Partial responders to intravenous corticosteroids at day 3 benefit from continued therapy in acute severe ulcerative colitis

Abstract

In acute severe ulcerative colitis (ASUC), early response patterns to intravenous corticosteroids (IVCS) during the first week may shape subsequent disease course. Characterizing these trajectories is crucial for guiding timely therapy and improving patient outcomes. We evaluated response patterns to IVCS in hospitalized patients with ASUC. Clinical response was categorized as complete response (CR), partial response (PR), or non-response (NR) on days 3, 5, and 7. Patients were followed up to day 90 for need for rescue therapy (RT), post-discharge escalation of therapy (EOT), and mortality. Seventy-nine patients were analyzed. By day 3, 12 (15%) achieved CR, 34 (43%) PR, and 33 (42%) NR. Among the 67 patients with PR or NR on day 3, 9 (13.4%) achieved CR by day 5 and 22 (32.8%) more by day 7. On day 7, CR, PR, and NR were observed in 42 (53%), 17 (21%), and 20 (25%) patients, respectively. Patients with CR on day 3 had a lower, but not statistically significant, likelihood of needing RT + EOT by day 90 compared to NR (29% vs. 47%; p = 0.15). However, by day 5, patients with CR or PR had significantly lower day 90 RT + EOT needs compared to NR (26% vs. 54%, p = 0.02; 23% vs. 54%, p = 0.03). By day 7, the day 90 probability of RT + EOT was higher in PR (45%) and NR (63%) versus CR (16%) (p = 0.001 and < 0.0001, respectively). Only a small proportion of patients with ASUC show CR at day 3; most are partial responders. Partial responders to IVCS at day 3 frequently achieve CR by day 5–7, and have relatively favorable short term disease course, highlighting the prognostic value of early response stratification.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-50171-4.

Introduction

Acute severe ulcerative colitis (ASUC) is a time-critical subset of ulcerative colitis (UC), characterized by rapid exacerbation of symptoms necessitating immediate medical intervention^1,2. ASUC, despite improved survival, carries high morbidity and poses significant challenges to both patients and healthcare providers, with nearly one-fifth of patients requiring colectomy³.

The timing of assessment of response in ASUC is a critical yet contentious issue. The landmark study by Travis et al., concluded that patients who exhibit persistent frequent stools (> 8 per day on day 3) or an elevated C-reactive protein (CRP) (> 45 mg/L) in conjunction with a stool frequency of 3–8 per day on day 3, have an 85% risk of requiring colectomy⁴. Since then day 3 is often regarded as the optimal time to evaluate the effectiveness of intravenous corticosteroids (IVCS) in treating ASUC. Among these patients, those with a partial response at day 3 represent a particularly challenging group. Partial responders have consistently been reported to have poor short- and long-term outcomes, with up to 40% undergoing colectomy within 6 months⁴. However, despite their high-risk profile, data on the clinical course and behavior of partial responders remain limited, leaving uncertainty around the optimal management strategy.

Additionally, in real-world practice, particularly in low- and middle-income countries (LMICs), the initiation of rescue therapy (RT) on day 3 among patients who fail to respond to IVCS may be influenced by several factors, including the high cost of biologic agents, concerns and stigma associated with colectomy and ostomy, and patient- and family-level reluctance to initiate advanced therapies due to financial constraints. Unlike in many Western healthcare systems, patients in these settings frequently bear a substantial proportion of treatment costs out of pocket, which may delay timely clinical decision-making. Furthermore, sociocultural factors play a significant role, as therapeutic decisions are often made collectively by family members rather than solely by the patient. Consequently, strict adherence to the day-3 decision point is variable, and RT is occasionally deferred to days 4–7 or later. The clinical implications of such delays remain incompletely characterized.

In this study, we prospectively analyzed a cohort of patients with ASUC, stratified according to their response to IVCS as complete response (CR), partial response (PR), or non-response (NR), and evaluated the association between response patterns within the first week of illness, the timing of RT, and medium-term disease outcomes.

Methods

This was a prospective observational cohort study conducted in patients with ASUC hospitalized at a tertiary care center in north India, between January 2023 and March 2024. The study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki. The study protocol was reviewed and approved by institutional ethics committee of Dayanand Medical College and Hospital, Ludhiana (Institutional Ethics Committee DMCH/R&D/2020/23). Informed consent was obtained from all participants prior to inclusion in the study.

Study population

Consecutive adult patients (≥ 18 years of age), hospitalized due to ASUC, were included. ASUC was defined by the modified Truelove and Witts (TW) criteria (≥ 6 stools with blood and at least one of the following: pulse > 90 beats per minute, temperature > 37.8 °C, hemoglobin < 10.5 g/dL, and C-reactive protein [CRP] ≥ 30 mg/L)^5,6. Patients with severe UC who did not meet the TW criteria, as well as those requiring emergency colectomy at admission for complications such as toxic megacolon, intestinal perforation, or massive hemorrhage, were excluded.

Demographic data and clinical profile

The patient demographics (including age, sex) and disease characteristics (including age at diagnosis, disease duration and extent, and previous treatments) were recorded at time of admission. The investigations, including hemogram, liver and renal function tests, CRP, chest and abdominal radiographs, and unprepared flexible sigmoidoscopy, were performed in all patients within 24 h of hospitalization. Endoscopic disease activity was assessed by the ulcerative colitis endoscopic index of severity (UCEIS)⁷. Evaluations for enteric infections included testing for Clostridioides difficile toxin A and toxin B (Vitassay^® Clostridium difficile GDH+Toxin A + B), and cytomegalovirus (CMV, histological assessment for owl’s eye inclusion bodies on hematoxylin and eosin-stained biopsies and immunohistochemistry). The blood investigations were repeated at the discretion of the treating physician.

Procedures

All patients received intravenous (IV) hydrocortisone (100 mg every 6 h) along with supportive measures, including correction of fluid and electrolyte imbalances; enteral nutrition in the form of soft, liquid-semisolid meals provided by the hospital kitchen, with dietitian-guided optimization of caloric and protein intake; and prophylaxis for venous thromboembolism. Daily clinical assessments, including stool frequency and rectal bleeding, were performed for each patient until day 7 of hospitalization or until discharge, whichever occurred earlier.

Clinical outcomes were evaluated on days 3, 5, and 7. The definitions of clinical response were adapted from the Oxford criteria⁴. A complete response (CR) was characterized by ≤ 3 stools/day without blood and partial response (PR) was defined as 3–8 stools/day with a CRP < 45 mg/L or a reduction in daily stool frequency and/or the number of blood-containing stools by at least 50% from baseline. Non-response (NR) was defined as > 8 stools/day or 3–8 stools/day with CRP > 45 mg/L⁴. These criteria were pragmatically applied at days 3, 5 and 7 to characterize the dynamic evolution of treatment response. However, clinical management was not restricted to these time points. Treating physicians were free to escalate or de-escalate therapy at any time based on the patient’s clinical condition, hemodynamic stability, laboratory parameters, complications, or other emergent considerations.

Patients with CR to IV corticosteroids were switched to per-oral prednisolone (40 mg/day) in a tapering dose schedule. Prednisolone was tapered by 5 mg per week until 20 mg/day, followed by a taper of 2.5 mg per week until discontinuation, resulting in a 12-week tapering regimen. Simultaneously, initiation or optimization of steroid-sparing therapy (thiopurines or advanced therapies) was undertaken in patients deemed to be steroid-dependent or at high risk of relapse. In partial-responders at day 3, IVCS were continued until day 5 or 7. Non responders were offered RT with infliximab, cyclosporine, tofacitinib, or colectomy. The timing of initiation of RT during hospitalization was determined through shared decision-making between the treating physician and the patient. All patients were followed up to day 90 to assess disease activity, post-discharge escalation of therapy (EOT), including colectomy, and mortality. Any increase in disease activity, defined as worsening stool frequency and/or recurrence of rectal bleeding, between days 8 and 90 was considered indicative of relapse, prompting evaluation for EOT. Standard dosing protocols were used for the pharmacological agents used for RT or EOT (infliximab: 5 mg/kg intravenously at weeks 0, 2, and 6; cyclosporine: continuous intravenous infusion at 2 mg/kg/day for 5–7 days, patients responding to intravenous cyclosporine were transitioned to oral cyclosporine [4–5 mg/kg/day in two divided doses] and continued for 12 weeks as bridge therapy, while azathioprine was initiated for long-term maintenance, cyclosporine was discontinued once therapeutic response to azathioprine was achieved; tofacitinib: 10 mg orally twice daily for 8–16 weeks, in patients achieving clinical response, dose was decreased to 5 mg twice daily.)

Statistical analysis

Descriptive statistics were used to summarize patient demographics, baseline characteristics, and clinical profiles. Baseline data were reported as number (%), mean (± standard deviation) or median (IQR) as appropriate and categorical variables summarized as frequencies with percentages. To determine whether the data were normally distributed, a Kolmogorov-Smirnov test was used. The Sankey diagram was used to visualize patient flow through different response categories. Kaplan-Meier survival analysis was used to estimate RT or EOT-free survival at day 90. Log rank test was used to compare the Kaplan-Meier curves. A multivariate logistic regression analysis was performed to identify parameters independently associated with response to IVCS. Statistical significance was set at p < 0.05. All analyses were performed using SPSS version 26.0 (IBM Corp., Armonk, NY, USA).

Results

Patient and disease characteristics

Out of the 120 UC patients admitted during the study period, 79 patients (median age 36 [28-51.5] years, 36 [45.5%] females) fulfilled the TW criteria and were included. In addition to experiencing six or more stools with blood per day, 33 (41.77%) patients had one, 35 (44.30%) patients had two, and 11 (13.92%) patients had three or more additional features of systemic toxicity according to the TW criteria. The majority of patients had left-sided colitis (n = 49, 62.02%), followed by pancolitis (n = 18, 22.7%), and proctitis (n = 12, 15.1%). The median duration of the disease was 2 years (IQR 0.5–4.5). Nearly half of the patients (n = 39, 49.36%) experienced ASUC within the first year of their UC diagnosis. Eight (10.12%) patients with new onset UC presented with ASUC. (Table 1) At the time of hospitalization, 66 (83.54%) patients had been symptomatic for ≥ 7 days, and 55 (69.62%) were already on oral corticosteroids. Approximately 30% of the patients (n = 24) were previously exposed to advanced therapies including biologics and small molecules.

Table 1.

Baseline characteristics (n = 79).

Patient characteristics
Age (years)	36 (28-51.5)
Age at diagnosis (years)	34 (24-46.5)
Females	36 (45.5)
Disease extent
Proctitis	12 (15.19)
Left sided colitis	49 (62.02)
Pancolitis	18 (22.79)
Disease duration (years)	2 (0.5–4.5)
≤ 1	39 (49.36)
> 1	40 (50.63)
Index presentation as ASUC	8 (10.12)
Previous medications (at the time of hospitalization)
5-Aminosalicylates	76 (96.20)
Prednisolone	61 (77.2)
Thiopurines	19 (24.05)
Tofacitinib	16 (20.2)
Anti TNFs (Infliximab)	5 (6.32)
Anti Integrins (Vedolizumab)	3 (3.79)
Use of oral corticosteroids at the time of hospitalization	55 (69.62)
Disease characteristics
Number of positive systemic features of toxicity
1	33 (41.77)
2	35 (44.30)
≥ 3	11 (13.92)
Duration of symptoms at the time of presentation
< 3 days	4 (5.06)
< 7 days	9 (11.39)
≥ 7 days	66 (83.54)
Nutritional assessment
MUST ≥ 2	43 (54.43)
BMI (kg/m2)	20.98 (17.87–23.88)
Hand grip strength (kg)	20 (13.5–25.5)
Investigations*
Hemoglobin (g/dL)	9.7 (8.45–11.45)
Albumin (g/dL)	3.16 (2.35–3.65)
CRP (mg/L)	29.12 (11.08–63.70)
UCEIS	5 (5–6)
Deep ulcers on endoscopy	29 (36.70)
Clostridioides difficile infection	8 (10.12)
Cytomegalovirus infection	12 (15.18)

The values are expressed as median (interquartile range) or n (%) as applicable.

ACE, Albumin, CRP, and Endoscopy; ADMIT-ASC, Admission Model for Intensification of Therapy in Acute Severe Colitis; ASUC, Acute severe ulcerative colitis; BMI, Body mass index; CRP, C-Reactive Protein, MUST, Malnutrition Universal Screening Tool; TNF, tumor necrosis factor; UCEIS, Ulcerative Colitis Endoscopic Index of Severity.

*At the time of hospitalization.

Investigations

The median levels for hemoglobin, albumin, and CRP were 9.7 g/dL (IQR 8.45–11.45), 3.16 g/L (IQR 2.35–3.65), and 29.12 mg/L (IQR 11.08–63.70), respectively. Elevated CRP (≥ 30 mg/L) was noted in 39 (49.36%) patients. Baseline characteristics stratified by admission CRP levels (< 30 mg/L vs. ≥ 30 mg/L) are shown in Supplementary Table 1. Clostridioides difficile infection was present in 4 patients (5.06%), and concomitant CMV infection was found in 7 patients (8.86%). The median UCEIS score was 5 (IQR 5–6).

Patterns of response

The response to therapy was evaluated on days 3, 5, and 7. On day 3, 46 (58.22%) patients were classified as responders; with CR observed in 12 (15.18%) patients and PR in 34 (43.03%). By day 5, 19 (24.05%) patients achieved CR and 35 (44.30%) patients achieved PR. Twenty-five (31.64%) patients were non-responders at day 5. By day 7, 42 (53.16%) patients had CR, 17 (21.51%) patients had PR, and 20 (25.31%) patients were non-responders. (Fig. 1)

Fig. 1.

Patterns of response in patients with ASUC. ASUC, Acute severe ulcerative colitis.

Among the 34 partial-responders on day 3, 6 (17.64%) patients achieved CR by day 5, and 15 (44.11%) patients by day 7. Additionally, of the 33 non-responders on day 3, 3 (9.09%) patients achieved CR by day 5, and 7 (21.21%) patients achieved CR by day 7. Furthermore, 11 (84.61%) of the 13 non-responders at day 5 remained non-responders by day 7. (Fig. 1)

Notably, two patients who initially achieved CR by day 3 experienced increased disease activity between days 3 and 5 and were reclassified as partial-responders at day 5. Of these, one patient regained CR by day 7, while the other remained a partial responder.

Clinical and biochemical trends

On day 3, patients with CR had low stool frequency (median 2/day) and low CRP (median 8.2 mg/L), compared with intermediate values in PR (median 5/day, 12.7 mg/L) and higher values in NR (median 8.5/day, 48.2 mg/L). This pattern persisted on day 5 and day 7, with patients with CR having lower stool frequency and CRP, compared to patients with PR and NR. (Supplementary Table 2)

Rescue therapy and post-discharge escalation of therapy

The median time to initiation of RT was 7 days (IQR, 5–7 days). Thirteen patients (16.45%) received medical RT (8 patients received infliximab, 3 received cyclosporine, and 2 received tofacitinib). Four patients (5.06%) underwent emergent colectomy. Between days 8 and 90, an additional 5 (6.32%) patients required EOT (infliximab, n = 4 and tofacitinib, n = 1). Another two patients (day 18, n = 1 and day 60, n = 1) underwent colectomy during this period. (Supplementary Fig. 1) Of the seven patients who required EOT between days 8 and 90, 3 patients had CR by day 3, and 1 patient each demonstrated CR by days 5 and 7. The remaining two patients were partial-responders at day 7.

Baseline use of advanced therapies did not differ between patients who achieved response to IVCS and those who required RT or EOT. (Supplementary Table 3)

Mortality

Three patients (3.79%) died during the study period. All three were non-responders to IVCS and had received RT (infliximab: n = 2 [both administered on day 7]; colectomy: n = 1 [done on day 7]). The causes of death were severe refractory disease and peritonitis (n = 1, day 13), multi-organ failure (n = 1, day 30), and post-surgery sepsis (n = 1, day 30).

Patterns of response and probability of receiving RT or EOT

We analyzed the cumulative probability of RT or EOT by day 90 based on the clinical status (CR, PR and NR) at days 3, 5 and 7. There was no significant difference in use of RT or EOT among complete responders, partial-responders and non-responders at day 3. However, the overall trends showed a significantly higher cumulative probability of RT or EOT use among the non-responders by day 5 and day 7, but not day 3. (Fig. 2)

Fig. 2.

Cumulative probability of rescue therapy + post discharge escalation of therapy free survival by day 90 stratified by the clinical status across various time-points of assessment, (a) day 3, (b) day 5, and (c) day 7.

A response-stratified analysis demonstrated that non-responders at day 5 had a significantly higher cumulative probability of requiring RT or EOT compared to complete responders (0.54 vs. 0.26, Log-Rank p = 0.02) and partial responders (0.54 vs. 0.23, Log-Rank p = 0.03). In contrast, the probability of RT or EOT was comparable between complete and partial responders at day 5 (0.26 vs. 0.23, Log-Rank p = 0.81).

Compared to complete responders at day 7, both partial responders (0.45 vs. 0.16, Log-Rank p = 0.001) and non-responders (0.63 vs. 0.16, Log-Rank p < 0.0001) exhibited a significantly higher cumulative probability of requiring RT or EOT. However, the probability of RT or EOT did not differ significantly between non-responders and partial responders (0.55 vs. 0.37, Log-Rank p = 0.80). (Figs. 3, 4 and 5).

Fig. 3.

Cumulative probability of rescue therapy + post discharge escalation of therapy free survival by day 90 based on response assessment at day 3 (a) complete response vs. non-response, (b) partial response vs. non-response, (c) complete response vs. partial response.

Fig. 4.

Cumulative probability of rescue therapy + post discharge escalation of therapy free survival by day 90 based on response assessment at day 5 (a) complete response vs. non-response, (b) partial response vs. non-response, (c) complete response vs. partial response.

Fig. 5.

Cumulative probability of rescue therapy + post discharge escalation of therapy free survival by day 90 based on response assessment at day 7 (a) complete response vs. non-response, (b) partial response vs. non-response, (c) complete response vs. partial response.

Discussion

ASUC is a medical emergency that requires prompt recognition, urgent hospitalization and timely management to improve patient outcomes. Since the publication of the landmark study by Travis et al., day 3 is widely utilized to assess the response to IVCS^1,4,8. However, we found CR rates were low (15%) on day 3 (majority were partial responders) but gradually increased on days 5 and 7. This trend of gradual increase in the response rates suggests the presence of patients who respond more slowly to IVCS^9,10. In the current study, the treatment outcomes stabilized by day 5, allowing distinction between non-responders and complete or partial-responders. The majority of patients who were non-responders on day 5 continued to be non-responders on day 7 highlighting no benefit of prolonging IVCS in this group.

Our findings also draw attention to partial responders as a clinically relevant subgroup. Importantly, a PR at day 3 did not equate to treatment failure, as a substantial proportion of these patients subsequently achieved CR by day 5 or 7. Historically, partial or incomplete responders have been considered at higher risk of adverse outcomes, with Travis et al. reporting colectomy rates approaching 40% within six months⁴. However, our cohort demonstrated a distinctly different course as none of the patients who were partial responders on day 3 required colectomy up to day 90. While this may suggest a relatively favourable short-term course, these should be interpreted cautiously given the sample size and follow up duration. Nevertheless, they support the consideration that immediate escalation decisions in partial responders at day 3 may not always be warranted and that a short period of continued IVCS could be reasonable in selected patients.

It was also interesting to observe the dynamic nature of the response. Some patients who initially showed a CR by day 3 could not sustain the same response until day 7, whereas certain non-responders on day 3 were reclassified as complete responders or partial-responders by days 5 and 7. Additionally, CR at day 3 does not preclude the need for subsequent escalation of therapy, as a proportion of these patients later required rescue therapy or EOT. This contrasts with the observations made by Travis et al. where only one complete responder required subsequent colectomy⁴. This variability in patient categorization at each assessment point highlights that the clinical status at day 3 is not definitive, and does not reflect the true narrative of response in patients with ASUC. It also indicates that clinical implications of ASUC are not confined to the first week of illness and the disease activity persists beyond, regardless of the patient’s clinical status during the initial week; necessitating extended follow-up.

Only 50% of the patients in our study exhibited elevated CRP levels (≥ 30 mg/L). Similar findings have been previously reported in studies on ASUC conducted in this region^11,12. Disease extent alone did not fully account for the observed distribution of CRP values. These findings indicate overlap in systemic inflammatory burden across disease phenotypes and suggest that anatomical extent alone is insufficient to explain inter-individual variability in CRP response. CRP is synthesized by hepatocytes and its release is predominantly stimulated by the cytokine interleukin-6 (IL-6)¹³. Ethnic variations in the IL-6 genotype have been documented, influencing IL-6 expression and consequently CRP secretion^14,15. These genetic differences could account for the proportion of patients with ASUC with normal CRP levels (< 30 mg/L). In addition, differences in disease biology, including predominantly localized mucosal inflammation with limited systemic inflammatory spillover, may further contribute to attenuated CRP responses in some patients. Alternatively, these observations might also suggest that the currently used CRP cut-off levels for diagnosing ASUC may need to be reconsidered and possibly lowered to enhance diagnostic accuracy¹⁶. It also indicates that CRP-based indices for assessing response may be prone to bias, highlighting the need for more precise and universally applicable indices for predicting therapeutic response.

A subset of patients were receiving biologics or small molecules as maintenance therapy at the time of hospitalization for ASUC, which may reflect a more treatment-refractory disease phenotype. Although this could potentially influence outcomes, prior use of advanced therapy was not associated with an increased requirement for RT or EOT in our cohort. Three patients (3.5%) died during the study period, a rate comparable to that reported in the TACOS trial¹⁷. These outcomes appeared related to factors such as late presentation, ongoing oral corticosteroid use, malnutrition, and the need for emergent colectomy^8,18.

Our findings should also be interpreted in the context of emerging data suggesting regional heterogeneity in the clinical course and outcomes of ASUC. A recent multinational study by Fernandes et al. reported significant differences in short-term outcomes and treatment trajectories between patients from Asia and Australasia¹⁹. These data suggest that the natural history and response to therapy in ASUC may vary across geographic settings, and caution is warranted when extrapolating findings across populations.

It may be argued that there is a circular reasoning on the premise that we have defined the outcome (need for RT or EOT) and then evaluated the measures (patterns of response during the first week of the illness) that define the outcome. In our cohort, we observed dynamic changes. Some initial partial responders and non-responders improved by day 5/day7, and some complete responders at day 3 eventually required EOT. These observations demonstrate that day 5 response is a risk stratifier rather than a surrogate of the final outcome, supporting its role as an independent predictor. The day 5 assessment, therefore, provides actionable information that is clinically meaningful and is distinct from merely restating the outcome itself.

The strengths of the study include a prospective real-world analysis of the patterns of response, treatment outcomes, and short-term disease trajectory in patients with ASUC. To our knowledge, this is the first study to analyze the disease course of ASUC based on the patient outcomes at different time-points of assessment. However, recruitment of patients from a single-center and cohort characteristics (majority of the patients being biologic-naïve) limit the generalizability of the findings. Response classification at days 5 and 7 was based on extrapolation of the criteria originally validated for day 3 assessment. Although this approach allowed longitudinal characterization of clinical trajectories, it has not been formally validated and should be interpreted accordingly. Also, follow-up beyond hospitalization was primarily focused on clinically relevant outcomes, including the need for EOT, colectomy, and mortality. Standardized clinical, biochemical, and endoscopic assessments were not routinely performed at day 90 or during longer-term follow-up, precluding evaluation of mucosal healing, or biomarker normalization. The study is, additionally, subject to the inherent limitations of an observational design. Potential confounders may influence both treatment response and clinicians’ decisions regarding RT/EOT. Although we performed multivariate analyses to account for these factors and none emerged as independent predictors of the need for RT or early escalation, residual confounding cannot be entirely excluded.

Conclusion

The response to IVCS in ASUC appears to be dynamic. While only a small proportion of patients achieve CR by day 3, a larger proportion demonstrate PR, some of whom may improve further by days 5–7 to achieve CR. These findings suggest that a PR at day 3 should be interpreted with caution and not automatically equated with treatment failure. Deferring definitive decision-making to day 5 in selected patients may allow for a more complete assessment of therapeutic response. However, these observations require validation in larger, multicenter studies. Future research should aim to identify predictors of delayed response to facilitate individualized treatment strategies while minimizing delays in escalation for true non-responders.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

1. ArS: Conceptualization, Methodology, Validation, Formal analysis, Resources, Writing - Original Draft, Writing - Review & Editing, Visualization2. AB: Methodology, Validation, Formal analysis, Resources, Writing - Original Draft, Writing - Review & Editing, Visualization3. RS: Validation, Formal analysis, Resources, Writing - Original Draft, Writing - Review & Editing, Visualization4. DS: Methodology, Data Curation, Writing - Review & Editing5. DJ: Methodology, Data Curation, Writing - Review & Editing6. NB: Formal analysis, Writing - Review & Editing7. SS: Visualization, Writing - Review & Editing8. LPB: Visualization, Writing - Review & Editing9. VM: Conceptualization, Methodology, Validation, Formal analysis, Resources, Writing - Original Draft, Writing - Review & Editing, Visualization, Supervision, Project Administration10. AS: Conceptualization, Methodology, Validation, Formal analysis, Resources, Writing - Original Draft, Writing - Review & Editing, Visualization, Supervision, Project Administration.

Data availability

The data that support the findings of this study are available from the corresponding author, [AS], upon reasonable request.

Declarations

Competing interests

Shaji Sebastian has received consulting fees from Celltrion, BMS Takeda, AbbVie, Merck, Ferring, Pharmacocosmos, Warner Chilcott, Janssen, Falk Pharma, Biohit, TriGenix, Celgene, Lilly and Tillots Pharma; has received payment or honoraria for lectures from AbbVie, Takeda, Celltrion, Pfizer, Biogen, AbbVie, Janssen, Merck, Warner Chilcott, Falk Pharma, Janssen and Lilly. Laurent Peyrin-Biroulet consults for Abbvie, Abivax, Adacyte, Alimentiv, Amgen, Applied Molecular Transport, Arena, Banook, Biogen, BMS, Celltrion, Connect Biopharm, Cytoki Pharma, Enthera, Ferring, Fresenius Kabi, Galapagos, Genentech, Gilead, Gossamer Bio, GSK, IAC Image Analysis, Index Pharmaceuticals, Inotrem, Janssen, Lilly, Medac, Mopac, Morphic, MSD, Nordic Pharma, Novartis, Oncodesign Precision Medicine, ONO Pharma, OSE Immunotherapeuthics, Pandion Therapeuthics, Par’ Immune, Pfizer, Prometheus, Protagonist, Roche, Samsung, Sandoz, Sanofi, Satisfay, Takeda, Telavant, Theravance, Thermo Fischer, Tigenix, Tillots, Viatris, Vectivbio, Ventyx, Ysopia and has received payment or honoraria for lectures from Abbvie, Amgen, Arena, Biogen, Celltrion, Ferring, Galapagos, Genentech, Gilead, Janssen, Lilly, Medac, MSD, Nordic Pharma, Pfizer, Sandoz, Takeda, Tillots, Viatris. Ajit Sood received honorarium for speaker events from Pfizer India and Takeda India. The remaining authors disclose no conflicts.

Ethics approval

The study was approved by institutional ethics committee (DMCH/R&D/2020/23).