https://orcid.org/0000-0002-8351-4626
Abstract
Several therapies, including sphingosine 1-phosphate receptor modulators, have been approved for ulcerative colitis (UC). Safety findings should be complemented with real-world data (RWD) as rare events may be underestimated in clinical trials due to populations not fully reflecting real-world practice.
We used RWD to investigate safety outcomes in 4 cohorts: (1) those receiving any UC treatment (UC overall), (2) those receiving advanced therapies (UC advanced therapy), (3) those meeting the selection criteria of the etrasimod ELEVATE UC clinical program (UC trial-similar), (4) individuals without UC (Non-UC). Data were extracted (1/2016–12/2022) from the Optum® de-identified Electronic Health Record data set.
Data from 32 170 (UC overall), 3332 (UC advanced therapy), 1435 (UC trial-similar), and 160 795 (Non-UC) individuals were included. In the UC overall cohort, <11% of patients used advanced therapy. The UC overall cohort had significantly higher incidence rates (IRs) across most safety outcomes compared with the Non-UC cohort (highest: malignancy [excluding non-melanoma skin cancer] at 31.1 compared with 14.1 IRs per 1000 patient-years, respectively), regardless of age or oral corticosteroid use. Incidence rates of safety outcomes for the UC advanced therapy and UC trial-similar (also requiring prior advanced therapy use) cohorts were generally similar or lower than the UC overall cohort. The UC trial-similar and UC advanced therapy cohorts had generally comparable safety outcomes.
These findings assist in understanding the background risks of safety events in patients with UC and suggest that the incidence of select safety outcomes is comparable between the ELEVATE UC trials and RWD.
Lay Summary
This large real-world data study found that patients with vs without ulcerative colitis had a higher incidence of most safety outcomes assessed, regardless of age or corticosteroid use. These findings may inform clinicians on background safety risks in these populations.
It is important that drug safety findings be complemented with real-world data in patients with ulcerative colitis.
This large real-world study found that patients with UC had higher incidence rates for most safety outcomes, regardless of age or corticosteroid use compared with individuals without UC. Patients using advanced therapies tended to have lower incidence rates than those who did not, and applying trial-like selection criteria did not significantly impact findings.
This study contributes to a better understanding of the safety profile of available therapies.
Introduction
Ulcerative colitis (UC) is a chronic inflammatory bowel disease (IBD) commonly presenting with frequent bloody bowel movements, abdominal pain, bowel urgency, incontinence, and fatigue.1,2 Due to the availability of novel advanced UC therapies, with varying efficacy profiles,3 different treat-to-target strategies are possible. Currently, these strategies aim for patients to achieve more immediate goals of symptomatic response and remission, while endoscopic healing, normalized quality of life, and absence of disability tend to be longer-term targets.4,5 As such, appropriate management of UC often requires the prolonged or chronic use of effective anti-inflammatory therapies.6
Treatment regimens for UC may comprise conventional therapies (ie, 5-aminosalicylates, corticosteroids [CS], and immunomodulators [eg, thiopurines]) or advanced therapies such as biologics (eg, tumor necrosis factor inhibitors), anti-integrins, anti-interleukins, and small molecules (eg, Janus kinase inhibitors and sphingosine 1-phosphate [S1P] receptor modulators).7 The treatment and management of UC usually require these agents to be administered as monotherapy or combined therapies.5 Generally patients initiate with conventional anti-inflammatory therapies; however, inadequate response to these treatments results in many patients being prescribed advanced therapies as subsequent lines of treatment.8 With approximately 1.0% of the US population having UC,9 and most available therapies requiring chronic treatment, it is crucial to continue monitoring safety events of available UC medications beyond their approvals.
Individuals with IBD are at an increased risk of developing certain comorbidities, which may or may not be related to active disease, and include, among others, arterial thromboembolic events, anxiety and depression, coronary artery disease, stroke, and colorectal cancer.10–12 Additionally, many UC treatments, such as immunosuppressive agents (eg, CS and thiopurines), may also increase the risk of patients developing certain comorbidities,13 which may be exacerbated with older age.11,14,15 Although phase 3 clinical trials aim to assess drug safety along with efficacy, limitations in regards to inclusion/exclusion criteria may hinder the generalizability of findings to real-world populations.16 Additionally, many rare safety events are not observed in phase 3 clinical trials due to relatively small population sizes and limited duration of follow-up. As safety is a key attribute for treatment selection, and with the availability of new therapy options such as the S1P receptor modulators (ie, etrasimod and ozanimod),17–19 it is crucial to contextualize clinical trial safety data20,21 with real-world findings.
Utilizing data from the large, national Optum® de-identified Electronic Health Record data set (Optum® EHR), we investigated safety outcomes of interest associated with S1P receptor modulator therapy in patients with UC compared with a non-UC population. Additionally, we also evaluated select safety outcomes in a sub-population of patients with UC who initiated advanced therapy or who mirrored the selection criteria of the ELEVATE UC clinical program.20
Methods
Study Design and Data Collection
This retrospective, non-interventional, observational study extracted data from the Optum® EHR database between January 1, 2016 and December 31, 2022. Optum® EHR contains medical data of over a 100 million US-based patients treated at more than 2000 hospitals and 7000 clinics across 150 000 providers. Medical, demographic, and clinical characteristic data were obtained from inpatient and ambulatory EHRs as well as practice management and other internal systems. Data were anonymized and certified by an independent statistical expert according to the Health Insurance Portability and Accountability Act of 1996 and managed as per Optum® EHR’s customer data use agreements. Following anonymization, data were processed, normalized, and standardized across the continuum of care from both acute inpatient stays and outpatient visits.
Cohorts
Two primary data sets were sampled: a general population without UC diagnosis (Non-UC cohort) and a population with UC diagnosis (UC overall cohort; Figure 1). Treatment with ≥1 prior UC therapy was an inclusion criterion for patients in the UC overall cohort (Figure 1). An additional subset within the UC population was created using inclusion/exclusion criteria to select patients who received advanced therapies (UC advanced therapy cohort; Figure 1). From the advanced therapy cohort, further selection criteria were applied to mirror that of a major, phase 3, clinical trial evaluating an S1P modulator (etrasimod), ie, the ELEVATE UC clinical program (Figure 1).20 Selection of the UC trial-similar cohort comprised age, UC characteristics, prior UC treatment, concomitant treatment, and general health/biomarker criteria.20 Detailed inclusion/exclusion criteria for all cohorts are shown in Figure 1. This final aim assists in evaluating the comparability of safety outcomes between large clinical trials and real-world findings to aid clinicians in treatment selection.
Attrition flow diagram indicating inclusion and exclusion criteria for cohort selection. aCodes could be from 2 outpatients (≥30 to ≤365 days apart) or ≥1 inpatient (≥30 to ≤365 days apart) ICD-9/10 diagnosis code for UC (K51.X). bTreatment must have occurred ±30 days from index date; conventional therapies comprised: 5-ASA, oral systemic CS, and thiopurines (azathioprine/6-mercaptopurine); advanced therapies comprised: infliximab, adalimumab, golimumab, vedolizumab, ustekinumab, and tofacitinib. cRandom sampling was performed at a frequency of approximately 5 to 1 matching the UC overall cohort based on age, sex, race, geographic location, and index year. dSelect cardiac conditions included myocardial infarction, unstable angina, stroke/transient ischemic attack, any heart failure requiring hospitalization ≤ 6 months prior to index date or history or presence of atrioventricular block/sick sinus syndrome. eDetermined by hemoglobin A1c > 9%. fOpportunistic infections included Pneumocystis jirovecii, Cryptococcal meningitis, progressive multifocal leukoencephalopathy, or history of disseminated herpes simplex or disseminated herpes zoster. gExcludes basal cell carcinoma and in situ squamous cell carcinomas of the skin. hBiologic agents include: infliximab, adalimumab, golimumab, vedolizumab, and ustekinumab; JAKi included: tofacitinib, upadacitinib. iS1P therapy comprised fingolimod, ozanimod, siponimod, or ponesimod. jLymphocyte-depleting therapies included: alemtuzumab, cladribine, rituximab, ocrelizumab, cyclophosphamide, mitoxantrone, total body irradiation, bone marrow transplantation, and daclizumab. 5-ASA, 5-aminosalicylates; AIDS, acquired immune deficiency syndrome; CS, corticosteroid; HIV, human immunodeficiency virus; ICD, International Classification of Diseases; IL, interleukin; IV, intravenous; JAKi, Janus kinase inhibitor; n, number of patients; N, the number of patients in the group; Optum® EHR, Optum® de-identified Electronic Health Record data set; S1P, sphingosine 1-phosphate; TB, tuberculosis; TNFi, tumor necrosis factor inhibitor; UC, ulcerative colitis.
Outcomes
For all cohorts, demographic and clinical characteristics were collected. Additionally, based on literature analyses, safety outcomes of interest for patients with UC receiving advanced therapies, including S1P receptor modulator therapy, were identified and categorized into fungal and viral infections (cryptococcal meningitis, cytomegalovirus, herpes simplex virus, progressive multifocal leukoencephalopathy, and varicella zoster virus); cardiac, vascular, and circulatory disorders (atrial fibrillation, atrioventricular block, bradycardia, heart failure, hypercholesteremia, myocardial infarction, stroke, and venous thromboembolic events); hepatobiliary injury/disorders (liver injury); malignancies and benign growths (malignancy [excluding non-melanoma skin cancer [NMSC] and malignancy [cutaneous]); and eye disorders (retinal disease). International Classification of Diseases (ICD)-9/10 codes assessed for all safety outcomes are summarized in Table S1.
Statistical Analyses
Descriptive statistics were performed for demographics and clinical characteristic data. Continuous variables were described using means, medians, and standard deviations (SDs), while dichotomous variables were described using frequencies and percentages which were categorical.
Patients at risk, number of events, person-years (PYs) at risk, and crude incidence rates (IRs) with 95% confidence intervals (CIs) were calculated for all safety outcomes across cohorts. Incidence rates for safety outcomes were calculated with person-time at risk starting on the index date (defined as the date of the first ICD-9/10 code for patients with UC; or enrollment date or January 1, 2016 if enrollment date was before the analysis start date for the Non-UC cohort individuals) and ending on the date of a censoring event (date until death, loss of activity, occurrence of the disease event of interest, occurrence of colectomy [for UC patients], loss of enrollment, or end of the study period). The washout period for incident outcomes was ≤365 days prior to the index date, consistent with baseline assessment of variables. If a prevalent event for a given patient was identified during the washout period, the patient was not eligible for that incident event. Incident rates per 1000 PY were calculated based on the number of new events divided by the sum of the duration of patient exposures from the index date to the censoring date during the risk period. Different safety events did not censor each other’s follow-up. Safety data were also stratified by age (18–39 years of age, 40–59 years of age, ≥60 years of age) and baseline use of oral CS. Non-overlapping 95% CI values were considered significantly different for this analysis. All analyses were executed using the statistical software SAS (version 9.4).
Results
Demographic and Clinical Characteristics
In total, data from 32 170 (UC overall cohort), 3332 (UC advanced therapy cohort), 1435 (UC trial-similar cohort), and 160 795 (Non-UC cohort) individuals were included. The mean age of the UC advanced therapy and UC trial-similar cohorts was lower than the UC overall cohort (44.0–45.0 years compared with 52.3 years; Table 1). Length of follow-up, sex, and race were generally well-balanced between all cohorts (Table 1 and Table S2). The mean Charlson Comorbidity Index score, among all cohorts, was highest in the UC overall cohort (1.0 [SD, 2.1]; Table 1 and Table S2). The UC advanced therapy and UC trial-similar cohorts tended to have a lower incidence of comorbid medical conditions compared with the UC overall cohort (Table 1). Among the UC overall cohort, <11% of patients used a form of advanced therapy (mostly infliximab [4.9%]), while 38.3% received oral CS within ±30 days of the index date (Table 1).
Demographic and clinical characteristics of the UC overall, UC advanced therapy, UC trial-similar, and Non-UC cohorts.
| UC overall (N = 32 170) | UC advanced therapy (N = 3332) | UC trial-similar (N = 1435) | Non-UC (N = 160 795) | |
| Demographics | ||||
| Age at index date, mean (SD) | 52.3 (17.9) | 45.0 (16.8) | 44.0 (16.6) | 52.3 (17.9) |
| Length of follow-up, years; mean (SD) | 2.8 (1.9) | 2.9 (2.0) | 2.7 (1.9) | 2.7 (2.0) |
| Sex, n (%) | ||||
| Female | 17 519 (54.5) | 1637 (49.1) | 696 (48.5) | 87 550 (54.4) |
| Male | 14 651 (45.5) | 1695 (50.9) | 739 (51.5) | 73 245 (45.6) |
| Race, n (%) | ||||
| White | 27 776 (86.3) | 2876 (86.3) | 1246 (86.8) | 138 865 (86.4) |
| Black | 2299 (7.1) | 164 (4.9) | 62 (4.3) | 11 480 (7.1) |
| Asian | 520 (1.6) | 68 (2.0) | 30 (2.1) | 2575 (1.6) |
| Other/Unknown | 1575 (4.9) | 224 (6.7) | 97 (6.8) | 7875 (4.9) |
| Baseline characteristics | ||||
| Charlson comorbidity index score (1 year prior to index date), mean (SD) | 1.0 (2.1) | 0.4 (1.4) | 0.1 (0.4) | 0.4 (1.2) |
| Medical history (1 year prior to index date), n (%) | ||||
| Diabetes (Type 1/Type 2) | 3368 (10.5) | 158 (4.7) | 23 (1.6) | 10 559 (6.6) |
| Pulmonary disease (COPD or asthma) | 3821 (11.9) | 186 (5.6) | 36 (2.5) | 8190 (5.1) |
| Chronic kidney disease | 1939 (6.0) | 54 (1.6) | 7 (0.5) | 3806 (2.4) |
| Nonalcoholic fatty liver disease | 652 (2.0) | 63 (1.9) | 11 (0.8) | 1132 (0.7) |
| Other immune diagnosis | 1081 (3.4) | 221 (6.6) | 75 (5.2) | 1695 (1.1) |
| Hypertension | 8808 (27.4) | 421 (12.6) | 95 (6.6) | 27 632 (17.2) |
| Hyperlipidemia | 7482 (23.3) | 372 (11.2) | 98 (6.8) | 21 190 (13.2) |
| Opportunistic infection | 1146 (3.6) | 87 (2.6) | 0 | 1789 (1.1) |
| Medical history (any time before index date), n (%) | ||||
| Extraintestinal manifestations | 4066 (12.6) | 367 (11.0) | 84 (5.9) | 4529 (2.8) |
| Malignancies (excluding NMSC) | 3703 (11.5) | 169 (5.1) | <5 (<0.4) | 6470 (4.0) |
| MACE (inpatient) | 2069 (6.4) | 70 (2.1) | 7 (0.5) | 3636 (2.3) |
| AV block | 388 (1.2) | 15 (0.5) | 0 | 650 (0.4) |
| Bradycardia | 1152 (3.6) | 43 (1.3) | 0 | 1995 (1.2) |
| Venous thromboembolism: PE | 411 (1.3) | 29 (0.9) | <5 (<0.4) | 472 (0.3) |
| Venous thromboembolism: DVT | 1291 (4.0) | 91 (2.7) | 17 (1.2) | 1418 (0.9) |
| Macular edema or retinopathy | 49 (0.2) | <5 (<0.4) | 0 | 66 (<0.1) |
| Medication history | ||||
| Prior conventional treatments (6 months prior to index date), n (%) | ||||
| Thiopurines | 1447 (4.5) | 233 (7.0) | 73 (5.1) | 70 (<0.1) |
| Oral CS | 21 970 (68.3) | 2321 (69.7) | 759 (52.9) | 29 530 (18.4) |
| Oral 5-ASA | 8448 (26.3) | 496 (14.9) | 142 (9.9) | 100 (0.1) |
| Prior conventional treatments (±30 days to index date), n (%) | ||||
| Thiopurines | 2247 (7.0) | 258 (7.7) | 85 (5.9) | 36 (<0.1) |
| Oral CS | 12 334 (38.3) | 1107 (33.2) | 307 (21.4) | 2258 (1.4) |
| Oral 5-ASA | 19 131 (59.5) | 660 (19.8) | 205 (14.3) | 47 (<0.1) |
| Prior advanced treatments (± 30 days to index date), n (%) | ||||
| Adalimumab | 884 (2.7) | 884 (26.5) | 380 (26.5) | 46 (<0.1) |
| Infliximab | 1575 (4.9) | 1575 (47.3) | 608 (42.4) | 22 (<0.1) |
| Golimumab | 78 (0.2) | 78 (2.3) | 27 (1.9) | <5 (<0.4) |
| Vedolizumab | 695 (2.2) | 695 (20.9) | 363 (25.3) | <5 (<0.4) |
| Ustekinumab | 64 (0.2) | 64 (1.9) | 29 (2.0) | 0 |
| Tofacitinib | 77 (0.2) | 77 (2.3) | 36 (2.5) | <5 (<0.4) |
| Upadacitinib | 9 (< 0.1) | 9 (0.3) | 7 (0.5) | <5 (<0.4) |
| UC overall (N = 32 170) | UC advanced therapy (N = 3332) | UC trial-similar (N = 1435) | Non-UC (N = 160 795) | |
| Demographics | ||||
| Age at index date, mean (SD) | 52.3 (17.9) | 45.0 (16.8) | 44.0 (16.6) | 52.3 (17.9) |
| Length of follow-up, years; mean (SD) | 2.8 (1.9) | 2.9 (2.0) | 2.7 (1.9) | 2.7 (2.0) |
| Sex, n (%) | ||||
| Female | 17 519 (54.5) | 1637 (49.1) | 696 (48.5) | 87 550 (54.4) |
| Male | 14 651 (45.5) | 1695 (50.9) | 739 (51.5) | 73 245 (45.6) |
| Race, n (%) | ||||
| White | 27 776 (86.3) | 2876 (86.3) | 1246 (86.8) | 138 865 (86.4) |
| Black | 2299 (7.1) | 164 (4.9) | 62 (4.3) | 11 480 (7.1) |
| Asian | 520 (1.6) | 68 (2.0) | 30 (2.1) | 2575 (1.6) |
| Other/Unknown | 1575 (4.9) | 224 (6.7) | 97 (6.8) | 7875 (4.9) |
| Baseline characteristics | ||||
| Charlson comorbidity index score (1 year prior to index date), mean (SD) | 1.0 (2.1) | 0.4 (1.4) | 0.1 (0.4) | 0.4 (1.2) |
| Medical history (1 year prior to index date), n (%) | ||||
| Diabetes (Type 1/Type 2) | 3368 (10.5) | 158 (4.7) | 23 (1.6) | 10 559 (6.6) |
| Pulmonary disease (COPD or asthma) | 3821 (11.9) | 186 (5.6) | 36 (2.5) | 8190 (5.1) |
| Chronic kidney disease | 1939 (6.0) | 54 (1.6) | 7 (0.5) | 3806 (2.4) |
| Nonalcoholic fatty liver disease | 652 (2.0) | 63 (1.9) | 11 (0.8) | 1132 (0.7) |
| Other immune diagnosis | 1081 (3.4) | 221 (6.6) | 75 (5.2) | 1695 (1.1) |
| Hypertension | 8808 (27.4) | 421 (12.6) | 95 (6.6) | 27 632 (17.2) |
| Hyperlipidemia | 7482 (23.3) | 372 (11.2) | 98 (6.8) | 21 190 (13.2) |
| Opportunistic infection | 1146 (3.6) | 87 (2.6) | 0 | 1789 (1.1) |
| Medical history (any time before index date), n (%) | ||||
| Extraintestinal manifestations | 4066 (12.6) | 367 (11.0) | 84 (5.9) | 4529 (2.8) |
| Malignancies (excluding NMSC) | 3703 (11.5) | 169 (5.1) | <5 (<0.4) | 6470 (4.0) |
| MACE (inpatient) | 2069 (6.4) | 70 (2.1) | 7 (0.5) | 3636 (2.3) |
| AV block | 388 (1.2) | 15 (0.5) | 0 | 650 (0.4) |
| Bradycardia | 1152 (3.6) | 43 (1.3) | 0 | 1995 (1.2) |
| Venous thromboembolism: PE | 411 (1.3) | 29 (0.9) | <5 (<0.4) | 472 (0.3) |
| Venous thromboembolism: DVT | 1291 (4.0) | 91 (2.7) | 17 (1.2) | 1418 (0.9) |
| Macular edema or retinopathy | 49 (0.2) | <5 (<0.4) | 0 | 66 (<0.1) |
| Medication history | ||||
| Prior conventional treatments (6 months prior to index date), n (%) | ||||
| Thiopurines | 1447 (4.5) | 233 (7.0) | 73 (5.1) | 70 (<0.1) |
| Oral CS | 21 970 (68.3) | 2321 (69.7) | 759 (52.9) | 29 530 (18.4) |
| Oral 5-ASA | 8448 (26.3) | 496 (14.9) | 142 (9.9) | 100 (0.1) |
| Prior conventional treatments (±30 days to index date), n (%) | ||||
| Thiopurines | 2247 (7.0) | 258 (7.7) | 85 (5.9) | 36 (<0.1) |
| Oral CS | 12 334 (38.3) | 1107 (33.2) | 307 (21.4) | 2258 (1.4) |
| Oral 5-ASA | 19 131 (59.5) | 660 (19.8) | 205 (14.3) | 47 (<0.1) |
| Prior advanced treatments (± 30 days to index date), n (%) | ||||
| Adalimumab | 884 (2.7) | 884 (26.5) | 380 (26.5) | 46 (<0.1) |
| Infliximab | 1575 (4.9) | 1575 (47.3) | 608 (42.4) | 22 (<0.1) |
| Golimumab | 78 (0.2) | 78 (2.3) | 27 (1.9) | <5 (<0.4) |
| Vedolizumab | 695 (2.2) | 695 (20.9) | 363 (25.3) | <5 (<0.4) |
| Ustekinumab | 64 (0.2) | 64 (1.9) | 29 (2.0) | 0 |
| Tofacitinib | 77 (0.2) | 77 (2.3) | 36 (2.5) | <5 (<0.4) |
| Upadacitinib | 9 (< 0.1) | 9 (0.3) | 7 (0.5) | <5 (<0.4) |
Abbreviations: 5-ASA, 5-aminosalicylic acid; AV, atrioventricular; COPD, chronic obstructive pulmonary disease; CS, corticosteroids; DVT, deep vein thrombosis; MACE, major adverse cardiovascular event; n, number of patients; N, the number of patients in the group; NMSC, non-melanoma skin cancer; PE, pulmonary embolism; UC, ulcerative colitis.
Demographic and clinical characteristics of the UC overall, UC advanced therapy, UC trial-similar, and Non-UC cohorts.
| UC overall (N = 32 170) | UC advanced therapy (N = 3332) | UC trial-similar (N = 1435) | Non-UC (N = 160 795) | |
| Demographics | ||||
| Age at index date, mean (SD) | 52.3 (17.9) | 45.0 (16.8) | 44.0 (16.6) | 52.3 (17.9) |
| Length of follow-up, years; mean (SD) | 2.8 (1.9) | 2.9 (2.0) | 2.7 (1.9) | 2.7 (2.0) |
| Sex, n (%) | ||||
| Female | 17 519 (54.5) | 1637 (49.1) | 696 (48.5) | 87 550 (54.4) |
| Male | 14 651 (45.5) | 1695 (50.9) | 739 (51.5) | 73 245 (45.6) |
| Race, n (%) | ||||
| White | 27 776 (86.3) | 2876 (86.3) | 1246 (86.8) | 138 865 (86.4) |
| Black | 2299 (7.1) | 164 (4.9) | 62 (4.3) | 11 480 (7.1) |
| Asian | 520 (1.6) | 68 (2.0) | 30 (2.1) | 2575 (1.6) |
| Other/Unknown | 1575 (4.9) | 224 (6.7) | 97 (6.8) | 7875 (4.9) |
| Baseline characteristics | ||||
| Charlson comorbidity index score (1 year prior to index date), mean (SD) | 1.0 (2.1) | 0.4 (1.4) | 0.1 (0.4) | 0.4 (1.2) |
| Medical history (1 year prior to index date), n (%) | ||||
| Diabetes (Type 1/Type 2) | 3368 (10.5) | 158 (4.7) | 23 (1.6) | 10 559 (6.6) |
| Pulmonary disease (COPD or asthma) | 3821 (11.9) | 186 (5.6) | 36 (2.5) | 8190 (5.1) |
| Chronic kidney disease | 1939 (6.0) | 54 (1.6) | 7 (0.5) | 3806 (2.4) |
| Nonalcoholic fatty liver disease | 652 (2.0) | 63 (1.9) | 11 (0.8) | 1132 (0.7) |
| Other immune diagnosis | 1081 (3.4) | 221 (6.6) | 75 (5.2) | 1695 (1.1) |
| Hypertension | 8808 (27.4) | 421 (12.6) | 95 (6.6) | 27 632 (17.2) |
| Hyperlipidemia | 7482 (23.3) | 372 (11.2) | 98 (6.8) | 21 190 (13.2) |
| Opportunistic infection | 1146 (3.6) | 87 (2.6) | 0 | 1789 (1.1) |
| Medical history (any time before index date), n (%) | ||||
| Extraintestinal manifestations | 4066 (12.6) | 367 (11.0) | 84 (5.9) | 4529 (2.8) |
| Malignancies (excluding NMSC) | 3703 (11.5) | 169 (5.1) | <5 (<0.4) | 6470 (4.0) |
| MACE (inpatient) | 2069 (6.4) | 70 (2.1) | 7 (0.5) | 3636 (2.3) |
| AV block | 388 (1.2) | 15 (0.5) | 0 | 650 (0.4) |
| Bradycardia | 1152 (3.6) | 43 (1.3) | 0 | 1995 (1.2) |
| Venous thromboembolism: PE | 411 (1.3) | 29 (0.9) | <5 (<0.4) | 472 (0.3) |
| Venous thromboembolism: DVT | 1291 (4.0) | 91 (2.7) | 17 (1.2) | 1418 (0.9) |
| Macular edema or retinopathy | 49 (0.2) | <5 (<0.4) | 0 | 66 (<0.1) |
| Medication history | ||||
| Prior conventional treatments (6 months prior to index date), n (%) | ||||
| Thiopurines | 1447 (4.5) | 233 (7.0) | 73 (5.1) | 70 (<0.1) |
| Oral CS | 21 970 (68.3) | 2321 (69.7) | 759 (52.9) | 29 530 (18.4) |
| Oral 5-ASA | 8448 (26.3) | 496 (14.9) | 142 (9.9) | 100 (0.1) |
| Prior conventional treatments (±30 days to index date), n (%) | ||||
| Thiopurines | 2247 (7.0) | 258 (7.7) | 85 (5.9) | 36 (<0.1) |
| Oral CS | 12 334 (38.3) | 1107 (33.2) | 307 (21.4) | 2258 (1.4) |
| Oral 5-ASA | 19 131 (59.5) | 660 (19.8) | 205 (14.3) | 47 (<0.1) |
| Prior advanced treatments (± 30 days to index date), n (%) | ||||
| Adalimumab | 884 (2.7) | 884 (26.5) | 380 (26.5) | 46 (<0.1) |
| Infliximab | 1575 (4.9) | 1575 (47.3) | 608 (42.4) | 22 (<0.1) |
| Golimumab | 78 (0.2) | 78 (2.3) | 27 (1.9) | <5 (<0.4) |
| Vedolizumab | 695 (2.2) | 695 (20.9) | 363 (25.3) | <5 (<0.4) |
| Ustekinumab | 64 (0.2) | 64 (1.9) | 29 (2.0) | 0 |
| Tofacitinib | 77 (0.2) | 77 (2.3) | 36 (2.5) | <5 (<0.4) |
| Upadacitinib | 9 (< 0.1) | 9 (0.3) | 7 (0.5) | <5 (<0.4) |
| UC overall (N = 32 170) | UC advanced therapy (N = 3332) | UC trial-similar (N = 1435) | Non-UC (N = 160 795) | |
| Demographics | ||||
| Age at index date, mean (SD) | 52.3 (17.9) | 45.0 (16.8) | 44.0 (16.6) | 52.3 (17.9) |
| Length of follow-up, years; mean (SD) | 2.8 (1.9) | 2.9 (2.0) | 2.7 (1.9) | 2.7 (2.0) |
| Sex, n (%) | ||||
| Female | 17 519 (54.5) | 1637 (49.1) | 696 (48.5) | 87 550 (54.4) |
| Male | 14 651 (45.5) | 1695 (50.9) | 739 (51.5) | 73 245 (45.6) |
| Race, n (%) | ||||
| White | 27 776 (86.3) | 2876 (86.3) | 1246 (86.8) | 138 865 (86.4) |
| Black | 2299 (7.1) | 164 (4.9) | 62 (4.3) | 11 480 (7.1) |
| Asian | 520 (1.6) | 68 (2.0) | 30 (2.1) | 2575 (1.6) |
| Other/Unknown | 1575 (4.9) | 224 (6.7) | 97 (6.8) | 7875 (4.9) |
| Baseline characteristics | ||||
| Charlson comorbidity index score (1 year prior to index date), mean (SD) | 1.0 (2.1) | 0.4 (1.4) | 0.1 (0.4) | 0.4 (1.2) |
| Medical history (1 year prior to index date), n (%) | ||||
| Diabetes (Type 1/Type 2) | 3368 (10.5) | 158 (4.7) | 23 (1.6) | 10 559 (6.6) |
| Pulmonary disease (COPD or asthma) | 3821 (11.9) | 186 (5.6) | 36 (2.5) | 8190 (5.1) |
| Chronic kidney disease | 1939 (6.0) | 54 (1.6) | 7 (0.5) | 3806 (2.4) |
| Nonalcoholic fatty liver disease | 652 (2.0) | 63 (1.9) | 11 (0.8) | 1132 (0.7) |
| Other immune diagnosis | 1081 (3.4) | 221 (6.6) | 75 (5.2) | 1695 (1.1) |
| Hypertension | 8808 (27.4) | 421 (12.6) | 95 (6.6) | 27 632 (17.2) |
| Hyperlipidemia | 7482 (23.3) | 372 (11.2) | 98 (6.8) | 21 190 (13.2) |
| Opportunistic infection | 1146 (3.6) | 87 (2.6) | 0 | 1789 (1.1) |
| Medical history (any time before index date), n (%) | ||||
| Extraintestinal manifestations | 4066 (12.6) | 367 (11.0) | 84 (5.9) | 4529 (2.8) |
| Malignancies (excluding NMSC) | 3703 (11.5) | 169 (5.1) | <5 (<0.4) | 6470 (4.0) |
| MACE (inpatient) | 2069 (6.4) | 70 (2.1) | 7 (0.5) | 3636 (2.3) |
| AV block | 388 (1.2) | 15 (0.5) | 0 | 650 (0.4) |
| Bradycardia | 1152 (3.6) | 43 (1.3) | 0 | 1995 (1.2) |
| Venous thromboembolism: PE | 411 (1.3) | 29 (0.9) | <5 (<0.4) | 472 (0.3) |
| Venous thromboembolism: DVT | 1291 (4.0) | 91 (2.7) | 17 (1.2) | 1418 (0.9) |
| Macular edema or retinopathy | 49 (0.2) | <5 (<0.4) | 0 | 66 (<0.1) |
| Medication history | ||||
| Prior conventional treatments (6 months prior to index date), n (%) | ||||
| Thiopurines | 1447 (4.5) | 233 (7.0) | 73 (5.1) | 70 (<0.1) |
| Oral CS | 21 970 (68.3) | 2321 (69.7) | 759 (52.9) | 29 530 (18.4) |
| Oral 5-ASA | 8448 (26.3) | 496 (14.9) | 142 (9.9) | 100 (0.1) |
| Prior conventional treatments (±30 days to index date), n (%) | ||||
| Thiopurines | 2247 (7.0) | 258 (7.7) | 85 (5.9) | 36 (<0.1) |
| Oral CS | 12 334 (38.3) | 1107 (33.2) | 307 (21.4) | 2258 (1.4) |
| Oral 5-ASA | 19 131 (59.5) | 660 (19.8) | 205 (14.3) | 47 (<0.1) |
| Prior advanced treatments (± 30 days to index date), n (%) | ||||
| Adalimumab | 884 (2.7) | 884 (26.5) | 380 (26.5) | 46 (<0.1) |
| Infliximab | 1575 (4.9) | 1575 (47.3) | 608 (42.4) | 22 (<0.1) |
| Golimumab | 78 (0.2) | 78 (2.3) | 27 (1.9) | <5 (<0.4) |
| Vedolizumab | 695 (2.2) | 695 (20.9) | 363 (25.3) | <5 (<0.4) |
| Ustekinumab | 64 (0.2) | 64 (1.9) | 29 (2.0) | 0 |
| Tofacitinib | 77 (0.2) | 77 (2.3) | 36 (2.5) | <5 (<0.4) |
| Upadacitinib | 9 (< 0.1) | 9 (0.3) | 7 (0.5) | <5 (<0.4) |
Abbreviations: 5-ASA, 5-aminosalicylic acid; AV, atrioventricular; COPD, chronic obstructive pulmonary disease; CS, corticosteroids; DVT, deep vein thrombosis; MACE, major adverse cardiovascular event; n, number of patients; N, the number of patients in the group; NMSC, non-melanoma skin cancer; PE, pulmonary embolism; UC, ulcerative colitis.
Safety Outcomes
Overall analyses
The UC overall cohort had significantly higher IRs per 1000 PYs for all safety outcomes reported compared with the Non-UC cohort (Figure 2A-E). Significantly lower IRs per 1000 PYs were reported for the UC advanced therapy and UC trial-similar cohorts compared with the UC overall cohort for cryptococcal meningitis and progressive multifocal leukoencephalopathy; significant differences were not seen between UC cohorts for other fungal and viral infections assessed (Figure 2A). The UC advanced therapy and UC trial-similar cohorts had significantly lower IRs per 1000 PYs for all cardiac, vascular, and circulatory disorders (Figure 2B), hepatobiliary injury/disorders (Figure 2C), malignancy (excluding NMSC; Figure 2D), and eye disorders assessed (Figure 2E) compared with the UC overall cohort. Across all safety outcomes assessed, a significant difference between the UC advanced therapy and UC trial-similar cohorts was only seen for malignancy (excluding NMSC; Figure 2D). Corresponding IR and 95% CI values are shown in Table S3. Sensitivity analyses revealed comparable results when the patients in the UC advanced therapy cohort were omitted from the UC overall cohort (Table S4).
IR per 1000 PYs for (A) fungal and viral infections, (B) cardiac, vascular, and circulatory disorders, (C) hepatobiliary injury/disorders, (D) malignancies, (E) eye disorders for the UC overall, UC advanced therapy, UC trial-similar, and Non-UC cohorts. *Indicates significant differences as determined by non-overlapping 95% CI values. CI, confidence interval; IR, incidence rate; NMSC, non-melanoma skin cancer; PY, patient year; UC, ulcerative colitis.
Analyses stratified by age
Overall, significant differences for safety outcomes assessed between age groups were reported for the majority of fungal and viral infections for the UC overall and Non-UC cohorts (Table S5). Other than cytomegalovirus and herpes simplex virus, significant differences were not observed between age groups for the UC advanced therapy and UC trial-similar cohorts (Table S5). The majority of IRs per 1000 PYs reported for cardiac, vascular, and circulatory disorders and malignancies assessed increased significantly with increasing age, regardless of cohort (Table S5).
For fungal and viral infections, the UC overall cohort had significantly higher IRs per 1000 PYs compared with the Non-UC cohort for cytomegalovirus, herpes simplex virus, and varicella zoster virus, regardless of age; other significant differences were only seen for cryptococcal meningitis (40–59 years) and progressive multifocal leukoencephalopathy (≥60 years; Figure 3A and Tables S6–S8). For cardiac, vascular, and circulatory disorders, the UC overall cohort had significantly higher IRs per 1000 PYs compared with the Non-UC cohort for all safety outcomes assessed for the 40–59 year and ≥60-year groups; for the 18–39 year group, significant differences were seen for bradycardia, heart failure, myocardial infarction, and venous thromboembolic events (Figure 3B and Tables S6–S8). For all hepatobiliary injury/disorders and malignancies assessed, the UC overall cohort had significantly higher IRs per 1000 PYs compared with the Non-UC cohort, regardless of age (Figure 3C and D, and Tables S6–S8). For eye disorders, the UC overall cohort had significantly higher IRs per 1000 PYs compared with the Non-UC cohort for only the ≥60-year group (Figure 3E and Tables S6–S8).
IR per 1000 PYs for (A) fungal and viral infections, (B) cardiac, vascular, and circulatory disorders, (C) hepatobiliary injury/disorders, (D) malignancies, (E) eye disorders for the UC overall, UC advanced therapy, UC trial-similar, and Non-UC cohorts assessed, stratified by age. *Indicates significant differences as determined by non-overlapping 95% CI values. CI, confidence interval; IR, incidence rate; NMSC, non-melanoma skin cancer; PY, patient year; UC, ulcerative colitis.
When comparing the UC advanced therapy and UC trial-similar cohorts with the UC overall cohort for fungal and viral infections, only cryptococcal meningitis showed significantly lower IRs per 1000 PYs, regardless of age; significant decreases were also seen for cytomegalovirus, herpes simplex virus, and progressive multifocal leukoencephalopathy for the ≥60-year group (Figure 3A and Tables S6–S8). For cardiac, vascular, and circulatory disorders, both UC advanced therapy and UC trial-similar cohorts had significantly lower IRs per 1000 PYs compared with the UC overall cohort for atrial fibrillation (≥60-year group), atrioventricular block (≥60-year group), bradycardia (40–59 year group), heart failure (≥60-year group), and myocardial infarction (≥60-year group; Figure 3B and Tables S6–S8). Significantly lower IRs per 1000 PYs for both UC advanced therapy and UC trial-similar cohorts compared with the UC overall cohort were seen for hepatobiliary injury/disorders (≥60-year group), malignancy (excluding NMSC; ≥60-year group), and eye disorders (all age groups; Figure 3C–E and Tables S6–S8).
Across all safety outcomes assessed, significantly lower IRs per 1000 PYs were reported for the UC trial-similar cohort compared with the UC advanced therapy cohort for cytomegalovirus and herpes simplex virus (both for the ≥60-year group), and atrial fibrillation, atrioventricular block, bradycardia, myocardial infarction, and stroke (all for the 18–39 year group; Figure 3A and B and Tables S6–S8).
Analyses stratified by baseline oral corticosteroid use
Overall, most IR per 1000 PYs for fungal and viral infections were significantly greater for patients with CS use than without for the Non-UC cohort with minimal differences for the UC cohorts (Table S9). The majority of IR per 1000 PYs for cardiac, vascular, and circulatory disorders, liver injury, and malignancy (excluding NMSC) were significantly greater for patients with CS use compared with those without CS use for the UC overall, UC advanced therapy, and Non-UC cohorts; the same trend was not seen for the UC trial-similar cohort (Table S9).
Among patients with oral CS use, the UC overall cohort had significantly higher IRs per 1000 PYs compared with the Non-UC cohort for cryptococcal meningitis, cytomegalovirus, progressive multifocal leukoencephalopathy, myocardial infarction, venous thromboembolic events, liver injury, malignancy (excluding NMSC), and retinal disease (Figure 4 and Tables S10 and S11). For patients without oral CS use, all IRs per 1000 PYs for safety outcomes assessed (excluding cryptococcal meningitis) were significantly increased in the UC overall cohort compared with the Non-UC cohort (Figure 4 and Tables S10 and S11).
IR per 1000 PYs for (A) fungal and viral infections, (B) cardiac, vascular, and circulatory disorders, (C) hepatobiliary injury/disorders, (D) malignancies, (E) eye disorders for the UC overall, UC advanced therapy, UC trial-similar, and Non-UC cohorts assessed, stratified by baseline oral CS use. *Indicates significant differences as determined by non-overlapping 95% CI values. CI, confidence interval; CS, corticosteroid; IR, incidence rate; NMSC, non-melanoma skin cancer; PY, patient year; UC, ulcerative colitis.
Significantly lower IR per 1000 PYs were reported among patients with oral CS use in the UC advanced therapy and UC trial-similar cohorts compared with the UC overall cohort for cryptococcal meningitis, progressive multifocal leukoencephalopathy, atrial fibrillation, heart failure, hypercholesteremia, and retinal disease (Figure 4 and Tables S10 and S11). For patients without CS use, significantly lower IRs per 1000 PYs were reported for the UC advanced therapy and UC trial-similar cohorts compared with the UC overall cohort for cryptococcal meningitis, progressive multifocal leukoencephalopathy, most cardiac, vascular, and circulatory disorders, liver injury, malignancy (excluding NMSC), and retinal disease (Figure 4 and Tables S10 and S11).
Across all safety outcomes assessed, regardless of CS use, no significant difference was reported between the UC advanced therapy and UC trial-similar cohorts (Figure 4 and Tables S10 and S11).
Discussion
This large cohort real-world data (RWD) analysis, evaluating data from over 32 000 patients with UC, provides important insight regarding the incidence of safety outcomes in patients with UC. The rich data set and large population size allow for robust IR determination for the comparison of UC subpopulations against a non-UC population, while also allowing contextualization of the results of the ELEVATE UC clinical program. To our knowledge, this is the first study to examine a breadth of demographic and clinical characteristics and safety outcomes of interest across different subgroups of patients with UC and the general population in a large, national dataset.
Some patient characteristics differed between cohorts, with the UC advanced therapy and UC trial-similar cohorts tending to be younger with fewer comorbid conditions than the UC overall and Non-UC cohorts. This aligns with RWD studies finding similar age discrepancies between these populations.22,23 The differences may be attributed to younger patients (<40 years of age) being almost twice as likely to initiate advanced or targeted therapies than older patients (≥40 years of age).24 Additionally, multi-morbidities are more likely in older patients with IBD than younger ones.25 As active UC directly correlates with the onset of UC-related comorbidities,26 more effective disease control with advanced therapies may also explain the lower rate of comorbidities seen for cohorts taking these treatments. Among the UC overall cohort, up to 60% had used a form of conventional therapy 30 days prior to index, which reflects generally high usage; however, as shown previously,27 there was a low rate of advanced therapy use in this cohort. Comparatively, there were generally low rates of concomitant conventional therapy use in the UC advanced therapy and UC trial-similar cohorts.
The UC overall cohort had higher IRs for all safety outcomes assessed when compared with the Non-UC cohort, consistent with previous findings.10–14,28–31 For UC populations receiving advanced therapies (ie, UC advanced therapy and UC trial-similar cohorts), lower IRs were reported for most safety outcomes, excluding fungal and viral infections, compared with the UC overall cohort. Although these discrepancies between UC cohorts may be attributed to the use of advanced therapies, other factors may impact interpretations. Firstly, the UC overall cohort had a higher mean age with a greater incidence of comorbidities than the UC advanced therapy and UC trial-similar cohorts; secondly, the UC overall cohort had a higher incidence of patients using oral CS than the UC advanced therapy and UC trial-similar cohorts (which has been associated with an increased risk of several adverse events)32,33; thirdly, the selection criteria of the UC trial-similar cohort may have excluded certain at-risk populations. As such, analyses according to specific population subsets may improve the interpretation of the findings.
To ensure that age was not a variable for incidence of safety outcomes, we stratified analyses by 3 age groups. Most IRs of safety outcomes were higher in the UC overall cohort compared with the Non-UC cohort, regardless of age. In this analysis, we found that patients receiving advanced therapies did not have an increased risk of infections, cardiac risks, or malignancies when compared with the UC overall cohort. Most IRs of safety outcomes for cohorts receiving advanced therapies were lower than the UC overall cohort for the ≥60 year group. Importantly, some clinical studies of advanced therapies have shown a potential increased risk of infections, cardiac disorders, malignancies, and hepatobiliary injury in older patients.34,35 Our findings did not support these risks in real-world populations. For most safety outcomes assessed, oral CS use was associated with higher IR than without oral CS use, supporting established findings.32 For cohorts receiving advanced therapies, lower IRs for safety outcomes were seen for certain infections, cardiac disorders, and retinal disease when compared with the UC overall cohort, regardless of oral CS use, consistent with previous findings.36 Taken together, these sub-analyses support the trends seen between cohorts for the unstratified analysis. Additionally, these analyses also support older age and oral CS use as risk factors for the onset of certain safety outcomes.
Safety outcomes from clinical trials provide the basis for identifying adverse events that could be experienced by patients during therapy use; however, applicability to real-world populations may be limited due to the strict selection criteria applied in clinical trials. As such, assessing the findings of the UC trial-similar cohort may help to interpret clinical trial safety results. For the overall analysis, the IRs of safety outcomes for the UC trial-similar cohort were generally comparable with similar outcomes in patients receiving etrasimod in the ELEVATE UC clinical trial program. The assessed outcomes included key events such as atrioventricular block and macular edema, as well as select infections, bradycardia, and liver enzyme elevations.20,37 Additionally, findings of long-latency events, such as malignancies, remained similar between the ELEVATE UC trial patients and the UC trial-similar cohort.20,37 Although selection criteria for the UC trial-similar cohort were not identical to the ozanimod True-North phase 3 trial,38 safety outcomes from this analysis showed comparable rates to those receiving ozanimod after a 3-year follow-up, such as liver enzyme elevations and herpes zoster infection.39
Importantly, it should be noted that despite the strict selection criteria employed for the generation of the UC trial-similar cohort, safety outcomes were comparable to the UC advanced therapy cohort, suggesting that the clinical trial selection criteria did not significantly impact contextualization to similar real-world populations. This analysis supports the generalizability of clinical trial outcomes with regard to select safety outcomes associated with S1P receptor modulator therapy.
The findings of this analysis should be considered in line with several limitations: (1) non-overlapping 95% CI values were regarded as statistically significant; however, analyses adjusting for confounding variables were not evaluated; (2) as this analysis is based on EHR data, the completeness is limited to what has been captured/provided within structured input fields (eg, common safety events, including infections, can be difficult to capture) and the population included within the system (including some trial-similar characteristics); (3) the infection analysis was limited to 5 specific infectious agents (cryptococcal meningitis, cytomegalovirus, herpes simplex virus, progressive multifocal leukoencephalopathy, and varicella zoster virus) and may not be representative of the infectious burden in the UC population; (4) a longer follow-up period may have provided additional insights into long-latency safety events; (5) results may not be fully generalizable to all populations of patients with UC within, or outside of, the United States; and (6) non-significant findings between the UC trial-similar and UC advanced therapy cohorts may be attributed to underpowering. Overall, the size of the population assessed, comprising rich data sets, allows for robust IR determination.
This large US-based RWD assessment, comprising common and rarely observed safety events, found that patients with UC compared with those without UC had higher IRs for most safety outcomes assessed, regardless of age or oral CS use. Additionally, patients using advanced therapies tended to have lower IR for safety outcomes assessed than the UC overall cohort, and these were generally comparable to clinical trial findings. These findings will assist clinicians by providing valuable information on the background risks of safety outcomes in patients with UC to inform shared decision-making. This study may also contribute to a better understanding of the safety profile of available treatment options.
Supplementary Data
Supplementary data is available at Inflammatory Bowel Diseases online.
Acknowledgments
Medical writing support, under the direction of the authors, was provided by Wynand van Losenoord, MSc, CMC Connect, a division of IPG Health Medical Communications was funded by Pfizer, New York, NY, USA, in accordance with Good Publication Practice (GPP 2022) guidelines (Ann Intern Med 2022;175:1298–1304).
Author Contributions
M.D.R.: Conceptualization; writing—review and editing.
M.B.: Conceptualization; writing—review and editing.
M.C.D.: Conceptualization; writing—review and editing.
C.F.: Formal analysis, methodology; writing—review and editing.
D.B.: Conceptualization; writing—review and editing.
A.M.: Conceptualization; writing—review and editing.
M.G.: Conceptualization; methodology; writing—original draft; writing—review and editing.
Funding
The authors disclosed receipt of the following financial support for the research, authorship and/or publication of this article: this study was sponsored by Pfizer. Funding for medical writing support was provided by Pfizer.
Conflicts of Interest
M.D.R. has served on advisory boards and as a consultant for AbbVie, ALFASIGMA, S.p.A., Allergan, Amgen, Bristol Meyer Squibb, Celgene, Genentech, Gilead, Janssen, Lilly, Miraca Labs, Pfizer Inc., Pharma Solutions, Prometheus, Salix, Seres, Takeda, TARGET, and UCB; declares CME companies for CME Outfitters, Cornerstones, GI Health Foundation (GiHF), iMedeX, MJH Life Sciences, and Remedy; royalties for Wolters Kluwer Health as author/editor of UpToDate; and unrestricted educational grants for AbbVie, Bristol Myers Squibb, Celgene, Genentech, Gilead, Janssen, Pfizer Inc., Takeda, and UCB. M.B. reports grant/research funding for GlaxoSmithKline, Iterative Scopes, Janssen, and Takeda; serving as a consultant for AbbVie, Bristol Myers Squibb, Johnson & Johnson, MedEd, and Pfizer Inc.; and received honorarium from CME program (sponsored by AbbVie). M.C.D. has received consulting fees from AbbVie, Abivax, Arena Pharmaceuticals, AstraZeneca, Bristol Myers Squibb, Eli Lilly, Galapagos, Genentech, Gilead Sciences, Janssen Pharmaceuticals, Merck, Pfizer Inc., Prometheus Laboratories, Prometheus Biosciences, Takeda, S-PHYRE, and UCB; declares a directorship/ownership interest in Trellus Health; and shareholder/royalties from Trellus Health. D.B., A.M., and M.G. are employees and shareholders of Pfizer Inc. C.F. was an employee of Pfizer Inc. at the time of this analysis; and is a shareholder of Pfizer Inc.
Data Availability
Upon request, and subject to review, Pfizer will provide the data that support the findings of this study. Subject to certain criteria, conditions, and exceptions, Pfizer may also provide access to the related individual de-identified participant data. See https://www.pfizer.com/science/clinical-trials/trial-data-and-results for more information.
