PubMedCentral-PMC4917292.pdf

(1)

Comparative Effectiveness and Safety of Infliximab and

Adalimumab in Patients with Ulcerative Colitis

Siddharth Singh, M.D., M.S.1,2,3, Herbert C. Heien, M.S.4, Lindsey R. Sangaralingham, M.P.H.4_{, Stephanie R. Schilz, B.A.}4_{, Michael D. Kappelman, M.D., M.P.H.}5_{, Nilay D. Shah,} Ph.D.4,6,7, and Edward V. Loftus Jr., M.D.1

1_{Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN}

2_{Division of Gastroenterology, University of California San Diego, La Jolla, CA}

3_{Division of Biomedical Informatics, University of California San Diego, La Jolla, CA}

4_{Robert D and Patricia E Kern Center for the Science of Health Care Delivery, Mayo Clinic,}

Rochester, MN

5_{Division of Pediatric Gastroenterology and Hepatology, University of North Carolina, Chapel Hill,}

NC

6_{Division of Health Care Policy and Research, Department of Health Services Research, Mayo}

Clinic, Rochester, MN

7_{Optum Labs, Cambridge, MA}

Abstract

Background—Real-world comparative benefits and risks of infliximab (IFX) and adalimumab (ADA) in patients with ulcerative colitis (UC) is unclear.

Corresponding author: Siddharth Singh, M.D., M.S., Assistant Professor of Medicine, Division of Gastroenterology, Clinical Informatics Fellow, Division of Biomedical Informatics, University of California San Diego, 9500 Gilman Drive, La Jolla, CA 92093, [email protected], Phone: 858-657-5271, Fax: 858-657-7259.

Disclosures: This study was made possible by support from (a) the American College of Gastroenterology Clinical Research Award

2014 to SS, and (b) the Center for the Science of Healthcare Delivery, Mayo Clinic, and a CTSA Grant UL1 TR000135 from the National Center for Advancing Translational Sciences (NCATS), a component of the National Institutes of Health (NIH). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Institutes of Health. Dr. Singh is supported by the National Library of Medicine training grant T15LM011271. Dr. Loftus has consulted for and has received research support from Janssen Biotech, AbbVie, and UCB Pharma. Dr. Kappelman has consulted for and has received research support from Janssen Biotech and AbbVie. None of the other authors have any relevant financial disclosures.

Author Contributions:

• Study concept and design: SS, NDS, EVL

• Acquisition of data: SS, HCH, LRS, SRS

• Analysis and interpretation of data: SS, HCH, LRS, SRS

• Drafting of the manuscript: SS

• Critical revision of the manuscript for important intellectual content: HCH, LRS, SRS, MDK, NDS, EVL

• Approval of the final manuscript: SS, HCH, LRS, SRS, MDK, NDS, EVL

• Study supervision: SS

HHS Public Access

Author manuscript

Aliment Pharmacol Ther

. Author manuscript; available in PMC 2017 May 01.

Published in final edited form as:

Aliment Pharmacol Ther. 2016 May ; 43(9): 994–1003. doi:10.1111/apt.13580.

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

(2)

Aims—To evaluate the comparative effectiveness and safety of IFX and ADA in patients with UC who were new users of anti-TNF agents.

Methods—Using an administrative claims database (Optum Labs Data Warehouse), we identified patients who received first anti-TNF (IFX, ADA) prescription after a 12-month period without any anti-TNF treatment (baseline), and with a minimum 6-month follow-up after anti-TNF initiation. Primary outcome measures were: all-cause and UC-related hospitalization, abdominal surgery, corticosteroid use >60 days after starting anti-TNF, and serious infections. We performed 2:1 propensity-score matched Cox proportional hazard analysis, and inverse probability-of-treatment weight (IPTW) analysis, accounting for healthcare utilization, comorbidities and use of UC-related medication.

Results—We included 1,400 new users of anti-TNF agents (age, 43±15 years; 52% males), from 2006-14. On propensity-score matched analysis, there was no significant difference in the risk of UC-related hospitalization (IFX vs. ADA; adjusted hazard ratio [aHR], 1.04; 95%

confidence interval [CI], 0.71-1.51), corticosteroid use (aHR, 0.85; 95% CI, 0.68-1.06) and serious infections (aHR, 0.62; 95% CI, 0.29-1.34) between IFX- and ADA-treated patients; the number of surgical events was very small. On IPTW analysis, risk of corticosteroid use was significantly lower in IFX- as compared to ADA-treated patients (aHR, 0.82; 95% CI, 0.68-0.99). Results were stable on multiple sensitivity analyses.

Conclusion—In a large retrospective cohort of patients with UC who were new users of anti-TNF agents, IFX-treated patients may have lower corticosteroid use than ADA-treated patients, but risk of hospitalization and serious infections was comparable.

Keywords

Biologics; outcomes; real-world effectiveness; administrative database; comparative; propensity matching

Introduction

Biologic therapy with anti-tumor necrosis factor-α (anti-TNF) agents such as infliximab (IFX) and adalimumab (ADA), alone or in combination with immunomodulators, is one of the most effective treatments in inducing and maintaining clinical remission in patients with ulcerative colitis (UC), and has been shown to decrease risk of hospitalization and surgery.1 In the absence of head-to-head trials, there is a unmet need among patients and clinicians to better understand the relative effectiveness and safety of different anti-TNF medications. Current decisions on the choice of anti-TNF agent are primarily driven by insurance coverage and patient and clinician preferences; however, there are differences in the

molecular construct, dosing and route of administration of these agents; hence, there may be differences in effectiveness.2

Therefore, we sought to study the real-world comparative effectiveness and safety of different anti-TNF agents in adult patients with UC who were new users of anti-TNF agents, using a propensity-score matched retrospective cohort study, in a nationally representative administrative database of privately insured individuals derived from the Optum Labs Data Warehouse.3 Using patient-important outcomes of all-cause and UC-related hospitalizations,

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

(3)

abdominal surgery, need for corticosteroids, and risk of serious infections, the results of this study might assist consumers, clinicians, purchasers, and policy makers to make informed decisions that will improve health care in IBD, both at the individual and the population level.

Methods

Data Sources

We conducted a retrospective analysis of medical and pharmacy administrative claims from a large database, Optum Labs Data Warehouse, which includes privately insured and Medicare Advantage enrollees throughout the United States.3 The database contains data on more than 100 million enrollees, from geographically diverse regions across the United States, with greatest representation from the South and Midwest. Medical claims include International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis codes; ICD-9 procedure codes; Current Procedural Terminology, Fourth Edition (CPT-4) procedure codes; Healthcare Common Procedure Coding System (HCPCS) procedure codes; site of service codes; and provider specialty codes. All study data were accessed using techniques compliant with the Health Insurance Portability and

Accountability Act of 1996, and because this study involved analysis of preexisting de-identified data, it was exempted from institutional review board approval.

Study Population

We identified all patients who filled a prescription for IFX or ADA or received an infusion for an anti-TNF agent in the clinic setting between January 1, 2006 and June 30, 2014. Our study cohort comprised of adult patients (≥18 years) with: (a) at least one ICD-9 diagnosis code of UC (ICD 556.x) in the baseline period (prior to index date of anti-TNF prescription), either from an inpatient or outpatient visit, (b) continuous health plan enrollment with pharmacy benefits, with no anti-TNF prescription in the 12 months prior to index date (to identify a group of anti-TNF-naïve patients), and at least a 6-month minimum enrollment in health plan after index date (patients who received anti-TNFs for <6m, and discontinued due to intolerance or primary non-response, but still remained in the health plan were included and considered treatment failures). These new users of anti-TNF agents were considered biologic-naïve since they had not received a prior prescription of an anti-TNF agent in the preceding 12m. We excluded patients with a concomitant diagnosis of rheumatoid arthritis, ankylosing spondylitis, psoriasis or psoriatic arthritis within the previous 12 months of anti-TNF prescription date, as competing causes for prescribing anti-anti-TNF agents. In case a patient received diagnostic codes for both UC and Crohn's disease, then the patient was classified as having UC if the majority of last nine diagnostic codes were for UC.4

Golimumab was not approved in the United States for commercial use in moderate to severe ulcerative colitis until May 2013; therefore, we did not have sufficient data on the use of this anti-TNF agent in our cohort. Figure 1 shows the flow of patients for identification of the cohort, and Supplementary Figure 1 outlines the general study scheme.

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

(4)

Exposures and Outcomes of Interest

The primary exposures of interest were prescription of IFX or ADA for UC. We considered patients as being continuously exposed from the index date (date of first prescription of anti-TNF agent) for the duration of their prescription. Patients were followed until occurrence of the outcome of interest (see below), disenrollment from healthcare plan, treatment

discontinuation or switching (absence of new prescription for a period of >4 months [for IFX] or >3 months [for ADA], either discontinuing all anti-TNF agent use or switching to another anti-TNF agent), or completion of the study (last date of follow-up, December 31, 2014).

The primary outcomes of interest were:

Effectiveness Outcomes

a. All-cause hospitalization

b. UC-related hospitalization, with UC either as the primary diagnosis, or as a secondary diagnosis if the primary diagnosis was related to a gastrointestinal symptom (abdominal pain, diarrhea, nausea, vomiting, constipation,

gastrointestinal bleeding)

c. Colectomy (identified using established procedural codes)5

d. Corticosteroid prescription, occurring at least 60 days after the start of anti-TNF therapy (to minimize confounding by disease severity)

e. Persistence on the index anti-TNF agent at 6-months (prescription of index anti-TNF agent between 140-220 days after index date)6 and 12-months (prescription of index anti-TNF agent between 325-405 days after index date)

Safety Outcomes

f Hospitalization for serious or opportunistic infections as primary diagnosis (Supplementary Appendix).7

If the index anti-TNF agent was started during an inpatient hospitalization, then that hospitalization was not counted as an outcome; only inpatient admissions occurring >23 hours after drug initiation were regarded as outcomes (to avoid misclassifying observation visits for IFX infusions as outcome).

Covariates of Interest

Independent variables of interest included measures of healthcare utilization, comorbidities including surrogate markers of disease severity, and overall medication burden, including use of IBD-related medications, in the baseline period (prior to initiation of the index anti-TNF agent). We assessed baseline healthcare utilization based on the number of all-cause hospitalizations, outpatient and emergency department visits, number of endoscopic procedures (upper endoscopy, colonoscopy, sigmoidoscopy, capsule endoscopy) identified via CPT codes, number of abdominal/pelvic radiologic procedures (magnetic resonance imaging, computed tomography, small bowel follow-through, barium enemas) and number

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

(5)

of different medication classes filled, both in the preceding 3 and 12 months. Comorbidity burden was measured using the validated Charlson–Deyo comorbidity index for

administrative data.8 While we did not have access to individual participant data medical records, endoscopy reports or biochemical parameters, we evaluated surrogate markers of disease severity, based on diagnostic codes for weight loss, malnutrition and anemia. Medication burden was assessed using overall generic and specific medication class count. We also evaluated UC-related medication use in the baseline 3- and 12-months including use of 5-aminosalicylates (mesalamine, sulfasalazine, balsalazide, olsalazine), oral

corticosteroids (prednisone, budesonide), immunomodulators (azathioprine, 6-mercaptopurine, methotrexate), other immunosuppressive medications (tacrolimus, cyclosporine, mycophenolate mofetil), and use of narcotic pain medications.

Patients were classified as being on anti-TNF-based combination immunomodulator therapy if they received immunomodulator prescriptions within 30 days before and/or after anti-TNF index start date. Based on corticosteroid use, patients were classified as: (a) remote users (if they received prescriptions for corticosteroids in the preceding 90 to 365 prior to index anti-TNF start date, but not within 90 days prior to start of anti-anti-TNF), and (b) recent users (if they received prescriptions for corticosteroids in the preceding 90 days prior to index anti-TNF start date).6

Statistical Analysis

We examined the relative effectiveness of IFX and ADA on the risk of all-cause and UC-related hospitalization, abdominal surgery, corticosteroid use and serious infections, using two statistical approaches. First, our primary analysis was performed using 2:1 propensity score matching without replacement to adjust for differences in baseline covariates, comparing patients exposed to IFX vs. ADA.9 The propensity score model included demographic variables (age categories, sex, census region), date of initiation of index anti-TNF, comorbidity index, healthcare utilization (as described above), surrogate markers of disease severity, medication class count, and IBD-related medications. We performed a paired t-test for continuous variables, a McNemar test for dichotomous variables, and a Bowker's test for categorical variables with more than two levels; then, we measured the standardized difference of each covariate in the propensity score model, and variables were considered to be different across treatment if after propensity score matching the

standardized difference was greater than 10%. In order to correct for any remaining imbalance after the propensity score analysis was performed, we included remaining covariates that were shown to be different across treatment groups into the final multivariate Cox proportional hazard models for assessment of the outcomes of interest. Second, we performed an inverse probability-of-treatment weight (IPTW) analysis where the IPTW was applied to each observation in the Cox model in order to assess the relative effectiveness of IFX and ADA. The IPTW analysis was derived by utilizing the propensity score on all observations before matching.10 In contrast to propensity-score matching in which the sample size decreases, this type of modeling allowed us to retain all identified patients in the analysis and increased power.

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

(6)

We calculated hazard ratios (HR) with 95% confidence intervals (CI) for each outcome of interest separately, and patients were censored at time of treatment discontinuation or switching (to another anti-TNF agent), health plan disenrollment or end of observation period. We created the analytic dataset in SAS 9.3 and used Stata SE software (version 13.0) for statistical analysis.

Sensitivity Analysis

We performed multiple sensitivity analyses to assess the robustness of our findings. First, to minimize the risk of being misclassified as anti-TNF-naïve, we analyzed only patients with at least 2- and 3-year baseline anti-TNF-free period in our cohort (i.e., no anti-TNF prescriptions for at least 2 or 3 years prior to start date of index anti-TNF agent). Second, because inpatient biologic use is not consistently captured by inpatient claims, we excluded patients who had an inpatient hospitalization within 30 days prior to the observed initiation of anti-TNF agents to minimize confounding by disease severity and misclassifying non-responders. Third, since ADA was approved by the US Food and Drug Administration (FDA) in 2012, we excluded patients with UC who's anti-TNF initiation date was before 2012, limiting only to patients started on anti-TNF agent between 2012-14. We also

performed stratified analysis based on use of anti-TNF monotherapy or combination therapy (concomitant use of immunomodulators).

Results

Characteristics of Patients

We identified 1400 patients with UC who started on their first anti-TNF agent between 2006 and 2014 and who had at least 18 months' continuous enrollment in the health plan (12 months prior to start of anti-TNF agent, and 6 months after initiating index anti-TNF agent). Of these, 1112 were treated with IFX, and 288 were treated with ADA as the index anti-TNF agent; on limiting only to patients with initiation date of index anti-TNF agent after 2012, 308 were treated with IFX and 171 were treated with ADA. Table 1 describes the baseline demographic, clinical, treatment characteristics and healthcare utilization of the overall cohort and the three subgroups. Median follow-up (and interquartile range [IQR]) after starting index anti-TNF agent was 19 months (IQR, 10-35), with significantly longer follow-up in IFX-treated patients as compared to ADA-treated patients (p<0.001). Only 9.4% of patients were >65 years old. Overall, the groups were comparable with regard to

demographic and clinical variables, as well as healthcare utilization in the baseline period, except significantly higher rates of hospitalization in the 12 month prior to starting anti-TNF in IFX-treated patients (40%) than ADA-treated patients (30%) (p<0.001)

The patients were also comparable with regard to baseline UC-related medication use. About 82% of patients received corticosteroids anytime in the 12-month baseline period, with a median of 3 prescriptions (IQR, 1-5); ADA-treated (86%) patients were more likely to have received corticosteroids compared to IFX-treated (81%) patients. However, there was no significant difference in the proportion of patients receiving corticosteroids in the 90 days prior to initiation of anti-TNF agent (total, 66%). Approximately 55% patients received a prescription of immunomodulator or immunosuppressive therapy in the 1 year prior to

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

(7)

index agent, with no difference across groups. The proportion of patients on anti-TNF-based combination therapy (immunomodulator prescriptions within 30 days before and/or after anti-TNF index start date) was comparable between groups (IFX vs. ADA: 33% vs. 32; p=0.71). Approximately 42% patients received narcotics in the preceding year, with comparable rates between IFX- and ADA-treated patients.

After propensity score matching, these groups were more balanced, with no significant difference in clinical variables, healthcare utilization or UC-related medication use (Supplementary Table 1). Only a small number of variables fell outside a standardized difference of 0.10 (Supplementary Figure 2); these variables were additionally adjusted for in the propensity score-matched multivariable Cox proportional hazard analysis.

Comparative effectiveness and safety of infliximab vs. adalimumab – Propensity Matched Analysis

A total of 544 patients treated with IFX were propensity score-matched to 272 ADA-treated patients. On unadjusted analysis, the risk of all-cause hospitalization and corticosteroid prescription was lower in IFX-treated patients compared to ADA-treated patients (Figure 2A-D). On Cox proportional hazard analysis, after additional adjustment for variables not balanced through propensity score matching, we observed that the risks of all-cause (aHR, 1.05; 95% CI, 0.78-1.43) and UC-related hospitalization (aHR, 1.04; 95% CI, 0.71-1.51), corticosteroid use > 60 days after initiation of anti-TNF therapy (aHR, 0.85; 95% CI, 0.68-1.06) and risk of serious infections (aHR, 0.62; 95% CI, 0.29-1.34) was not significantly different between IFX- and ADA-treated patients (Table 2). The number of surgical events was very small, precluding meaningful analysis. Persistence on index anti-TNF was comparable for IFX and ADA at 6-months (IFX vs. ADA, 79% vs. 74%, p=0.08), but significantly higher for IFX at 12-months (57% vs. 42%, p<0.001); after adjusting for variables that were unbalanced after propensity score matching, the probability of persistence on IFX and ADA at 6-months was 78% and 75%, respectively (p<0.001), and the corresponding rates at 12-months were 54% and 47%, respectively (p<0.001).

The overall results were similar on stratified analysis by anti-TNF monotherapy or concomitant immunomodulator therapy (Table 3). The results were also similar to the primary analysis on sensitivity analysis restricting only to (a) patients with at least a 3-year baseline anti-TNF-free period (Table 2), and (b) patients without inpatient hospitalization within 30 days prior to initiation of anti-TNF agents (Supplementary Table 2), and (c) patients with index date after 2012 (Table 4).

Comparative effectiveness and safety of infliximab vs. adalimumab – Inverse Probability-of-Treatment Weight Analysis

The overall results were largely comparable to the primary propensity-score matched analysis, when using IPTW analysis. There were no significant differences in the risks of all-cause and UC-related hospitalization or risk of serious infections between IFX- and ADA-treated patients (Table 2). However, the need for corticosteroids any time after initiation of index anti-TNF agent (aHR, 0.83; 95% CI, 0.69-1.00) and >60 days after anti-TNF initiation (aHR, 0.82; 95% CI, 0.68-0.99) was significantly lower in IFX-treated patients as compared

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

(8)

to ADA-treated patients. The adjusted probability of persistence on IFX and ADA at 6-months was 78% and 72%, respectively (p<0.001), and the corresponding rates at 12-6-months were 54% and 41%, respectively (p<0.001).

The overall IPTW results were similar on stratified analysis by anti-TNF monotherapy or concomitant immunomodulator therapy (Table 3). The results were also similar to the primary IPTW analysis on sensitivity analysis restricting only to (a) patients with at least a 3-year baseline anti-TNF-free period (Table 2), (b) patients without inpatient hospitalization within 30 days prior to initiation of anti-TNF agents (Supplementary Table 2), and (c) patients with index date after 2012 (Table 4).

Discussion

While anti-TNF-based therapy is one of the most effective treatments for UC, there are limited data on the comparative effectiveness and safety of different anti-TNF agents. In this nationally representative, propensity score-matched retrospective cohort study of 1,400 patients with UC who were new users of anti-TNF agents, we made several key

observations. First, we observed that there were no significant differences in the risk of all-cause or UC-related hospitalization after starting IFX or ADA as the first anti-TNF agent; the number of surgical events was small precluding meaningful analysis. Second, we observed that there was IFX-treated patients may have lower need for corticosteroids (after starting index anti-TNF agent) as compared to ADA-treated patients, using IPTW analysis, although this was not statistically significant in the propensity-matched analysis. Third, there was no significant difference in the risk of serious infections requiring hospitalization among IFX- and ADA-treated patients, although there was a consistent but non-significant trend for fewer serious infections among the IFX-treated patients. Based on the findings of our observational study, IFX and ADA appear comparable for most key patient-related effectiveness and safety outcomes for treatment of UC, although IFX-treated patients may be less likely to require corticosteroids and are likely to stay on their index anti-TNF longer. This is one of the largest observational comparative effectiveness studies in a contemporary cohort of UC patients, with patient-important effectiveness and safety outcomes, assessed with robust complementary statistical approaches, with several stratified and/or sensitivity analyses.

These results were consistent on sensitivity analysis after (a) excluding patients with inpatient hospitalization in the preceding 30 days (to minimize misclassification of non-responders and confounding by disease severity), (b) on restricting analysis to patients with at least a 3-year baseline free period (to minimize misclassification of anti-TNF-naivety), and (c) on restricting only patients started on anti-TNF agent after 2012 (after US-FDA approval of ADA). The results were also similar on analysis stratified by use of anti-TNF monotherapy and combination immunomodulator therapy, with comparable summary estimates.

Indirect treatment comparison network meta-analyses have suggested that in a subset of biologic-naïve patients with UC, IFX may be superior to ADA for induction of clinical response and mucosal healing,11, 12 but comparable for maintenance of remission.13

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

(9)

However, there are subtle differences in trial designs and co-interventions, with no head-to-head trials, which decreases the quality of evidence from these network meta-analyses. Moreover, all of these trials had restrictive inclusion criteria and were short-term (maximum maintenance therapy, 1 year), and hence, not representative of real-world patient-important outcomes.

There have been a limited number of observational studies on the comparative effectiveness of different anti-TNF agents in IBD. In a retrospective multi-site cohort study in which 170 physicians reported on 6-month outcomes of biologic-naïve IFX- (n=380) or ADA-treated (n=424) UC patients from their practices, there were no significant differences in clinical response (based on physician global assessment), healthcare utilization and partial Mayo score between the two groups, similar to our findings.14

In our study, there were subtle differences in the results depending on choice of the statistical approach, particularly for one of the outcomes (need for corticosteroids). Propensity-score matching resulted in loss of sample size, such that we included only 816 out of 1400 eligible patients. IPTW analysis, on the other hand, allowed us to overcome this limitation and include all 1400 patients in the analysis, increasing statistical power to detect small differences. In the IPTW analysis (but not the propensity score-matched analysis), we observed a lower risk of steroid requirement in IFX-treated patients compared to those treated with ADA, However, no significant differences were observed in the propensity-score matched and IPTW analysis for other outcomes.

Our findings must be interpreted with caution, given the limitations associated with our study design. First, this was an observational, not an interventional study, and hence, there may be unmeasured confounders across groups. The potential for unmeasured confounding by severity is of particular importance, as administrative data do not include objective measures of disease severity such as endoscopic or biochemical markers; patients with acute severe colitis were excluded. Anecdotally, in our clinical practice, patients with more severe disease may preferentially be recommended IFX over ADA. This may bias results against IFX, since sicker patients (at highest risk of patient-important outcomes) may be more likely to be given IFX. We also did not have data on baseline cigarette smoking status and body weight, both of which can influence outcomes; however, there is no clear reason why tobacco use would be different across exposure groups. Second, there is potential for misclassification of patients as new users of anti-TNF agents, since the baseline period required that patients not have received another anti-TNF agent only 1 year prior to index date. It is well known that the response to a second anti-TNF agent is generally inferior to that of the first anti-TNF agent, and likewise, response decreases for the 3rd anti-TNF agent as compared to the second.15 However, on sensitivity analysis increasing the baseline period to 3 years, we observed similar summary estimates for patient-important outcomes,

supporting the robustness of our findings. Additionally, on restricting to a subset of patients who received their index anti-TNF agent after 2012 (after FDA approval of ADA for UC), we observed similar results. We were unable to account for the impact of therapeutic drug monitoring or dose escalation, practices which have become more common recently, given inaccurate estimation using administrative databases. However, results were similar when restricted to 2012 onwards, and this would capture recent practices in UC management.

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

(10)

Third, both baseline covariates and outcomes were measured using administrative claims codes and may be subject to errors. Definitions for some covariates and outcomes, such as combination immunosuppressive therapy, treatment discontinuation, etc. were chosen using best investigator judgment, and have not been validated. The number of surgical events and serious infections was small, precluding meaningful analysis; the incidence of serious infections was low and hence, this analysis may have been underpowered. Fourth, by restricting inclusion only to patients with at least 6-month follow-up in the same healthcare plan after initiation of anti-TNF agents (regardless of whether they continued on medication or not), we may have potentially missed some patients with poor response to induction therapy and left the plan before 6 months. There may have been early mortality, a vital outcome that was missed; however, mortality is very uncommon in patients with UC. By choosing a 6-month cut-off, we were able to ensure a more homogenous cohort for accurate estimation of outcomes. Finally, the possibility of observing statistically significant results may be related to chance due to multiple statistical testing for related outcomes.

In conclusion, using a propensity-matched retrospective observational cohort study, we observed that IFX and ADA may be comparable for key patient-important outcomes such as risk of hospitalization and serious infections, in patients with UC who were new users of anti-TNF agents; the risk of surgery was low in both cohorts. The need for corticosteroids may be modestly lower, and persistence on index anti-TNF agent higher, in IFX-treated patients as compared to ADA-treated patients with UC. Future prospective cohort studies with adjustment for baseline objective measures of disease severity and pragmatic randomized controlled trials are warranted to confirm these observations.

Supplementary Material

Refer to Web version on PubMed Central for supplementary material.

References

1. Nielsen OH, Ainsworth MA. Tumor necrosis factor inhibitors for inflammatory bowel disease. N Engl J Med. 2013; 369:754–62. [PubMed: 23964937]

2. Nestorov I. Clinical pharmacokinetics of TNF antagonists: how do they differ? Semin Arthritis Rheum. 2005; 34:12–8. [PubMed: 15852249]

3. Wallace PJ, Shah ND, Dennen T, et al. Optum Labs: building a novel node in the learning health care system. Health Aff (Millwood). 2014; 33:1187–94. [PubMed: 25006145]

4. Benchimol EI, Guttmann A, Mack DR, et al. Validation of international algorithms to identify adults with inflammatory bowel disease in health administrative data from Ontario, Canada. J Clin Epidemiol. 2014; 67:887–96. [PubMed: 24774473]

5. Loftus EV Jr, Delgado DJ, Friedman HS, et al. Colectomy and the incidence of postsurgical complications among ulcerative colitis patients with private health insurance in the United States. Am J Gastroenterol. 2008; 103:1737–45. [PubMed: 18564126]

6. Osterman MT, Haynes K, Delzell E, et al. Comparative effectiveness of infliximab and adalimumab for Crohn's disease. Clin Gastroenterol Hepatol. 2014; 12:811–817 e3. [PubMed: 23811254] 7. Grijalva CG, Chen L, Delzell E, et al. Initiation of tumor necrosis factor-alpha antagonists and the

risk of hospitalization for infection in patients with autoimmune diseases. JAMA. 2011; 306:2331– 9. [PubMed: 22056398]

8. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992; 45:613–9. [PubMed: 1607900]

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

(11)

9. Borah BJ, Moriarty JP, Crown WH, et al. Applications of propensity score methods in observational comparative effectiveness and safety research: where have we come and where should we go? J Comp Eff Res. 2014; 3:63–78. [PubMed: 24266593]

10. Sugihara M. Survival analysis using inverse probability of treatment weighted methods based on the generalized propensity score. Pharm Stat. 2010; 9:21–34. [PubMed: 19199275]

11. Danese S, Fiorino G, Peyrin-Biroulet L, et al. Biological agents for moderately to severely active ulcerative colitis: a systematic review and network meta-analysis. Ann Intern Med. 2014; 160:704–11. [PubMed: 24842416]

12. Singh S, Garg SK, Wang Z, et al. Letter: comparative efficacy of biological therapy in patients with ulcerative colitis. Aliment Pharmacol Ther. 2014; 39:1432–3. [PubMed: 24849155]

13. Stidham RW, Lee TC, Higgins PD, et al. Systematic review with network meta-analysis: the efficacy of anti-tumour necrosis factor-alpha agents for the treatment of ulcerative colitis. Aliment Pharmacol Ther. 2014; 39:660–71. [PubMed: 24506179]

14. Sandborn WSA, Burak Ozbay A, Sorg RA, Reichmann R, Chao J, Skup M. Comparison of Real-World Outcomes of Adalimumab (ADA) and Infliximab (IFX) for Patients With Ulcerative Colitis (UC) in the United States. Gastroenterology. 2015; 148:S272.

15. Gisbert JP, Marin AC, McNicholl AG, et al. Systematic review with meta-analysis: the efficacy of a second anti-TNF in patients with inflammatory bowel disease whose previous anti-TNF treatment has failed. Aliment Pharmacol Ther. 2015; 41:613–23. [PubMed: 25652884]

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

(12)

Figure 1.

Flow of patients for identification of the ulcerative colitis cohort who were new users of anti-TNF agents.

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

(13)

Figure 2a

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

(14)

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

(15)

Figure 2b

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

(16)

Figure 2d

Figure 2.

Survival free of (a) all-cause hospitalization, (b) UC-related hospitalization, (c) corticosteroid prescription (at least 60 days after index anti-TNF agent) and (d) hospitalization for serious infection, in comparing propensity-score matched cohort of patients treated with infliximab vs. adalimumab as first-line index anti-TNF agent. IP refers to inpatient hospitalization.

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

(17)

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

Table 1

Baseline demographic characteristics, healthcare utilization and IBD-related medication use in the 12 months prior to initiation of anti-TNF agents, in the entire cohort (figures in bold represent statistically significant difference between groups)

Variable Infliximab

(n=1112)

Adalimumab (n=288)

p-value

Demographic variables

Mean age ± SD, years 43 ± 16 42 ± 14 0.14

Sex (% males) 52 52 1.00

Median follow-up after starting anti-TNF, months (IQR) 21.8 (11.0-38.1)

13.6 (9.1-22.5)

<0.001

Healthcare utilization (in 12 months prior to starting anti-TNF)

Median outpatient visits (IQR) 9 (5-13) 9 (6-14) 0.22

Emergency room visits (% pts with ≥1) 41 41 0.81

Inpatient visits (% pts with ≥1) 40 30 <0.001

Imaging (% of pts with ≥1) 24 26 0.49

Endoscopic procedures (% pts with ≥1) 76 76 0.91

Median generic medication count (IQR) 6 (4-9) 6 (4-9) 0.91

IBD-related medication use (in 12 months prior to starting anti-TNF)

Mesalamine, oral, % 77 81 0.16

Steroids

• Any prior use, % 81 86 0.04

• Recent (in 90 days prior to anti-TNF), % 66 68 0.35

Immunomodulators

• Any prior use, % 51 46 0.14

• Concurrent (index date ± 30 days), % 33 32 0.71

Narcotics, % 41 45 0.22

(18)

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

T ab le 2 Comparati v e ef fecti v

eness and safety of infliximab vs. adalimumab in patients with ulcerati

v

e colitis who were ne

w users of anti-TNF agents, with 1-year

baseline anti-TNF-free period (primary analysis) and 2- and 3-year baseline anti-TNF-free period (sensiti

vity analysis), using 2:1 IFX:AD

A

propensity-score matched analysis and in

v

erse probability-of-treatment weighted analysis. The number of sur

gical e

v

ents w

as v

ery small limiting statistical analysis.

Figures in bold represent statistically signif

icant results.

Outcomes of inter

est

Pr

opensity-scor

e Matched Analysis

1-y

ear baseline (N=816)

2-y

3-y

HR (95% CI), IFX vs. AD

A

p-v

alue

A

p-v

alue

A

p-v

alue

All-cause hospitalization

1.05 (0.78, 1.43)

0.737

1.03 (0.72, 1.48)

0.883

0.99 (0.64, 1.51)

0.946

UC-related hospitalization

1.04 (0.71, 1.51)

0.854

1.00 (0.63, 1.57)

0.992

0.98 (0.57, 1.70)

0.946

Steroid use (>60d after anti-TNF initiation)

0.85 (0.68, 1.06)

0.157

0.98 (0.76, 1.27)

0.898

0.78 (0.59, 1.05)

0.097

Serious Infection

0.62 (0.29, 1.34)

0.223

0.88 (0.33, 2.37)

0.807

0.74 (0.23, 2.36)

0.607 In v erse Pr obability-of-T reatment W eighted Analysis 1-y

2-y

3-y

A

p-v

alue

A

p-v

alue

A

p-v

alue

All-cause hospitalization

1.00 (0.76, 1.32)

0.979

0.83 (0.60, 1.15)

0.258

1.08 (0.89, 1.31)

0.465

UC-related hospitalization

1.06 (0.76, 1.50)

0.723

0.92 (0.61, 1.39)

0.695

0.92 (0.71, 1.19)

0.537

Steroid use (>60d after anti-TNF initiation)

0.82 (0.68, 0.99)

0.043

0.85 (0.69, 1.05)

0.143

0.82 (0.71, 0.95)

0.010

Serious Infection

0.69 (0.35, 1.35)

0.282

0.91 (0.38, 2.17)

0.825

1.06 (0.49, 2.29)

0.883

[Abbre

viations: CI-conf

idence interv

al, HR-hazard ratio, N=number of patients, UC-ulcerati

v

(19)

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

Table 3

Stratified analysis, by baseline use of anti-TNF monotherapy and combination immunomodulator therapy.

Outcomes of interest

Propensity Score Matched Analysis Inverse Probability-of-Treatment Weighted Analysis

HR (95% CI), IFX vs. ADA p-value HR (95% CI), IFX vs. ADA p-value

Anti-TNF monotherapy (within ±30 days of index date)

All-cause hospitalization 1.09 (0.77, 1.55) 0.631 0.98 (0.72, 1.34) 0.918

UC-related hospitalization 1.08 (0.70, 1.66) 0.733 1.09 (0.73, 1.63) 0.662

Steroid use (>60 days after anti-TNF

initiation) 0.88 (0.68, 1.13) 0.312 0.84 (0.66, 1.07) 0.165

Serious infections 0.60 (0.25, 1.41) 0.229 0.72 (0.34, 1.51) 0.382

Anti-TNF combination therapy (within ±30 days of index date)

Steroid use (>60d after anti-TNF initiation) 0.78 (0.49, 1.23) 0.283 0.72 (0.49, 1.04) 0.083

(20)

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

A

uthor Man

uscr

ipt

Table 4

Sensitivity analysis, on restricting only to patients with initiation of index anti-TNF agent after 2012 (approval year of ADA for UC)

Comparative effectiveness and safety of infliximab vs. adalimumab in patients with ulcerative colitis who were new users of anti-TNF agents (with 1-year baseline anti-TNF-free period), using 2:1 IFX:ADA propensity-score matched analysis (138 IFX-treated patients vs. 69 ADA-treated patients) and inverse probability-of-treatment weighted analysis (308 IFX-treated and 171 ADA-treated patients). The number of surgical events was very small limiting statistical analysis.

Outcomes of interest

Propensity Score Matched Analysis Inverse Probability-of-Treatment Weighted Analysis

HR (95% CI) (IFX vs. ADA) p-value HR (95% CI) (IFX vs. ADA) p-value

Steroid use (>60 days after anti-TNF

initiation) 0.81 (0.54, 1.20) 0.286 0.85 (0.62, 1.15) 0.290