Adaptive Steroid Tapering Impedes Corticosteroid-free Remissions Compared with Forced Tapering in Clinical Trials of Ulcerative Colitis

Abstract

Introduction

It is unclear if steroid tapering protocols can affect clinical trial outcomes in ulcerative colitis [UC], particularly fixed versus adaptive steroid tapering. Fixed steroid tapering involves incremental dose decreases at prespecified intervals, and adaptive steroid tapering uses investigator discretion as determined by the patient’s response.

Methods

In this post-hoc analysis from six clinical trials of UC [VARSITY, ACT 1, PURSUIT, GEMINI1, OCTAVE, and ULTRA2], responders to induction therapy with baseline corticosteroid use were considered as the primary population of interest. Adjustments were made to account for treat-through versus re-randomisation designs, and multivariate regression was performed to account for other potential confounding variables. The primary outcome was corticosteroid-free clinical remission [CR] at 1 year, and secondary outcomes were CR and endoscopic improvement.

Results

There was a total of 861 patients who had achieved clinical response after induction and were using corticosteroids. Within multivariate analysis, patients using adaptive steroid tapering regimens were less likely to achieve corticosteroid-free CR at 1 year (odds ratio [OR] 0.66 [95% confidence interval, CI, 0.48-0.92], p = 0.015) but had increased odds for achieving CR at 1 year (OR 1.9 [95% CI 1.43-2.52], p < 0.001). The steroid tapering regimen was not associated with achievement of endoscopic improvement at 1 year.

Conclusions

Among patients with UC on corticosteroids in clinical trials, patients using adaptive steroid weaning regimens were less likely to achieve corticosteroid-free CR at 1 year but more likely to achieve CR at 1 year. Consideration should be given to implementing mandatory fixed steroid weaning protocols in future clinical trials of UC.

Graphical Abstract

Corticosteroid, steroid, weaning, ulcerative colitis

1. Introduction

Ulcerative colitis [UC] is a chronic inflammatory bowel disease characterised by mucosal inflammation affecting the large intestine. UC is characterised by blood in the stool, diarrhoea, increased frequency of bowel movements, tenesmus, and/or urgency.¹ Corticosteroids remain relevant for treatment of patients with moderate to severe disease, but in those with refractory or corticosteroid-dependent disease, advanced therapies are recommended. The discovery of novel biologics and small molecules has been a breakthrough for improving long-term outcomes, including corticosteroid-free clinical remission.

Many patients with UC enrolled in clinical trials of biologic and small molecule therapies are using corticosteroids at the time of trial entry. However, protocols for managing steroids during clinical trials lack standardisation. Most studies use a fixed steroid tapering regimen, which entails a mandated and structured incremental decrease in the corticosteroid dose starting at a prespecified time point. Within many UC trials that use a fixed regimen, patients whose clinical symptoms worsen during steroid weaning are permitted to increase the steroid dose back to their initial baseline dose and then resume a mandatory taper, with failure of weaning considered treatment failure.^2–5 Some studies consider need for a one-time steroid re-escalation as a treatment failure.⁶ Recent trials of moderate to severe UC have permitted adaptive corticosteroid tapering regimens.^7,8

The primary objective of this study was to evaluate whether steroid tapering regimens in clinical trials affect the rate of corticosteroid-free CR at 1 year among patients with ulcerative colitis using corticosteroids.

2. Methods

2.1. Study design

Participant-level data were obtained from multiple, placebo-controlled, clinical trials of moderate to severe UC. Data from GEMINI-1 [Clinicaltrials.gov: NCT00783718], ULTRA-2 [NCT00408629], OCTAVE-1 [NCT01465763], OCTAVE-2 [NCT01458951], OCTAVE-SUSTAIN [NCT01458574], and VARSITY [NCT02497469] were obtained through Vivli [protocol #00007656], and by permission from Takeda, AbbVie, and Pfizer. In addition, data from ACT-1 [NCT00036439] and PURSUIT [NCT00488631] were obtained through the YODA [Yale Open Data Access #2022-4882] Project,⁹ and by permission from Janssen. Details of each study, including steroid tapering regimen used, are provided in Supplementary Appendix 1. The Hamilton Integrated Research Ethics Board determined that local ethics review was unnecessary, as data used in this analysis were previously collected and deidentified, and therefore no informed consent was required.

2.2. Exposures

Patients included in this analysis who received corticosteroids throughout the period of induction and after achievement of clinical response, underwent an attempt to wean the steroids while in maintenance and continuing to receive an active biologic or small molecule therapy. This included patients from ‘treat-straight-through’ studies who received active therapy [VARSITY, ACT-1, and ULTRA-2], and patients who responded to induction therapy and were subsequently re-randomised during the maintenance phase to an active therapy [PURSUIT, GEMINI-1, and OCTAVE]. Patients in the re-randomisation studies who were re-randomised to placebo were omitted. Adjustment was made for patients in treat-straight-through studies, VARSITY, ACT-1 and ULTRA-2, to account for those who had clinical response post-induction and were using corticosteroids, in order to create a cohort of patients comparable to a re-randomisation study.

2.3. Variables

The main variable of interest was the steroid weaning regimen used in the clinical trial [adaptive vs fixed]. Other variables that may have an association with the outcomes of interest were considered, including patient demographics [age, sex, race, weight, and current smoking], disease characteristics [prior anti-tumour necrosis factor [TNF] exposure, prior immunomodulator], treatment received within the clinical trial, steroid dose at time of steroid weaning initiation, biochemical markers of disease (C-reactive protein [CRP], serum albumin), and disease activity [partial Mayo score, Mayo endoscopic score]. All predictors were evaluated as binary or continuous variables.

2.4. Outcomes

The primary outcome of this study was corticosteroid-free clinical remission [CR] at 1 year. This was defined as the absence of corticosteroid use at the 1-year assessment, with a total Mayo score of ≤ 2 and no subscore > 1, among those using corticosteroids at baseline. The secondary outcomes of interest were CR at 1 year, defined as total Mayo score of ≤ 2 with no subscore > 1, and endoscopic improvement [Mayo endoscopic subscore of 0 or 1] at 1 year.

2.5. Statistical analysis

This was a post-hoc analysis of data from six clinical trial programmes. Descriptive statistics were provided for each comparison. Continuous variables were presented as means (and standard deviations [SD]) or medians (and interquartile ranges [IQR]). Dichotomous variables were summarised as proportions or percentages. Univariate analyses were conducted to evaluate the relationship between variables of interest at the time of steroid wean initiation [‘baseline’] and achievement of the primary and secondary outcomes. Subsequently, multivariate logistic regression using backward stepwise selection was performed, and only those baseline variables with p < 0.05 on univariate analysis were considered for inclusion. Unadjusted odds ratios [ORs] and adjusted ORs [aORs] are presented with associated 95% confidence intervals [CIs]. Statistical significance was chosen to be at p < 0.05. Data were analysed using Stata IC 15.0.

3. Results

3.1. Demographics

There was a total of 1723 patients across the included UC trials who had achieved clinical response after induction using advanced therapy. Baseline characteristics for all these patients are provided in Supplementary Table 1. From ‘treat-straight-through’ studies VARSITY, ACT 1, and ULTRA-2, 133, 78, and 83 patients, respectively, were on corticosteroids at baseline and achieved clinical response at Week 6 or 8. From re-randomisation studies PURSUIT, GEMINI-1, and OCTAVE, 78, 301, and 188 patients respectively: i] achieved clinical response after the induction phase of the study; ii] were re-randomised to continue active therapy; and iii] were on corticosteroids at baseline. Further, the 154 patients treated with golimumab 50 mg every 4 weeks were excluded, since 100 mg every 4 weeks was selected as the approved dosing per the US product monograph.¹⁰ As such, 861 patients were included in the primary analysis. Supplementary Table 2 summarises the individual trial results from the included studies, and Supplementary Figure 1 provides insight into how the cohorts of 1723 [total] and 861 [steroid-using] patients were derived. Table 1 provides baseline characteristics of patients in this analysis at the time of steroid weaning initiation. Of note, the median steroid dose of included patients was 17.5 mg [IQR 15] and patients had a mean Mayo score of 8.67 [SD 1.59]. Details of the steroid weaning regimens used within the included clinical trials are provided in Table 2.

Table 1

Demographics and disease characteristics of all patients at baseline.

Variable	Overall [n = 1723]
Age in years, mean [SD]	39.9 [13.1]
Sex, male, n [%]	1030 [59.8]
Race, White, n [%]	1066 [61.8]
Weight in kg, mean [SD]	74.4 [17.7]
Smoking, n [%]	125 [7.3]
Steroid dose, mg, median [IQR]	16 [17]
CRP mg/L, median [IQR]	3.61 [9.61]
Albumin g/L, mean [SD]	29.2 [18.2]
Previous TNF inhibitor, n [%]	382 [22.2]
Previous immunomodulator, n [%]	448 [26]
Total Mayo score, mean [SD]	8.69 [1.61]
Endoscopic Mayo score 3, n [%]	915 [53.1]
Clinical trial [biologic treatment], n [%]a	ACT 1 [infliximab]*	128 [7.4]
	VARSITY [adalimumab]a	164 [9.5]
	VARSITY [vedolizumab]a	201 [11.7]
	OCTAVE3 [tofacitinib]	395 [22.9]
	GEMINI1 [vedolizumab]	579 [33.7]
	PURSUIT [golimumab, 100 mg dose]	154 [8.9]
	ULTRA2 [adalimumab]a	102 [5.9]

SD, standard deviation; CRP, C-reactive protein; IQR, interquartile range; TNF, tumour necrosis factor.

^aNumber of people with clinical response post-induction and included in the maintenance study.

Table 2

Steroid weaning regimens.

Clinical trial	Steroid regimen	Adaptive vs fixed
GEMINI 1 and ULTRA 2	Incremental decrease by 5 mg starting Week 6 until achieving 10 mg, then decrease by 2.5 mg	Fixed
ACT 1, PURSUIT and OCTAVE	Incremental decrease by 5 mg starting Week 8 until achieving 20 mg, then decrease by 2.5 mg	Fixed
VARSITY	Adaptive steroid tapering starting Week 6	Adaptive

3.2. Univariate and multivariate model for prediction of corticosteroid-free clinical remission at 1 year

In the univariate analysis of baseline variables, smoking, albumin, steroid tapering regimen [adaptive vs fixed], partial Mayo score, and baseline endoscopic Mayo score had statistically significant associations with the primary outcome of corticosteroid-free CR at 1 year [Table 3]. After running backward stepwise selection, only steroid tapering regimen and partial Mayo score had an association with 1-year corticosteroid-free CR. Patients in studies where an adaptive steroid weaning protocol was used had lower odds of achieving corticosteroid-free CR at 1 year (aOR 0.66 [95% CI 0.48-0.92, p = 0.015]) and patients with a higher baseline partial Mayo score were also less likely to achieve corticosteroid-free CR at 1 year (aOR 0.88 [95% CI 0.80-0.97, p = 0.012]). The other variables were no longer significant, so were not selected for inclusion into the multivariate model.

Table 3

Univariate and multivariate analyses of baseline variables and association with Week 52 corticosteroid-free remission.

	Univariate		Multivariate
Predictor Variable	Odds ratio [95% CI]	p-value	Odds ratio [95% CI]	p-value
Age	0.99 [0.98-1.01]	0.56
Sex	0.99 [0.59-1.05]	0.94
Race [White]	0.73 [0.56–0.96]	0.026*	N/A	0.51
Smoking	0.29 [0.13–0.165]	<0.001*	N/A	0.27
Weight	1 [0.99–1.01]	0.39
Albumin	1.01 [1–1.02]	0.039*	N/A	0.08
CRP	0.99 [0.98–1.01]	0.92
Treatment allocation [reference: adalimumab]
Infliximab	1.4 [0.81-2.4]	0.23
Tofacitinib	0.69 [0.45-1.1]	0.095
Vedolizumab	0.71 [0.49-1.03]	0.51
Golimumab	0.79 [0.44-1.4]	0.42
Steroid tapering regimen [adaptive vs fixed]	0.5 [0.37–0.69]	<0.001*	0.66 [0.48-0.92]	0.015*
Steroid dose [mg]	0.99 [0.98–1]	0.64
Previous TNF inhibitor	0.48 [0.18-1.28]	0.14
Previous immunomodulator	0.89 [0.58 -1.38]	0.61
Partial Mayo score	0.91 [0.83-0.99]	0.035*	0.88 [0.8-0.97]	0.012*
Endoscopic Mayo score	0.72 [0.54–0.93]	0.016*	N/A	0.112

CRP, C-reactive protein; TNF, tumour necrosis factor; N/A, not available.

^*Bolded values indicate statistical significance with p-value <0.05.

3.3. Univariate and multivariate model for prediction of clinical remission at 1 year

In the univariate analysis of baseline variables, treatment allocation to infliximab, vedolizumab, and golimumab, steroid tapering regimen [adaptive vs fixed], CRP, baseline partial Mayo score, and endoscopic Mayo score had statistically significant associations with the secondary outcome of CR at 1 year [Table 4]. These variables were considered for inclusion into the multivariate model, where treatment allocations, steroid weaning regimen, baseline endoscopic Mayo score, and partial Mayo score were selected based on backward stepwise selection. Patients on an adaptive steroid tapering regimen were more likely to achieve CR at 1 year (aOR 1.9 [95% CI 1.43-2.52, p < 0.001]). Patients with a higher baseline endoscopic Mayo score and higher partial Mayo score were less likely to achieve CR at 1 year. CRP was no longer significant once the other variables were included in the multivariate model.

Table 4

Univariate and multivariate analyses of baseline variables and association with Week 52 clinical remission.

	Univariate		Multivariate
Predictor variable	Odds ratio [95% CI]	p-value	Odds ratio [95% CI]	p-value
Age	1 [0.99-1.01]	0.968
Sex	1.17 [0.97-1.41]	0.098
Race [White]	1.03 [0.88-1.22]	0.679
Smoking	0.77 [0.58-1.01]	0.059
Weight	1 [0.99-1.01]	0.259
Albumin	1 [0.99-1.01]	0.274
CRP	0.98 [0.97-0.99]	0.001*	N/A	0.33
Treatment allocation [reference: adalimumab]
Infliximab	1.8 [1.31-2.47]	<0.001*	2.96 [1.88-4.66]	<0.001*
Tofacitinib	1.36 [0.95-1.94]	0.096
Vedolizumab	1.36 [1.08-1.71]	0.007*	1.85 [1.36-2.53]	<0.001*
Golimumab	2.3 [1.57-3.38]	<0.001*	3.93 [2.38-6.48]	<0.001*
Steroid tapering regimen [adaptive vs fixed]	1.19 [1.01-1.43]	0.043*	1.9 [1.43-2.52]	<0.001*
Steroid dose [mg]	0.99 [0.99-1]	0.697
Previous TNF inhibitor	1.25 [0.33-4.76]	0.746
Previous immunomodulator	1.15[0.77-1.73]	0.494
Partial Mayo score	0.88 [0.83-0.93]	<0.001*	0.91 [0.85-0.98]	0.015*
Endoscopic Mayo score	0.69 [0.59-0.81]	<0.001*	0.66 [0.53-0.81]	<0.001*

CRP, C-reactive protein; TNF, tumour necrosis factor; N/A, not available.

^*Bolded values indicate statistical significance with p-value <0.05.

3.4. Univariate and multivariate models for prediction of endoscopic improvement at 1 year

In the univariate analysis of baseline variables, treatment allocation to infliximab, tofacitinib, and vedolizumab, CRP, baseline partial Mayo score, and endoscopic Mayo score had statistically significant associations with the secondary outcome of endoscopic improvement at 1 year [Table 5]. These variables were considered for inclusion into the multivariate model, where treatment allocations, baseline endoscopic Mayo score, and partial Mayo score were selected based on backward stepwise selection. Patients with a higher baseline endoscopic Mayo score and higher partial Mayo score were less likely to achieve endoscopic improvement at 1 year. CRP was no longer significant once the other variables were included in the multivariate model. The steroid tapering regimen did not have a significant association with endoscopic improvement at 1 year.

Table 5

Univariate and multivariate analyses of baseline variables and association with Week 52 endoscopic improvement.

	Univariate		Multivariate
Predictor variable	Odds ratio [95% CI]	p-value	Odds ratio [95% CI]	p-value
Age	0.99 [0.98-1]	0.263
Sex	1.08 [0.83-1.4]	0.573
Race [White]	0.98 [0.79-1.21]	0.884
Smoking	0.79 [0.55-1.14]	0.215
Weight	1.006 [0.99-1.01]	0.054
Albumin	1.01 [0.99-1.02]	0.337
CRP	0.98 [0.96-0.99]	0.049*	N/A	0.52
Treatment allocation [reference: adalimumab]
Infliximab	2.4 [1.52-3.7]	<0.001*	2.79 [1.47-5.28]	0.002*
Tofacitinib	0.59 [0.4-0.87]	0.008*	N/A	0.157
Vedolizumab	0.71 [0.53-0.94]	0.015*	N/A	0.512
Golimumab	1.34 [0.92-2.12]	0.116	N/A	N/A
Steroid tapering regimen [adaptive vs fixed]	1.23 [0.97-1.56]	0.087
Steroid dose [mg]	1 [0.99-1]	0.455
Previous TNF inhibitor	0.62 [0.24-1.61]	0.332
Previous immunomodulator	0.85 [0.58-1.26]	0.423
Partial Mayo score	0.92 [0.85-0.99]	0.018*	0.84 [0.74-0.95]	0.007*
Endoscopic Mayo score	0.66 [0.53-0.81]	<0.001*	0.57 [0.39-0.81]	<0.002*

CRP, C-reactive protein; TNF, tumour necrosis factor; N/A, not available.

^*Bolded values indicate statistical significance with p-value <0.05.

4. Discussion

This study demonstrates that patients with moderate to severe UC, managed with an adaptive corticosteroid tapering regimen, are less likely to achieve corticosteroid-free CR at 1 year of maintenance, but more likely to achieve CR after 1 year. This illustrates how corticosteroid management may affect clinical trial outcomes.

Inconsistencies in corticosteroid weaning protocols across different studies may obscure interpretation of clinical trials. The VARSITY study was the only trial included in our study which implemented an adaptive steroid tapering regimen. In VARSITY, more patients experienced CR at 1 year with vedolizumab compared with adalimumab, but among patients using corticosteroids at baseline, fewer patients experienced 1-year corticosteroid-free CR with vedolizumab compared with adalimumab. Continuation of corticosteroids for prolonged periods can mask clinical symptoms of active disease and may not permit assessment of the true efficacy of a treatment.

From a patient and ethical perspective, it may be argued that use of adaptive steroid weaning regimens is beneficial. Patients whose symptoms increase as corticosteroids are weaned should be offered re-escalation of corticosteroids for patient comfort, as is sometimes the case in clinical practice. However within clinical trials, this should be restricted to only one attempt and these patients should be required to reinitiate steroid weaning shortly after the re-escalation. Failure to have success with steroid weaning upon the second attempt should be classified as failure of the underlying advanced therapy being studied, at which time patients can continue steroid re-escalation if needed outside the setting of the clinical trial, and/or switch to an alternative therapy.

The strengths of our study include the use of multiple, high-quality, prospective datasets and a large number of patients. Limitations include the fact that only one trial used adaptive weaning and there was inconsistent collection of steroid doses between studies during the maintenance phase. Additionally, in more recent trials of IBD, steroid tapering was introduced during the induction phase,¹¹ but we were unable to assess the impact of this, as none of the studies included within our analysis had steroid weaning during induction. There were limited data on when complete tapering off corticosteroids was achieved [eg, may have occurred between two follow-up visits] and the cumulative dose of corticosteroids received by patients. Among the studies using a fixed steroid tapering regimen, discrepancies across data collection in the studies made it infeasible to ascertain which fixed steroid wean protocol may be preferable. Furthermore, there was some variability in the definition of corticosteroid-free CR between clinical trials [eg, being off steroids for at least 90 days vs being off steroids just at time of 1-year assessment]. For the purposes of this study, corticosteroid-free CR was calculated based on absence of steroids at time of the 1-year assessment, which may not exactly correspond to the corticosteroid-free CR rates reported within each of the individual trials, depending on the definition used with each trial. Last, the impact of steroid tapering regimen on Week 52 corticosteroid-free CR may be influenced by advanced therapies used, since data from multiple studies of advanced therapies were combined in this analysis. Univariate analysis did not find any association of advanced therapies used with Week 52 corticosteroid-free CR, and as such, these were not included in the multivariate analysis. However, we found that treatment received was significantly associated with Week 52 CR in both the univariate and multivariate analyses. It remains possible that there is some association of advanced therapy used with Week 52 corticosteroid-free CR, which was not found significant on univariate analysis due to smaller sample size of patients using corticosteroids at baseline [n = 861] as compared with the entire population studied [n = 1723].

Conclusion

In conclusion, our study suggests adaptive steroid tapering was associated with higher rates of CR at 1 year, but lower corticosteroid-free CR when compared with fixed steroid tapering in clinical trials of moderate-severe UC. These findings emphasise the impact of steroid weaning and its effects on clinical trial interpretability and patient outcomes. Moving forward, we recommend that standardised, fixed, steroid weaning protocols be implemented in maintenance during UC clinical trials, as it will enhance the quality of future clinical trials, improve interpretability of trial data, mitigate the risks associated with prolonged corticosteroid exposure, and optimise patient management and outcomes in UC.

Supplementary Data

Supplementary data are available at ECCO-JCC online.

Funding

No authors have received support for the submitted manuscript.

Conflict of Interest

NN has received honoraria from Janssen, Abbvie, Takeda, Pfizer, Sandoz, Novartis, Iterative Health, Innomar Strategies, Fresinius Kabi, Amgen, Organon, Eli Lilly, and Ferring. PSD has received research support, consulting, and/or speaker fees from Abbvie, Abivax, Adiso, Bristol Meyers Squibb, GSK, Janssen, Lilly, Pfizer, Roivant, Takeda; and royalties from University of California San Diego. JKM has receive consulting and/or speaking fees from AbbVie, Alimentiv, Amgen, Astra Zeneca, Bausch Health, Bristol Myers Squibb, Celltrion, Ferring, Fresenius Kabi, Janssen, Lilly, Lupin, Organon, Paladin, Pfizer, Pharmascience, Qu Biologics, Roche, Sandoz, SCOPE, Takeda, Teva, Viatris. WR has served as a speaker for AbbVie, Celltrion, Falk Pharma GmbH, Ferring, Janssen, Galapagos Medice, MSD, Roche, Pfizer, Pharmacosmos, Shire, Takeda, Therakos; as a consultant for AbbVie, Amgen, AOP Orphan, Arena Pharmaceuticals, Astellas, Astra Zeneca, Bioclinica, Boehringer Ingelheim, Bristol Myers Squibb, Calyx, Celgene, Celltrion, Eli Lilly, Falk Pharma GmbH, Ferring, Galapagos, Gatehouse Bio, Genentech, Gilead, Grünenthal, ICON, Index Pharma, Inova, Janssen, Landos Biopharma, Medahead, MedImmune, Microbiotica, Mitsubishi Tanabe Pharma Corporation, MSD, Novartis, OMass, Otsuka, Parexel, Periconsulting, Pharmacosmos, Pfizer, Protagonist, Provention, Quell Therapeutics, Sandoz, Seres Therapeutics, Setpointmedical, Sigmoid, Sublimity, Takeda, Teva Pharma, Therakos, Theravance, Zealand; as an advisory board member for AbbVie, Amgen, Astra Zeneca, Boehringer Ingelheim, Bristol Myers Squibb, Celgene, Celltrion, Galapagos, Janssen, Mitsubishi Tanabe Pharma Corporation, MSD, Pharmacosmos, Pfizer, Sandoz, Takeda; and has received research funding from AbbVie, Janssen, MSD, Sandoz, Sanofi, Takeda. VJ has received consulting/advisory board fees from AbbVie, Alimentiv, Arena Pharmaceuticals, Asahi Kasei Pharma, Asieris, Astra Zeneca, Avoro Capital, Bristol Myers Squibb, Celltrion, Eli Lilly, Endpoint Health, Enthera, Ferring, Flagship Pioneering, Fresenius Kabi, Galapagos, Gilde Healthcare, GlaxoSmithKline, Genentech, Gilead, Innomar, JAMP, Janssen, Merck, Metacrine, Mylan, Pandion, Pendopharm, Pfizer, Protagonist, Prometheus Biosciences, Reistone Biopharma, Roche, Roivant, Sandoz, SCOPE, Second Genome, Sorriso, Takeda, TD Securities, Teva, Topivert, Ventyx, Vividion; speaker’s fees from Abbvie, Ferring, Bristol Myers Squibb, Galapagos, Janssen Pfizer Shire, Takeda, Fresenius Kabi. SBH declares potential conflicts of interest including: Abbvie consultant, Clinical Research [Institution], speaker,; Amgen consultant, Clinical Research [Institution]; Boehringer-Ingelheim consultant; BMS consultant, DSMB; Celltrion consultant; Fresnius-Kabi consultant; Genentech consultant, Clinical Research, [Institution]; Gilead consultant, Clinical Research [Institution]; Gossamer DSMB, GSK consultant, Clinical Research [Institution]; Immunic consultant; Intercept Pharmaceiticals; Janssen consultant, Clinical Research [Institution], speaker; Lilly consultant, Clinical Research [Institution]; Merck consultant; Novartis consultant, Clinical Research [Institution]; Organon consultant; Pfizer consultant, Clinical Research [Institution], Speaker; Progenity consultant; Prometheus consultant, Clinical Research [Institution]; Protagonist consultant, DSMB; Receptos consultant, Clinical Research [Institution]; Salix consultant; Samsung Bioepis consultant; Seres Therapeutics consultant, Clinical Research [Institution]; Takeda consultant, Clinical Research [Institution], speaker; UCB consultant, Clinical Research [Institution]; Ventyx DSMB VHsquared consultant. No other authors have any relevant conflicts of interest.

Acknowledgements

Parts of this study were carried out under YODA Project # 2022-4882, and used data obtained from the Yale University Open Data Access Project, which has an agreement with Janssen Research & Development. The interpretation and reporting of research using this data are solely the responsibility of the authors and do not necessarily represent the official views of the Yale University Open Data Access Project or Janssen Research & Development. This publication [Vivli protocol #00007656] is based on research using data from AbbVie, Pfizer, and Takeda, which has been made available through Vivli. Vivli has not contributed to, or approved, and is not in any way responsible for, the contents of this publication.

Author Contributions

NN—study concept and design; acquisition and compilation of data; statistical analysis; data interpretation; drafting of the manuscript. JL, HH, ECW—acquisition and compilation of data; statistical analysis; drafting of the manuscript. JKM, VP, SBH, WR, PSD—study design; drafting of the manuscript. All authors approved the final version of the manuscript.

Data Availability

Data will be shared on request to the corresponding author with permission of Vivli/YODA.