Rate of Colectomy Increases within Twelve Months of Sequential Therapy for Ulcerative Colitis

Abstract

Introduction:

Patients with medically refractory ulcerative colitis (UC) who previously would have undergone surgery can now elect for subsequent medical therapy.

Methods:

In a commercially insured population, we evaluated the proportion of patients initiating second, third, or fourth-line treatment who underwent colectomy in the following 12 months.

Results:

Among 3,325 patients with UC, the colectomy rate within 12 months of a switch in therapy increased from 12% with the first switch to 17% and 19% with the second and third switch (p<0.001).

Conclusion:

Treatment effectiveness declines with successive switching; however, even after initiating fourth line therapy most patients remain surgery-free.

INTRODUCTION

One of the primary indications for surgery in patients with ulcerative colitis (UC) is medically-refractory disease. Historically, medically-refractory disease included patients who were refractory to corticosteroids, thiopurines, calcineurin inhibitors, and/or anti-tumor necrosis factor alpha (anti-TNF) therapies. However, the recent expansion of the UC treatment armamentarium¹ has created the need to re-define medically-refractory disease and determine whether colectomy should be offered after failure of one therapy/mechanism, multiple mechanisms, or all available mechanisms.

Subsequent therapy offers the potential to achieve clinical remission and avoid surgery. However, if unsuccessful, subsequent medical therapy will delay surgery, prolong suffering, and increase the risk of complications. Decisions about whether to pursue subsequent medical therapy or colectomy must be based on the absolute effectiveness and risks of these strategies. We sought to evaluate the real-world effectiveness of sequential medical therapy in patients with UC and determine whether there is a point at which subsequent treatment becomes futile.

METHODS

We performed a retrospective cohort study of patients with UC age 18-64 in the IBM^® MarketScan^® Research Databases (MarketScan) between the years 2007 and 2019.² Full details regarding the database, eligibility, and outcomes are provided as Supplemental Methods. We evaluated baseline UC medication and patient demographics including age, sex, United States census region, and year at the time of each switch. The primary outcome was the proportion of patients requiring colectomy in the 12 months following each switch in biologic or small molecule therapy. We defined colectomy using a previously described algorithm.³ Secondary outcomes included all-cause and UC-related hospitalizations, which were evaluated as a cumulative event rate among all patients regardless of colectomy status. To evaluate the proportion of patients undergoing colectomy after each switch in therapy we conducted bivariate analyses using the Cochran-Armitage test for trend and multivariate analyses using Generalized Estimating Equations.

RESULTS

We identified 3,325 undergoing a switch in biologic or small molecule therapy for UC. Demographic and treatment characteristics are provided in Table 1. The median follow up was 891 days (interquartile range 592-1381). Overall, 538 patients (16%) underwent colectomy during the study period. The percentage of patients undergoing colectomy in the 12 months following a switch in therapy was 12% after the first switch, 17% with the second switch, and 19% with the third switch, p<0.001, Figure 1, Supplemental Table 2. This increase in the likelihood of undergoing colectomy with each switch in therapy was significant after adjusting for age at switch date, sex, and switch year (p<0.001).

Table 1.

Clinical and demographic characteristics of patients undergoing at least one switch in biologic or small molecule therapy for ulcerative colitis

	First Switch n=3325		Second Switch n=695		Third Switch n=161
	median	IQR	median	IQR	median	IQR
Age at switch date (years)	41	31-52	42	32-53	45	36-56
	n	%	n	%	n	%
Female sex	1660	50	338	49	71	44
US Census Region
Northeast	662	20	138	20	36	22
Midwest	733	22	148	21	35	22
South	1431	43	308	44	69	43
West	486	15	97	14	20	12
Year of switch
2007-2014	1009	44	107	18	9	6
2015-2019	2316	56	588	82	152	94

Interquartile range (IQR)

Figure 1. Proportion of patients undergoing colectomy or hospitalization, evaluated at 1 year after switch in biologic or small molecule therapy for ulcerative colitis.

p-values generated using Generalized Estimating Equation model adjusting for age (Age 18-35, 36-55, >55), sex and index year (year 2007-2014 and 2015-2019)

When comparing all-cause hospitalizations, the proportion of patients requiring hospitalization in the 12 months after a switch in therapy increased from 23% with the first switch to 28% and 31% after the second and third switches respectively, p=0.002. Similarly, the proportion of patients requiring UC-related hospitalizations increased with each subsequent switch, from 11% within the 12 months following the first switch to 15% and 19% after the second and third switches respectively, p=0.001. The proportion of patients demonstrating corticosteroid-free remission decreased from 62% after a first switch to 56% and 50% after a second and third switch respectively, p=0.002.

The most frequent initial biologic switches were from an anti-tumor necrosis factor alpha (TNF) to an anti-TNF (1,895 patients [57%]) and anti-TNF to vedolizumab (980 patients [30%], Table 2). The proportion of patients undergoing colectomy did not differ by switch pattern.

Table 2.

Proportion of patients undergoing colectomy after specific drug switches, evaluated at 1 year after switch in each sequence

	Colectomy after First Switch		Colectomy after Second Switch		Colectomy after Third Switch
	n	%	n	%	n	%
anti-TNF to anti-TNF	236/1895	12	32/193	17	3/27	11
anti-TNF to vedolizumab	103/980	11	44/251	18	5/29	17
anti-TNF to tofacitinib	7/101	7	5/29	17	2/17	12
anti-TNF to ustekinumab	5/107	5	2/41	5	1/7	14
Vedolizumab to other	39/214	18	25/145	17	17/58	29
Tofacitinib to other	1/17	6	6/24	25	2/15	13
Ustekinumab to other	1/11	9	2/12	17	0/8	0

Anti-tumor necrosis factor alpha (Anti-TNF)

Note: tofacitinib approved for ulcerative colitis in 2018, ustekinumab approved for ulcerative colitis 2019

DISCUSSION

We demonstrated that the proportion of patients with UC undergoing colectomy within 12 months of switching biologic or small molecule therapy significantly increases with each subsequent switch, indicating a progressively more refractory disease state. Although the use of subsequent medical therapy is not futile, the proportion of patients hospitalized within 12 months of a switch increased to over 30% with a third switch, which may have significant implications for a patient’s quality of life, as well as one’s work and/or school and family responsibilities. Earlier conversations regarding the potential for colectomy may allow for avoidance of prolonged suffering and unnecessary hospitalizations among the highest risk individuals. We believe that the current data are important for anchoring discussions with patients considering therapy transitions and thus should help inform individualized decision making.

Given the continued emergence new UC therapies, finding the most appropriate treatment at the earliest opportunity may be paramount. The correct sequence of therapy remains an ongoing topic of research for many groups.^{1, 4-7} Of note, the majority of patients remained colectomy-free at 12 months after each switch. Thus, while providing important data for discussions regarding the risks of immediate colectomy after a switch in therapy, this study should also promote future investigation into those patients who have a successful transition in therapy. Several recently approved therapies appear to be highly effective therapies after anti-TNF failure.^{1, 8} Identifying individualized predictors for response to current and emerging therapeutics will be an important for future clinical and translational research, along with studies to further evaluate the most appropriate timing of switches.

Our study has limitations. First, our study identified patients who switched therapies as outpatients, and thus these results should not be generalized to hospitalized patients with acute severe UC. In a small number of patients, the switch may have occurred in the inpatient setting and our observed switch date may actually have been a continuation date. To the degree that continuation occurs following successful switches, we may have over-estimated colectomy-free survival following treatment switching. Furthermore, by requiring a full 12 months of enrollment following a treatment switch we may have slightly over-estimated the effectiveness of switching, as patients who died or lost insurance due to disability or other reasons would have been excluded. Our study population contained commercially-insured individuals and those ≤64 years of age. It is unknown whether these results should be generalized to other populations. While we would assume that disease severity would influence both the decision to switch therapy and the pursuit of colectomy as a treatment option, there is a lack of granular clinical, laboratory, and endoscopic data available in administrative claims to expand upon this hypothesis. Additionally, we were unable to assess the duration of UC prior to colectomy.

In conclusion, we demonstrated that the proportion of patients undergoing colectomy within the first 12 months after a switch in biologic or small molecule therapy significantly increases with each subsequent switch, yet does not reach a point of futility. These results may inform both physicians and patients in their decision making, and should serve to promote future research into appropriate sequencing of therapies to maximize effectiveness.