Briefing Document. Food and Drug Administration. Center for Drug Evaluation and Research

Transcription

1 Briefing Document Food and Drug Administration Center for Drug Evaluation and Research Meeting: Joint Meeting of the Gastrointestinal Drugs Advisory Committee (GIDAC) and the Drug Safety and Risk Management Advisory Committee (DSaRMAC) December 9,

2 Disclaimer: The attached package contains background information prepared by the Food and Drug Administration (FDA) for the panel members of the advisory committee. The FDA background package often contains assessments and/or conclusions and recommendations written by individual FDA reviewers. Such conclusions and recommendations do not necessarily represent the final position of the individual reviewers, nor do they necessarily represent the final position of the Review Division or Office. We have brought vedolizumab to this Advisory Committee in order to gain the Committee s insights and opinions, and the background package may not include all issues relevant to the final regulatory recommendation and instead is intended to focus on issues identified by the Agency for discussion by the advisory committee. The FDA will not issue a final determination on the issues at hand until input from the advisory committee process has been considered and all reviews have been finalized. The final determination may be affected by issues not discussed at the advisory committee meeting. 2

3 Food and Drug Administration Center for Drug Evaluation and Research December 9, 2013 Joint GIDAC / DSARM Advisory Committee Meeting Briefing Materials - Table of Contents Sections Tab Executive Summary 1 Background and Regulatory History 2 Ulcerative Colitis Clinical Efficacy and Statistics 3 Crohn s Disease Clinical Efficacy 4 Crohn s Disease Statistics 5 Clinical Safety 6 Natalizumab Experience 7 PML Discussion 8 Risk Management Strategies 9 Nonclinical 10 Clinical Pharmacology 11 CMC 12 3

4 TAB 1 EXECUTIVE SUMMARY 1. Executive Summary 1.1. Statement of Purpose The purpose of the Advisory Committee meeting is to obtain advice from the Committee regarding the efficacy and safety of Entyvio (vedolizumab) for the proposed ulcerative colitis (UC) and Crohn's disease (CD) indications based on data from the Phase 3 clinical trials (one induction trial and one maintenance trial in UC; two induction trials and one maintenance trial in CD). The Applicant proposes the following indications: UC: " for reducing signs and symptoms, inducing and maintaining clinical response and remission, and mucosal healing, and achieving corticosteroid-free remission in adult patients with moderately to severely active ulcerative colitis who have had an inadequate response with, lost response to, or were intolerant to either conventional therapy or a tumor necrosis factor-alpha (TNFα) antagonist." CD: " for reducing signs and symptoms, inducing and maintaining clinical response and remission, and achieving corticosteroid-free remission in adult patients with moderately to severely active Crohn s disease who have had an inadequate response with, lost response to, or were intolerant to either conventional therapy or a tumor necrosis factor-alpha (TNFα) antagonist." The key efficacy issue is whether substantial evidence of efficacy has been established for induction of clinical remission in CD. Of the two induction clinical trials in CD, only one met its primary endpoint. The key safety issue is the risk of progressive multifocal leukoenceophalopathy (PML) which could potentially be caused by vedolizumab. There is uncertainty about the adequacy of the safety database to provide an acceptable pre-marketing assessment of this risk of PML. The Applicant's proposed indicated population is patients that have failed either "conventional therapy" (which includes steroids and immunosuppressants) or TNFαantagonists. We question if, given the potential risk of PML, the appropriate indicated population should be patients that have been tried on steroids only or patients that have been tried on at least immunosuppressants. Although a relationship between concomitant immunosuppressive therapies with infections was not found, there remains the concern that the risk of infections and of PML might be higher with concomitant immunosuppressive therapies. In the vedolizumab trials, these considerations led to the requirement that concomitant immunosuppressants will not be allowed beyond the induction phase (e.g., 6 weeks) in the US trials. We question whether the labeling should have similar restrictions. We also question if, based on the efficacy data from the UC and CD clinical trials, the benefit of vedolizumab (in each of these populations) outweighs the risks (especially the 4

5 TAB 1 EXECUTIVE SUMMARY potential risk of PML), and if, based on the current exposure and safety data, risk management strategies for PML beyond labeling are needed Potential Risk of PML Although there were no PML cases in the vedolizumab clinical development program, there is concern about the risk of PML with vedolizumab because its purported mechanism of action (disruption of integrin function) is similar to that of natalizumab, a product that has been associated with PML. It should be noted that with natalizumab, longer treatment duration (particularly greater than 24 months), prior treatment with an immunosuppressant, and presence of anti-jcv antibodies are associated with increased risk of PML. However, for vedolizumab, the risk factors for the potential risk of PML are not known at this time as no cases have been reported in patients taking vedolizumab. The Applicant asserts that vedolizumab does not have the same risk of PML as natalizumab because of mechanistic differences between the two products (mainly differences in receptor binding targets), and that in vitro activity data, animal models, and human pharmacodynamic (PD) data suggest a lower PML risk than for natalizumab. FDA Reviewers have questioned if this data provide sufficient evidence of less risk of PML than natalizumab Approach to Risk Evaluation and Management in Vedolizumab Clinical Trials The key elements of the Division's approach to managing the potential risk of PML in the vedolizumab clinical trials was to select an appropriate study population (one for whom the potential risk of PML would be more acceptable) and to limit concomitant immunosuppressive therapies during the trials (because of concerns of increased risk with concomitant immunosuppression). The Division required (in the US version of the protocols) that patient enrollment be limited to patients who failed immunosuppressants or TNFα antagonists (patients who failed steroids only could not enter the trials), and that concomitant immunosuppressants be limited to the induction phase only. Other key protocol requirements that were aimed at evaluating and managing the potential risk of PML included neurological examinations (baseline and periodically during and after the study), a case evaluation algorithm for suspected PML (that included referral to a panel of PML experts), and post-study follow-up for two years. There were also requirements for informed consent of subjects, and for education of subjects and site personnel. Many of these features were implemented by the Applicant in the Risk Assessment and Minimization for PML (RAMP) Program; this is described in detail in the PML Discussion section. 5

6 TAB 1 EXECUTIVE SUMMARY 1.4. Nonclinical Data and Human PD Data Nonclinical data suggest that the mechanism of action of vedolizumab differs from that of natalizumab. In vitro studies show that vedolizumab selectively binds to α4β7 integrin, and not to α4β1 integrin. Also, in vivo data from an experimental autoimmune encephalomyelitis (EAE) model showed that unlike natalizumab, vedolizumab did not inhibit any component of immune surveillance of the CNS. See the Nonclinical section of this background document. In a study of 14 healthy subjects, a single dose of vedolizumab did not affect the CD4+:CD8+ ratio in the cerebrospinal fluid (CSF); a previous study showed that the CSF CD4+:CD8+ ratio was reduced in MS patients receiving natalizumab. See the Clinical Safety section of this background document Clinical Summary The Phase 3 UC and CD clinical trials are summarized in the table below. Table 1. Phase 3 UC and CD Clinical Trials Clinical Trials Arms Primary Endpoint N* UC C13006 Induction Trial PBO VDZ 300 mg at Wks 0 and 2 Clinical Response at Wk PBO C13006 Maintenance Trial # VDZ 300 mg Q4W (start at Wk 6) Clinical Remission at Wk VDZ 300 mg Q8W (start at Wk 6) CD C13007 Induction Trial C13007 Maintenance Trial # C13011 Induction Trial PBO VDZ 300 mg at Wks 0 and 2 PBO VDZ 300 mg Q4W (start at Wk 6) VDZ 300 mg Q8W (start at Wk 6) PBO CDAI-100 Response at Wk 6 or Clinical Remission at Wk Clinical Remission at Wk Clinical Remission at Wk VDZ 300 mg at Wks 0 and 2 PBO: Placebo; VDZ: Vedolizumab; *ITT # For each Maintenance Trial (C13006 and C13007), patients must have achieved Clinical Response at Wk 6 in the corresponding Induction Phase (see UC and CD Clinical Efficacy sections for details) Alternative endpoints: at least one of the two alternative primary endpoints must be met to declare success (see CD Clinical Efficacy section for details) Analysis population for the primary endpoint was the TNFα-antagonist-failure population (n=315) UC: Clinical Response = Complete Mayo Score of 3 points and 30% from baseline with an accompanying decrease in rectal bleeding subscore of 1 point or absolute rectal bleeding subscore of 1 point Clinical Remission = Complete Mayo Score of 2 points and no individual subscore > 1 point. CD: CDAI-100 Response = Crohn's Disease Activity Index (CDAI) decrease from baseline by 100 points Clinical Remission = CDAI 150 points Table modified from UC and CD Clinical Efficacy sections. 6

7 TAB 1 EXECUTIVE SUMMARY Exposure data are summarized in the table below (based on a data cutoff date of June 27, 2013). Table 2. Total Number of Patients by Duration of Dosing and Number of Infusions (Vedolizumab) Duration of Dosing (Total Number of Patients): 1 dose 6 mo. 12 mo. 18 mo. 24 mo. 36 mo. 48 mo Number of Infusions with 4-Week Follow-up (Total Number of Patients): Table modified from Clinical Safety section Efficacy The results of the Phase 3 clinical trials are in the tables below. Table 3. Induction Clinical Trials Results - UC (C13006) and CD (C13007 and C13011) Clinical Trial Primary Endpoint PBO VDZ VDZ-PBO p-value UC Induction C13006 Clinical Response at Wk 6 CD Induction 25.5% (38/149) C13007 Clinical Remission at Wk 6* 6.8% (10/148) C13011 Clinical Remission at Wk 6 # 12.1% 47.1% (106/225) 14.5% (32/220) 15.2% (24/158) 21.7% < % % (19/157) PBO: Placebo; VDZ: Vedolizumab *There were two alternative primary endpoints: Clinical Remission at Wk 6 and CDAI-100 Response at Wk 6. At least one of the two alternative primary endpoints must be met to declare success. Of the two, only Clinical Remission at Wk 6 was met (see CD Clinical Efficacy section for details). # The analysis population for the primary endpoint was the TNFα-antagonist-failure population (76% of the ITT population). Table modified from UC and CD Clinical Efficacy sections. Table 4. Maintenance Clinical Trials Results - UC (C13006) and CD (C13007) Clinical Trial UC Maintenance Primary Endpoint C13006 Clinical Remission at Wk 52 CD Maintenance PBO 15.9% (20/126) 21.6% C13007 Clinical Remission at Wk 52 (33/153) PBO: Placebo; VDZ: Vedolizumab Table modified from UC and CD Clinical Efficacy sections. VDZ Q8W 41.8% (51/122) 39.0% (60/154) VDZ Q4W 44.8% (56/125) 36.4% (56/154) Q8W- PBO (p-value) 26.1% (<0.0001) 17.4% (0.0007) Q4W- PBO (p-value) 29.1% (<0.0001) 14.7% (0.0042) Each of the proposed indications (UC induction, UC maintenance, CD induction, and CD maintenance) and the strength of evidence for these indications are discussed below. 7

8 TAB 1 EXECUTIVE SUMMARY UC Induction: The single UC induction trial (C13006 Induction) demonstrated superiority of vedolizumab over placebo for Clinical Response at Week 6. In addition, both of the prespecified secondary endpoints (Clinical Remission at Week 6 and "Mucosal Healing" 1 at Week 6) were met. Note that a "mucosal healing" labeling claim would require histologic data, and the applicant provided no histologic data (see the UC Efficacy section). UC Maintenance: The single UC maintenance trial (C13006 Maintenance) demonstrated superiority of both vedolizumab arms over placebo for Clinical Remission at Week 52. In addition, all four of the pre-specified secondary endpoints (Durable Clinical Response 2, "Mucosal Healing" 1, Durable Clinical Remission 3, and Corticosteroid-Free Remission 4 ) were met. Note that a "mucosal healing" labeling claim would require histologic data, and the applicant provided no histologic data (see the UC Efficacy section). CD Induction: Of the two CD induction trials (C13007 and C13011), only one (C13007) showed superiority of vedolizumab over placebo for induction of clinical remission. Neither the pre-specified alternative primary endpoint nor the pre-specified secondary endpoint was met. We question whether the level of evidence requirements for a single trial were met to establish substantial evidence of efficacy for induction of clinical remission in CD based on the Evidence of Effectiveness Guidance. 5 The Guidance states, in general terms, that a single trial would be acceptable for approval of an indication if it provided strength of evidence equal to two adequate and well-controlled trials. Key considerations (also described in the Guidance) include: (1) whether the observed outcome of the C13007 induction trial was statistically very persuasive; and (2) whether multiple prospectively identified endpoints involving different events (each of which represents a beneficial, but different, effect) were met. Study C13011 did not meet its primary endpoint. Two trends in the data are worth noting however, and may serve as hypothesis generating for future trials. The FDA Clinical 1 Mucosal Healing was defined as a Mayo endoscopic subscore of 1 point. Note that a "mucosal healing" labeling claim would require histologic data, and the applicant provided no histologic data (see the UC Efficacy section). 2 Durable Clinical Response was defined as Clinical Response both at Week 6 and Week 52 3 Durable Clinical Remission was defined as Clinical Remission both at Week 6 and Week 52 4 Corticosteroid-Free Remission was defined as Clinical remission in patients using oral corticosteroids at baseline (Week 0) who have discontinued corticosteroids and are in clinical remission at Week FDA Guidance "Providing Clinical Evidence of Effectiveness for Human Drugs and Biological Products" (available at pdf) 8

9 TAB 1 EXECUTIVE SUMMARY Reviewer noted that exploratory analyses in the overall study population (i.e., not limited to TNFα Antagonist Failures) at Week 6 suggested that there may be a treatment effect in the overall study population. Additionally, since a third dose at Week 6 was administered, the FDA Clinical Reviewer noted that exploratory efficacy analyses at a later timepoint (Week 10) (in both the overall and the TNFα Antagonist Failures study populations) suggested that there may be a treatment effect at the later timepoint. However, these exploratory analyses need to be interpreted with caution because these were not pre-specified. (See the CD Clinical Efficacy section.) CD Maintenance: The single maintenance trial (C13007 Maintenance) demonstrated superiority of both vedolizumab arms over placebo for Clinical Remission at Week 52. In addition, two of the three pre-specified secondary endpoints were met for the Q8W arm; only of the prespecified secondary endpoints was met for the Q4W arm. We question if substantial evidence of efficacy for maintenance of clinical remission can be demonstrated without first having demonstrated substantial evidence of efficacy for induction of clinical remission. The efficacy standard that has been used in prior drug approvals (consistent with the Evidence of Effectiveness Guidance) is that if there is substantial evidence of efficacy for induction (in a disease population; e.g., UC or CD), a single successful maintenance trial (in that same disease population) could be sufficient to extend the claim to maintenance. Other Key Exploratory Subgroup Analyses: For each of the trials, exploratory subgroup analyses in patients that met the US protocol criteria (i.e., failure of immunosuppressants or TNFα-antagonists requirement for entry; and concomitant immunosuppressants not allowed beyond the induction phase) suggest that the treatment differences for the induction and maintenance endpoints are similar to the overall study population. This suggests that the treatment effect would be expected to be preserved for these subgroups; however, the analyses were post-hoc and need to be interpreted with caution. In summary, we ask the Committee to consider the strength of evidence that vedolizumab is effective for each of the proposed indications (UC and CD) given the efficacy data above Safety Overall safety data are presented from the total of 3,326 subjects that received 1 dose of vedolizumab. Data from UC and CD patients are combined in these analyses. Comparative safety data from the 1,434 patients who received vedolizumab only throughout the induction and maintenance trials for UC (C13006) and CD (C13007) (VDZ/VDZ) were compared to the 297 patients who received only placebo (PBO/PBO) 9

10 TAB 1 EXECUTIVE SUMMARY and the 279 patients who received vedolizumab during induction and placebo during the maintenance phase (VDZ/PBO). Deaths: There were 12 deaths in patients receiving vedolizumab; the FDA Clinical Reviewer concluded that none of the deaths were related to the study drug. Infections: As there have been no cases of PML reported, one focus of the safety review has been to determine if infections, particularly serious infections, are increased with vedolizumab treatment versus placebo or with cumulative vedolizumab dosing. In the comparative safety data, infections overall were higher in the VDZ/VDZ group than the PBO/PBO group (43% vs. 35%), but serious infections were similar across groups (4% in VDZ/VDZ, 3% in PBO/PBO, and 3% in VDZ/PBO). The most commonly reported infections were classified as upper respiratory tract infections (high level term) (24% VDZ/VDZ vs. 17% PBO/PBO) and appear to have driven the difference in frequency of overall infections between the VDZ/VDZ and PBO/PBO groups. The safety database was also evaluated for opportunistic infections. Systemic infections from enteric pathogens occurred in very small numbers, so comparisons were difficult to make. Fifty-one patients reported Herpes viral infections, but none were reported as serious, all were considered mild to moderate in intensity, and the majority were oral herpes; the rates of herpes infections were similar between treatment groups (3% VDZ/PBO, 2% PBO/PBO, and 3% VDZ/VDZ). No clear relation of these infections to number of infusions or to concomitant immunosuppressant use was found. PML Risk Estimation: Using the "Rule of Three," 6 the worst possible scenario (i.e., the 95% upper bound of the true rate of PML) can be calculated based on the size of the safety database if no events are observed. Since no PML cases were observed in the 3,326 subjects that received one or more infusions, the true rate of PML will be lower than 0.9 in 1,000 with 95% confidence in patients that received one or more infusions. Similarly, since no PML cases were observed in the 1,004 patients that received 24 or more infusions, the true rate of PML will be lower than 2.99 in 1,000 with 95% confidence in patients that received 24 or more infusions. [Note that if the calculation is based on the number of patients that were exposed for 24 or more months (i.e., 906 patients) (instead of the number of patients 6 The rule of three states that in a study where no events are observed, the 95% confidence upper bound for the true event rate is approximately 3/n, where n is the study sample size (Jovanovic, B.D. and Levy, P.S. A Look at the Rule of Three. The American Statistician 1997;51(2): ). 10

11 TAB 1 EXECUTIVE SUMMARY that received 24 or more infusions), the true rate of PML will be lower than 3.31 in 1,000 with 95% confidence in patients that were exposed for 24 or more months.] Other Safety Issues: Hypersensitivity (several cases of urticaria and at least one case of anaphylaxis) were reported with vedolizumab use. Infusion-related reactions occurred at a rate of approximately 4% in patients receiving vedolizumab. There was no clear increase in risk in carcinogenicity; however, long-term follow-up data would be necessary to reliably assess the risk of carcinogenicity. Risk Assessment and Minimization for PML (RAMP) Program Findings: In clinical trials, no cases of PML have been identified in the 2,927 patients monitored through the RAMP program (as of June 28, 2013) (see PML Discussion section of this background document). A total of 290 (10%) patients reported at least one abnormality on the subjective PML checklist; 64 had abnormal findings identified on the objective PML checklist. Fifty-eight MRIs were performed and 86 cases have been adjudicated by the independent adjudication committee (IAC). Five lumbar punctures have been performed. Risk Management Options: If vedolizumab is approved, a risk management strategy (beyond labeling) may be required to address the potential risk of PML. The applicant has proposed a strategy to inform patients and healthcare professionals of the potential risk of PML, and of the need to monitor for signs and symptoms of PML. Additional risk mitigation strategies may be needed to ensure early detection of PML and timely discontinuation if PML occurs until there is a reasonable level of certainty that the potential risk of PML has been ruled out. In summary, we ask the Committee to consider if the premarketing risk assessment is adequate (in particular, the premarketing risk assessment for PML) or if additional premarketing safety studies are needed. If the Committee believes that the risk assessment is adequate, we seek input on potential strategies to manage the risk in the post-market setting. Benefit in UC: Benefit-Risk Considerations For induction in UC, only a single induction trial was conducted. Statistically significant treatment differences were observed for the primary endpoint (Clinical Response treatment difference=21.7%) and both secondary endpoints (Clinical Remission treatment difference=11.5%; "Mucosal Healing" 1 treatment difference=16.1%). Note that a 11

12 TAB 1 EXECUTIVE SUMMARY "mucosal healing" labeling claim would require histologic data, and the applicant provided no histologic data (see the UC Efficacy section). For maintenance in UC, statistically significant treatment differences were observed with the proposed dosing regimen of Q8W for the primary endpoint (Clinical Remission treatment difference=26.1%) and all four secondary endpoints (Durable Clinical Response 2 treatment difference=32.8%; "Mucosal Healing" 1 treatment difference=32.0%; Durable Clinical Remission 3 treatment difference = 25.3%; Corticosteroid-Free Remission 4 treatment difference=17.6%). Note that a "mucosal healing" labeling claim would require histologic data, and the applicant provided no histologic data (see the UC Efficacy section). Other treatment options in this population of moderately to severely active UC include TNFα-antagonists (infliximab, adalimumab, and golimumab) and surgery (potentially curative). Benefit in CD: For induction in CD, one induction trial was successful; a statistically significant treatment difference was observed for the primary endpoint (Clinical Remission treatment difference=7.8%) but not for the alternative primary endpoint (CDAI-100 Response) or for the secondary endpoint (Change in CRP). The results were not replicated in a second induction trial in a TNFα-antagonist failure population. For maintenance in CD, statistically significant treatment differences were observed with the proposed dosing regimen of Q8W for the primary endpoint (Clinical Remission treatment difference=17.4%) and for two secondary endpoints (CDAI-100 Response treatment difference=13.4%; Corticosteroid-Free Remission 4 treatment difference=15.9%) of the three tested (the secondary endpoint of Durable Clinical Remission 3 was not met). Other treatment options in this population of moderately to severely active CD include TNFα-antagonists (infliximab, adalimumab, and certolizumab) and natalizumab. The number of patients that have received natalizumab for CD is very small (approximately 1,100). The natalizumab indication is limited to patients that have failed TNFαantagonists. Thus, for patients currently being treated with natalizumab, vedolizumab could be considered as a possible alternative. However, given the discussion above, it is not known if vedolizumab would be efficacious in TNFα-antagonist failures; the CD induction trial with a primary analysis population of TNFα-antagonist failures did not demonstrate superiority of vedolizumab over placebo. Risk of PML: It is important to note that the safety database provides a comparison of the PML risk of vedolizumab to a benchmark rate (e.g., 2.99 in 1,000), but does not provide a comparison of the PML risk with vedolizumab to the PML risk with natalizumab. Thus, it will be 12

13 TAB 1 EXECUTIVE SUMMARY difficult to infer that one drug has a more desirable risk profile than the other; any comparisons of risk between vedolizumab and natalizumab will be crude and should be interpreted with caution. Additional limitations are that vedolizumab Phase 3 trials sampled from a different population (CD or UC patients) than that from which natalizumab s PML risk was estimated in the natalizumab clinical trials (approximately two-thirds were MS patients, and one-third were CD patients) and that from which natalizumab s PML risk was estimated based on natalizumab postmarketing data (approximately 99% MS patients and 1% CD patients). In summary, we ask the Committee to consider if the benefits for the proposed UC and/or CD indications outweigh the potential risks (in particular, PML). We also ask the Committee if risk management strategies beyond labeling are needed. 2. Points for Consideration by the Advisory Committee The FDA requests that the Advisory Committee consider the following topics during review of this (and the Applicant s) briefing document. These topics are intended to frame the major review issues that will be the foundation for more specific questions posed to the Committee at the meeting. Efficacy - UC: 1. Evidence for vedolizumab efficacy for UC induction and for UC maintenance is provided by one trial each. Discuss if the available data support the efficacy of vedolizumab for the proposed UC induction and maintenance indications. Efficacy - CD: 2. Evidence for vedolizumab efficacy for CD induction is provided by one trial but not replicated in a second trial in a refractory population. Evidence for vedolizumab efficacy for CD maintenance is provided in one trial. Discuss if the available data support the efficacy of vedolizumab for the proposed CD induction and maintenance indications. 3. Considering the currently available data, discuss whether additional efficacy studies should be obtained prior to approving vedolizumab for the proposed CD induction and maintenance indications. Safety: 4. Consider whether the nonclinical data and human PD data presented for vedolizumab (e.g., specific α4β7 receptor binding target) provide sufficient evidence of less risk of PML than natalizumab. 13

14 TAB 1 EXECUTIVE SUMMARY 5. Considering the currently available data, discuss whether additional safety data or studies should be obtained prior to approving vedolizumab for the proposed UC and/or CD indications. Benefit-Risk Assessment for UC: 6. Based on currently available efficacy and safety data, discuss if the benefits outweigh the potential risks of vedolizumab (in particular, PML) for the proposed UC indications. Benefit-Risk Assessment for CD: 7. Based on currently available efficacy and safety data, discuss if the benefits outweigh the potential risks of vedolizumab (in particular, PML) for the proposed CD indications. Labeling and Risk Management Strategies: 8. Discuss if the indicated population should be limited to patients that have failed immunosuppressants or TNFα-antagonists or if the indicated population should include patients that failed steroids only. 9. Discuss if concomitant immunosuppressants should be limited to a specific duration (e.g., during induction only). 10. If you believe vedolizumab should be approved for the proposed UC or CD indications, discuss if risk management strategies beyond labeling are needed, and discuss the particular strategies. Post-Approval Studies: 11. If you believe vedolizumab should be approved for the proposed UC or CD indications, discuss if any additional studies should be recommended post-approval. 14

15 TAB 2 BACKGROUND AND REGULATORY HISTORY Background and Regulatory History 1. Background Information 1.1. Inflammatory Bowel Disease Inflammatory Bowel Disease (IBD) is defined as chronic intestinal inflammation without identifiable etiology. There are two major types of IBD, Crohn s disease (CD) and ulcerative colitis (UC). Both are considered autoimmune diseases and have some similarities in clinical presentation and course; however, they have distinguishing features based on histology, extraintestinal involvement, and complications (see Table 5). Table 5. List of Features to Distinguish Crohn s Disease and Ulcerative Colitis Crohn s Disease Ulcerative Colitis Mouth to anus Limited to colon Transmural inflammation Mucosal inflammation Discontinuous Continuous Fistulae/abscesses common No fistulae/abscesses rare Strictures common Strictures rare Perianal disease ~30% No perianal disease Granulomas in ~50% No granulomas Recurs after surgery Curable with surgery Source: Reviewer s table adopted from Dr. Jessica Lee s presentation at DGIEP Medical Officer Rounds April 12, 2011 Ulcerative colitis (UC) is a chronic relapsing inflammatory disease of the rectal and colonic mucosa, which is characterized by clinical remissions and exacerbations resulting from intestinal inflammation. Crohn's disease (CD) involves chronic inflammation of all layers of the bowel and may affect any segment of the gastrointestinal (GI) tract. Symptoms can vary depending on the severity of inflammation and extent of disease; however, both UC and CD patients typically experience recurrent episodes of rectal bleeding and diarrhea, often associated with crampy abdominal pain and tenesmus. Fever and weight loss are also commonly seen. The transmural inflammation seen in CD can involve the wall of the bowel, leading to the development of strictures, fistulae between diseased parts of the bowel and adjacent structures (e.g., bladder or other bowel segments) and abscesses. Perianal disease and granulomas are common. Extraintestinal manifestations may also be seen and frequently involve the skin, eyes, and joints; and patients may experience sequelae due to malabsorption. CD usually has a chronic relapsing course with acute clinical exacerbations. Patients with UC also generally present with chronic relapsing disease, and periods of bloody mucoid diarrhea and abdominal pain may be followed by long quiescent periods between episodes. Patients may also exhibit systemic symptoms including fever, 15

16 TAB 2 BACKGROUND AND REGULATORY HISTORY malaise, and weight loss; and severe colitis can result in ischemic colitis requiring surgical colectomy. Colectomy is considered curative in UC, but it is associated with significant morbidity, including recurrent pouchitis in up to 25% of patients, fecal incontinence, and female infertility. Finally, patients with long-standing UC are at increased risk for colorectal cancer. It is estimated that that between 7,000 and 46,000 persons are newly diagnosed with ulcerative colitis each year and that as many as 780,000 persons may have ulcerative colitis. Between 10,000 and 47,000 residents of the United States and Canada are diagnosed with Crohn s disease annually, and as many as 630,000 persons may suffer from Crohn s disease (Loftus Jr EV. Gastroenterology (2004): ). Decisions about treatment of CD and UC weigh such factors as disease activity, disease extent and duration, previous treatment attempts and the patient s preference. The goal is to stop the patient's active acute disease (induction of remission) and then maintain the patient in remission. Aminosalicylates in oral and rectal formulations are approved for the treatment of mild to moderate UC, but are also used in the treatment of CD, though not approved. Corticosteroids are used to treat acute exacerbations and induce remission in both UC and CD. Immunomosuppressants, including azathioprine and mercaptopurine, are widely used, but are not approved for either UC or CD. Use of any of the preceding has come to be considered part of conventional therapy. TNFα antagonist biologic therapies (infliximab and adalimumab) are approved for the induction and maintenance of UC and CD in moderate to severe disease, while golimumab is approved only for UC and certolizumab pegol is approved only for CD. The integrin antagonist natalizumab (Tysabri) is approved for the treatment of moderate to severe CD, in patients who have failed other available therapies. Despite multiple available therapies, limitations remain in the treatment of IBD, and patients continue to have symptoms or develop intolerance to or side effects from their treatment regimens. 2. Regulatory History Clinical development of vedolizumab began in 1998, and IND 009,125 was opened in June 2000 to initiate clinical development in the United States. In January 2006, development of vedolizumab was placed on clinical hold due to concerns that integrin antagonists might predispose patients to progressive multifocal leukoencephalopathy (PML). This stemmed from the market withdrawal of natalizumab, following 2 cases of confirmed PML in patients receiving the drug to treat MS and one reported case in a patient treated for Crohn s disease. All integrin antagonists under development in the US at that time were placed on clinical hold. The clinical hold on IND 009,125 was lifted in July 2007 with the implementation of an active screening and monitoring program. 16

17 TAB 2 BACKGROUND AND REGULATORY HISTORY Multiple subsequent regulatory meetings, including an Advisory Committee (AC) meeting, have focused on risk minimization and safety monitoring related to potential PML risk. Clinical Considerations A Joint Meeting of the Gastrointestinal Advisory Committee (GIDAC) and the Drug Safety and Risk Management (DSaRM) Advisory Committee to evaluate intravenous vedolizumab for treatment of Inflammatory Bowel Disease (induction and maintenance of Crohn s Disease and Ulcerative Colitis) and the risk of PML was held on July 20, The purpose of this closed session Advisory committee meeting was to seek the committee s recommendations regarding the Phase 3 study design for vedolizumab, including the number of patients and duration of study needed to exclude the risk of PML. The following recommendations and responses were provided by the expert committee, in response to 4 questions: The committee voted 12 to 5, with one abstention, that the available nonclinical and human pharmacodynamic data for vedolizumab do not provide assurance of less risk of PML than natalizumab. The committee commented on an acceptable safety database size for pre-approval assessment of PML risk in patients with CD and UC. No consensus was reached; however, the AC strongly expressed that the duration of exposure is important and that 24 months could be considered as the minimum duration timeframe. The majority of the committee felt that increasing the sample size has merit. The committee voted 15 to 2, with one abstention, that the available nonclinical and clinical data do not support making the entry criteria less stringent for vedolizumab Phase 3 studies (i.e., allow entry of patients that have not yet been treated with TNFα antagonists or immunosuppressants). The committee voted 17 to 0, with one abstention, that restrictions on concomitant immunosuppressants (prohibited beyond the induction phase of vedolizumab treatment) should not be made less stringent. Over the course of several meetings between the sponsor and FDA, and based on the AC recommendations, the following major agreements were made relating to the management and assessment of the risk of PML with this product: Entry Criteria: Patients enrolled in Phase 3 studies were required to meet the stricter requirement of inadequate response or intolerance to immunosuppressants or TNFα antagonists (rather than inadequate response or intolerance to immunosuppressants, TNFα antagonists, or corticosteroids). Concomitant Immunosuppressive Therapies: Concomitant immunosuppressant and corticosteroid use was limited during the clinical trials. Concomitant immunosuppressants were limited to the induction phase only (e.g., six weeks). There were protocol-defined provisions to mandate tapering of corticosteroids after six weeks in patients that were in clinical response or at the subsequent visit when clinical response was achieved, and to limit corticosteroid dose increases to no greater than baseline dose; there were specified maximum steroid doses allowed on entry into the study. 17

18 TAB 2 BACKGROUND AND REGULATORY HISTORY Informed Consent / Education of Subjects and Site Personnel: Subjects were to be advised of the risks of PML, how to recognize symptoms of possible PML, and who to contact should they experience symptoms suggestive of PML. Prior to the start of the study, investigators and site staff must have been provided with education about the signs and symptoms of PML, and the procedures to follow if a case of PML was suspected. Neurological Examinations: Neurological examinations were to be conducted at entry with exclusion of patients with an abnormal finding. Follow-up neurologic exams were to be conducted approximately every three months while on treatment, and approximately three months after stopping treatment. PML Case Evaluation Algorithm: There was a PML case evaluation algorithm that indicated what further studies were to be conducted if a physician could not rule out PML. If, after further testing, PML still could not be ruled out, the subject was to be referred to an outside panel of PML experts, including at least one neurologist, for final determination of whether or not the subject had PML. Post-study Follow-up: Subjects were to be followed for two years after treatment was completed. Interim assessments were at 3, 6, 12, and 18 months. Subjects were to be questioned regarding the presence of any signs or symptoms of opportunistic infections and PML. Any positive findings on questioning were to be expeditiously referred to a physician for additional evaluation, and the PML case evaluation algorithm followed. Safety Database: The safety database at the time of original BLA submission was required to include data on at least 900 patients that received 24 infusions, with a minimum of 4 weeks of follow-up after the last infusion (in order to provide a pre-approval assessment of PML risk in patients with UC and CD that would be adequate to take to Advisory Committee for consideration). It should be noted that the Division only reviewed the US versions of the protocols. Some of the above protocol provisions, most notably restrictions on entry and restrictions on concomitant immunosuppressive therapies, are not part of the protocols outside the US. Product Quality Considerations Several formal meetings also occurred between the sponsor and FDA to discuss manufacturing changes. Vedolizumab was initially manufactured utilizing a mouse myeloma (NS0) cell line, and initial clinical studies used drug product from this process (MLN02, Process A). A Chinese hamster ovary (CHO) cell line was developed to improve productivity, and drug product from this process (MLN0002, Process B) was used in multiple Phase 1 and 2 clinical studies. Further manufacturing improvements to the CHO-based process were then implemented to establish a commercially representative process (MLN0002, Process C) that was used to supply Phase 3 clinical trials. The effects of these changes on product potency, purity, safety and efficacy were determined by a variety of comprehensive biochemical, biophysical, immunological and pharmacological assessments; the results of these analytical studies demonstrate the comparability of vedolizumab across all these process iterations (see CMC section). For 18

19 TAB 2 BACKGROUND AND REGULATORY HISTORY simplicity, vedolizumab will be used throughout this background document to refer to the drug product throughout its development. Regulatory Considerations Presubmission regulatory activities related to this submission included an Advisory Committee meeting and 14 formal face-to-face meetings between the sponsor and FDA. In addition, there were a number of teleconferences and written correspondences exchanged during the development program for ulcerative colitis. The sponsor was granted Fast Track Designation in February Table 6 below summarizes presubmission regulatory meetings and submissions and highlights key clinical agreements. Table 6. Pre-submission Regulatory History Date Regulatory Action(s) June 7, 2000 Original IND 9125 submitted for MLN02 June 2004 Type C Meeting to discuss the clinical development outcomes from two Phase 2 studies, M and L IND 9125 placed on clinical hold for insufficient information to allow the January 24, 2006 Agency to assess the risk of progressive multifocal leukoencephalopathy (PML) to subjects with MLN02 April 4, 2006 Type A Meeting to discuss options for removing clinical hold, including PML risk minimization and safety monitoring. July 26, 2006 Type C Meeting to discuss manufacturing changes from MLN02 to MLN0002 Sponsor submitted an amendment which was a complete response to the June 18, 2007 clinical hold and included the Risk Assessment and Minimization of PML (RAMP) program. July19, 2007 Removal of clinical hold based on additional safety measures related to potential PML risk December 11, 2007 Type C Meeting to continue discussions about PML risk management program Type C meeting to discuss overall development plan for MLN0002, April 18, 2008 specifically dose selection, CMC, and nonclinical data to support Phase 3 studies. June 5, 2008 Type C, End of Phase 2 Meeting to discuss pivotal studies for the proposed IBD indications September 16, 2008 Type B, End of Phase 2 meeting to discuss the CMC development plan September 26, 2008 Type C End of Phase 2 Teleconference to discuss outstanding clinical questions and issues for Phase 3 activities. September 10, 2009 Type C, Phase 3 meeting to discuss the Statistical Analysis Plan for the Phase 3 Crohn s disease study, C13007 July 13, 2010 Meeting to discuss Phase 3 development plan Closed Joint Meeting of the Gastrointestinal Drugs Advisory Committee July 20, 2011 (GIDAC) and the Drug Safety and Risk Management (DSaRM) Advisory Committee 19

20 TAB 2 BACKGROUND AND REGULATORY HISTORY Date September 6, 2011 July 24-25, 2012 November 6, 2012 November 13, 2012 February 21, 2013 Regulatory Action(s) Type C follow-up Meeting to discuss the outcomes from the Joint GIDAC/DSaRM meeting Type C, post-phase 3 meeting to discuss pivotal study data and clinical plan to support registration Type C, Pre-BLA meeting to discuss clinical, nonclinical, and regulatory aspects of the BLA Type B, Pre-BLA meeting to discuss CMC aspects of the BLA Fast track designation granted for vedolizumab in the treatment of ulcerative colitis and Crohn s disease 20

21 TAB 3 ULCERATIVE COLITIS CLINICAL EFFICACY AND STATISTICS Ulcerative Colitis Clinical Efficacy and Statistics 1. Clinical Efficacy 1.1.Ulcerative Colitis Efficacy Summary The indication sought by the Applicant is that vedolizumab is effective at inducing and maintaining clinical response and remission in adult patients with moderately to severely active ulcerative colitis who have had an inadequate response with, lost response to, or were intolerant to either conventional therapy or a tumor necrosis factor-alpha (TNFα) antagonist. The primary efficacy data to support this indication were from Trial C13006, a phase 3, randomized, double-blind, placebo-controlled trial in 895 patients with moderately to severely active UC. This trial was conducted under a single protocol but designed and analyzed as two separate studies: C13006 Induction Study and C13006 Maintenance Study. C13006 was a multicenter, multinational trial, and approximately a quarter of the patients were from the United States. The US population varied from the ex-us population, both in entry criteria and in its allowance of concomitant immunosuppressant use. In the US, patients were required to have failed either an immunomodulator (5-mercaptopurine or azathioprine) or a TNFα agent, while outside the US, failing corticosteroids alone was sufficient for study entry. In addition, in the US, patients were required to discontinue immunomodulators at Week 6, while those outside the US could continue concomitant immunomodulator therapy over the course of the trial. Primary and secondary efficacy assessments for the Induction and Maintenance Studies were based on Mayo scores. The key study endpoint definitions are provided in Table 7, below and will be used throughout the document for this indication: Table 7: C13006 Study Endpoint Definitions Endpoint Definition Clinical Remission A complete Mayo score of 2 points and no individual subscore > 1 point Clinical Response A reduction in complete Mayo score of 3 points and 30% from baseline (or a partial Mayo score of 2 points and 25% from baseline, if the complete Mayo score was not performed at the visit) with an accompanying decrease in rectal bleeding subscore of 1 point or absolute rectal bleeding subscore of 1 point Corticosteroid-Free Clinical remission in patients using oral corticosteroids at Remission baseline (Week 0) who have discontinued corticosteroids and are in clinical remission at Week 52. Durable Clinical Remission Clinical remission at Weeks 6 and 52 Durable Clinical Response Clinical response at Weeks 6 and 52 Mucosal Healing Mayo endoscopic subscore of 1 point Source: Clinical Study Protocol C

22 TAB 3 ULCERATIVE COLITIS CLINICAL EFFICACY AND STATISTICS Induction Study Results Summary The primary efficacy endpoint for the C13006 Induction Study was the proportion of subjects in clinical response at Week 6. The proportion of patients in clinical response at Week 6 was significantly greater in the vedolizumab group (47.1%) relative to placebo (25.5%). The difference from placebo was 21.7% (95% CI: 11.6, 31.7; p< ). Vedolizumab had numerically lower response rates in the subgroup of patients who previously failed TNFα agents; however, even in the more difficult to treat patients (i.e., patients who previously failed TNFα agents), the results favor vedolizumab. The results were generally consistent across a variety of subgroups, including age, gender, race, duration of disease, geographic region, and baseline disease activity, as well as in anti- TNF users and failures. The first tested secondary induction endpoint was the proportion of patients with clinical remission at Week 6. Thirty-eight (38) patients (16.9%) in the vedolizumab group achieved clinical remission, compared to eight patients (5.4%) in the placebo group. The difference from placebo was 11.5% (95% CI: 1.5, 6.6; p = ). The other secondary endpoint for the Induction Study was "mucosal healing" which was defined as a Mayo endoscopic subscore of 1 point. Using this prespecified definition, 40.9% of patients in the vedolizumab treatment group achieved "mucosal healing", compared with 24.8% of patients receiving placebo, a 16.1% treatment difference (95% CI: 1.2, 2.3; p = ). However, the FDA believes to establish "mucosal healing" requires histologic data, and the applicant provided no histologic data. Maintenance Study Results Summary Patients that received vedolizumab and achieved clinical response at Week 6 in the Induction Study were randomized to vedolizumab Q4W, vedolizumab Q8W, or placebo on entry into the Maintenance Study. The primary efficacy endpoint for the Maintenance Study was the proportion of these subjects who were in clinical remission at Week 52. In comparison to placebo, each of the two vedolizumab dosing regimens showed significant improvement on the prespecified definition for clinical remission. In the Q8W dosing group, the difference from placebo was 26.1%, and in the Q4W group, the difference was 29.1%, with p < for both groups. Again, sensitivity analyses and subgroup analyses (e.g., age, gender, race, duration of disease, geographic region, and baseline disease activity, anti-tnf users and failures) were generally consistent. The analysis population for the primary endpoint of clinical remission at Week 52 included all patients in clinical response at Week 6. When looking at the subset of patients who were in clinical remission at Week 6, the results were consistent. A significantly higher proportion of vedolizumab treated patients in both dosing regimens were in clinical remission both at Week 6 and 52 compared with patients who received placebo (8.7% in the placebo group, 20.5% in the vedolizumab Q8W group, and 24.0% 22