Primary Non-Response to Tumor Necrosis Factor Antagonists is Associated with Inferior Response to Second-line Biologics in Patients with Inflammatory Bowel Diseases: A Systematic Review and Meta-analysis

Primary Non-Response to Tumor Necrosis Factor Antagonists is Associated with Inferior Response to... Abstract Background and Aims We sought to analyze whether response to a second-line biologic varies depending on the reason for discontinuation of the primary anti-TNF agent (primary non-response [PNR], secondary loss of response [LOR] after initial response, or intolerance), through a systematic review and meta-analysis. Methods Through a systematic search through May 31, 2017, we identified eight randomized controlled trials [RCTs] of biologics in patients with IBD with prior exposure to anti-TNF agents, that stratified response to second-line therapy by reason for discontinuing primary anti-TNF therapy [PNR vs. LOR vs. intolerance]. We estimated relative risk [RR] (and 95% confidence interval [CI]) of achieving clinical remission in patients with PNR as compared with patients with LOR, and intolerance, through random effects meta-analysis. Results As compared with patients who discontinued prior anti-TNF due to intolerance, patients with prior PNR were 24% less likely to achieve remission with second-line biologics (RR,0.76 [0.61–0.96]). As compared with patients who discontinued prior anti-TNF due to LOR, patients with prior PNR were 27% less likely to achieve remission with induction therapy with second-line biologics (RR,0.73 [0.56–0.97]), particularly to ustekinumab (RR,0.64 [0.52–0.80]). There was no difference in response to vedolizumab in patients with prior PNR or LOR to anti-TNF agents (RR,1.16 [0.85–1.58]). Conclusion Patients with PNR to anti-TNF agents are less likely to respond to second-line non-TNF biologics, as compared with patients who discontinued therapy due to secondary LOR or intolerance. This may be attributed to underlying pharmacokinetics and pharmacodynamics of anti-TNF agents in patients with PNR. Prediction, anti-integrin, anti-interleukin, second-line, inflammatory bowel diseases 1. Introduction Anti-tumor necrosis factor-α [TNF] agents have been the mainstay for managing patients with moderate to severe inflammatory bowel diseases [IBDs], inducing and maintaining remission, and decreasing the risk of surgery, hospitalization and disease-related complications.1–3 However, ~30–40% patients may not respond to anti-TNF agents (primary non-response [PNR]); additionally, ~30–40% patients may lose response over time (secondary loss of response [LOR]) or may be intolerant to anti-TNFs.4 Previous studies have suggested that response to a second anti-TNF agent after discontinuation of the index anti-TNF agent varies depending on the reason for discontinuation. In a systematic review of 46 studies, Gisbert and colleagues observed that clinical remission rates with the second anti-TNF were highest in patients who discontinued the primary anti-TNF due to intolerance [61%], followed by those with LOR [45%].5 Patients with PNR to the index anti-TNF are least likely to respond to a second anti-TNF agent. The treatment landscape of moderate–severe IBD is expanding, with several non-TNF biologics recently approved, and other biologics and small molecules in advanced stages of development.6–9 Among approved agents, both vedolizumab and ustekinumab are more effective in anti-TNF–naïve patients, than in anti-TNF–exposed patients.7,10,11 This may suggest that the latter group of patients are intrinsically more treatment-resistant, or may have been “primed” by prior anti-TNF exposure for inadequate response to a second agent. Moreover, all patients with prior anti-TNF exposure are not equivalent, and may have intrinsic differences in response to second-line biologics. For example, patients with PNR may have altered pharmacokinetics [rapid drug clearance resulting in low trough levels] or pharmacodynamics [mechanistic failure with non-TNF–mediated inflammation], resulting in decreased likelihood of response to a second biologic agent;4,12,13 in contrast, patients with secondary LOR after initial response potentially due to immunogenicity may be more treatment responsive. Through a systematic review and meta-analysis of published clinical trials, we analyzed whether response to a second-line biologic or small molecule differs according to the reason for discontinuation of the primary anti-TNF agent [PNR vs. LOR vs. intolerance]. 2. Methods This systematic review followed the preferred reporting items for systematic reviews and meta-analysis [PRISMA] standards, and followed an a priori protocol.14 2.1. Selection criteria Studies included in this meta-analysis were Phase II or III RCTs that met the following inclusion criteria: [1] Patients: adults [age >18 years] with moderate to severe ulcerative colitis [UC] (Mayo Clinic Score [MCS] 6–12, with an endoscopic subscore of 2 or 3] or Crohn’s disease [CD] (Crohn’s Disease Activity Index [CDAI] >220 but <450), who had previously been exposed to anti-TNF agents; [2] Intervention: biologic therapy [anti-TNF agents, anti-integrin agents, anti-interleukin-12 and/or -23], or small molecules [janus kinase inhibitors, sphingosine-1 phosphate receptor agonist or SMAD7 antisense oligonucleotide], with a minimum duration of therapy of 14 days; [3] Comparator: another biologic agent or placebo; [4] Outcome: achievement of clinical remission or response, stratified by reason for discontinuation [PNR vs. LOR vs. intolerance] of index anti-TNF agent. We excluded the following studies: [1] trials conducted exclusively in biologic-naïve patients, [2] trials where results were not stratified by reason for discontinuation of prior anti-TNF, [3] Phase I trials, [4] pediatric studies, or [5] trials conducted in patients with acute severe colitis. 2.2. Search strategy We conducted a comprehensive search of multiple electronic databases through May 31, 2017, about adults with no language restrictions. The databases included Ovid MEDLINE In-Process & Other Non-Indexed Citations, Ovid MEDLINE, Ovid EMBASE, Ovid Cochrane Central Register of Controlled Trials, Ovid Cochrane Database of Systematic Reviews, Web of Science, and Scopus. The search terms used included a combination of phrases indicating the diseases of interest “Crohn[s] disease”, “Ulcerative colitis”, “inflammatory bowel disease”, “regional enteritis” and treatments including biologics [“infliximab”, “adalimumab”, “certolizumab pegol”, “golimumab”, “anti-TNF”, “TNF-antagonist”, “vedolizumab”, “natalizumab”, “etrolizumab”, “monoclonal antibod*”, “anti-integrin”, “anti-interleukin”, “ustekinumab”, “risankizumab”] and small molecules [“tofacitinib”, “janus kinase”, “ozanimod”, “trafficking”, “mongersen”, “SMAD7”]. Two study investigators [SS and JG] independently reviewed the title and abstract of studies identified in the search to exclude studies that did not address the research question of interest on the basis of pre-specified inclusion and exclusion criteria. The full text of the remaining articles was examined to determine whether it contained relevant information. Conflicts in study selection at this stage were resolved by consensus, referring back to the original article, in consultation with a senior investigator [WJS]. Second, we searched the bibliographies of these selected articles, systematic reviews and clinical trial registries [www.clinicaltrials.gov] to identify any additional studies. Third, we conducted a manual search of abstracts from major gastroenterology conferences [Digestive Disease Week, American College of Gastroenterology annual meeting, Advances in Inflammatory Bowel Diseases meeting organized by the Crohn’s and Colitis Foundation of America, European Crohn’s and Colitis Organization annual meeting and United European Gastroenterology Week] from 2012 to 2017 to identify additional abstracts on the topic. Finally, we contacted experts in the field to identify other unpublished studies. 2.3. Data abstraction and quality assessment Data on study-, participant-, disease- and treatment-related characteristics were abstracted onto a standardized form, by two authors [SS and JG] independently and discrepancies were resolved by consensus, referring to the original article, in consultation with a third reviewer. We focused only on outcomes in patients receiving active intervention. We abstracted data on the definitions of PNR, LOR and intolerance in included trials, definition of clinical remission or response, and rates of clinical remission [or response] in patients receiving active intervention across these strata. Two study investigators [SS and JG] independently rated the quality of included studies by using the Cochrane Risk of Bias Tool.15 2.4. Outcomes assessed The primary outcome measure was the proportion of patients achieving clinical remission in patients in different strata based on reason for discontinuation of index anti-TNF agent. Clinical remission was defined as Mayo Clinic Score [MCS] ≤2 with no individual subscore of >1 [for patients with UC], and Crohn’s disease activity index [CDAI] <150 [for patients with CD]; clinical response was defined as decline in CDAI by 100 points [CR-100], and was used if clinical remission was not reported. We a priori hypothesized that potential impact of prior anti-TNF exposure on response to second-line agent would be most apparent in trials of induction therapy, and may be less apparent in trials of maintenance therapy that re-randomized only responders to induction therapy. Hence, we performed subgroup analyses based on trial design [induction vs. maintenance therapy]. Other subgroups included: disease type [CD vs. UC] and class of second-line agent. 2.5. Statistical analysis We used the random-effects model described by DerSimonian and Laird to calculate relative risk [RR] (and 95% confidence interval [CI]) of achieving clinical remission from active intervention with second-line agent in patients who discontinued prior anti-TNF due to PNR [vs. intolerance], PNR [vs. LOR] and LOR [vs. intolerance].16 We assessed heterogeneity between study-specific estimates using the inconsistency index [I2], and used cut-offs of <30%, 30–59%, 60–75% and >75% to suggest low, moderate, substantial and considerable heterogeneity, respectively.17 Due to the small number of studies, a reliable assessment of publication bias could not be estimated. All analysis was performed using Comprehensive Meta-Analysis [CMA] version 2 [Biostat, Englewood, NJ]. 3. Results From 1866 unique studies identified using our search strategy, we included eight RCTs in patients with moderate–severe IBD with prior exposure to an anti-TNF, in which response to the second-line agent was stratified according to the reason for discontinuation of the index anti-TNF agent. These included six RCTs in patients with moderate–severe CD [GAIN,18 WELCOME,19 GEMINI-II Induction and Maintenance,20 GEMINI-III,10 CERTIFI,21 UNITI-17] and two RCTs in patients with moderate–severe UC [GEMINI I-Induction and Maintenance22,23] [Figure 1]. No trial of small molecules reported differences in response stratified according to the reason for discontinuation of the prior anti-TNF agent. Figure 1. View largeDownload slide Study selection flowsheet. Figure 1. View largeDownload slide Study selection flowsheet. The trial characteristics are summarized in Table 1. Across trials, the reason for discontinuation of the index anti-TNF agent was defined clinically, and not based on therapeutic drug monitoring. PNR was defined as lack of initial clinical response to the index anti-TNF agent, and was reported in a median of 43.1% [interquartile range (IQR), 28.9–49.6] of patients with prior anti-TNF exposure. Secondary LOR was defined based on loss of response in patients with initial response to the anti-TNF agent, and a median of 64.7% [IQR, 41.0–64.7] were classified as having LOR; of note, a higher proportion of patients in ustekinumab trials were deemed to have secondary LOR as compared with participants in trials of vedolizumab. Intolerance was defined based on unacceptable side effects due to the index anti-TNF agent, warranting discontinuation of therapy, and was reported in a median of 35.4% [IQR, 26.5–42.2] of participants [Table 2]. Across trials, some patients were classified in more than one category of reason for discontinuation of prior therapy, since this was often based on patient recall. Of note, trials of adalimumab and certolizumab pegol included only patients with LOR and intolerance, and excluded patients with PNR.18,19 Overall, these RCTs were at low risk of bias. Table 1. Trial characteristics focusing on patients with prior exposure to anti-TNF therapy. [Abbreviations: C = Control; CS = corticosteroids; CDAI-Crohn’s disease activity index; I = intervention; IS = immunomodulator; MCS = Mayo Clinic Score.] Trial name; active intervention, condition  Proportion of patients with prior exposure to anti- TNF agents [%]  Intervention vs. Control  Outcome; timing of assessment  Co-interventions [proportion on immunomodulators and corticosteroids]  Response to prior therapy  Overall proportion of patients achieving remission  Primary non-response  Secondary loss of response  Intolerance  GEMINI 1 INDUCTION;22,23Vedolizumab, ulcerative colitis  367/895 [41.0%]  I: 82/367 [22.3%] C: 63/367 [17.2%]  MCS<2 Week 6  I: CS 30/82 [36.6%], IS 5/82 [6.1%] C: CS 27/63 [42.9%], IS 6/63 [9.5%]  I: 44/82 [53.7%] C: 29/63 [46.0%]  I: 32/82 [39.0%] C: 26/63 [41.3%]  I: 13/82 [15.6%] C: 10/63 [15.9%]  I: 8/82 [9.8%] C: 2/63 [3.2%]  GEMINI 1 MAINTENANCE;22,23Vedolizumab, ulcerative colitis  121/345 [35.1%]  I: 83/121 [68.6%] C: 38/121 [31.4%]  MCS<2 Week 52  I: CS 32/83 [38.6%], IS 7/83 [8.4%] C: CS 16/38 [42.1%], IS 6/38 [15.8%]  I: 33/83 [39.8%] C: 19/38 [50.0%]  I: 34/83 [41.0%] C: 13/38 [34.2%]  I: 22/83 [26.5%] C: 10/38 [26.3%]  I: 30/83 [36.1%] C: 2/38 [5.3%]  GEMINI 2 & 3 INDUCTION;10,11,20 Vedolizumab, Crohn’s disease  960/1476 [65.0%]  I: 263/960 [27.4%] C: 227/960 [23.65]  CDAI < 150 Week 10  I: CS 94/263 [35.7%], IS 35/263 [13.3%] C: CS 85/227 [37.4%], IS 30/227 [13.2%]  I: 109/263 [41.4%] C: 110/227 [48.5%]  I: 143/263 [54.4%] C: 141/227 [62.1%]  I: 94/263 [35.7%] C: 95/227 [41.9%]  I: 52/263 [19.8%] C: 25/227 [11.0%]  GEMINI 2 MAINTENANCE;11,20 Vedolizumab; Crohn’s disease  237/445 [53.3%]  I: 159/237 [67.1%] C: 78/237 [32.9%]  CDAI < 150 Week 52  I: CS 65/159 [40.9%], IS 27/159 [17.0%] C: CS 29/78 [37.2%], IS 10/78 [12.8%]  I: 68/159 [42.8%] C: 35/78 [44.9%]  I: 94/159 [59.2%] C: 46/78 [59.0%]  I: 64/159 [40.3%] C: 30/78 [38.5%]  I: 44/159 [27.7%] C: 10/78 [12.8%]  CERTIFI;21 Ustekinumab, Crohn’s disease  526/526 [100.0%]  I: 394/526 [74.9%] C: 132/526 [25.1%]  CR-100 Week 6  I: CS 189/394 [48.0%], IS 96/394 [24.4%] C: CS 73/132 [55.3%], IS 30/132 [22.7%]  I: 116/394 [29.4%] C: 44/132 [33.3%]  I: 289/394 [73.4%] C: 91/132 [68.9%]  I: 135/394 [41.9%] C: 41/132 [31.1%]  I: 58/394 [14.7%] C: 14/132 [10.6%]  UNITI 1;7 Ustekinumab, Crohn’s disease  492/496 [99.2%]  I: 246/249 [98.85] C: 246/247 [99.6%]  CR-100 Week 6  I: CS 108/249 [43.4%], IS 78/249 [31.3%] C: CS 111/247 [44.9%], IS 81/247 [32.8%]  I: 72/249 [28.9%] C: 74/247  I: 171/249 [68.7%] C: 170/247 [35.2%]  I: 105/249 [42.2%] C: 87/247 [35.2%]  I: 46/249 [18.5%] C: 22/247 [8.9%]  GAIN;18 Adalimumab, Crohn’s disease  325/325 [100.0%]  I: 159/325 [48.9%] C: 166/325 [51.1%]  CDAI < 150 Week 4  I: CS 55/159 [34.6%], IS 73/159 [45.9%] C: CS 73/166 [44.0%], IS 85/166 [51.2%]  0 [0.0%]  I: 77/159 [48.4%] C: 87/166 [52.4%]  I: 95/159 [59.7%] C: 95/166 [57.2%]  I: 34/159 [21.4%] C: 12/166 [7.2%]  WELCOME;19 Certolizumab pegol, Crohn’s disease  539/539 [100.0%]  I: 539/539 [100.0%] C: 0 [0.0%]  CR-100 Week 6  I: CS 67/168 [39.9%], IS 77/168 [45.8%]  0 [0.0%]  I: 79/168 [47.0%]  I: 66/168 [39.3%]  I: 212/539 [39.3%]  Trial name; active intervention, condition  Proportion of patients with prior exposure to anti- TNF agents [%]  Intervention vs. Control  Outcome; timing of assessment  Co-interventions [proportion on immunomodulators and corticosteroids]  Response to prior therapy  Overall proportion of patients achieving remission  Primary non-response  Secondary loss of response  Intolerance  GEMINI 1 INDUCTION;22,23Vedolizumab, ulcerative colitis  367/895 [41.0%]  I: 82/367 [22.3%] C: 63/367 [17.2%]  MCS<2 Week 6  I: CS 30/82 [36.6%], IS 5/82 [6.1%] C: CS 27/63 [42.9%], IS 6/63 [9.5%]  I: 44/82 [53.7%] C: 29/63 [46.0%]  I: 32/82 [39.0%] C: 26/63 [41.3%]  I: 13/82 [15.6%] C: 10/63 [15.9%]  I: 8/82 [9.8%] C: 2/63 [3.2%]  GEMINI 1 MAINTENANCE;22,23Vedolizumab, ulcerative colitis  121/345 [35.1%]  I: 83/121 [68.6%] C: 38/121 [31.4%]  MCS<2 Week 52  I: CS 32/83 [38.6%], IS 7/83 [8.4%] C: CS 16/38 [42.1%], IS 6/38 [15.8%]  I: 33/83 [39.8%] C: 19/38 [50.0%]  I: 34/83 [41.0%] C: 13/38 [34.2%]  I: 22/83 [26.5%] C: 10/38 [26.3%]  I: 30/83 [36.1%] C: 2/38 [5.3%]  GEMINI 2 & 3 INDUCTION;10,11,20 Vedolizumab, Crohn’s disease  960/1476 [65.0%]  I: 263/960 [27.4%] C: 227/960 [23.65]  CDAI < 150 Week 10  I: CS 94/263 [35.7%], IS 35/263 [13.3%] C: CS 85/227 [37.4%], IS 30/227 [13.2%]  I: 109/263 [41.4%] C: 110/227 [48.5%]  I: 143/263 [54.4%] C: 141/227 [62.1%]  I: 94/263 [35.7%] C: 95/227 [41.9%]  I: 52/263 [19.8%] C: 25/227 [11.0%]  GEMINI 2 MAINTENANCE;11,20 Vedolizumab; Crohn’s disease  237/445 [53.3%]  I: 159/237 [67.1%] C: 78/237 [32.9%]  CDAI < 150 Week 52  I: CS 65/159 [40.9%], IS 27/159 [17.0%] C: CS 29/78 [37.2%], IS 10/78 [12.8%]  I: 68/159 [42.8%] C: 35/78 [44.9%]  I: 94/159 [59.2%] C: 46/78 [59.0%]  I: 64/159 [40.3%] C: 30/78 [38.5%]  I: 44/159 [27.7%] C: 10/78 [12.8%]  CERTIFI;21 Ustekinumab, Crohn’s disease  526/526 [100.0%]  I: 394/526 [74.9%] C: 132/526 [25.1%]  CR-100 Week 6  I: CS 189/394 [48.0%], IS 96/394 [24.4%] C: CS 73/132 [55.3%], IS 30/132 [22.7%]  I: 116/394 [29.4%] C: 44/132 [33.3%]  I: 289/394 [73.4%] C: 91/132 [68.9%]  I: 135/394 [41.9%] C: 41/132 [31.1%]  I: 58/394 [14.7%] C: 14/132 [10.6%]  UNITI 1;7 Ustekinumab, Crohn’s disease  492/496 [99.2%]  I: 246/249 [98.85] C: 246/247 [99.6%]  CR-100 Week 6  I: CS 108/249 [43.4%], IS 78/249 [31.3%] C: CS 111/247 [44.9%], IS 81/247 [32.8%]  I: 72/249 [28.9%] C: 74/247  I: 171/249 [68.7%] C: 170/247 [35.2%]  I: 105/249 [42.2%] C: 87/247 [35.2%]  I: 46/249 [18.5%] C: 22/247 [8.9%]  GAIN;18 Adalimumab, Crohn’s disease  325/325 [100.0%]  I: 159/325 [48.9%] C: 166/325 [51.1%]  CDAI < 150 Week 4  I: CS 55/159 [34.6%], IS 73/159 [45.9%] C: CS 73/166 [44.0%], IS 85/166 [51.2%]  0 [0.0%]  I: 77/159 [48.4%] C: 87/166 [52.4%]  I: 95/159 [59.7%] C: 95/166 [57.2%]  I: 34/159 [21.4%] C: 12/166 [7.2%]  WELCOME;19 Certolizumab pegol, Crohn’s disease  539/539 [100.0%]  I: 539/539 [100.0%] C: 0 [0.0%]  CR-100 Week 6  I: CS 67/168 [39.9%], IS 77/168 [45.8%]  0 [0.0%]  I: 79/168 [47.0%]  I: 66/168 [39.3%]  I: 212/539 [39.3%]  View Large Table 1. Trial characteristics focusing on patients with prior exposure to anti-TNF therapy. [Abbreviations: C = Control; CS = corticosteroids; CDAI-Crohn’s disease activity index; I = intervention; IS = immunomodulator; MCS = Mayo Clinic Score.] Trial name; active intervention, condition  Proportion of patients with prior exposure to anti- TNF agents [%]  Intervention vs. Control  Outcome; timing of assessment  Co-interventions [proportion on immunomodulators and corticosteroids]  Response to prior therapy  Overall proportion of patients achieving remission  Primary non-response  Secondary loss of response  Intolerance  GEMINI 1 INDUCTION;22,23Vedolizumab, ulcerative colitis  367/895 [41.0%]  I: 82/367 [22.3%] C: 63/367 [17.2%]  MCS<2 Week 6  I: CS 30/82 [36.6%], IS 5/82 [6.1%] C: CS 27/63 [42.9%], IS 6/63 [9.5%]  I: 44/82 [53.7%] C: 29/63 [46.0%]  I: 32/82 [39.0%] C: 26/63 [41.3%]  I: 13/82 [15.6%] C: 10/63 [15.9%]  I: 8/82 [9.8%] C: 2/63 [3.2%]  GEMINI 1 MAINTENANCE;22,23Vedolizumab, ulcerative colitis  121/345 [35.1%]  I: 83/121 [68.6%] C: 38/121 [31.4%]  MCS<2 Week 52  I: CS 32/83 [38.6%], IS 7/83 [8.4%] C: CS 16/38 [42.1%], IS 6/38 [15.8%]  I: 33/83 [39.8%] C: 19/38 [50.0%]  I: 34/83 [41.0%] C: 13/38 [34.2%]  I: 22/83 [26.5%] C: 10/38 [26.3%]  I: 30/83 [36.1%] C: 2/38 [5.3%]  GEMINI 2 & 3 INDUCTION;10,11,20 Vedolizumab, Crohn’s disease  960/1476 [65.0%]  I: 263/960 [27.4%] C: 227/960 [23.65]  CDAI < 150 Week 10  I: CS 94/263 [35.7%], IS 35/263 [13.3%] C: CS 85/227 [37.4%], IS 30/227 [13.2%]  I: 109/263 [41.4%] C: 110/227 [48.5%]  I: 143/263 [54.4%] C: 141/227 [62.1%]  I: 94/263 [35.7%] C: 95/227 [41.9%]  I: 52/263 [19.8%] C: 25/227 [11.0%]  GEMINI 2 MAINTENANCE;11,20 Vedolizumab; Crohn’s disease  237/445 [53.3%]  I: 159/237 [67.1%] C: 78/237 [32.9%]  CDAI < 150 Week 52  I: CS 65/159 [40.9%], IS 27/159 [17.0%] C: CS 29/78 [37.2%], IS 10/78 [12.8%]  I: 68/159 [42.8%] C: 35/78 [44.9%]  I: 94/159 [59.2%] C: 46/78 [59.0%]  I: 64/159 [40.3%] C: 30/78 [38.5%]  I: 44/159 [27.7%] C: 10/78 [12.8%]  CERTIFI;21 Ustekinumab, Crohn’s disease  526/526 [100.0%]  I: 394/526 [74.9%] C: 132/526 [25.1%]  CR-100 Week 6  I: CS 189/394 [48.0%], IS 96/394 [24.4%] C: CS 73/132 [55.3%], IS 30/132 [22.7%]  I: 116/394 [29.4%] C: 44/132 [33.3%]  I: 289/394 [73.4%] C: 91/132 [68.9%]  I: 135/394 [41.9%] C: 41/132 [31.1%]  I: 58/394 [14.7%] C: 14/132 [10.6%]  UNITI 1;7 Ustekinumab, Crohn’s disease  492/496 [99.2%]  I: 246/249 [98.85] C: 246/247 [99.6%]  CR-100 Week 6  I: CS 108/249 [43.4%], IS 78/249 [31.3%] C: CS 111/247 [44.9%], IS 81/247 [32.8%]  I: 72/249 [28.9%] C: 74/247  I: 171/249 [68.7%] C: 170/247 [35.2%]  I: 105/249 [42.2%] C: 87/247 [35.2%]  I: 46/249 [18.5%] C: 22/247 [8.9%]  GAIN;18 Adalimumab, Crohn’s disease  325/325 [100.0%]  I: 159/325 [48.9%] C: 166/325 [51.1%]  CDAI < 150 Week 4  I: CS 55/159 [34.6%], IS 73/159 [45.9%] C: CS 73/166 [44.0%], IS 85/166 [51.2%]  0 [0.0%]  I: 77/159 [48.4%] C: 87/166 [52.4%]  I: 95/159 [59.7%] C: 95/166 [57.2%]  I: 34/159 [21.4%] C: 12/166 [7.2%]  WELCOME;19 Certolizumab pegol, Crohn’s disease  539/539 [100.0%]  I: 539/539 [100.0%] C: 0 [0.0%]  CR-100 Week 6  I: CS 67/168 [39.9%], IS 77/168 [45.8%]  0 [0.0%]  I: 79/168 [47.0%]  I: 66/168 [39.3%]  I: 212/539 [39.3%]  Trial name; active intervention, condition  Proportion of patients with prior exposure to anti- TNF agents [%]  Intervention vs. Control  Outcome; timing of assessment  Co-interventions [proportion on immunomodulators and corticosteroids]  Response to prior therapy  Overall proportion of patients achieving remission  Primary non-response  Secondary loss of response  Intolerance  GEMINI 1 INDUCTION;22,23Vedolizumab, ulcerative colitis  367/895 [41.0%]  I: 82/367 [22.3%] C: 63/367 [17.2%]  MCS<2 Week 6  I: CS 30/82 [36.6%], IS 5/82 [6.1%] C: CS 27/63 [42.9%], IS 6/63 [9.5%]  I: 44/82 [53.7%] C: 29/63 [46.0%]  I: 32/82 [39.0%] C: 26/63 [41.3%]  I: 13/82 [15.6%] C: 10/63 [15.9%]  I: 8/82 [9.8%] C: 2/63 [3.2%]  GEMINI 1 MAINTENANCE;22,23Vedolizumab, ulcerative colitis  121/345 [35.1%]  I: 83/121 [68.6%] C: 38/121 [31.4%]  MCS<2 Week 52  I: CS 32/83 [38.6%], IS 7/83 [8.4%] C: CS 16/38 [42.1%], IS 6/38 [15.8%]  I: 33/83 [39.8%] C: 19/38 [50.0%]  I: 34/83 [41.0%] C: 13/38 [34.2%]  I: 22/83 [26.5%] C: 10/38 [26.3%]  I: 30/83 [36.1%] C: 2/38 [5.3%]  GEMINI 2 & 3 INDUCTION;10,11,20 Vedolizumab, Crohn’s disease  960/1476 [65.0%]  I: 263/960 [27.4%] C: 227/960 [23.65]  CDAI < 150 Week 10  I: CS 94/263 [35.7%], IS 35/263 [13.3%] C: CS 85/227 [37.4%], IS 30/227 [13.2%]  I: 109/263 [41.4%] C: 110/227 [48.5%]  I: 143/263 [54.4%] C: 141/227 [62.1%]  I: 94/263 [35.7%] C: 95/227 [41.9%]  I: 52/263 [19.8%] C: 25/227 [11.0%]  GEMINI 2 MAINTENANCE;11,20 Vedolizumab; Crohn’s disease  237/445 [53.3%]  I: 159/237 [67.1%] C: 78/237 [32.9%]  CDAI < 150 Week 52  I: CS 65/159 [40.9%], IS 27/159 [17.0%] C: CS 29/78 [37.2%], IS 10/78 [12.8%]  I: 68/159 [42.8%] C: 35/78 [44.9%]  I: 94/159 [59.2%] C: 46/78 [59.0%]  I: 64/159 [40.3%] C: 30/78 [38.5%]  I: 44/159 [27.7%] C: 10/78 [12.8%]  CERTIFI;21 Ustekinumab, Crohn’s disease  526/526 [100.0%]  I: 394/526 [74.9%] C: 132/526 [25.1%]  CR-100 Week 6  I: CS 189/394 [48.0%], IS 96/394 [24.4%] C: CS 73/132 [55.3%], IS 30/132 [22.7%]  I: 116/394 [29.4%] C: 44/132 [33.3%]  I: 289/394 [73.4%] C: 91/132 [68.9%]  I: 135/394 [41.9%] C: 41/132 [31.1%]  I: 58/394 [14.7%] C: 14/132 [10.6%]  UNITI 1;7 Ustekinumab, Crohn’s disease  492/496 [99.2%]  I: 246/249 [98.85] C: 246/247 [99.6%]  CR-100 Week 6  I: CS 108/249 [43.4%], IS 78/249 [31.3%] C: CS 111/247 [44.9%], IS 81/247 [32.8%]  I: 72/249 [28.9%] C: 74/247  I: 171/249 [68.7%] C: 170/247 [35.2%]  I: 105/249 [42.2%] C: 87/247 [35.2%]  I: 46/249 [18.5%] C: 22/247 [8.9%]  GAIN;18 Adalimumab, Crohn’s disease  325/325 [100.0%]  I: 159/325 [48.9%] C: 166/325 [51.1%]  CDAI < 150 Week 4  I: CS 55/159 [34.6%], IS 73/159 [45.9%] C: CS 73/166 [44.0%], IS 85/166 [51.2%]  0 [0.0%]  I: 77/159 [48.4%] C: 87/166 [52.4%]  I: 95/159 [59.7%] C: 95/166 [57.2%]  I: 34/159 [21.4%] C: 12/166 [7.2%]  WELCOME;19 Certolizumab pegol, Crohn’s disease  539/539 [100.0%]  I: 539/539 [100.0%] C: 0 [0.0%]  CR-100 Week 6  I: CS 67/168 [39.9%], IS 77/168 [45.8%]  0 [0.0%]  I: 79/168 [47.0%]  I: 66/168 [39.3%]  I: 212/539 [39.3%]  View Large Table 2. Definition of primary non-response, secondary loss of response and intolerance to index anti-TNF agent in included trials. [*Signs and symptoms of active disease = lack of improvement or worsening in stool frequency, rectal bleeding, fever, abdominal pain draining fistula or increase/initiation of antidiarrheal medication.] Trial Acronym  Primary non-response or Inadequate Response  Secondary loss of response  Intolerance  GEMINI 1, 2 & 3  *Signs and symptoms of persistently active disease despite at least one 4 week induction regimen of infliximab[5mg/kg IV, 2doses, > 2 weeks apart],adalimumab [one 80mg dose followed by one 40mg dose 2 weeks apart] or certolizumab [400mg 2doses> 2 weeks apart]  Recurrence of symptoms during maintenance in a patient who had previously benefited from treatment  Patients experiencing treatment related toxicity [infusion reaction, psoriasiform skin lesion, demyelination, congestive heart failure, infection]  CERTIFI UNITI  Presence of at least 1 active sign/symptom* despite infliximab[2 or 3 doses of 5mg/kg], adalimumab [160mg dose followed by 80mg or 80 mg dose followed by 40mg] or certolizumab [2 or 3 doses of 400mg] and have documentation available to investigator  Recurrence of at least 1 sign/symptom after patients initially respond to induction therapy and have received at least 2 maintenance doses of infliximab 5mg/ kg, adalimumab 40mg/kg or certolizumab 400mg and have documentation available to investigator  Significant acute [< 24 hours] or delayed infusion [>24 & <15 days] reaction or injection site reaction [within 24 hours] and have documentation available to investigator  GAIN WELCOME  Not included  Patients have history of an initial response to infliximab, as defined by the investigator; had received at least 2 doses of infliximab of 5 mg/kg of body weight or more every 8 weeks; and had lacked improvement or had worsening signs and symptoms  History of discontinuing infliximab treatment because of a clinically significant acute or delayed infusion reaction.  Trial Acronym  Primary non-response or Inadequate Response  Secondary loss of response  Intolerance  GEMINI 1, 2 & 3  *Signs and symptoms of persistently active disease despite at least one 4 week induction regimen of infliximab[5mg/kg IV, 2doses, > 2 weeks apart],adalimumab [one 80mg dose followed by one 40mg dose 2 weeks apart] or certolizumab [400mg 2doses> 2 weeks apart]  Recurrence of symptoms during maintenance in a patient who had previously benefited from treatment  Patients experiencing treatment related toxicity [infusion reaction, psoriasiform skin lesion, demyelination, congestive heart failure, infection]  CERTIFI UNITI  Presence of at least 1 active sign/symptom* despite infliximab[2 or 3 doses of 5mg/kg], adalimumab [160mg dose followed by 80mg or 80 mg dose followed by 40mg] or certolizumab [2 or 3 doses of 400mg] and have documentation available to investigator  Recurrence of at least 1 sign/symptom after patients initially respond to induction therapy and have received at least 2 maintenance doses of infliximab 5mg/ kg, adalimumab 40mg/kg or certolizumab 400mg and have documentation available to investigator  Significant acute [< 24 hours] or delayed infusion [>24 & <15 days] reaction or injection site reaction [within 24 hours] and have documentation available to investigator  GAIN WELCOME  Not included  Patients have history of an initial response to infliximab, as defined by the investigator; had received at least 2 doses of infliximab of 5 mg/kg of body weight or more every 8 weeks; and had lacked improvement or had worsening signs and symptoms  History of discontinuing infliximab treatment because of a clinically significant acute or delayed infusion reaction.  View Large Table 2. Definition of primary non-response, secondary loss of response and intolerance to index anti-TNF agent in included trials. [*Signs and symptoms of active disease = lack of improvement or worsening in stool frequency, rectal bleeding, fever, abdominal pain draining fistula or increase/initiation of antidiarrheal medication.] Trial Acronym  Primary non-response or Inadequate Response  Secondary loss of response  Intolerance  GEMINI 1, 2 & 3  *Signs and symptoms of persistently active disease despite at least one 4 week induction regimen of infliximab[5mg/kg IV, 2doses, > 2 weeks apart],adalimumab [one 80mg dose followed by one 40mg dose 2 weeks apart] or certolizumab [400mg 2doses> 2 weeks apart]  Recurrence of symptoms during maintenance in a patient who had previously benefited from treatment  Patients experiencing treatment related toxicity [infusion reaction, psoriasiform skin lesion, demyelination, congestive heart failure, infection]  CERTIFI UNITI  Presence of at least 1 active sign/symptom* despite infliximab[2 or 3 doses of 5mg/kg], adalimumab [160mg dose followed by 80mg or 80 mg dose followed by 40mg] or certolizumab [2 or 3 doses of 400mg] and have documentation available to investigator  Recurrence of at least 1 sign/symptom after patients initially respond to induction therapy and have received at least 2 maintenance doses of infliximab 5mg/ kg, adalimumab 40mg/kg or certolizumab 400mg and have documentation available to investigator  Significant acute [< 24 hours] or delayed infusion [>24 & <15 days] reaction or injection site reaction [within 24 hours] and have documentation available to investigator  GAIN WELCOME  Not included  Patients have history of an initial response to infliximab, as defined by the investigator; had received at least 2 doses of infliximab of 5 mg/kg of body weight or more every 8 weeks; and had lacked improvement or had worsening signs and symptoms  History of discontinuing infliximab treatment because of a clinically significant acute or delayed infusion reaction.  Trial Acronym  Primary non-response or Inadequate Response  Secondary loss of response  Intolerance  GEMINI 1, 2 & 3  *Signs and symptoms of persistently active disease despite at least one 4 week induction regimen of infliximab[5mg/kg IV, 2doses, > 2 weeks apart],adalimumab [one 80mg dose followed by one 40mg dose 2 weeks apart] or certolizumab [400mg 2doses> 2 weeks apart]  Recurrence of symptoms during maintenance in a patient who had previously benefited from treatment  Patients experiencing treatment related toxicity [infusion reaction, psoriasiform skin lesion, demyelination, congestive heart failure, infection]  CERTIFI UNITI  Presence of at least 1 active sign/symptom* despite infliximab[2 or 3 doses of 5mg/kg], adalimumab [160mg dose followed by 80mg or 80 mg dose followed by 40mg] or certolizumab [2 or 3 doses of 400mg] and have documentation available to investigator  Recurrence of at least 1 sign/symptom after patients initially respond to induction therapy and have received at least 2 maintenance doses of infliximab 5mg/ kg, adalimumab 40mg/kg or certolizumab 400mg and have documentation available to investigator  Significant acute [< 24 hours] or delayed infusion [>24 & <15 days] reaction or injection site reaction [within 24 hours] and have documentation available to investigator  GAIN WELCOME  Not included  Patients have history of an initial response to infliximab, as defined by the investigator; had received at least 2 doses of infliximab of 5 mg/kg of body weight or more every 8 weeks; and had lacked improvement or had worsening signs and symptoms  History of discontinuing infliximab treatment because of a clinically significant acute or delayed infusion reaction.  View Large 3.1. Primary non-response vs. intolerance On meta-analysis of the active intervention arm of six RCTs, patients with PNR to the index anti-TNF agent were 24% less likely to achieve remission with the second-line biologic agent as compared with patients who discontinued the index anti-TNF therapy due to intolerance [RR, 0.76; 95% CI, 0.61–0.96] with minimal heterogeneity [I2 = 18%] [Figure 2A]. These results were stable on subgroup analyses based on study design, class of second-line intervention and disease type [Table 3]. Figure 2. View largeDownload slide Response to second-line biologic in patients who discontinued prior anti-TNF therapy due to [A] primary non-response vs. intolerance, [B] primary non-response vs. secondary loss of response and [C] secondary loss of response vs. intolerance. GEMINI 2 and 3, UNITI 1, CERTIFI, GAIN and WELCOME were conducted in patients with Crohn’s disease, and GEMINI 1 was conducted in patients with ulcerative colitis. Figure 2. View largeDownload slide Response to second-line biologic in patients who discontinued prior anti-TNF therapy due to [A] primary non-response vs. intolerance, [B] primary non-response vs. secondary loss of response and [C] secondary loss of response vs. intolerance. GEMINI 2 and 3, UNITI 1, CERTIFI, GAIN and WELCOME were conducted in patients with Crohn’s disease, and GEMINI 1 was conducted in patients with ulcerative colitis. Table 3. Subgroup analysis. Subgroups  Categories  RR  95% CI  I2  P-interaction  Prior primary non-response vs. intolerance  Overall  6 trials  0.76  0.61–0.96  18    Trial design  Induction [4]  0.74  0.60–0.92  0  0.93  Maintenance [2]  0.77  0.37–1.59  70  Second-line agent  VDZ [4]  0.83  0.54–1.28  40  0.46  UST [2]  0.69  0.55–0.88  0  Disease type  CD [4]  0.82  0.64–1.07  26  0.13  UC [2]  0.52  0.30–0.89  0  Prior primary non-response vs. secondary loss of response  Overall  6 trials  0.88  0.64–1.21  52    Trial design  Induction [4]  0.73  0.56–0.97  27  0.037  Maintenance [2]  1.27  0.82–1.94  0  Second-line agent  VDZ [4]  1.16  0.85–1.58  0  0.002  UST [2]  0.64  0.52–0.80  0  Disease type  CD [4]  0.82  0.59–1.15  59  0.22  UC [2]  1.36  0.66–2.79  0  Secondary loss of response vs. intolerance  Overall  8 trials  0.96  0.78–1.18  35    Trial design  Induction [6]  1.06  0.91–1.23  0  0.27  Maintenance [2]  0.58  0.20–1.67  82  Second-line agent  VDZ [4]  0.71  0.41–1.22  58  0.34  UST [2]  1.08  0.91–1.30  0  Anti-TNF [2]  1.03  0.74–1.43  0  Disease type  CD [6]  1.06  0.91–1.22  0  0.002  UC [2]  0.36  0.18–0.70  0  Subgroups  Categories  RR  95% CI  I2  P-interaction  Prior primary non-response vs. intolerance  Overall  6 trials  0.76  0.61–0.96  18    Trial design  Induction [4]  0.74  0.60–0.92  0  0.93  Maintenance [2]  0.77  0.37–1.59  70  Second-line agent  VDZ [4]  0.83  0.54–1.28  40  0.46  UST [2]  0.69  0.55–0.88  0  Disease type  CD [4]  0.82  0.64–1.07  26  0.13  UC [2]  0.52  0.30–0.89  0  Prior primary non-response vs. secondary loss of response  Overall  6 trials  0.88  0.64–1.21  52    Trial design  Induction [4]  0.73  0.56–0.97  27  0.037  Maintenance [2]  1.27  0.82–1.94  0  Second-line agent  VDZ [4]  1.16  0.85–1.58  0  0.002  UST [2]  0.64  0.52–0.80  0  Disease type  CD [4]  0.82  0.59–1.15  59  0.22  UC [2]  1.36  0.66–2.79  0  Secondary loss of response vs. intolerance  Overall  8 trials  0.96  0.78–1.18  35    Trial design  Induction [6]  1.06  0.91–1.23  0  0.27  Maintenance [2]  0.58  0.20–1.67  82  Second-line agent  VDZ [4]  0.71  0.41–1.22  58  0.34  UST [2]  1.08  0.91–1.30  0  Anti-TNF [2]  1.03  0.74–1.43  0  Disease type  CD [6]  1.06  0.91–1.22  0  0.002  UC [2]  0.36  0.18–0.70  0  View Large Table 3. Subgroup analysis. Subgroups  Categories  RR  95% CI  I2  P-interaction  Prior primary non-response vs. intolerance  Overall  6 trials  0.76  0.61–0.96  18    Trial design  Induction [4]  0.74  0.60–0.92  0  0.93  Maintenance [2]  0.77  0.37–1.59  70  Second-line agent  VDZ [4]  0.83  0.54–1.28  40  0.46  UST [2]  0.69  0.55–0.88  0  Disease type  CD [4]  0.82  0.64–1.07  26  0.13  UC [2]  0.52  0.30–0.89  0  Prior primary non-response vs. secondary loss of response  Overall  6 trials  0.88  0.64–1.21  52    Trial design  Induction [4]  0.73  0.56–0.97  27  0.037  Maintenance [2]  1.27  0.82–1.94  0  Second-line agent  VDZ [4]  1.16  0.85–1.58  0  0.002  UST [2]  0.64  0.52–0.80  0  Disease type  CD [4]  0.82  0.59–1.15  59  0.22  UC [2]  1.36  0.66–2.79  0  Secondary loss of response vs. intolerance  Overall  8 trials  0.96  0.78–1.18  35    Trial design  Induction [6]  1.06  0.91–1.23  0  0.27  Maintenance [2]  0.58  0.20–1.67  82  Second-line agent  VDZ [4]  0.71  0.41–1.22  58  0.34  UST [2]  1.08  0.91–1.30  0  Anti-TNF [2]  1.03  0.74–1.43  0  Disease type  CD [6]  1.06  0.91–1.22  0  0.002  UC [2]  0.36  0.18–0.70  0  Subgroups  Categories  RR  95% CI  I2  P-interaction  Prior primary non-response vs. intolerance  Overall  6 trials  0.76  0.61–0.96  18    Trial design  Induction [4]  0.74  0.60–0.92  0  0.93  Maintenance [2]  0.77  0.37–1.59  70  Second-line agent  VDZ [4]  0.83  0.54–1.28  40  0.46  UST [2]  0.69  0.55–0.88  0  Disease type  CD [4]  0.82  0.64–1.07  26  0.13  UC [2]  0.52  0.30–0.89  0  Prior primary non-response vs. secondary loss of response  Overall  6 trials  0.88  0.64–1.21  52    Trial design  Induction [4]  0.73  0.56–0.97  27  0.037  Maintenance [2]  1.27  0.82–1.94  0  Second-line agent  VDZ [4]  1.16  0.85–1.58  0  0.002  UST [2]  0.64  0.52–0.80  0  Disease type  CD [4]  0.82  0.59–1.15  59  0.22  UC [2]  1.36  0.66–2.79  0  Secondary loss of response vs. intolerance  Overall  8 trials  0.96  0.78–1.18  35    Trial design  Induction [6]  1.06  0.91–1.23  0  0.27  Maintenance [2]  0.58  0.20–1.67  82  Second-line agent  VDZ [4]  0.71  0.41–1.22  58  0.34  UST [2]  1.08  0.91–1.30  0  Anti-TNF [2]  1.03  0.74–1.43  0  Disease type  CD [6]  1.06  0.91–1.22  0  0.002  UC [2]  0.36  0.18–0.70  0  View Large 3.2. Primary non-response vs. secondary loss of response There was no significant difference in response to second-line non-TNF biologic in patients who discontinued prior anti-TNF due to PNR vs. LOR [RR, 0.88; 95% CI, 0.64–1.21], with considerable heterogeneity [I2 = 52%] [Figure 2B]. On analyses stratified by study design, trials of induction therapy demonstrated a lower likelihood of achieving remission with the second-line non-TNF biologic in patients with prior PNR to anti-TNF as compared with patients with secondary LOR to the anti-TNF [RR, 0.73; 95% CI, 0.56–0.97], with minimal heterogeneity [I2 = 27%]; no significant difference in response was observed in trials of maintenance therapy, in which only patients with response to induction therapy were included [Table 3]. The risk of achieving remission among patients with prior PNR to anti-TNF as compared with patients with prior LOR was observed only in patients treated with ustekinumab [RR, 0.53; 95% CI, 0.39–0.71]; there was no difference in response to vedolizumab in patients with prior PNR or LOR [RR, 1.06; 95% CI, 0.60–1.87]. These results were stable when stratified based on disease type [CD or UC]. 3.3. Secondary loss of response vs. intolerance On meta-analysis of eight RCTs, there was no significant difference in response to the second-line biologic in patients who discontinued the prior anti-TNF due to LOR or intolerance [RR, 0.96; 95% CI, 0.78–1.18], with moderate heterogeneity [I2 = 35%] [Figure 2C]. Results were stable on subgroup analyses based on trial design and class of medication, though in trials of UC, patients with secondary LOR showed an inferior response as compared with patients who discontinued the primary anti-TNF due to intolerance [Table 3]. 3.4. Publication bias Due to the small number of studies, formal assessment of funnel plot asymmetry was not performed. 4. Discussion With availability of newer non-TNF biologics and small molecules for treating patients with IBD, assessment of likelihood of response to second-line agents in patients who have previously been exposed to anti-TNF agents is important in order to position subsequent therapies. In this systematic review and meta-analysis, we made several key observations: [a] patients with prior PNR to anti-TNF agents were less likely to respond to treatment with another non-anti-TNF biologic, as compared with patients who discontinued their index anti-TNF agent due to intolerance; [b] patients with prior PNR to anti-TNF agents may have been less likely to achieve induction of remission with another non–anti-TNF biologic as compared with patients with LOR; once they achieve response to induction therapy, however, the likelihood of maintaining response was comparable; [c] patients treated with ustekinumab, which blocks IL-12/23, but not those treated with vedolizumab, which blocks lymphocyte recruitment to the gastrointestinal tract, were less likely to respond to induction therapy if they had PNR to anti-TNF agents as compared with prior secondary LOR to anti-TNF agents. Overall, these results suggest that patients with IBD with PNR to anti-TNF agents may be intrinsically more difficult to treat with second-line biologics. These findings directly inform clinical practice. In particular, patients with PNR to initial anti-TNF therapy attributed to high drug clearance are potentially at increased risk of rapid drug clearance on second-line biologics, and attempts should be made to minimize this risk [such as through use of higher doses for induction therapy, proactive drug monitoring during the induction phase to allow early optimization, and combination therapy with immunomodulators to minimize the risk of immunogenicity]. Trials of all approved biologic therapies, as well as those in development, suggest lower response rates in the subset of patients with prior exposure to anti-TNF agents as compared with biologic-naïve patients.7,10,11,24 However, within this group of anti-TNF–exposed patients, we observed that response rates vary depending on the specific reason for discontinuing the anti-TNF therapy. Intuitively, we observed that patients with PNR to anti-TNFs were less likely to respond to a second line agent, as compared with patients who were intolerant to therapies. Recent observational studies have confirmed that patients with PNR to first-line therapy infliximab have a >50% chance of progressing to surgery over the next 1 year.25,26 It is not surprising that patients with intolerance to one medication, may tolerate the next medication better, without compromising possible efficacy of the second-line therapy. The observation that patients with PNR to anti-TNF agents may not respond as well as patients with secondary LOR may be due to a combination of pharmacokinetic as well as pharmacodynamic properties of biologics. An exposure–response relationship has been observed with all biologics in IBD, wherein patients with higher drug concentrations respond better than patients with low trough levels.7,12,27–29 Patients may have PNR to first-line anti-TNF agents either because of inadequate trough concentrations due to rapid, non–immune-mediated drug clearance [pharmacokinetics] or due to mechanistic failure wherein patients fail to respond despite adequate trough concentrations, perhaps because the disease is mediated through a non–TNF-mediated pathway.4 Unfortunately, the included trials did not classify patients as having PNR based on trough concentrations, but rather based on clinical history of prior response or not. Patients with PNR to the first-line anti-TNF due to inadequate trough concentrations may be prone to PNR to the second-line biologic due to the same factors that led to rapid drug clearance with the first agent. Available population pharmacokinetic analyses for biologics in IBD for infliximab, certolizumab pegol and vedolizumab have identified similar covariates to be associated with drug clearance, such as body weight, albumin concentration and anti-drug antibodies.30–33 In these patients with pharmacokinetically determined PNR, it is possible that upfront dose optimization with the second-line anti-TNF or non-TNF biologic may help overcome the pharmacokinetic problem. On the other hand, patients mechanistically resistant to a cytokine-based treatment focusing on blocking TNF may be more difficult to treat with any other agent, resulting in lower response to a second-line biologic. It is conceivable that with use of therapeutic drug monitoring [TDM] and more accurate classification of PNR vs. LOR based on adequacy of drug exposure these distinctions in response to second-line therapy would be exaggerated.4 The observation of a differential response in patients with prior PNR and LOR to anti-TNF, with ustekinumab but not vedolizumab was unexpected. These findings are hypothesis-generating, suggesting that though there may be biologically distinct signaling pathways, functionally there are shared cellular and downstream immunologic effects such that patients who fail to respond to one TNF blockade may be less likely to respond to another IL-12/23 blockade. This may inform personalized clinical decision-making, providing a framework and rationale for switching between therapies for IBD based on the downstream immunologic effects. It is very important to note that these results reflect lack of difference in response regardless of PNR vs. LOR in a cohort of patients treated with vedolizumab [which may be due to the overall low response rate observed in vedolizumab-treated patients with prior anti-TNF exposure], and does not represent differences in response rate between ustekinumab and vedolizumab within a cohort of patients with PNR; in the absence of head-to-head comparative trials, the latter data is not available. Prospective mechanistic studies interrogating these hypotheses, in which drug clearance and drug concentrations are incorporated into the analyses, are warranted. There are several limitations to our study. First, response rate stratified by reason for discontinuation of prior anti-TNF therapy was inconsistently reported in trials. When reported, some patients within trials were classified in more than one category for reason for discontinuation of therapy, since this was primarily based on patient recall, not objective confirmation. This may also introduce bias, since exposure types are not explicitly determined. There were no trials evaluating response to second-line anti-TNF agent in patients with a PNR to the first anti-TNF agent. As noted above, it is conceivable that patients with reported PNR to the first anti-TNF agent may respond well to a second anti-TNF agent, particularly those with a pharmacokinetically determined PNR, due to differences in affinity-dependent target-mediated drug disposition between different anti-TNF agents and early appearance anti-drug antibodies. Second, in the absence of individual participant–level data, only unadjusted analysis could be performed, and we were unable to account for other potential differences in patients with prior PNR, LOR or intolerance. Third, for our overall analysis, we combined trials of UC and CD; conceptually, we believed this would be acceptable for our study question, since specific IBD phenotype–related factors are unlikely to influence response to second-line biologic based on reason for discontinuation of primary anti-TNF therapy. There were differences in definition and time point of assessment of outcome measure, and trials of induction and maintenance therapy were combined for primary analysis. However, we confirmed our findings on relevant subgroup analyses. Fourth, only clinical outcomes [remission/response] were reported, and endoscopic or biochemical response/remission rates stratified according to reason for primary anti-TNF failure were not reported in trials. Along the same lines, since objective confirmation of active inflammatory disease was not warranted at enrollment in all trials, it is possible that a subset of patients [patients with non-inflammatory etiology of symptoms, such as patients with irritable bowel syndrome, fibrostenotic disease, etc.] characterized as having PNR to anti-TNF agents may not be responsive to any IBD-related therapy. In conclusion, we observed considerable differences in response to the second-line biologic, based on reason for discontinuation of the primary anti-TNF therapy. Patients with prior PNR were less likely to respond to second-line therapy as compared with patients with prior LOR or intolerance. Future studies are warranted to confirm these findings in clinical practice, particularly with classification of PNR and LOR based on TDM. Mechanistic studies exploring potential reasons for discrepancy in response to ustekinumab and vedolizumab in a subset of patients with PNR are warranted. Funding This study did not receive any direct funding Conflict of Interest Dr Singh is supported by the American College of Gastroenterology, the Crohn’s and Colitis Foundation and Pfizer. Dr Boland has served as a consultant for Abbvie and has a research grant from Takeda and Janssen. Dr Vande Casteele has served as a consultant for Boehringer Ingelheim, UCB Pharma, Pfizer and Takeda. Dr Sandborn has served as a consultant and received research funding from Janssen, Abbvie, UCB Pharma, Takeda, and Pfizer. Author Contributions Study concept and design: SS Acquisition of data: SS, JG Analysis and interpretation of data: SS, JG Drafting of the manuscript: SS, JG Critical revision of the manuscript for important intellectual content: BSB, NVC, WJS Approval of the final manuscript: SS, JG, BSB, NVC, WJS Guarantor of the article: SS References 1. Nielsen OH, Ainsworth MA. Tumor necrosis factor inhibitors for inflammatory bowel disease. N Engl J Med  2013; 369: 754– 62. Google Scholar CrossRef Search ADS PubMed  2. Singh S, Pardi DS. Update on anti-tumor necrosis factor agents in Crohn disease. Gastroenterol Clin North Am  2014; 43: 457– 78. Google Scholar CrossRef Search ADS PubMed  3. Sandborn WJ, Rutgeerts P, Feagan BGet al.   Colectomy rate comparison after treatment of ulcerative colitis with placebo or infliximab. Gastroenterology  2009; 137: 1250– 60; quiz 1520. Google Scholar CrossRef Search ADS PubMed  4. Papamichael K, Gils A, Rutgeerts Pet al.   Role for therapeutic drug monitoring during induction therapy with TNF antagonists in IBD: evolution in the definition and management of primary nonresponse. Inflamm Bowel Dis  2015; 21: 182– 97. Google Scholar CrossRef Search ADS PubMed  5. Gisbert JP, Marín AC, McNicholl AG, Chaparro M. 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. Google Scholar CrossRef Search ADS PubMed  6. Sandborn WJ, Su C, Sands BEet al.  ; OCTAVE Induction 1, OCTAVE Induction 2, and OCTAVE Sustain Investigators. Tofacitinib as induction and maintenance therapy for ulcerative colitis. N Engl J Med  2017; 376: 1723– 36. Google Scholar CrossRef Search ADS PubMed  7. Feagan BG, Sandborn WJ, Gasink Cet al.  ; UNITI–IM-UNITI Study Group. Ustekinumab as induction and maintenance therapy for Crohn’s disease. N Engl J Med  2016; 375: 1946– 60. Google Scholar CrossRef Search ADS PubMed  8. Feagan BG, Sandborn WJ, D’Haens Get al.   Induction therapy with the selective interleukin-23 inhibitor risankizumab in patients with moderate-to-severe Crohn’s disease: a randomised, double-blind, placebo-controlled phase 2 study. Lancet  2017; 389: 1699– 709. Google Scholar CrossRef Search ADS PubMed  9. Sandborn WJ, Feagan BG, Wolf DCet al.  ; TOUCHSTONE Study Group. Ozanimod induction and maintenance treatment for ulcerative colitis. N Engl J Med  2016; 374: 1754– 62. Google Scholar CrossRef Search ADS PubMed  10. Sands BE, Feagan BG, Rutgeerts Pet al.   Effects of vedolizumab induction therapy for patients with Crohn’s disease in whom tumor necrosis factor antagonist treatment failed. Gastroenterology  2014; 147: 618– 627.e3. Google Scholar CrossRef Search ADS PubMed  11. Sands BE, Sandborn WJ, Van Assche Get al.   Vedolizumab as induction and maintenance therapy for Crohn’s disease in patients naïve to or who have failed tumor necrosis factor antagonist therapy. Inflamm Bowel Dis  2017; 23: 97– 106. Google Scholar CrossRef Search ADS PubMed  12. Vande Casteele N, Herfarth H, Katz J, Falck-Ytter Y, Singh S. American Gastroenterological Association Institute technical review on the role of therapeutic drug monitoring in the management of inflammatory bowel diseases. Gastroenterology  2017; 153: 835– 857.e6. Google Scholar CrossRef Search ADS PubMed  13. Feuerstein JD, Nguyen GC, Kupfer SS, Falck-Ytter Y, Singh S; American Gastroenterological Association Institute Clinical Guidelines Committee. American Gastroenterological Association Institute guideline on therapeutic drug monitoring in inflammatory bowel disease. Gastroenterology  2017; 153: 827– 34. Google Scholar CrossRef Search ADS PubMed  14. Liberati A, Altman DG, Tetzlaff Jet al.   The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ  2009; 339: b2700. Google Scholar CrossRef Search ADS PubMed  15. Higgins JPT, Altman DG, Sterne JAC. Assessing risk of bias in included studies . In: Higgins JPT Green S, editors. Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1.0  [updated March 2011]. The Cochrane Collaboration; 2011. http://handbook.cochrane.org. Accessed July 1, 2017. 16. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials  1986; 7: 177– 88. Google Scholar CrossRef Search ADS PubMed  17. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ  2003; 327: 557– 60. Google Scholar CrossRef Search ADS PubMed  18. Sandborn WJ, Rutgeerts P, Enns Ret al.   Adalimumab induction therapy for Crohn disease previously treated with infliximab: a randomized trial. Ann Intern Med  2007; 146: 829– 38. Google Scholar CrossRef Search ADS PubMed  19. Sandborn WJ, Vermeire S, D’Haens GRet al.   WELCOME: a randomized, double-blind, controlled trial comparing certolizumab pegol 400 mg every 2 weeks with every 4 weeks for maintenance of response and remission in patients with moderate to severe Crohn’s disease with secondary failure to infliximab. Gastroenterology  2009; 136: A27. 20. Sandborn WJ, Feagan BG, Rutgeerts Pet al.  ; GEMINI 2 Study Group. Vedolizumab as induction and maintenance therapy for Crohn’s disease. N Engl J Med  2013; 369: 711– 21. Google Scholar CrossRef Search ADS PubMed  21. Sandborn WJ, Gasink C, Gao LLet al.  ; CERTIFI Study Group. Ustekinumab induction and maintenance therapy in refractory Crohn’s disease. N Engl J Med  2012; 367: 1519– 28. Google Scholar CrossRef Search ADS PubMed  22. Feagan BG, Rutgeerts P, Sands BEet al.  ; GEMINI 1 Study Group. Vedolizumab as induction and maintenance therapy for ulcerative colitis. N Engl J Med  2013; 369: 699– 710. Google Scholar CrossRef Search ADS PubMed  23. Feagan BG, Rubin DT, Danese Set al.   Efficacy of vedolizumab induction and maintenance therapy in patients with ulcerative colitis, regardless of prior exposure to tumor necrosis factor antagonists. Clin Gastroenterol Hepatol  2017; 15: 229– 39.e5. Google Scholar CrossRef Search ADS PubMed  24. Vermeire S, O’Byrne S, Keir Met al.   Etrolizumab as induction therapy for ulcerative colitis: a randomised, controlled, phase 2 trial. Lancet  2014; 384: 309– 18. Google Scholar CrossRef Search ADS PubMed  25. Fernández-Salazar L, Muñoz F, Barrio Jet al.   Infliximab in ulcerative colitis: real-life analysis of factors predicting treatment discontinuation due to lack of response or colectomy: ECIA (ACAD Colitis and Infliximab Study). Scand J Gastroenterol  2016; 51: 186– 95. Google Scholar CrossRef Search ADS PubMed  26. Papamichael K, Rivals-Lerebours O, Billiet Tet al.   Long-term outcome of patients with ulcerative colitis and primary non-response to infliximab. J Crohns Colitis  2016; 10: 1015– 23. Google Scholar CrossRef Search ADS PubMed  27. Rosario M, French JL, Dirks NLet al.   Exposure-efficacy relationships for vedolizumab induction therapy in patients with ulcerative colitis or Crohn’s disease. J Crohns Colitis  2017; 11: 921– 9. Google Scholar CrossRef Search ADS PubMed  28. Battat R, Kopylov U, Bessissow Tet al.   Association between ustekinumab trough concentrations and clinical, biomarker, and endoscopic outcomes in patients with Crohn’s disease. Clin Gastroenterol Hepatol  2017; 15: 1427– 1434.e2. Google Scholar CrossRef Search ADS PubMed  29. Adedokun OJ, Sandborn WJ, Feagan BGet al.   Association between serum concentration of infliximab and efficacy in adult patients with ulcerative colitis. Gastroenterology  2014; 147: 1296– 307.e5. Google Scholar CrossRef Search ADS PubMed  30. Brandse JF, Mould D, Smeekes Oet al.   A real-life population pharmacokinetic study reveals factors associated with clearance and immunogenicity of infliximab in inflammatory bowel disease. Inflamm Bowel Dis  2017; 23: 650– 60. Google Scholar CrossRef Search ADS PubMed  31. Paul S, Moreau AC, Del Tedesco Eet al.   Pharmacokinetics of adalimumab in inflammatory bowel diseases: a systematic review and meta-analysis. Inflamm Bowel Dis  2014; 20: 1288– 95. Google Scholar CrossRef Search ADS PubMed  32. Rosario M, Dirks NL, Gastonguay MRet al.   Population pharmacokinetics–pharmacodynamics of vedolizumab in patients with ulcerative colitis and Crohn’s disease. Aliment Pharmacol Ther  2015; 42: 188– 202. Google Scholar CrossRef Search ADS PubMed  33. Wade JR, Parker G, Kosutic Get al.   Population pharmacokinetic analysis of certolizumab pegol in patients with Crohn’s disease. J Clin Pharmacol  2015; 55: 866– 74. Google Scholar CrossRef Search ADS PubMed  Copyright © 2018 European Crohn’s and Colitis Organisation (ECCO). Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Crohn's and Colitis Oxford University Press

Primary Non-Response to Tumor Necrosis Factor Antagonists is Associated with Inferior Response to Second-line Biologics in Patients with Inflammatory Bowel Diseases: A Systematic Review and Meta-analysis

Loading next page...
 
/lp/ou_press/primary-non-response-to-tumor-necrosis-factor-antagonists-is-MJ47EgvElZ
Publisher
Elsevier Science
Copyright
Copyright © 2018 European Crohn’s and Colitis Organisation (ECCO). Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com
ISSN
1873-9946
eISSN
1876-4479
D.O.I.
10.1093/ecco-jcc/jjy004
Publisher site
See Article on Publisher Site

Abstract

Abstract Background and Aims We sought to analyze whether response to a second-line biologic varies depending on the reason for discontinuation of the primary anti-TNF agent (primary non-response [PNR], secondary loss of response [LOR] after initial response, or intolerance), through a systematic review and meta-analysis. Methods Through a systematic search through May 31, 2017, we identified eight randomized controlled trials [RCTs] of biologics in patients with IBD with prior exposure to anti-TNF agents, that stratified response to second-line therapy by reason for discontinuing primary anti-TNF therapy [PNR vs. LOR vs. intolerance]. We estimated relative risk [RR] (and 95% confidence interval [CI]) of achieving clinical remission in patients with PNR as compared with patients with LOR, and intolerance, through random effects meta-analysis. Results As compared with patients who discontinued prior anti-TNF due to intolerance, patients with prior PNR were 24% less likely to achieve remission with second-line biologics (RR,0.76 [0.61–0.96]). As compared with patients who discontinued prior anti-TNF due to LOR, patients with prior PNR were 27% less likely to achieve remission with induction therapy with second-line biologics (RR,0.73 [0.56–0.97]), particularly to ustekinumab (RR,0.64 [0.52–0.80]). There was no difference in response to vedolizumab in patients with prior PNR or LOR to anti-TNF agents (RR,1.16 [0.85–1.58]). Conclusion Patients with PNR to anti-TNF agents are less likely to respond to second-line non-TNF biologics, as compared with patients who discontinued therapy due to secondary LOR or intolerance. This may be attributed to underlying pharmacokinetics and pharmacodynamics of anti-TNF agents in patients with PNR. Prediction, anti-integrin, anti-interleukin, second-line, inflammatory bowel diseases 1. Introduction Anti-tumor necrosis factor-α [TNF] agents have been the mainstay for managing patients with moderate to severe inflammatory bowel diseases [IBDs], inducing and maintaining remission, and decreasing the risk of surgery, hospitalization and disease-related complications.1–3 However, ~30–40% patients may not respond to anti-TNF agents (primary non-response [PNR]); additionally, ~30–40% patients may lose response over time (secondary loss of response [LOR]) or may be intolerant to anti-TNFs.4 Previous studies have suggested that response to a second anti-TNF agent after discontinuation of the index anti-TNF agent varies depending on the reason for discontinuation. In a systematic review of 46 studies, Gisbert and colleagues observed that clinical remission rates with the second anti-TNF were highest in patients who discontinued the primary anti-TNF due to intolerance [61%], followed by those with LOR [45%].5 Patients with PNR to the index anti-TNF are least likely to respond to a second anti-TNF agent. The treatment landscape of moderate–severe IBD is expanding, with several non-TNF biologics recently approved, and other biologics and small molecules in advanced stages of development.6–9 Among approved agents, both vedolizumab and ustekinumab are more effective in anti-TNF–naïve patients, than in anti-TNF–exposed patients.7,10,11 This may suggest that the latter group of patients are intrinsically more treatment-resistant, or may have been “primed” by prior anti-TNF exposure for inadequate response to a second agent. Moreover, all patients with prior anti-TNF exposure are not equivalent, and may have intrinsic differences in response to second-line biologics. For example, patients with PNR may have altered pharmacokinetics [rapid drug clearance resulting in low trough levels] or pharmacodynamics [mechanistic failure with non-TNF–mediated inflammation], resulting in decreased likelihood of response to a second biologic agent;4,12,13 in contrast, patients with secondary LOR after initial response potentially due to immunogenicity may be more treatment responsive. Through a systematic review and meta-analysis of published clinical trials, we analyzed whether response to a second-line biologic or small molecule differs according to the reason for discontinuation of the primary anti-TNF agent [PNR vs. LOR vs. intolerance]. 2. Methods This systematic review followed the preferred reporting items for systematic reviews and meta-analysis [PRISMA] standards, and followed an a priori protocol.14 2.1. Selection criteria Studies included in this meta-analysis were Phase II or III RCTs that met the following inclusion criteria: [1] Patients: adults [age >18 years] with moderate to severe ulcerative colitis [UC] (Mayo Clinic Score [MCS] 6–12, with an endoscopic subscore of 2 or 3] or Crohn’s disease [CD] (Crohn’s Disease Activity Index [CDAI] >220 but <450), who had previously been exposed to anti-TNF agents; [2] Intervention: biologic therapy [anti-TNF agents, anti-integrin agents, anti-interleukin-12 and/or -23], or small molecules [janus kinase inhibitors, sphingosine-1 phosphate receptor agonist or SMAD7 antisense oligonucleotide], with a minimum duration of therapy of 14 days; [3] Comparator: another biologic agent or placebo; [4] Outcome: achievement of clinical remission or response, stratified by reason for discontinuation [PNR vs. LOR vs. intolerance] of index anti-TNF agent. We excluded the following studies: [1] trials conducted exclusively in biologic-naïve patients, [2] trials where results were not stratified by reason for discontinuation of prior anti-TNF, [3] Phase I trials, [4] pediatric studies, or [5] trials conducted in patients with acute severe colitis. 2.2. Search strategy We conducted a comprehensive search of multiple electronic databases through May 31, 2017, about adults with no language restrictions. The databases included Ovid MEDLINE In-Process & Other Non-Indexed Citations, Ovid MEDLINE, Ovid EMBASE, Ovid Cochrane Central Register of Controlled Trials, Ovid Cochrane Database of Systematic Reviews, Web of Science, and Scopus. The search terms used included a combination of phrases indicating the diseases of interest “Crohn[s] disease”, “Ulcerative colitis”, “inflammatory bowel disease”, “regional enteritis” and treatments including biologics [“infliximab”, “adalimumab”, “certolizumab pegol”, “golimumab”, “anti-TNF”, “TNF-antagonist”, “vedolizumab”, “natalizumab”, “etrolizumab”, “monoclonal antibod*”, “anti-integrin”, “anti-interleukin”, “ustekinumab”, “risankizumab”] and small molecules [“tofacitinib”, “janus kinase”, “ozanimod”, “trafficking”, “mongersen”, “SMAD7”]. Two study investigators [SS and JG] independently reviewed the title and abstract of studies identified in the search to exclude studies that did not address the research question of interest on the basis of pre-specified inclusion and exclusion criteria. The full text of the remaining articles was examined to determine whether it contained relevant information. Conflicts in study selection at this stage were resolved by consensus, referring back to the original article, in consultation with a senior investigator [WJS]. Second, we searched the bibliographies of these selected articles, systematic reviews and clinical trial registries [www.clinicaltrials.gov] to identify any additional studies. Third, we conducted a manual search of abstracts from major gastroenterology conferences [Digestive Disease Week, American College of Gastroenterology annual meeting, Advances in Inflammatory Bowel Diseases meeting organized by the Crohn’s and Colitis Foundation of America, European Crohn’s and Colitis Organization annual meeting and United European Gastroenterology Week] from 2012 to 2017 to identify additional abstracts on the topic. Finally, we contacted experts in the field to identify other unpublished studies. 2.3. Data abstraction and quality assessment Data on study-, participant-, disease- and treatment-related characteristics were abstracted onto a standardized form, by two authors [SS and JG] independently and discrepancies were resolved by consensus, referring to the original article, in consultation with a third reviewer. We focused only on outcomes in patients receiving active intervention. We abstracted data on the definitions of PNR, LOR and intolerance in included trials, definition of clinical remission or response, and rates of clinical remission [or response] in patients receiving active intervention across these strata. Two study investigators [SS and JG] independently rated the quality of included studies by using the Cochrane Risk of Bias Tool.15 2.4. Outcomes assessed The primary outcome measure was the proportion of patients achieving clinical remission in patients in different strata based on reason for discontinuation of index anti-TNF agent. Clinical remission was defined as Mayo Clinic Score [MCS] ≤2 with no individual subscore of >1 [for patients with UC], and Crohn’s disease activity index [CDAI] <150 [for patients with CD]; clinical response was defined as decline in CDAI by 100 points [CR-100], and was used if clinical remission was not reported. We a priori hypothesized that potential impact of prior anti-TNF exposure on response to second-line agent would be most apparent in trials of induction therapy, and may be less apparent in trials of maintenance therapy that re-randomized only responders to induction therapy. Hence, we performed subgroup analyses based on trial design [induction vs. maintenance therapy]. Other subgroups included: disease type [CD vs. UC] and class of second-line agent. 2.5. Statistical analysis We used the random-effects model described by DerSimonian and Laird to calculate relative risk [RR] (and 95% confidence interval [CI]) of achieving clinical remission from active intervention with second-line agent in patients who discontinued prior anti-TNF due to PNR [vs. intolerance], PNR [vs. LOR] and LOR [vs. intolerance].16 We assessed heterogeneity between study-specific estimates using the inconsistency index [I2], and used cut-offs of <30%, 30–59%, 60–75% and >75% to suggest low, moderate, substantial and considerable heterogeneity, respectively.17 Due to the small number of studies, a reliable assessment of publication bias could not be estimated. All analysis was performed using Comprehensive Meta-Analysis [CMA] version 2 [Biostat, Englewood, NJ]. 3. Results From 1866 unique studies identified using our search strategy, we included eight RCTs in patients with moderate–severe IBD with prior exposure to an anti-TNF, in which response to the second-line agent was stratified according to the reason for discontinuation of the index anti-TNF agent. These included six RCTs in patients with moderate–severe CD [GAIN,18 WELCOME,19 GEMINI-II Induction and Maintenance,20 GEMINI-III,10 CERTIFI,21 UNITI-17] and two RCTs in patients with moderate–severe UC [GEMINI I-Induction and Maintenance22,23] [Figure 1]. No trial of small molecules reported differences in response stratified according to the reason for discontinuation of the prior anti-TNF agent. Figure 1. View largeDownload slide Study selection flowsheet. Figure 1. View largeDownload slide Study selection flowsheet. The trial characteristics are summarized in Table 1. Across trials, the reason for discontinuation of the index anti-TNF agent was defined clinically, and not based on therapeutic drug monitoring. PNR was defined as lack of initial clinical response to the index anti-TNF agent, and was reported in a median of 43.1% [interquartile range (IQR), 28.9–49.6] of patients with prior anti-TNF exposure. Secondary LOR was defined based on loss of response in patients with initial response to the anti-TNF agent, and a median of 64.7% [IQR, 41.0–64.7] were classified as having LOR; of note, a higher proportion of patients in ustekinumab trials were deemed to have secondary LOR as compared with participants in trials of vedolizumab. Intolerance was defined based on unacceptable side effects due to the index anti-TNF agent, warranting discontinuation of therapy, and was reported in a median of 35.4% [IQR, 26.5–42.2] of participants [Table 2]. Across trials, some patients were classified in more than one category of reason for discontinuation of prior therapy, since this was often based on patient recall. Of note, trials of adalimumab and certolizumab pegol included only patients with LOR and intolerance, and excluded patients with PNR.18,19 Overall, these RCTs were at low risk of bias. Table 1. Trial characteristics focusing on patients with prior exposure to anti-TNF therapy. [Abbreviations: C = Control; CS = corticosteroids; CDAI-Crohn’s disease activity index; I = intervention; IS = immunomodulator; MCS = Mayo Clinic Score.] Trial name; active intervention, condition  Proportion of patients with prior exposure to anti- TNF agents [%]  Intervention vs. Control  Outcome; timing of assessment  Co-interventions [proportion on immunomodulators and corticosteroids]  Response to prior therapy  Overall proportion of patients achieving remission  Primary non-response  Secondary loss of response  Intolerance  GEMINI 1 INDUCTION;22,23Vedolizumab, ulcerative colitis  367/895 [41.0%]  I: 82/367 [22.3%] C: 63/367 [17.2%]  MCS<2 Week 6  I: CS 30/82 [36.6%], IS 5/82 [6.1%] C: CS 27/63 [42.9%], IS 6/63 [9.5%]  I: 44/82 [53.7%] C: 29/63 [46.0%]  I: 32/82 [39.0%] C: 26/63 [41.3%]  I: 13/82 [15.6%] C: 10/63 [15.9%]  I: 8/82 [9.8%] C: 2/63 [3.2%]  GEMINI 1 MAINTENANCE;22,23Vedolizumab, ulcerative colitis  121/345 [35.1%]  I: 83/121 [68.6%] C: 38/121 [31.4%]  MCS<2 Week 52  I: CS 32/83 [38.6%], IS 7/83 [8.4%] C: CS 16/38 [42.1%], IS 6/38 [15.8%]  I: 33/83 [39.8%] C: 19/38 [50.0%]  I: 34/83 [41.0%] C: 13/38 [34.2%]  I: 22/83 [26.5%] C: 10/38 [26.3%]  I: 30/83 [36.1%] C: 2/38 [5.3%]  GEMINI 2 & 3 INDUCTION;10,11,20 Vedolizumab, Crohn’s disease  960/1476 [65.0%]  I: 263/960 [27.4%] C: 227/960 [23.65]  CDAI < 150 Week 10  I: CS 94/263 [35.7%], IS 35/263 [13.3%] C: CS 85/227 [37.4%], IS 30/227 [13.2%]  I: 109/263 [41.4%] C: 110/227 [48.5%]  I: 143/263 [54.4%] C: 141/227 [62.1%]  I: 94/263 [35.7%] C: 95/227 [41.9%]  I: 52/263 [19.8%] C: 25/227 [11.0%]  GEMINI 2 MAINTENANCE;11,20 Vedolizumab; Crohn’s disease  237/445 [53.3%]  I: 159/237 [67.1%] C: 78/237 [32.9%]  CDAI < 150 Week 52  I: CS 65/159 [40.9%], IS 27/159 [17.0%] C: CS 29/78 [37.2%], IS 10/78 [12.8%]  I: 68/159 [42.8%] C: 35/78 [44.9%]  I: 94/159 [59.2%] C: 46/78 [59.0%]  I: 64/159 [40.3%] C: 30/78 [38.5%]  I: 44/159 [27.7%] C: 10/78 [12.8%]  CERTIFI;21 Ustekinumab, Crohn’s disease  526/526 [100.0%]  I: 394/526 [74.9%] C: 132/526 [25.1%]  CR-100 Week 6  I: CS 189/394 [48.0%], IS 96/394 [24.4%] C: CS 73/132 [55.3%], IS 30/132 [22.7%]  I: 116/394 [29.4%] C: 44/132 [33.3%]  I: 289/394 [73.4%] C: 91/132 [68.9%]  I: 135/394 [41.9%] C: 41/132 [31.1%]  I: 58/394 [14.7%] C: 14/132 [10.6%]  UNITI 1;7 Ustekinumab, Crohn’s disease  492/496 [99.2%]  I: 246/249 [98.85] C: 246/247 [99.6%]  CR-100 Week 6  I: CS 108/249 [43.4%], IS 78/249 [31.3%] C: CS 111/247 [44.9%], IS 81/247 [32.8%]  I: 72/249 [28.9%] C: 74/247  I: 171/249 [68.7%] C: 170/247 [35.2%]  I: 105/249 [42.2%] C: 87/247 [35.2%]  I: 46/249 [18.5%] C: 22/247 [8.9%]  GAIN;18 Adalimumab, Crohn’s disease  325/325 [100.0%]  I: 159/325 [48.9%] C: 166/325 [51.1%]  CDAI < 150 Week 4  I: CS 55/159 [34.6%], IS 73/159 [45.9%] C: CS 73/166 [44.0%], IS 85/166 [51.2%]  0 [0.0%]  I: 77/159 [48.4%] C: 87/166 [52.4%]  I: 95/159 [59.7%] C: 95/166 [57.2%]  I: 34/159 [21.4%] C: 12/166 [7.2%]  WELCOME;19 Certolizumab pegol, Crohn’s disease  539/539 [100.0%]  I: 539/539 [100.0%] C: 0 [0.0%]  CR-100 Week 6  I: CS 67/168 [39.9%], IS 77/168 [45.8%]  0 [0.0%]  I: 79/168 [47.0%]  I: 66/168 [39.3%]  I: 212/539 [39.3%]  Trial name; active intervention, condition  Proportion of patients with prior exposure to anti- TNF agents [%]  Intervention vs. Control  Outcome; timing of assessment  Co-interventions [proportion on immunomodulators and corticosteroids]  Response to prior therapy  Overall proportion of patients achieving remission  Primary non-response  Secondary loss of response  Intolerance  GEMINI 1 INDUCTION;22,23Vedolizumab, ulcerative colitis  367/895 [41.0%]  I: 82/367 [22.3%] C: 63/367 [17.2%]  MCS<2 Week 6  I: CS 30/82 [36.6%], IS 5/82 [6.1%] C: CS 27/63 [42.9%], IS 6/63 [9.5%]  I: 44/82 [53.7%] C: 29/63 [46.0%]  I: 32/82 [39.0%] C: 26/63 [41.3%]  I: 13/82 [15.6%] C: 10/63 [15.9%]  I: 8/82 [9.8%] C: 2/63 [3.2%]  GEMINI 1 MAINTENANCE;22,23Vedolizumab, ulcerative colitis  121/345 [35.1%]  I: 83/121 [68.6%] C: 38/121 [31.4%]  MCS<2 Week 52  I: CS 32/83 [38.6%], IS 7/83 [8.4%] C: CS 16/38 [42.1%], IS 6/38 [15.8%]  I: 33/83 [39.8%] C: 19/38 [50.0%]  I: 34/83 [41.0%] C: 13/38 [34.2%]  I: 22/83 [26.5%] C: 10/38 [26.3%]  I: 30/83 [36.1%] C: 2/38 [5.3%]  GEMINI 2 & 3 INDUCTION;10,11,20 Vedolizumab, Crohn’s disease  960/1476 [65.0%]  I: 263/960 [27.4%] C: 227/960 [23.65]  CDAI < 150 Week 10  I: CS 94/263 [35.7%], IS 35/263 [13.3%] C: CS 85/227 [37.4%], IS 30/227 [13.2%]  I: 109/263 [41.4%] C: 110/227 [48.5%]  I: 143/263 [54.4%] C: 141/227 [62.1%]  I: 94/263 [35.7%] C: 95/227 [41.9%]  I: 52/263 [19.8%] C: 25/227 [11.0%]  GEMINI 2 MAINTENANCE;11,20 Vedolizumab; Crohn’s disease  237/445 [53.3%]  I: 159/237 [67.1%] C: 78/237 [32.9%]  CDAI < 150 Week 52  I: CS 65/159 [40.9%], IS 27/159 [17.0%] C: CS 29/78 [37.2%], IS 10/78 [12.8%]  I: 68/159 [42.8%] C: 35/78 [44.9%]  I: 94/159 [59.2%] C: 46/78 [59.0%]  I: 64/159 [40.3%] C: 30/78 [38.5%]  I: 44/159 [27.7%] C: 10/78 [12.8%]  CERTIFI;21 Ustekinumab, Crohn’s disease  526/526 [100.0%]  I: 394/526 [74.9%] C: 132/526 [25.1%]  CR-100 Week 6  I: CS 189/394 [48.0%], IS 96/394 [24.4%] C: CS 73/132 [55.3%], IS 30/132 [22.7%]  I: 116/394 [29.4%] C: 44/132 [33.3%]  I: 289/394 [73.4%] C: 91/132 [68.9%]  I: 135/394 [41.9%] C: 41/132 [31.1%]  I: 58/394 [14.7%] C: 14/132 [10.6%]  UNITI 1;7 Ustekinumab, Crohn’s disease  492/496 [99.2%]  I: 246/249 [98.85] C: 246/247 [99.6%]  CR-100 Week 6  I: CS 108/249 [43.4%], IS 78/249 [31.3%] C: CS 111/247 [44.9%], IS 81/247 [32.8%]  I: 72/249 [28.9%] C: 74/247  I: 171/249 [68.7%] C: 170/247 [35.2%]  I: 105/249 [42.2%] C: 87/247 [35.2%]  I: 46/249 [18.5%] C: 22/247 [8.9%]  GAIN;18 Adalimumab, Crohn’s disease  325/325 [100.0%]  I: 159/325 [48.9%] C: 166/325 [51.1%]  CDAI < 150 Week 4  I: CS 55/159 [34.6%], IS 73/159 [45.9%] C: CS 73/166 [44.0%], IS 85/166 [51.2%]  0 [0.0%]  I: 77/159 [48.4%] C: 87/166 [52.4%]  I: 95/159 [59.7%] C: 95/166 [57.2%]  I: 34/159 [21.4%] C: 12/166 [7.2%]  WELCOME;19 Certolizumab pegol, Crohn’s disease  539/539 [100.0%]  I: 539/539 [100.0%] C: 0 [0.0%]  CR-100 Week 6  I: CS 67/168 [39.9%], IS 77/168 [45.8%]  0 [0.0%]  I: 79/168 [47.0%]  I: 66/168 [39.3%]  I: 212/539 [39.3%]  View Large Table 1. Trial characteristics focusing on patients with prior exposure to anti-TNF therapy. [Abbreviations: C = Control; CS = corticosteroids; CDAI-Crohn’s disease activity index; I = intervention; IS = immunomodulator; MCS = Mayo Clinic Score.] Trial name; active intervention, condition  Proportion of patients with prior exposure to anti- TNF agents [%]  Intervention vs. Control  Outcome; timing of assessment  Co-interventions [proportion on immunomodulators and corticosteroids]  Response to prior therapy  Overall proportion of patients achieving remission  Primary non-response  Secondary loss of response  Intolerance  GEMINI 1 INDUCTION;22,23Vedolizumab, ulcerative colitis  367/895 [41.0%]  I: 82/367 [22.3%] C: 63/367 [17.2%]  MCS<2 Week 6  I: CS 30/82 [36.6%], IS 5/82 [6.1%] C: CS 27/63 [42.9%], IS 6/63 [9.5%]  I: 44/82 [53.7%] C: 29/63 [46.0%]  I: 32/82 [39.0%] C: 26/63 [41.3%]  I: 13/82 [15.6%] C: 10/63 [15.9%]  I: 8/82 [9.8%] C: 2/63 [3.2%]  GEMINI 1 MAINTENANCE;22,23Vedolizumab, ulcerative colitis  121/345 [35.1%]  I: 83/121 [68.6%] C: 38/121 [31.4%]  MCS<2 Week 52  I: CS 32/83 [38.6%], IS 7/83 [8.4%] C: CS 16/38 [42.1%], IS 6/38 [15.8%]  I: 33/83 [39.8%] C: 19/38 [50.0%]  I: 34/83 [41.0%] C: 13/38 [34.2%]  I: 22/83 [26.5%] C: 10/38 [26.3%]  I: 30/83 [36.1%] C: 2/38 [5.3%]  GEMINI 2 & 3 INDUCTION;10,11,20 Vedolizumab, Crohn’s disease  960/1476 [65.0%]  I: 263/960 [27.4%] C: 227/960 [23.65]  CDAI < 150 Week 10  I: CS 94/263 [35.7%], IS 35/263 [13.3%] C: CS 85/227 [37.4%], IS 30/227 [13.2%]  I: 109/263 [41.4%] C: 110/227 [48.5%]  I: 143/263 [54.4%] C: 141/227 [62.1%]  I: 94/263 [35.7%] C: 95/227 [41.9%]  I: 52/263 [19.8%] C: 25/227 [11.0%]  GEMINI 2 MAINTENANCE;11,20 Vedolizumab; Crohn’s disease  237/445 [53.3%]  I: 159/237 [67.1%] C: 78/237 [32.9%]  CDAI < 150 Week 52  I: CS 65/159 [40.9%], IS 27/159 [17.0%] C: CS 29/78 [37.2%], IS 10/78 [12.8%]  I: 68/159 [42.8%] C: 35/78 [44.9%]  I: 94/159 [59.2%] C: 46/78 [59.0%]  I: 64/159 [40.3%] C: 30/78 [38.5%]  I: 44/159 [27.7%] C: 10/78 [12.8%]  CERTIFI;21 Ustekinumab, Crohn’s disease  526/526 [100.0%]  I: 394/526 [74.9%] C: 132/526 [25.1%]  CR-100 Week 6  I: CS 189/394 [48.0%], IS 96/394 [24.4%] C: CS 73/132 [55.3%], IS 30/132 [22.7%]  I: 116/394 [29.4%] C: 44/132 [33.3%]  I: 289/394 [73.4%] C: 91/132 [68.9%]  I: 135/394 [41.9%] C: 41/132 [31.1%]  I: 58/394 [14.7%] C: 14/132 [10.6%]  UNITI 1;7 Ustekinumab, Crohn’s disease  492/496 [99.2%]  I: 246/249 [98.85] C: 246/247 [99.6%]  CR-100 Week 6  I: CS 108/249 [43.4%], IS 78/249 [31.3%] C: CS 111/247 [44.9%], IS 81/247 [32.8%]  I: 72/249 [28.9%] C: 74/247  I: 171/249 [68.7%] C: 170/247 [35.2%]  I: 105/249 [42.2%] C: 87/247 [35.2%]  I: 46/249 [18.5%] C: 22/247 [8.9%]  GAIN;18 Adalimumab, Crohn’s disease  325/325 [100.0%]  I: 159/325 [48.9%] C: 166/325 [51.1%]  CDAI < 150 Week 4  I: CS 55/159 [34.6%], IS 73/159 [45.9%] C: CS 73/166 [44.0%], IS 85/166 [51.2%]  0 [0.0%]  I: 77/159 [48.4%] C: 87/166 [52.4%]  I: 95/159 [59.7%] C: 95/166 [57.2%]  I: 34/159 [21.4%] C: 12/166 [7.2%]  WELCOME;19 Certolizumab pegol, Crohn’s disease  539/539 [100.0%]  I: 539/539 [100.0%] C: 0 [0.0%]  CR-100 Week 6  I: CS 67/168 [39.9%], IS 77/168 [45.8%]  0 [0.0%]  I: 79/168 [47.0%]  I: 66/168 [39.3%]  I: 212/539 [39.3%]  Trial name; active intervention, condition  Proportion of patients with prior exposure to anti- TNF agents [%]  Intervention vs. Control  Outcome; timing of assessment  Co-interventions [proportion on immunomodulators and corticosteroids]  Response to prior therapy  Overall proportion of patients achieving remission  Primary non-response  Secondary loss of response  Intolerance  GEMINI 1 INDUCTION;22,23Vedolizumab, ulcerative colitis  367/895 [41.0%]  I: 82/367 [22.3%] C: 63/367 [17.2%]  MCS<2 Week 6  I: CS 30/82 [36.6%], IS 5/82 [6.1%] C: CS 27/63 [42.9%], IS 6/63 [9.5%]  I: 44/82 [53.7%] C: 29/63 [46.0%]  I: 32/82 [39.0%] C: 26/63 [41.3%]  I: 13/82 [15.6%] C: 10/63 [15.9%]  I: 8/82 [9.8%] C: 2/63 [3.2%]  GEMINI 1 MAINTENANCE;22,23Vedolizumab, ulcerative colitis  121/345 [35.1%]  I: 83/121 [68.6%] C: 38/121 [31.4%]  MCS<2 Week 52  I: CS 32/83 [38.6%], IS 7/83 [8.4%] C: CS 16/38 [42.1%], IS 6/38 [15.8%]  I: 33/83 [39.8%] C: 19/38 [50.0%]  I: 34/83 [41.0%] C: 13/38 [34.2%]  I: 22/83 [26.5%] C: 10/38 [26.3%]  I: 30/83 [36.1%] C: 2/38 [5.3%]  GEMINI 2 & 3 INDUCTION;10,11,20 Vedolizumab, Crohn’s disease  960/1476 [65.0%]  I: 263/960 [27.4%] C: 227/960 [23.65]  CDAI < 150 Week 10  I: CS 94/263 [35.7%], IS 35/263 [13.3%] C: CS 85/227 [37.4%], IS 30/227 [13.2%]  I: 109/263 [41.4%] C: 110/227 [48.5%]  I: 143/263 [54.4%] C: 141/227 [62.1%]  I: 94/263 [35.7%] C: 95/227 [41.9%]  I: 52/263 [19.8%] C: 25/227 [11.0%]  GEMINI 2 MAINTENANCE;11,20 Vedolizumab; Crohn’s disease  237/445 [53.3%]  I: 159/237 [67.1%] C: 78/237 [32.9%]  CDAI < 150 Week 52  I: CS 65/159 [40.9%], IS 27/159 [17.0%] C: CS 29/78 [37.2%], IS 10/78 [12.8%]  I: 68/159 [42.8%] C: 35/78 [44.9%]  I: 94/159 [59.2%] C: 46/78 [59.0%]  I: 64/159 [40.3%] C: 30/78 [38.5%]  I: 44/159 [27.7%] C: 10/78 [12.8%]  CERTIFI;21 Ustekinumab, Crohn’s disease  526/526 [100.0%]  I: 394/526 [74.9%] C: 132/526 [25.1%]  CR-100 Week 6  I: CS 189/394 [48.0%], IS 96/394 [24.4%] C: CS 73/132 [55.3%], IS 30/132 [22.7%]  I: 116/394 [29.4%] C: 44/132 [33.3%]  I: 289/394 [73.4%] C: 91/132 [68.9%]  I: 135/394 [41.9%] C: 41/132 [31.1%]  I: 58/394 [14.7%] C: 14/132 [10.6%]  UNITI 1;7 Ustekinumab, Crohn’s disease  492/496 [99.2%]  I: 246/249 [98.85] C: 246/247 [99.6%]  CR-100 Week 6  I: CS 108/249 [43.4%], IS 78/249 [31.3%] C: CS 111/247 [44.9%], IS 81/247 [32.8%]  I: 72/249 [28.9%] C: 74/247  I: 171/249 [68.7%] C: 170/247 [35.2%]  I: 105/249 [42.2%] C: 87/247 [35.2%]  I: 46/249 [18.5%] C: 22/247 [8.9%]  GAIN;18 Adalimumab, Crohn’s disease  325/325 [100.0%]  I: 159/325 [48.9%] C: 166/325 [51.1%]  CDAI < 150 Week 4  I: CS 55/159 [34.6%], IS 73/159 [45.9%] C: CS 73/166 [44.0%], IS 85/166 [51.2%]  0 [0.0%]  I: 77/159 [48.4%] C: 87/166 [52.4%]  I: 95/159 [59.7%] C: 95/166 [57.2%]  I: 34/159 [21.4%] C: 12/166 [7.2%]  WELCOME;19 Certolizumab pegol, Crohn’s disease  539/539 [100.0%]  I: 539/539 [100.0%] C: 0 [0.0%]  CR-100 Week 6  I: CS 67/168 [39.9%], IS 77/168 [45.8%]  0 [0.0%]  I: 79/168 [47.0%]  I: 66/168 [39.3%]  I: 212/539 [39.3%]  View Large Table 2. Definition of primary non-response, secondary loss of response and intolerance to index anti-TNF agent in included trials. [*Signs and symptoms of active disease = lack of improvement or worsening in stool frequency, rectal bleeding, fever, abdominal pain draining fistula or increase/initiation of antidiarrheal medication.] Trial Acronym  Primary non-response or Inadequate Response  Secondary loss of response  Intolerance  GEMINI 1, 2 & 3  *Signs and symptoms of persistently active disease despite at least one 4 week induction regimen of infliximab[5mg/kg IV, 2doses, > 2 weeks apart],adalimumab [one 80mg dose followed by one 40mg dose 2 weeks apart] or certolizumab [400mg 2doses> 2 weeks apart]  Recurrence of symptoms during maintenance in a patient who had previously benefited from treatment  Patients experiencing treatment related toxicity [infusion reaction, psoriasiform skin lesion, demyelination, congestive heart failure, infection]  CERTIFI UNITI  Presence of at least 1 active sign/symptom* despite infliximab[2 or 3 doses of 5mg/kg], adalimumab [160mg dose followed by 80mg or 80 mg dose followed by 40mg] or certolizumab [2 or 3 doses of 400mg] and have documentation available to investigator  Recurrence of at least 1 sign/symptom after patients initially respond to induction therapy and have received at least 2 maintenance doses of infliximab 5mg/ kg, adalimumab 40mg/kg or certolizumab 400mg and have documentation available to investigator  Significant acute [< 24 hours] or delayed infusion [>24 & <15 days] reaction or injection site reaction [within 24 hours] and have documentation available to investigator  GAIN WELCOME  Not included  Patients have history of an initial response to infliximab, as defined by the investigator; had received at least 2 doses of infliximab of 5 mg/kg of body weight or more every 8 weeks; and had lacked improvement or had worsening signs and symptoms  History of discontinuing infliximab treatment because of a clinically significant acute or delayed infusion reaction.  Trial Acronym  Primary non-response or Inadequate Response  Secondary loss of response  Intolerance  GEMINI 1, 2 & 3  *Signs and symptoms of persistently active disease despite at least one 4 week induction regimen of infliximab[5mg/kg IV, 2doses, > 2 weeks apart],adalimumab [one 80mg dose followed by one 40mg dose 2 weeks apart] or certolizumab [400mg 2doses> 2 weeks apart]  Recurrence of symptoms during maintenance in a patient who had previously benefited from treatment  Patients experiencing treatment related toxicity [infusion reaction, psoriasiform skin lesion, demyelination, congestive heart failure, infection]  CERTIFI UNITI  Presence of at least 1 active sign/symptom* despite infliximab[2 or 3 doses of 5mg/kg], adalimumab [160mg dose followed by 80mg or 80 mg dose followed by 40mg] or certolizumab [2 or 3 doses of 400mg] and have documentation available to investigator  Recurrence of at least 1 sign/symptom after patients initially respond to induction therapy and have received at least 2 maintenance doses of infliximab 5mg/ kg, adalimumab 40mg/kg or certolizumab 400mg and have documentation available to investigator  Significant acute [< 24 hours] or delayed infusion [>24 & <15 days] reaction or injection site reaction [within 24 hours] and have documentation available to investigator  GAIN WELCOME  Not included  Patients have history of an initial response to infliximab, as defined by the investigator; had received at least 2 doses of infliximab of 5 mg/kg of body weight or more every 8 weeks; and had lacked improvement or had worsening signs and symptoms  History of discontinuing infliximab treatment because of a clinically significant acute or delayed infusion reaction.  View Large Table 2. Definition of primary non-response, secondary loss of response and intolerance to index anti-TNF agent in included trials. [*Signs and symptoms of active disease = lack of improvement or worsening in stool frequency, rectal bleeding, fever, abdominal pain draining fistula or increase/initiation of antidiarrheal medication.] Trial Acronym  Primary non-response or Inadequate Response  Secondary loss of response  Intolerance  GEMINI 1, 2 & 3  *Signs and symptoms of persistently active disease despite at least one 4 week induction regimen of infliximab[5mg/kg IV, 2doses, > 2 weeks apart],adalimumab [one 80mg dose followed by one 40mg dose 2 weeks apart] or certolizumab [400mg 2doses> 2 weeks apart]  Recurrence of symptoms during maintenance in a patient who had previously benefited from treatment  Patients experiencing treatment related toxicity [infusion reaction, psoriasiform skin lesion, demyelination, congestive heart failure, infection]  CERTIFI UNITI  Presence of at least 1 active sign/symptom* despite infliximab[2 or 3 doses of 5mg/kg], adalimumab [160mg dose followed by 80mg or 80 mg dose followed by 40mg] or certolizumab [2 or 3 doses of 400mg] and have documentation available to investigator  Recurrence of at least 1 sign/symptom after patients initially respond to induction therapy and have received at least 2 maintenance doses of infliximab 5mg/ kg, adalimumab 40mg/kg or certolizumab 400mg and have documentation available to investigator  Significant acute [< 24 hours] or delayed infusion [>24 & <15 days] reaction or injection site reaction [within 24 hours] and have documentation available to investigator  GAIN WELCOME  Not included  Patients have history of an initial response to infliximab, as defined by the investigator; had received at least 2 doses of infliximab of 5 mg/kg of body weight or more every 8 weeks; and had lacked improvement or had worsening signs and symptoms  History of discontinuing infliximab treatment because of a clinically significant acute or delayed infusion reaction.  Trial Acronym  Primary non-response or Inadequate Response  Secondary loss of response  Intolerance  GEMINI 1, 2 & 3  *Signs and symptoms of persistently active disease despite at least one 4 week induction regimen of infliximab[5mg/kg IV, 2doses, > 2 weeks apart],adalimumab [one 80mg dose followed by one 40mg dose 2 weeks apart] or certolizumab [400mg 2doses> 2 weeks apart]  Recurrence of symptoms during maintenance in a patient who had previously benefited from treatment  Patients experiencing treatment related toxicity [infusion reaction, psoriasiform skin lesion, demyelination, congestive heart failure, infection]  CERTIFI UNITI  Presence of at least 1 active sign/symptom* despite infliximab[2 or 3 doses of 5mg/kg], adalimumab [160mg dose followed by 80mg or 80 mg dose followed by 40mg] or certolizumab [2 or 3 doses of 400mg] and have documentation available to investigator  Recurrence of at least 1 sign/symptom after patients initially respond to induction therapy and have received at least 2 maintenance doses of infliximab 5mg/ kg, adalimumab 40mg/kg or certolizumab 400mg and have documentation available to investigator  Significant acute [< 24 hours] or delayed infusion [>24 & <15 days] reaction or injection site reaction [within 24 hours] and have documentation available to investigator  GAIN WELCOME  Not included  Patients have history of an initial response to infliximab, as defined by the investigator; had received at least 2 doses of infliximab of 5 mg/kg of body weight or more every 8 weeks; and had lacked improvement or had worsening signs and symptoms  History of discontinuing infliximab treatment because of a clinically significant acute or delayed infusion reaction.  View Large 3.1. Primary non-response vs. intolerance On meta-analysis of the active intervention arm of six RCTs, patients with PNR to the index anti-TNF agent were 24% less likely to achieve remission with the second-line biologic agent as compared with patients who discontinued the index anti-TNF therapy due to intolerance [RR, 0.76; 95% CI, 0.61–0.96] with minimal heterogeneity [I2 = 18%] [Figure 2A]. These results were stable on subgroup analyses based on study design, class of second-line intervention and disease type [Table 3]. Figure 2. View largeDownload slide Response to second-line biologic in patients who discontinued prior anti-TNF therapy due to [A] primary non-response vs. intolerance, [B] primary non-response vs. secondary loss of response and [C] secondary loss of response vs. intolerance. GEMINI 2 and 3, UNITI 1, CERTIFI, GAIN and WELCOME were conducted in patients with Crohn’s disease, and GEMINI 1 was conducted in patients with ulcerative colitis. Figure 2. View largeDownload slide Response to second-line biologic in patients who discontinued prior anti-TNF therapy due to [A] primary non-response vs. intolerance, [B] primary non-response vs. secondary loss of response and [C] secondary loss of response vs. intolerance. GEMINI 2 and 3, UNITI 1, CERTIFI, GAIN and WELCOME were conducted in patients with Crohn’s disease, and GEMINI 1 was conducted in patients with ulcerative colitis. Table 3. Subgroup analysis. Subgroups  Categories  RR  95% CI  I2  P-interaction  Prior primary non-response vs. intolerance  Overall  6 trials  0.76  0.61–0.96  18    Trial design  Induction [4]  0.74  0.60–0.92  0  0.93  Maintenance [2]  0.77  0.37–1.59  70  Second-line agent  VDZ [4]  0.83  0.54–1.28  40  0.46  UST [2]  0.69  0.55–0.88  0  Disease type  CD [4]  0.82  0.64–1.07  26  0.13  UC [2]  0.52  0.30–0.89  0  Prior primary non-response vs. secondary loss of response  Overall  6 trials  0.88  0.64–1.21  52    Trial design  Induction [4]  0.73  0.56–0.97  27  0.037  Maintenance [2]  1.27  0.82–1.94  0  Second-line agent  VDZ [4]  1.16  0.85–1.58  0  0.002  UST [2]  0.64  0.52–0.80  0  Disease type  CD [4]  0.82  0.59–1.15  59  0.22  UC [2]  1.36  0.66–2.79  0  Secondary loss of response vs. intolerance  Overall  8 trials  0.96  0.78–1.18  35    Trial design  Induction [6]  1.06  0.91–1.23  0  0.27  Maintenance [2]  0.58  0.20–1.67  82  Second-line agent  VDZ [4]  0.71  0.41–1.22  58  0.34  UST [2]  1.08  0.91–1.30  0  Anti-TNF [2]  1.03  0.74–1.43  0  Disease type  CD [6]  1.06  0.91–1.22  0  0.002  UC [2]  0.36  0.18–0.70  0  Subgroups  Categories  RR  95% CI  I2  P-interaction  Prior primary non-response vs. intolerance  Overall  6 trials  0.76  0.61–0.96  18    Trial design  Induction [4]  0.74  0.60–0.92  0  0.93  Maintenance [2]  0.77  0.37–1.59  70  Second-line agent  VDZ [4]  0.83  0.54–1.28  40  0.46  UST [2]  0.69  0.55–0.88  0  Disease type  CD [4]  0.82  0.64–1.07  26  0.13  UC [2]  0.52  0.30–0.89  0  Prior primary non-response vs. secondary loss of response  Overall  6 trials  0.88  0.64–1.21  52    Trial design  Induction [4]  0.73  0.56–0.97  27  0.037  Maintenance [2]  1.27  0.82–1.94  0  Second-line agent  VDZ [4]  1.16  0.85–1.58  0  0.002  UST [2]  0.64  0.52–0.80  0  Disease type  CD [4]  0.82  0.59–1.15  59  0.22  UC [2]  1.36  0.66–2.79  0  Secondary loss of response vs. intolerance  Overall  8 trials  0.96  0.78–1.18  35    Trial design  Induction [6]  1.06  0.91–1.23  0  0.27  Maintenance [2]  0.58  0.20–1.67  82  Second-line agent  VDZ [4]  0.71  0.41–1.22  58  0.34  UST [2]  1.08  0.91–1.30  0  Anti-TNF [2]  1.03  0.74–1.43  0  Disease type  CD [6]  1.06  0.91–1.22  0  0.002  UC [2]  0.36  0.18–0.70  0  View Large Table 3. Subgroup analysis. Subgroups  Categories  RR  95% CI  I2  P-interaction  Prior primary non-response vs. intolerance  Overall  6 trials  0.76  0.61–0.96  18    Trial design  Induction [4]  0.74  0.60–0.92  0  0.93  Maintenance [2]  0.77  0.37–1.59  70  Second-line agent  VDZ [4]  0.83  0.54–1.28  40  0.46  UST [2]  0.69  0.55–0.88  0  Disease type  CD [4]  0.82  0.64–1.07  26  0.13  UC [2]  0.52  0.30–0.89  0  Prior primary non-response vs. secondary loss of response  Overall  6 trials  0.88  0.64–1.21  52    Trial design  Induction [4]  0.73  0.56–0.97  27  0.037  Maintenance [2]  1.27  0.82–1.94  0  Second-line agent  VDZ [4]  1.16  0.85–1.58  0  0.002  UST [2]  0.64  0.52–0.80  0  Disease type  CD [4]  0.82  0.59–1.15  59  0.22  UC [2]  1.36  0.66–2.79  0  Secondary loss of response vs. intolerance  Overall  8 trials  0.96  0.78–1.18  35    Trial design  Induction [6]  1.06  0.91–1.23  0  0.27  Maintenance [2]  0.58  0.20–1.67  82  Second-line agent  VDZ [4]  0.71  0.41–1.22  58  0.34  UST [2]  1.08  0.91–1.30  0  Anti-TNF [2]  1.03  0.74–1.43  0  Disease type  CD [6]  1.06  0.91–1.22  0  0.002  UC [2]  0.36  0.18–0.70  0  Subgroups  Categories  RR  95% CI  I2  P-interaction  Prior primary non-response vs. intolerance  Overall  6 trials  0.76  0.61–0.96  18    Trial design  Induction [4]  0.74  0.60–0.92  0  0.93  Maintenance [2]  0.77  0.37–1.59  70  Second-line agent  VDZ [4]  0.83  0.54–1.28  40  0.46  UST [2]  0.69  0.55–0.88  0  Disease type  CD [4]  0.82  0.64–1.07  26  0.13  UC [2]  0.52  0.30–0.89  0  Prior primary non-response vs. secondary loss of response  Overall  6 trials  0.88  0.64–1.21  52    Trial design  Induction [4]  0.73  0.56–0.97  27  0.037  Maintenance [2]  1.27  0.82–1.94  0  Second-line agent  VDZ [4]  1.16  0.85–1.58  0  0.002  UST [2]  0.64  0.52–0.80  0  Disease type  CD [4]  0.82  0.59–1.15  59  0.22  UC [2]  1.36  0.66–2.79  0  Secondary loss of response vs. intolerance  Overall  8 trials  0.96  0.78–1.18  35    Trial design  Induction [6]  1.06  0.91–1.23  0  0.27  Maintenance [2]  0.58  0.20–1.67  82  Second-line agent  VDZ [4]  0.71  0.41–1.22  58  0.34  UST [2]  1.08  0.91–1.30  0  Anti-TNF [2]  1.03  0.74–1.43  0  Disease type  CD [6]  1.06  0.91–1.22  0  0.002  UC [2]  0.36  0.18–0.70  0  View Large 3.2. Primary non-response vs. secondary loss of response There was no significant difference in response to second-line non-TNF biologic in patients who discontinued prior anti-TNF due to PNR vs. LOR [RR, 0.88; 95% CI, 0.64–1.21], with considerable heterogeneity [I2 = 52%] [Figure 2B]. On analyses stratified by study design, trials of induction therapy demonstrated a lower likelihood of achieving remission with the second-line non-TNF biologic in patients with prior PNR to anti-TNF as compared with patients with secondary LOR to the anti-TNF [RR, 0.73; 95% CI, 0.56–0.97], with minimal heterogeneity [I2 = 27%]; no significant difference in response was observed in trials of maintenance therapy, in which only patients with response to induction therapy were included [Table 3]. The risk of achieving remission among patients with prior PNR to anti-TNF as compared with patients with prior LOR was observed only in patients treated with ustekinumab [RR, 0.53; 95% CI, 0.39–0.71]; there was no difference in response to vedolizumab in patients with prior PNR or LOR [RR, 1.06; 95% CI, 0.60–1.87]. These results were stable when stratified based on disease type [CD or UC]. 3.3. Secondary loss of response vs. intolerance On meta-analysis of eight RCTs, there was no significant difference in response to the second-line biologic in patients who discontinued the prior anti-TNF due to LOR or intolerance [RR, 0.96; 95% CI, 0.78–1.18], with moderate heterogeneity [I2 = 35%] [Figure 2C]. Results were stable on subgroup analyses based on trial design and class of medication, though in trials of UC, patients with secondary LOR showed an inferior response as compared with patients who discontinued the primary anti-TNF due to intolerance [Table 3]. 3.4. Publication bias Due to the small number of studies, formal assessment of funnel plot asymmetry was not performed. 4. Discussion With availability of newer non-TNF biologics and small molecules for treating patients with IBD, assessment of likelihood of response to second-line agents in patients who have previously been exposed to anti-TNF agents is important in order to position subsequent therapies. In this systematic review and meta-analysis, we made several key observations: [a] patients with prior PNR to anti-TNF agents were less likely to respond to treatment with another non-anti-TNF biologic, as compared with patients who discontinued their index anti-TNF agent due to intolerance; [b] patients with prior PNR to anti-TNF agents may have been less likely to achieve induction of remission with another non–anti-TNF biologic as compared with patients with LOR; once they achieve response to induction therapy, however, the likelihood of maintaining response was comparable; [c] patients treated with ustekinumab, which blocks IL-12/23, but not those treated with vedolizumab, which blocks lymphocyte recruitment to the gastrointestinal tract, were less likely to respond to induction therapy if they had PNR to anti-TNF agents as compared with prior secondary LOR to anti-TNF agents. Overall, these results suggest that patients with IBD with PNR to anti-TNF agents may be intrinsically more difficult to treat with second-line biologics. These findings directly inform clinical practice. In particular, patients with PNR to initial anti-TNF therapy attributed to high drug clearance are potentially at increased risk of rapid drug clearance on second-line biologics, and attempts should be made to minimize this risk [such as through use of higher doses for induction therapy, proactive drug monitoring during the induction phase to allow early optimization, and combination therapy with immunomodulators to minimize the risk of immunogenicity]. Trials of all approved biologic therapies, as well as those in development, suggest lower response rates in the subset of patients with prior exposure to anti-TNF agents as compared with biologic-naïve patients.7,10,11,24 However, within this group of anti-TNF–exposed patients, we observed that response rates vary depending on the specific reason for discontinuing the anti-TNF therapy. Intuitively, we observed that patients with PNR to anti-TNFs were less likely to respond to a second line agent, as compared with patients who were intolerant to therapies. Recent observational studies have confirmed that patients with PNR to first-line therapy infliximab have a >50% chance of progressing to surgery over the next 1 year.25,26 It is not surprising that patients with intolerance to one medication, may tolerate the next medication better, without compromising possible efficacy of the second-line therapy. The observation that patients with PNR to anti-TNF agents may not respond as well as patients with secondary LOR may be due to a combination of pharmacokinetic as well as pharmacodynamic properties of biologics. An exposure–response relationship has been observed with all biologics in IBD, wherein patients with higher drug concentrations respond better than patients with low trough levels.7,12,27–29 Patients may have PNR to first-line anti-TNF agents either because of inadequate trough concentrations due to rapid, non–immune-mediated drug clearance [pharmacokinetics] or due to mechanistic failure wherein patients fail to respond despite adequate trough concentrations, perhaps because the disease is mediated through a non–TNF-mediated pathway.4 Unfortunately, the included trials did not classify patients as having PNR based on trough concentrations, but rather based on clinical history of prior response or not. Patients with PNR to the first-line anti-TNF due to inadequate trough concentrations may be prone to PNR to the second-line biologic due to the same factors that led to rapid drug clearance with the first agent. Available population pharmacokinetic analyses for biologics in IBD for infliximab, certolizumab pegol and vedolizumab have identified similar covariates to be associated with drug clearance, such as body weight, albumin concentration and anti-drug antibodies.30–33 In these patients with pharmacokinetically determined PNR, it is possible that upfront dose optimization with the second-line anti-TNF or non-TNF biologic may help overcome the pharmacokinetic problem. On the other hand, patients mechanistically resistant to a cytokine-based treatment focusing on blocking TNF may be more difficult to treat with any other agent, resulting in lower response to a second-line biologic. It is conceivable that with use of therapeutic drug monitoring [TDM] and more accurate classification of PNR vs. LOR based on adequacy of drug exposure these distinctions in response to second-line therapy would be exaggerated.4 The observation of a differential response in patients with prior PNR and LOR to anti-TNF, with ustekinumab but not vedolizumab was unexpected. These findings are hypothesis-generating, suggesting that though there may be biologically distinct signaling pathways, functionally there are shared cellular and downstream immunologic effects such that patients who fail to respond to one TNF blockade may be less likely to respond to another IL-12/23 blockade. This may inform personalized clinical decision-making, providing a framework and rationale for switching between therapies for IBD based on the downstream immunologic effects. It is very important to note that these results reflect lack of difference in response regardless of PNR vs. LOR in a cohort of patients treated with vedolizumab [which may be due to the overall low response rate observed in vedolizumab-treated patients with prior anti-TNF exposure], and does not represent differences in response rate between ustekinumab and vedolizumab within a cohort of patients with PNR; in the absence of head-to-head comparative trials, the latter data is not available. Prospective mechanistic studies interrogating these hypotheses, in which drug clearance and drug concentrations are incorporated into the analyses, are warranted. There are several limitations to our study. First, response rate stratified by reason for discontinuation of prior anti-TNF therapy was inconsistently reported in trials. When reported, some patients within trials were classified in more than one category for reason for discontinuation of therapy, since this was primarily based on patient recall, not objective confirmation. This may also introduce bias, since exposure types are not explicitly determined. There were no trials evaluating response to second-line anti-TNF agent in patients with a PNR to the first anti-TNF agent. As noted above, it is conceivable that patients with reported PNR to the first anti-TNF agent may respond well to a second anti-TNF agent, particularly those with a pharmacokinetically determined PNR, due to differences in affinity-dependent target-mediated drug disposition between different anti-TNF agents and early appearance anti-drug antibodies. Second, in the absence of individual participant–level data, only unadjusted analysis could be performed, and we were unable to account for other potential differences in patients with prior PNR, LOR or intolerance. Third, for our overall analysis, we combined trials of UC and CD; conceptually, we believed this would be acceptable for our study question, since specific IBD phenotype–related factors are unlikely to influence response to second-line biologic based on reason for discontinuation of primary anti-TNF therapy. There were differences in definition and time point of assessment of outcome measure, and trials of induction and maintenance therapy were combined for primary analysis. However, we confirmed our findings on relevant subgroup analyses. Fourth, only clinical outcomes [remission/response] were reported, and endoscopic or biochemical response/remission rates stratified according to reason for primary anti-TNF failure were not reported in trials. Along the same lines, since objective confirmation of active inflammatory disease was not warranted at enrollment in all trials, it is possible that a subset of patients [patients with non-inflammatory etiology of symptoms, such as patients with irritable bowel syndrome, fibrostenotic disease, etc.] characterized as having PNR to anti-TNF agents may not be responsive to any IBD-related therapy. In conclusion, we observed considerable differences in response to the second-line biologic, based on reason for discontinuation of the primary anti-TNF therapy. Patients with prior PNR were less likely to respond to second-line therapy as compared with patients with prior LOR or intolerance. Future studies are warranted to confirm these findings in clinical practice, particularly with classification of PNR and LOR based on TDM. Mechanistic studies exploring potential reasons for discrepancy in response to ustekinumab and vedolizumab in a subset of patients with PNR are warranted. Funding This study did not receive any direct funding Conflict of Interest Dr Singh is supported by the American College of Gastroenterology, the Crohn’s and Colitis Foundation and Pfizer. Dr Boland has served as a consultant for Abbvie and has a research grant from Takeda and Janssen. Dr Vande Casteele has served as a consultant for Boehringer Ingelheim, UCB Pharma, Pfizer and Takeda. Dr Sandborn has served as a consultant and received research funding from Janssen, Abbvie, UCB Pharma, Takeda, and Pfizer. Author Contributions Study concept and design: SS Acquisition of data: SS, JG Analysis and interpretation of data: SS, JG Drafting of the manuscript: SS, JG Critical revision of the manuscript for important intellectual content: BSB, NVC, WJS Approval of the final manuscript: SS, JG, BSB, NVC, WJS Guarantor of the article: SS References 1. Nielsen OH, Ainsworth MA. Tumor necrosis factor inhibitors for inflammatory bowel disease. N Engl J Med  2013; 369: 754– 62. Google Scholar CrossRef Search ADS PubMed  2. Singh S, Pardi DS. Update on anti-tumor necrosis factor agents in Crohn disease. Gastroenterol Clin North Am  2014; 43: 457– 78. Google Scholar CrossRef Search ADS PubMed  3. Sandborn WJ, Rutgeerts P, Feagan BGet al.   Colectomy rate comparison after treatment of ulcerative colitis with placebo or infliximab. Gastroenterology  2009; 137: 1250– 60; quiz 1520. Google Scholar CrossRef Search ADS PubMed  4. Papamichael K, Gils A, Rutgeerts Pet al.   Role for therapeutic drug monitoring during induction therapy with TNF antagonists in IBD: evolution in the definition and management of primary nonresponse. Inflamm Bowel Dis  2015; 21: 182– 97. Google Scholar CrossRef Search ADS PubMed  5. Gisbert JP, Marín AC, McNicholl AG, Chaparro M. 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. Google Scholar CrossRef Search ADS PubMed  6. Sandborn WJ, Su C, Sands BEet al.  ; OCTAVE Induction 1, OCTAVE Induction 2, and OCTAVE Sustain Investigators. Tofacitinib as induction and maintenance therapy for ulcerative colitis. N Engl J Med  2017; 376: 1723– 36. Google Scholar CrossRef Search ADS PubMed  7. Feagan BG, Sandborn WJ, Gasink Cet al.  ; UNITI–IM-UNITI Study Group. Ustekinumab as induction and maintenance therapy for Crohn’s disease. N Engl J Med  2016; 375: 1946– 60. Google Scholar CrossRef Search ADS PubMed  8. Feagan BG, Sandborn WJ, D’Haens Get al.   Induction therapy with the selective interleukin-23 inhibitor risankizumab in patients with moderate-to-severe Crohn’s disease: a randomised, double-blind, placebo-controlled phase 2 study. Lancet  2017; 389: 1699– 709. Google Scholar CrossRef Search ADS PubMed  9. Sandborn WJ, Feagan BG, Wolf DCet al.  ; TOUCHSTONE Study Group. Ozanimod induction and maintenance treatment for ulcerative colitis. N Engl J Med  2016; 374: 1754– 62. Google Scholar CrossRef Search ADS PubMed  10. Sands BE, Feagan BG, Rutgeerts Pet al.   Effects of vedolizumab induction therapy for patients with Crohn’s disease in whom tumor necrosis factor antagonist treatment failed. Gastroenterology  2014; 147: 618– 627.e3. Google Scholar CrossRef Search ADS PubMed  11. Sands BE, Sandborn WJ, Van Assche Get al.   Vedolizumab as induction and maintenance therapy for Crohn’s disease in patients naïve to or who have failed tumor necrosis factor antagonist therapy. Inflamm Bowel Dis  2017; 23: 97– 106. Google Scholar CrossRef Search ADS PubMed  12. Vande Casteele N, Herfarth H, Katz J, Falck-Ytter Y, Singh S. American Gastroenterological Association Institute technical review on the role of therapeutic drug monitoring in the management of inflammatory bowel diseases. Gastroenterology  2017; 153: 835– 857.e6. Google Scholar CrossRef Search ADS PubMed  13. Feuerstein JD, Nguyen GC, Kupfer SS, Falck-Ytter Y, Singh S; American Gastroenterological Association Institute Clinical Guidelines Committee. American Gastroenterological Association Institute guideline on therapeutic drug monitoring in inflammatory bowel disease. Gastroenterology  2017; 153: 827– 34. Google Scholar CrossRef Search ADS PubMed  14. Liberati A, Altman DG, Tetzlaff Jet al.   The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ  2009; 339: b2700. Google Scholar CrossRef Search ADS PubMed  15. Higgins JPT, Altman DG, Sterne JAC. Assessing risk of bias in included studies . In: Higgins JPT Green S, editors. Cochrane Handbook for Systematic Reviews of Interventions, Version 5.1.0  [updated March 2011]. The Cochrane Collaboration; 2011. http://handbook.cochrane.org. Accessed July 1, 2017. 16. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials  1986; 7: 177– 88. Google Scholar CrossRef Search ADS PubMed  17. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ  2003; 327: 557– 60. Google Scholar CrossRef Search ADS PubMed  18. Sandborn WJ, Rutgeerts P, Enns Ret al.   Adalimumab induction therapy for Crohn disease previously treated with infliximab: a randomized trial. Ann Intern Med  2007; 146: 829– 38. Google Scholar CrossRef Search ADS PubMed  19. Sandborn WJ, Vermeire S, D’Haens GRet al.   WELCOME: a randomized, double-blind, controlled trial comparing certolizumab pegol 400 mg every 2 weeks with every 4 weeks for maintenance of response and remission in patients with moderate to severe Crohn’s disease with secondary failure to infliximab. Gastroenterology  2009; 136: A27. 20. Sandborn WJ, Feagan BG, Rutgeerts Pet al.  ; GEMINI 2 Study Group. Vedolizumab as induction and maintenance therapy for Crohn’s disease. N Engl J Med  2013; 369: 711– 21. Google Scholar CrossRef Search ADS PubMed  21. Sandborn WJ, Gasink C, Gao LLet al.  ; CERTIFI Study Group. Ustekinumab induction and maintenance therapy in refractory Crohn’s disease. N Engl J Med  2012; 367: 1519– 28. Google Scholar CrossRef Search ADS PubMed  22. Feagan BG, Rutgeerts P, Sands BEet al.  ; GEMINI 1 Study Group. Vedolizumab as induction and maintenance therapy for ulcerative colitis. N Engl J Med  2013; 369: 699– 710. Google Scholar CrossRef Search ADS PubMed  23. Feagan BG, Rubin DT, Danese Set al.   Efficacy of vedolizumab induction and maintenance therapy in patients with ulcerative colitis, regardless of prior exposure to tumor necrosis factor antagonists. Clin Gastroenterol Hepatol  2017; 15: 229– 39.e5. Google Scholar CrossRef Search ADS PubMed  24. Vermeire S, O’Byrne S, Keir Met al.   Etrolizumab as induction therapy for ulcerative colitis: a randomised, controlled, phase 2 trial. Lancet  2014; 384: 309– 18. Google Scholar CrossRef Search ADS PubMed  25. Fernández-Salazar L, Muñoz F, Barrio Jet al.   Infliximab in ulcerative colitis: real-life analysis of factors predicting treatment discontinuation due to lack of response or colectomy: ECIA (ACAD Colitis and Infliximab Study). Scand J Gastroenterol  2016; 51: 186– 95. Google Scholar CrossRef Search ADS PubMed  26. Papamichael K, Rivals-Lerebours O, Billiet Tet al.   Long-term outcome of patients with ulcerative colitis and primary non-response to infliximab. J Crohns Colitis  2016; 10: 1015– 23. Google Scholar CrossRef Search ADS PubMed  27. Rosario M, French JL, Dirks NLet al.   Exposure-efficacy relationships for vedolizumab induction therapy in patients with ulcerative colitis or Crohn’s disease. J Crohns Colitis  2017; 11: 921– 9. Google Scholar CrossRef Search ADS PubMed  28. Battat R, Kopylov U, Bessissow Tet al.   Association between ustekinumab trough concentrations and clinical, biomarker, and endoscopic outcomes in patients with Crohn’s disease. Clin Gastroenterol Hepatol  2017; 15: 1427– 1434.e2. Google Scholar CrossRef Search ADS PubMed  29. Adedokun OJ, Sandborn WJ, Feagan BGet al.   Association between serum concentration of infliximab and efficacy in adult patients with ulcerative colitis. Gastroenterology  2014; 147: 1296– 307.e5. Google Scholar CrossRef Search ADS PubMed  30. Brandse JF, Mould D, Smeekes Oet al.   A real-life population pharmacokinetic study reveals factors associated with clearance and immunogenicity of infliximab in inflammatory bowel disease. Inflamm Bowel Dis  2017; 23: 650– 60. Google Scholar CrossRef Search ADS PubMed  31. Paul S, Moreau AC, Del Tedesco Eet al.   Pharmacokinetics of adalimumab in inflammatory bowel diseases: a systematic review and meta-analysis. Inflamm Bowel Dis  2014; 20: 1288– 95. Google Scholar CrossRef Search ADS PubMed  32. Rosario M, Dirks NL, Gastonguay MRet al.   Population pharmacokinetics–pharmacodynamics of vedolizumab in patients with ulcerative colitis and Crohn’s disease. Aliment Pharmacol Ther  2015; 42: 188– 202. Google Scholar CrossRef Search ADS PubMed  33. Wade JR, Parker G, Kosutic Get al.   Population pharmacokinetic analysis of certolizumab pegol in patients with Crohn’s disease. J Clin Pharmacol  2015; 55: 866– 74. Google Scholar CrossRef Search ADS PubMed  Copyright © 2018 European Crohn’s and Colitis Organisation (ECCO). Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Journal of Crohn's and ColitisOxford University Press

Published: Jan 23, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off