for DNA (frozen in liquid nitrogen), RNA extraction (RNAlater), and histology (placed in formalin and subsequently embedded in paraffin within 24 hours of collection by the biobank).

In CD patients undergoing colonoscopy, up to 5 pinch biopsies using regular or jumbo forceps are obtained from the ileum or the most proximal extent of the exam and the rectosigmoid junction (at 20 cm from the anal verge). If both areas appear normal on insertion of the colonoscope, an additional 5 pinch biopsies are obtained from an area with macroscopically active disease, if present.

For those with UC or IBD-U, up to 5 pinch biopsies are obtained from the cecum (or most proximal extent of the colon exam) and the rectosigmoid junction (at 20 cm). In patients without pancolitis, if feasible and safe, biopsies from the normal area just adjacent to the transition area from abnormal to normal-appearing mucosa are obtained. If the only evidence of inflamed tissue on colonoscopy is located distal to 20 cm, biopsies are obtained more distally in the area of active inflammation.

Technology

A key challenge of a multi-institutional data repository and biobank is the technology underpinning its structure. Mechanisms to unify the intake of data and biosamples across the multiple sites are required to permit seamless integration of patient clinical and biological data with the disease course longitudinally. IBD Plexus, which manages SPARC IBD data and biosamples, uses a centralized infrastructure approach to store, curate, and harmonize large amounts of data on individuals with IBD and—critically—to link and mine these data for insights into causes and treatments through an easy to access user interface for data extraction. These data include clinical data from electronic case report forms, IBD SmartForm completed by the study sites, patient-reported data from surveys completed by the patients, and electronic health record data that are collected for clinical or billing purposes (eg, International Classification of Disease and Common Procedural Terminology codes). Data derived from biosamples are stored in the same central database.

Data Extraction

With the patient’s consent and authorization, participating institutions have agreed to transfer the participating patient’s electronic health record data to IBD Plexus. The Foundation has developed a computing infrastructure to extract and securely transmit electronic health data from consented patients to IBD Plexus on a weekly or biweekly basis (Fig. 2). Structured Query Language queries are used to accurately extract the appropriate electronic health record and SmartForm data elements. The files are securely transferred into IBD Plexus via secured protocols (SFTPs). Storage and transfer of data include data encryption both at rest and in transit. The extracted data are then normalized and transformed into common standards. De-identified data are stored in a research database platform within IBD Plexus. As of end of 2020, the data transfer has been implemented at 13 institutions for 2920 participants.

Figure 2.

IBD plexus and SPARC IBD patient data flow.

Confidentiality

The study protocol was approved by the University of Pennsylvania’s institutional review board.

Information about study participants is confidential and managed according to the requirements of the Health Insurance Portability and Accountability Act of 1996 (HIPAA). If a participant revokes authorization to collect or use protected health information (PHI), investigators, by regulation, retain the ability to use all information collected before the revocation of participant authorization but not after the revocation.

All data related to this research study that are released to investigators use a patient-assigned unique study number. Data are reported only in a confidential manner such that the personal identity of any subject is not identifiable. All study data are maintained under a double locked system.

Data Quality Evaluation

The quality of clinical data entry into the IBD SmartForm was assessed in a study involving all the SPARC Vanguard sites after the first year of initiating patient enrollment. The goal of the study was to validate the accuracy of key data elements. Ten investigators from the 7 Vanguard SPARC sites reviewed 12 de-identified extracts of consented patients’ medical records and completed a simplified version of the SmartForm focused on disease phenotype. One investigator reviewed 7 charts, 2 reviewed 11 charts, and the other 7 reviewed 12 charts each. The medical records reviewed included the first patient evaluation note and/or progress note related to the IBD diagnosis, the progress note completed at the time of the initial SmartForm completion, the 2 most recent progress notes previous to the completion of the initial IBD SmartForm, all pathology reports, all endoscopy and video capsule endoscopy reports, all cross-sectional imaging reports, and all operative reports. After review of available medical records, investigators were asked to record the patient’s IBD subtype, year of diagnosis, disease location and behavior according to Montreal classification, and histologic extent of disease (UC only). Crohn’s disease behavior was categorized as inflammatory, penetrating, stricturing, or both stricturing and penetrating. Descriptive statistics (ie, frequencies and percentages) were calculated for each item. Agreement was defined as the proportion of the reviewers who agreed with the majority opinion. For example, if 7 of 10 reviewers categorized extent of disease for a patient with ulcerative colitis as pancolitis and the other 3 reviewers as left-sided colitis, agreement was computed as 70% (7 of 10).

Statistical Analysis

Descriptive statistics (including frequencies and proportions) have been used to describe the patient cohort and examine reviewer agreement for the quality control exercise including median and means as appropriate.

Results

Quality control of phenotyping procedures

The results of the data quality exercise are summarized in Table 2. There was 96% agreement on year of diagnosis and 97% agreement on IBD subtype. Among 4 patients with UC, agreement on extent of disease based on macroscopic appearance was 100%, 75%, 70%, and 56%. When dichotomizing UC extent as proctitis or left-sided vs extensive or pancolitis, overall agreement was 91%. By chance, 7 of the 8 patients with CD selected for review had stricturing disease phenotype. For individual patients, agreement in CD behavior ranged from 70% to 100%, with an overall agreement for CD behavior of 83%. For individual patients, agreement on a history of disease involving the colon ranged from 70% to 100%, with an overall agreement of 92%. For individual patients, agreement on disease involving the ileum ranged from 78% to 100%, with an overall agreement of 96%. Agreement on having a history of perianal fistula ranged from 89% to 100% for individual patients, with an overall agreement of 96%.

Table 2.

Rates of Agreement in the Data Quality Exercise

Characteristic	Agreement at Patient Level (range)	Agreement Overall
Year of Diagnosis	70%–100%	96%
IBD subtypeb	78%- 100%	97%
Ulcerative colitis macroscopic extent of disease	56%–100%	91%a
Crohn’s disease behavior	70%–100%	83%
Crohn’s disease of colon	70%–100%	92%
Crohn’s disease of ileum	78%–100%	96%
Crohn’s disease perianal fistula	89%–100%	96%

^aProctitis or left-sided colitis vs extensive colitis or pancolitis

^bThe 3 reviews with disagreement were all categorized as IBD-U, whereas the majority are categorized as UC.

Description of the cohort

Between November 2016 and January 1, 2020, 3029 patients have been enrolled across 16 SPARC IBD sites (Table 3). Enrollment at the time of the COVID-19 pandemic was stopped in April 2020 and slowly resumed in July 2020. Of the participants enrolled, 2003 (66.1%) have CD, 974 (32.2%) have UC, and 52 (1.7%) have IBD-U. Across all 3 IBD diagnoses, there are 1678 females (55.4%) and 1351 males (44.6%) enrolled. Age at enrollment ranged from 18 to 89 years old, with median age of 39 years. Disease duration ranged from less than 1 year to 77 years, with a median of 12 years. The majority are white (77.3%) and non-Hispanic or Latino (83.0%). Black patients comprise 6.1% of the cohort. Inflammatory bowel disease medication use spans multiple classes at enrollment, with the majority (n = 1533) on biologic therapy (Table 3). Of all IBD patients enrolled, 19.2% had a history of IBD-related surgery at time of enrollment.

Table 3.

Clinical Characteristics & Phenotype of IBD Participants

No. participants	3029
Age at time of enrollment in years (Range / Median)	18–89/39
Race
White (%)	2341 (77.3)
Black (%)	186 (6.1)
Asian (%)	39 (1.3)
American Indian or Alaska Native	6 (0.2)
Other (%)	42 (1.4)
Unknown (%)	415 (13.7)
Ethnicity
Hispanic or Latino (%)	49 (1.6)
Non-Hispanic or Latino (%)	2513 (83)
Unknown (%)	467 (15.4)
Sex
Male (%)	1351 (44.6)
Female (%)	1678 (55.4)
Diagnosis
Crohn’s disease (%)	2003 (66.1)
Ulcerative colitis (%)	974 (32.2)
IBD-U (%)	52 (1.7)
Age of diagnosis (Range / Median)	1–79/25
Disease duration at enrollment (years; Range / Median)	0–77/12
Disease behavior (CD)
B1 Inflammatory, nonstricturing, nonpenetrating (%)	661 (33)
B2 Stricturing (%)	465 (23.2)
B3 Penetrating (%)	266 (13.3)
Both stricturing and penetrating	201 (10)
Unknown (%)	410 (20.5)
Perianal disease	461 (23)
Disease location (CD)
L1 Ileal	439 (21.9)
L2 Colonic	268 (13.4)
L3 Ileocolonic	894 (44.6)
L4 Upper track	142 (7.1)
Unknown	396 (19.8)
Disease extent (UC)
E1 Ulcerative proctitis	76 (7.8)
E2 Left-sided ulcerative colitis	191 (19.6)
E3 Extensive ulcerative colitis	74 (7.6)
Unknown	209 (21.5)
Pancolitis	424 (43.5)
History of IBD related Surgery	582 (19.2)
Tobacco use (%) n = 2355	296 (12.6)

Disease phenotype and location

Most patients with Crohn’s disease had ileocolonic disease at enrollment (44.6%), whereas 21.9% had isolated small bowel disease, and 13.4% had isolated colonic disease. Upper tract disease was present in 7.1% of patients. Disease location was unknown (not reported) for 19.8% of patients. Disease behavior was recorded as inflammatory (nonstricturing, nonpenetrating) in 33.0% patients, stricturing in 23.2%, penetrating in 13.3%, and both stricturing and penetrating in 10% patients. A history of perianal disease was recorded in 23.0% of patients (Table 3).

In the UC group, 43.5% of patients had pancolitis as their macroscopic extent of disease, followed by 19.6% with left-sided colitis, 7.6% with extensive colitis, and 7.8% with proctitis. Macroscopic disease extent was unknown (not reported) in 21.5% of patients (Table 3).

Disease activity

Using the short Crohn’s Disease Activity Index (sCDAI), 63.9% of patients in the Crohn’s disease group (63.9%) were in remission (score <150) at the time of enrollment. Using the 6-point Mayo score to assess UC activity, 80.6% of patients were in remission with a Mayo score ≤2 and no individual score >1 (Table 4).

Table 4.

Disease Severity at Time of Enrollment

sCDAI (CD) n = 1902
Remission	1216 (63.9)
Mild	353 (18.6)
Moderate	311 (16.4)
Severe	22 (1.2)
Physician Global Assessment (CD) n = 1360
Quiescent	800 (58.8)
Mild	343 (25.2)
Moderate	172 (12.6)
Severe	45 (3.3)
6-point Mayo (UC) n = 891
<=1	636 (71.4)
2	82 (9.2)
3	85 (9.5)
4	51 (5.7)
5	22 (2.5)
>5	15 (1.7)
Physician Global Assessment (UC) n = 655
Quiescent	416 (63.5)
Mild	133 (20.3)
Moderate	82 (12.5)
Severe	24 (3.7)

Among the patients enrolled in SPARC IBD, 990 underwent colonoscopy during a mean follow-up time of 7.7 months. For patients with CD, the median (interquartile range [IQR]) SES-CD was 3 (0–6), and for patients with UC, the median (IQR) Mayo endoscopic score was 1 (0–2).

Medication use

Medication history stratified by the SPARC IBD cohort is summarized in Table 5. Biologic medications were used by 50.9% of participants at the time of enrollment in SPARC IBD, with antitumor necrosis factor (anti-TNF) use being most common (33.2%). Thiopurines or methotrexate were used by 21.9% of participants at the time of enrollment.

Table 5.

Medication Use at Enrollment and Before Enrollment in SPARC IBD

	Crohn’s Disease		Ulcerative Colitis		IBD Unclassified		Total
	N	%	N	%	N	%	N	%
	2003	100.0%	974	100.0%	52	100.0%	3029	100.0%
At time of enrollment
Aminosalicylates	169	8.4%	350	35.9%	17	32.7%	536	17.7%
Mesalamine	132	6.6%	301	30.9%	15	28.8%	448	14.8%
Sulfasalazine	31	1.5%	37	3.8%	2	3.8%	70	2.3%
Balsalazide	9	0.4%	28	2.9%	0	0.0%	37	1.2%
Biologics	1147	57.3%	372	38.2%	19	36.5%	1538	50.8%
Infliximab	365	18.2%	173	17.8%	6	11.5%	544	18.0%
Adalimumab	335	16.7%	64	6.6%	6	11.5%	405	13.4%
Vedolizumab	167	8.3%	115	11.8%	4	7.7%	286	9.4%
Ustekinumab	236	11.8%	9	0.9%	2	3.8%	247	8.2%
Certolizumab Pegol	39	1.9%	2	0.2%	1	1.9%	42	1.4%
Golimumab	3	0.1%	9	0.9%	0	0.0%	12	0.4%
Natalizumab	2	0.1%	0	0.0%	0	0.0%	2	0.1%
Immunomodulators	458	22.9%	221	22.7%	13	25.0%	692	22.8%
Azathioprine	254	12.7%	135	13.9%	5	9.6%	394	13.0%
Methotrexate	115	5.7%	34	3.5%	6	11.5%	155	5.1%
Mercaptopurine	86	4.3%	28	2.9%	0	0.0%	114	3.8%
Tofacitinib	4	0.2%	22	2.3%	1	1.9%	27	0.9%
Tacrolimus	1	0.0%	3	0.3%	1	1.9%	5	0.2%
Steroid Therapies	151	7.5%	111	11.4%	5	9.6%	267	8.8
Exposure
Aminosalicylates	898	44.8%	612	62.8%	38	73.1%	1548	51.1%
Mesalamine	829	41.4%	587	60.3%	37	71.2%	1453	48.0%
Sulfasalazine	195	9.7%	134	13.8%	12	23.1%	341	11.3%
Biologics	1018	50.8%	313	32.1%	22	42.3%	1353	44.7%
Infliximab	887	44.3%	338	34.7%	20	38.5%	1245	41.1%
Adalimumab	814	40.6%	200	20.5%	16	30.8%	1030	34.0%
Vedolizumab	323	16.1%	190	19.5%	10	19.2%	523	17.3%
Ustekinumab	213	10.6%	15	1.5%	3	5.8%	231	7.6%
Certolizumab Pegol	202	10.1%	8	0.8%	5	9.6%	215	7.1%
Golimumab	16	0.8%	24	2.5%	0	0.0%	40	1.3%
Natalizumab	34	1.7%	1	0.1%	0	0.0%	35	1.2%
Immunomodulators	395	19.7%	111	11.4%	11	21.2%	517	17.1%
Thiopurine	945	47.2%	373	38.3%	20	38.5%	1338	44.2%
Methotrexate	390	19.5%	90	9.2%	8	15.4%	488	16.1%
Tofacitinib	8	0.4%	25	2.6%	5	9.6%	38	1.3%
Steroid Therapies	1205	60.2%	629	64.6%	37	71.2	1871	61.8%

Discussion

We demonstrate the feasibility of creating a prospective cohort of adult IBD Patients with standardized phenotype descriptors, longitudinal clinical and patient-reported data, and biosample collection focused on events often associated with change in therapy across multiple institutions while leveraging a common electronic health record system. This resource has grown to date to include more than 3000 patients across sites with a recruitment goal of 7000 patients. Electronic health record data extraction to the Plexus platform has been implemented for over 2920 SPARC patients with IBD, which is accessible to researchers in academia and industry.

One of the greatest challenges of a multisite data and biorepository is standardization of data collection. We were able to demonstrate through a quality control study of phenotyping procedures excellent agreement across investigators at 7 separate sites regarding diagnosis, extent of disease, and disease behavior. There have been great efforts for SPARC IBD to continue to optimize accurate data collection through the development of phenotyping manuals and regularly scheduled conference calls for study coordinators and in-person coordinator meetings including phenotyping workshops. After the completion of the data quality exercise, a manual was created by the IBD SmartForm work group to improve standardization of phenotyping of participants. The manual details the criteria for diagnosing CD, UC, and IBD-U. The manual also describes criteria for determining CD disease behavior (ie, inflammatory, stricturing, penetrating, and structuring), and macroscopic disease location including microscopic extent in UC and IBD-U patients. All new investigators and their research coordinators enrolling patients in SPARC IBD are required to review the manual before entering data into the IBD SmartForm.

CONCLUSION

The IBD Plexus seeks to meet the need for precision medicine in IBD through building the world’s largest IBD data and biosample repository. IBD Plexus is the first of its kind and scope for the IBD community, enabled by the Foundation partnering with academic and industry researchers, to share data and promote collaboration and allowing studies to be performed that could never be achieved at any one institution or company. The SPARC IBD is a key component of this effort to advance precision medicine in IBD. By leveraging the Epic electronic health record, SPARC IBD has streamlined clinical data collection, reducing the burden for individual sites and investigators and, in turn, improving the long-term feasibility of the study. This collaborative approach addresses a significant gap in our approach to advancing the care of patients with IBD.

Conflicts of Interest

LR is a consultant for Alivio Therapeutics. SS is a consultant for Eli Lily. MB has received research funding from Janssen, GlaxoSmithKline, and Takeda, received honorarium for participation in a CME program sponsored by AbbVie, and is a consultant for Janssen, BMS, and AbbVie. SB has received grant funding from the Crohn’s & Colitis Foundation (CDA 598997) and NIH (K08 DK1234. RC has received consulting fees from Abbvie, LabCorp, Janssen, Pfizer, Prometheus, Samsung Bioepis, and Takeda. MF is on the advisory board of Takeda, BMS, Janssen, and dSMB member Ferring. AY is on the consulting/advisory board for Takeda, Prometheus Bioscience, Arena Pharmaceuticals, and Bristol Myers Squibb. DH has consulted for Abbvie, BMS, Pfizer, Janssen, and Takeda and received research support from Pfizer. PD is a consultant or advisory board member for Janssen, Pfizer, BMS-Celgene, Arena Pharmaceuticals, and Prometheus Biosciences and has received research grants from Takeda. FC has received research support from Takeda Pharmaceuticals and Sanofi. JK is a consultant for Corrona, Thetis Pharma, and Roche. JP is on the scientific advisory board of Pfizer, Janssen, and Takeda, is a consultant for Verastem and CVS Caremark, and has received research grants from Abbvie and Takeda. MC is a consultant for Pfizer and Bristol Myers Squibb and has received a research grant from Pfizer. SS is on the scientific advisory board for Pandion (for which he also has ownership/stock options), Lilly, Takeda, Cosmo, and Pfizer and is a consultant for Roche, Amgen, Kyverna, Merck, and Third Rock. UW is a consultant for AbbVie and Samsung Bioepis. JL is a consultant for Janssen, Samsung Bioepis, UCB, Bristol-Meyers Squibb, Nestle Health Science, Merck, Celgene, Bridge Biotherapeutics, Pfizer, Gilead, Arena Pharmaceuticals, Protagonist Therapeutics, and Entasis Therapeutics and has received research grants from Janssen, Nestle Health Sciences, and Takeda.

Supported by

Crohn’s & Colitis Foundation, and The Leona M. and Harry B. Helmsley Charitable Trust.