CENTER FOR DRUG EVALUATION AND
`RESEARCH
`A.
`
`APPLICA TION NUMBER:
`
`22-350
`
`SUMMARY REVIEW
`
`

`

`Division Director Review
`
`Summary Review for Regulatory Action
`
`
`
`
`Jul_L27, 2009
`
`
`Mary H. Parks, MD.
`
`__I Division Director Summary Review
`
`22-350
`NDA/BLA #
`'
`
`Supflment #
`i
`
`
`
`A lieant Name
`Bristol Myer Sfluibb
`
`
`
`Date of Submission
`June 30, 2008
`
`PDUFA Goal Date
`At July 31, 2009 (including 3-month extension for major
`
`
`
`
`amendment)
`
`
`Proprietary Name /
`Onglyza® (saxagliptin)
`
`
`
`Established @AN) Name
`
`
`
`Dosage Forms / Strength
`25 and 5.0 tablets
`
`
`Proposed Indication(s)
`i743 an adjunct to diet and exercise to improve glycemic
`
`
`
`control in adults with type 2 diabetes mellitus
`
`
`Approval
`
`
`
`Action/Recommended Action for
`NME:
`
`
`
`
`
`Material Reviewed/Consulted
`
`0ND Action Package, including:
`Names of discflne reviewers.
`
`Medical Officer Review
`Naomi Lowy, M.D.
`Statistical Review
`Roswitha Kelly, M.S. (CMC Stats Review)
`Karl Lin, Ph.D. (Carci Stats Review)
`Joy Mele, M.S.
`Todd Sahlroot,‘ Ph.D.
`Atair Rahman, Ph.D. (carci Stats Review)
`Thomas Permutt, Ph.D.
`
`‘
`
`
`
`
`CMC Review/OBP Review
`
`Ali Al-Hakim, Ph.D.
`Shamista Chatterjee, Ph.D.
`Blair Fraser, Ph.D.
`John Hill, Ph.D.
`Christine Moore, Ph.D.
`Prafull Shiromani, Ph.D.
`
`Su Tran, Ph.D.
`NA
`Sally Choe, Ph.D.
`Justin C. Earp, Ph.D.
`Wei Qiu, Ph.D.
`
`Microbiolo y Review
`Clinical Pharmacology Review
`
`Page ] of 2
`
`

`

`_
`Sam Skariah, PharmD.
`San eeta Vaswani, Pharm D.
`Susan Leibenhaut, M.D.
`
`Constance Lewin, M.D., M.P.H.
`
`Hylton Joffe, M.D., M.M.Sc.
`Kristina Amwine, PharmD.
`Anne Crandall, PharmD.
`Melina Griffis, R.Ph;
`Carol Holquist, R.Ph.
`Denise To er, Pharm. D.
`
`
`
`-
`
`-
`
`'
`
`I
`
`Division Director Review
`
`
`
`Christoffer Tomoe, Ph.D.
`Jaya Vaidyanathan, Ph.D.
`
`Immo Zdrojewski, Ph.D.
`Robert Dean, M.B.A
`Kendra Jones, B.S.
`
`Jodi Duckhorn, M.A.
`I Lina Aljuburi, Pharmt D.
`Laurie Burke, (SEALD)
`Jeanne Delasko, RN, MS (SEALD)
`Abby Jacobs, Ph.D. (ExecutiveCAC)
`David Jacobson-Kram, Ph.D. (Executive CAC)
`
`Barr Rosloff, PhLD. Executive CAC
`
`CDTL Review
`OSE/DMETS
`
`OSE’DSRCS
`
`-
`
`‘
`
`'
`Other
`
`OND=0ffice of New Drugs
`DDMAC=Division of Drug Marketing, Advertising and Communication
`OSE= Office of Surveillance and Epidemiology
`DMETS=Division of Medication Errors and Technical Support
`DSI=Division of Scientific Investigations
`DDRE= Division of Drug Risk Evaluation
`DSRCS=Division of Surveillance, Research, and Communication Support
`CDTL=Cross-Discipline Team Leader
`
`Page 2 of 2
`
`

`

`Division Director Review
`
`Division Director Review
`
`1. Introduction
`
`Saxagliptin is a dipeptidyl peptidase-4 enzyme inhibitor (DPP4—inhibitor) developed for the
`management of hyperglycemia in patients with type 2 diabetes mellitus (T2DM). This is a
`relatively new class of anti—diabetic therapy whose mechanism of action targets the impaired
`release and availability of the incretin hormone, glucagon-like peptide—l (GLP- 1)1n patients
`with type 2 diabetes. GLP- l and another incretin hormone, glucose-dependent insulinotropic
`polypeptide (GIP), are released from the gastrointestinal tract in response to meals to flirther
`stimulate insulin release. Because GLP- 1IS rapidly degraded by the serine protease,
`dipeptidyl peptidase 4, an inhibitor of this enzyme will prolong the half—life of this incretin
`hormone allowing for a more sustained effect on glucose control.
`
`Unlike other anti-diabetic therapies, which control hyperglycemia through stimulation Of
`insulin release from the pancreas (e.g. sulfonylureas or glinides), incretin-based therapies
`control hyperglycemia through a glucose-dependent manner thereby mitigating the risk of
`hypoglycemia. GLP-1 receptor agonists are another class of incretin-based therapies. These
`agents are manufactured to avoid susceptibility to enzyme degradation while maintaining
`sufficient cross-reactivity with the GLP- 1 receptor to impart similar effects on glucose control
`as the native hormone.
`
`Currently, Januvia (sitagliptin) is the only marketed DPP4-inhibitor in the United States. /
`1/
`'
`5
`9
`
`I
`
`114
`D i)
`
`.
`
`.
`
`§
`
`2. Background
`
`Over the past two to three years, concerns regarding the cardiovascular safety profile of certain
`anti-diabetics have resulted in much debate within the scientific and regulatory community on
`the adequacy of the development programs for anti—diabetic therapies to ensure that these
`drugs do not contribute to excess cardiovascular mortality and morbidity in a patient
`population that is already at 2- to 4-fold risk of dying from heart disease.
`
`On July 1 and 2, 2008, the FDA convened a public advisory committee meeting to discuss the
`role of CV assessment in the pre- and postrnarket settings. The pivotal question raised to the
`panel members was:
`
`Page 3 of 2
`
`

`

`Division Director Review
`
`It should be assumed that an anti-diabetic therapy with a concerning CV safety signal
`during Phase 2/3 development will be required to conduct a long-term cardiovascular
`trial. For those drugs or biologics without such a signal, should there be a
`requirement to conduct a long-term cardiovascular trial or to provide other equivalent
`evidence to rule out an unacceptable cardiovascular risk. (vote yes/n0 requested).
`
`The outcome was 14 “yes” and 2 “no” votes.
`
`Following this advisory committee meeting, the FDA issued a Final Guidance to Industry in
`December 2008 titled, Diabetes Mellitus — Evaluating Cardiovascular Risk in New
`Antidiabetic Therapies to Treat Type 2 Diabetes. With its release, the FDA also publicly
`announced that the recommendations in this guidance will be applied to all ongoing diabetes
`development programs and marketing applications pending before the agency. In order to gain
`approval, applicants must compare the incidence of important cardiovascular events occurring
`with the investigational agent to the incidence of the same types of events occurring with the
`control group to show that the upper bound of the two-sided 95 percent confidence interval for
`the estimated risk ratio is less than 1.8.
`
`(‘
`At the time of its issuance, the FDA had three NDAs under review:
`saxagliptin (Onglyza), and liraglutide (Victoza). Saxagliptin and liraglutide were each
`presented at a public advisory committee meeting on April 1 and 2, 2009, respectively.
`/
`j
`K
`.
`7‘
`Because none ofthese NDAs were conducted with knowledge of
`these new recommendations, the review division applied a uniform approach to assessing risk
`for these NDAs. This approach is clearly described by the clinical and statistical reviewers in
`their finalized review of this NDA and also in the advisory committee briefing materials. This
`memo will summarize how this applicant has met the new regulatory requirements for
`establishing sufficient cardiovascular safety for approval under Section 8.0.
`
`)
`
`M4)
`
`The advisory committee meeting for saxagliptin focused only on the cardiovascular risk
`assessment. An in-depth review of efficacy was not presented by FDA at that time; however,
`the applicant did provide data supporting a conclusion that therapy with saxagliptin results in
`significant reductions in HbAlc, as both monotherapy and in combination with several other
`anti-diabetic agents. The finalized statistical review by Ms. Mele provides greater detail of the
`efficacy findings, including variables which may have influenced efficacy and flaws in the
`study designs which must be considered in the interpretation of efficacy. Section 7.0 of my
`memo will present the highlights of her findings.
`’
`
`In addition to cardiovascular safety, signals identified in the nonclinical program that have also
`directed the clinical safety review are summarized in this memo. Some of these safety signals
`appear to be a class effect observed in several clinical development programs (e.g.,
`hypersensitivity reactions) or in the nonclinical toxicology programs (e.g., cutaneous lesions).
`Spontaneous postmarketing adverse event reports of pancreatitis for other incretin—based
`therapies have also necessitated a careful evaluation in this NDA.
`
`Page 4 of 2
`
`

`

`Division Director Review
`
`3. CMC/Device
`
`Saxagliptin tablets are available as 2.5 or 5 mg film-coated tablets. There are no outstanding
`CMC issues. Manufacturing site inspections were acceptable. Stability testing supports an
`expiry of 36 months for the 2.5 mg tablets supplied in 30— and 90-count bottles containing
`dessicant. The 5 mg tablets supplied in 30-, 90-, and SOD—count bottles containing dessicant or
`when stored in aluminum/aluminum blisters also have an expiry of 36 months. Recommended
`Storage conditions for all presentations is 25°C (77°F) with excursions permitted to 15°-30°C
`(59°-86°F).
`
`4. Nonclinical Pharmacology/Toxicology
`
`The pharmacology/toxicology reviewers have recommended approval of this NDA. In Dr.
`Alavi’s review he stated the following:
`
`“Subchronic and chronic toxicology studies in mice, rats, dogs, and monkeys identified several
`areas ofpotential human concern: a) brain lesions in male rats, [2) cutaneous lesions in
`I
`- cynomolgus monkeys andfootpaa’ cracks in dogs, c) maIfierations in embryofetal
`development in rats with saxagliptin/metformin combination, d) saxagliptin-related decreased
`in lymphocytes andplatelets and immune system. “
`
`For each of these areas of concern, both Drs. Alavi and Bourcier have provided a thorough
`scientific review, including relevance of findings to humans. I concur with'their conclusion
`that these concerns do not preclude the approval of this NDA but labeling will reflect these
`findings and two post—marketing required studies will be necessary under FDAAA to address
`teratogenicity concerns with the combined use of saxagliptin and metformin. In this section I
`will only highlight the first 3 issues. The clinical setting was deemed to be more appropriate
`for assessing the effects of saxagliptin on the hematopoietic and immune system.
`
`.
`4.1 Brain Lesions in Male Rats
`Brain lesions (predominantly in the corpus callosum) were noted o_nly in male Sprague-
`Dawley rats and at high doses (355x themaximum therapeutic dose of saxagliptin, based on
`_ AUC); From a series of mechanistic studies it was concluded that these lesions were the result
`of a gender and species-specific metabolism of saxagliptin. Rats express CYP2C11, an .
`androgen—regulated liver enzyme which causes the release of cyanide from saxagliptin
`resulting in the histopathological findings resembling what has been described in the literature
`for cyanide poisoning. Support for the conclusion that this toxicity is specific to the
`expression of this androgen-regulated liver enzyme was the absence of such findings when the
`study was conducted in castrated rats or in rats receiving cimetidine, a CYP2C11 inhibitor.
`
`, In humans, saxagliptin is predominantly metabolized in the liver by CYP3A4/5. Incubation
`studies of saxagliptin in human liver microsomes (CYP2C8, 2C18, and 2Cl9) did reveal small
`amOunts cf cyanide formation that were below the lower limits of quantitation. Given the
`absence of CYPZCll in humans and no notable detection of cyanide in HLM studies, the _
`
`Page 5 of 2
`
`

`

`Division Director Review
`
`pharmtox reviewers have concluded that the brain lesions noted in male rats at very high
`multiples of drug exposure have no clinical relevance.
`
`4.2 Cutaneous Lesions in Monkeys and Dogs
`
`Some other DPP4-inhibitors in development have been associated with peripheral skin lesions,
`cutaneous sores, peripheral edema, and severe swelling associated with CK and LFT
`elevations. As a result, all manufacturers are required to conduct a 13-week monkey study to
`evaluate the potential for causing the peripheral lesions which may be due to non—selectivity of
`the compound for other dipeptidyl peptidases. Minimal and reversible non-necrotizing
`cutaneous lesions were observed in several animals treated with saxagliption at exposures 2
`20x the clinical dose. Severe necrotizing lesions were observed only at 60x the clinical dose.
`In a 12-month dog study, minimal erosive lesions were noted on the paws but this was at
`exposures Z 35x the clinical dose.
`
`Given the high multiples of clinical exposures before any of these cutaneous lesions were
`noted, these findings are not considered to be of sufficient clinical risk.
`
`4.3 Einbryofetal Malformations
`
`Co-administration of saxagliptin and metformin was associated with a rare and serious neural
`tube defect (craniorachischisis) in two fetuses from a single dam in a rat reproductive
`toxicology study. This was not an expected finding as saxagliptin alone was not associated
`with any malformations at doses exceeding 1500x clinical exposure. In addition, such a
`malformation has not been observed in the nonclinical studies submitted for approval of
`metformin. The applicant provided literature to suggest this finding was related to metformin
`and its effect on folate metabolism; however, the applicant did not include a metformin-only
`arm to adequately assess this hypothesis. Given that most anti-diabetic drugs are co-
`administered with metformin, the pharmtox reviewers are recommending a more appropriate
`embryofetal toxicology study be conducted in both rats and rabbits involving a metformin-
`only, saxagliptin—only, and a combination arm as post—marketing required studies under
`FDAAA. Labeling will include the findings from the current study.
`I
`
`Several points which require discussion is the timing of these two embyrofetal toxicology
`studies. The pharmtox and clinical review disciplines did not feel that these two studies were
`necessary pre-approval. It was felt that while the neural tube defect finding is a serious
`finding, it occurred in only 2 fetuses from the same litter; Given the absence of teratogenicity
`findings for both saxagliptin and metformin monotherapies, there is a strong possibility that
`this was a spurious finding for which labeling could provide adequate data to inform
`prescribers on a theoretical risk such that a decision can be made regarding co-prescribing with
`metformin in women of childbearing potential.
`'
`
`There were discussions regarding whether the label should include a contraindication against
`the co-administration of saxagliptin and metformin. FDA’s Reproductive and Development
`Toxicblogical Subcommittee and the Director and Associated Director of
`pharmacology/toxicology did not deem this to be necessary. I concur with this
`
`Page 6 of 2
`
`

`

`Division Director Review
`
`_
`recommendation. The repeat reproductive toxicology studies will be completed by C
`J and submitted to the FDA by April 30, 20 C ) Upon review of these data, labeling
`will be updated if warranted.
`'
`
`5. Clinical PharmacologyIBiopharmaceutics
`
`Clinical pharmacology reviewers recommend approval of this NDA. As there are no internal
`disagreements regarding any recommendations made by this discipline, my memo will only
`' highlight relevant clinical findings, particularly th0se requiring special emphasis in labeling
`with respect to dosing instructions or special monitoring. '
`
`Saxagliptin is metabolized predominantly by CYP3A4/5 to BMS-S 10849 which is present in
`human plasma at 2 to 7—fold higher levels than the parent drug. Although EMS-510849 is an
`active metabolite, it is less potent than the parent drug but has greater selectivity for DPP4
`overDPP8 than the parent drug. These attributes of the metabolite are reassuring as there is
`less concern for off-target toxicity with respect to cutaneous lesions. Animal toxicology
`studies have also included assessments of this metabolite and other minor metabolites.
`
`The kidney is the major route of elimination of the parent compound and the metabolite. As
`such, renal function affects the exposure 'of saxagliptin and its metabolite. Patients with severe
`renal impairment had a 2.1-fold increase in saxagliptin exposure compared to control subjects
`and both severe renal impairment and hemodialysis were associated with an approximate 4—
`fold increased in EMS-510849. Patients with moderate renal impairment had a less
`pronounced increased in exposure to saxagliptin (40%) and its metabolite (~3-fold); however,
`this was felt to be clinically relevant such that the lowest. proposed dose of 2.5 mg is
`recommended for patients with moderate and severe renal impairment and with ESRD. The
`applicant is currently conducting a dedicated safety trial in patients with renal impairment.
`
`Several drug-drug interaction studies were performed and discussed in detail in Dr.
`Vaidyanathan’s review. Interestingly, two DDI studies were performed with the strong
`CYP3A4/5 inhibitor, ketoconazole. The first one utilized a single high dose of saxagliptin 100
`mg with ketoconazole 200 mg q12 given for 6 days. This study resulted in a 25-fold increase
`in saxagliptin exposure and a 1.62-fold increase in Cmax. Not surprisingly, the metabolite
`exposure decreased. Because 14 out of 1'5 patients experienced a decline in lymphocyte
`counts on Day 10 following the co-administration of saxagliptin and ketoconazole, a second
`PK study was conducted which used saxagliptin 20 mg which now resulted in a 3.8-fold
`.
`increase in saxagliptin exposures; This study also showed a 30.6% decrease in absolute
`lymphocyte count which returned to baseline 72 hrs after study drug discontinuation; Since a
`near 4—fold exposure increase was observed with a strong CYP3A4/5 inhibitor that results in
`drug levels not evaluated in the Phase 3 program, I concur with clinical pharmacology’s
`recommendation to limit dosing of saxagliptin to 2.5 mg in patients receiving concurrent
`strong CYP3A4/5 inhibitors.
`
`Several DDP4 inhibitors in development have selected doses based on the drug’s ability at
`maintaining DPP4 inhibitory activity > 80% after 24 hours. For the two doses selected for
`
`Page 7 of 2
`
`

`

`Division Director Review
`
`' marketing, saxagliptin 2.5 and 5 mg administered for 14 days in patients with T2DM resulted
`in DPP4 inhibitory'activity of 37% and 65%, respectively, after 24 hours. Despite this, the
`degree of HbAlc reduction in the pivotal Phase 3 trials appears similar between saxagliptin
`and other DPP4 inhibitors (e.g., Sitagliptin) achieving a greater degree of DPP4 inhibitory
`activity. This would suggest that DPP4 inhibitory activity is not a reliable predictor of
`efficacy, particularly in Phase 2 dose-selection studies.
`.
`-
`
`Ms. Mele’s statiStical review observed a statistically significant interaction between Asian race
`and efficacy raising the'possibility of PK differences and possibly safety in Asians. From
`Figure 25 in Dr. Vaidyanathan’s review there was no difference in the clearance of saxagliptin
`between Asians and other races evaluated. The clearance of the metabolite was slightly
`elevated compared to other races. None of these changes would explain the significant
`interaction between Asian ethnicity and efficacy.
`
`6. Clinical Microbiology
`
`Not applicable.
`
`7. ClinicallStatistical-Efficacy
`Similar to other clinical development programs for anti-diabetic therapies, the primary efficacy
`endpoint for all pivotal Phase 3 trials was HbAlc, a validated surrogate for the reduction of
`microvascular complications associated with both type 1 and type 2 diabetes. Secondary
`endpoints which further evaluate the effect of drug On glycemic parameters included fasting
`plasma glucose (FPG), proportion of patients reaching a HbA 10 < 7%, and AUC from 0-180
`minutes for post-prandial glucose (PPG) in response to an oral glucose tolerance test (OGTT).
`
`' ThisNDA included clinical efficacy and safety from 8 Phase 2 and 3 trials. For the purpose of
`this section, my memo will primarily focus on the 6 pivotal Phase 3 trials summarized in Table
`7.1 below. Another study not discussed at length in my memo is Study CV181008 (008).
`This was a 12-wk, Phase 2 dose—ranging study which evaluated saxagliptin 2.5 mg, 5 mg, 10
`mg, 20 mg, 40 mg, and 100 mg which found statistically significant differences in mean
`percent change form Baseline in HbAl c relative to placebo but no dose response.
`
`Table 7.1 Summary of Pivotal Phase 3 Trials
`Study Number
`Treatment Groups
`(number randomized)
`
`Patient Population
`
`Study Duration (primary
`efficacy assessment
`duration listed first)
`
`~
`Monotherapy Trials
`CV181011
`
`CV181038
`
`Page 8 of 2
`
`Saxa 2.5 (n=102)
`Saxa 5 (n=107)
`Saxa 10 (n=98)
`Pbo (n=96)
`
`Saxa 2.5 qAM (n=n=74)
`Saxa 2.5 titrate to 5 qAM
`(n=71)
`Saxa 5 qAM (n=74)
`Saxa 5 qPM (n=72)
`
`Drug—na'l've
`Mean Baseline HbAlc (7.8-
`8.0)
`
`. Drug-naive
`Mean Baseline HbAlc (7 .8—
`8.0)
`
`24 weeks
`
`18 months+
`LT ongoing
`
`24 weeks
`'
`12 months+
`LT ongoing
`
`

`

`Division Director Review
`
`Add—on Trials
`
`CV181013
`
`CV181014
`
`CV181040
`
`'
`
`.
`
`Combination Trials
`
`CV181039
`
`Pbo (n=74)
`
`.
`
`-
`
`TZD+Saxa 2.5 (n=195)
`TZD+Saxa 5 (n=186)
`TZD+Pbo (n=184)
`
`TZD failures
`Mean Baseline HbAlc = 8.3
`.
`
`Met+Saxa 2.5 (n=192)
`Met+Saxa 5 (n=191)
`Met+Saxa 10 (n=181)
`Met+Pbo (n=179)
`
`’
`
`'
`
`'
`
`Metformin failures
`Mean Baseline HbAlc = 8.0
`.
`
`Gly7.5+Saxa 2.5 (n=248)
`GlyIO-I-Saxa 5 (n=253)
`Gly10+Pbo (n=267)
`
`SU failures
`Mean Baseline HbAlc = 8.4
`'
`
`Saxa 5+Met (n=320)
`Saxa 10+Met (n=323)
`Saxa 10 (n=335)
`Met (n=328).
`
`Drug-naive
`Mean Baseline HbAlc (9.4-
`9.6)
`
`24 weeks
`
`12 months+
`LT ongoing
`
`24 weeks
`
`12 months+
`LT ongoing
`
`24 weeks
`
`12 months+
`LT ongoing
`
`.
`
`24 weeks
`
`12 months+
`LT ongoing
`
`The primary efficacy analysis for all ofthese studies was at Week 24 and the study designs
`were similar in that they were all randomized and double-blinded trials and continuation into
`an extension period was not voluntary and included patients who required rescue therapy for
`glycemic control. The double-blind, randomized treatments were continued into the extension
`period, a characteristic of these studies which enable a better evaluation of safety (See Section
`8.0).
`
`As in other trials of anti-diabetic therapies, there are rescue criteria incorporated into the study
`designs to address progressive worsening of glycemic control, particularly for studies of < 6
`months duration. While this has become standard practice due to the notion that it is unethical
`to ignore worsening glycemic control in a clinical investigation, the addition of other anti—
`diabetic therapies or the discontinuation of study participants (not done in the saxagliptin
`program) presents challenges to the interpretation of drug efficacy. This is extensively
`discussed in Ms. Mele’s'review for each pivotal trial studied and separately in Section 3.1.4 of
`her review. The criteria for initiating glycemic rescue therapy can be found in Tables 5.4 and
`5 .5 of Dr. Lowy’s review. Noteworthy is that the time point for determining whether
`additional therapy is necessary is before Week 26, the time point for the primary efficacy
`analysis. Hence, there will be some proportion of patients at Week 26 who will have data
`from their last measured HbAlc prior to rescue therapy contributing to the overall efficacy
`analysis. Across all trials, poorer control of diabetes at Baseline (HbAlc, FPG) and a higher
`BMI predicted a higher incidence of rescue therapy.
`
`Across the monotherapy and add—on pivotal Phase 3 trials, the different doses of saxagliptin
`studied achieved statistically greater reduction in HbAlc from Baseline relative to placebo.
`The difference in mean adjusted HbAlc change ranged from -0.4 to -0.8%. There was no
`greater HbAlc reduction observed with saxagliptin 10 mg daily dosing. The applicant is
`proposing to market both the 2.5 and 5 mg doses with the 5 mg dose to be used in the general
`
`Page 9 of 2
`
`

`

`Division Director Review
`
`diabetic population while the 2.5 mg dose is reserved for patients with renal insufficiency.
`Although there is little evidence of a dose response between the 2.5 and 5 mg doses, Ms. Mele
`wrote a separate memo dated June 25, 2009 which specifically evaluated the 71 patients in
`Study 38 who had their dose titrated fiom saXagliptin 2.5 mg to 5.0 mg. Her review suggests
`that for some patients initiated on therapy at the 2.5 mg dose and who do not have an adequate
`glycemic response, upward titration to 5 mg may provide additional reductions in HbAlc.
`These data would support having both doses available but the applicant should not be allowed
`to promote the 5 mg dose as the recommended start dose for the majority of patients.
`
`'
`
`In Study 39 which evaluated the use of saxagliptin+metformin as initial therapy compared to
`the individual Components, the use of the two agents in combination provided greater
`reductiOns than saxagliptin 10 mg or metformin monotherapy. Ms. Mele noted that the
`absence of a saxagliptin 5 mg monotherapy would require a comparison of the saxagliptin 5
`mg + metformin treatment group to saxagliptin 10 mg monotherapy. Givenlthat all other
`Phase 3 trials have shown similar efficacy between saxagliptin Sand 10 mg, I believe the
`results from this comparison would yield a similar finding if the applicant had included a
`saxagliptin 5 mg, especially since the saxagliptin 5 mg + metformin efficacy is superior to
`saxagliptin 10 mg monotherapy (LS Mean 0.84; p<0.0001). A noteWorthy point made by Ms.
`Mele is that there was a significantly higher percentage of patients requiring rescue therapy in
`the saxagliptin 10 mg treatment arm (20%) than observed in the saxagliptin arms in other trials
`- despite the enrollment of treatment-naive patients. This might reflect the higher Baseline
`HbAl c in this trial. Regardless, it would suggest that saxagliptin 2.5 or 5 mg monotherapy
`would not be a reasonable initial therapy for patients with poor glycemic control.
`
`8. Safety -
`
`Drs. Lowy and Joffe have provided a detailed assessment on the safety of saxagliptin and have
`not identified a safety issue that will preclude the approval of this NDA. However, there are
`safety issues which have contributed to the postmarketing requirements outlined in the action
`letter which merit discussion in my memo.
`
`8.1 Cardiovascular Safety .
`As discussed under the Background section of this memo, this applicant was required to'show
`adequate CV safety with saxagliptin based on the recently implemented requirements outlined
`in the December 2008 FDA Guidance to Industry. The entire clinical development program
`intended for support of an NDA for saxagliptin'was designed and completed prior to the
`issuance of this Guidance. As a result, prospective adjudication of CV events was not possible
`for this application. Instead, a method for post-hoe evaluation of CV events collected in this
`NDA was proposed by FDA for saxagliptin and other NDAs pending before the FDA at the
`time the guidance was made public. Dr. Lowy’s review'and the background materials
`provided for the April 1, 2009 advisory committee meeting have outlined the details of the
`methodology for selecting CV events for a risk assessment. Table 8.1 summarizes the
`PreferredTerms which would be selected for analyses of “Broad MACE SMQ” and the more
`specific “FDA Custom MACE”.
`
`Page 10 of 2
`
`

`

`Division Director Review
`
`Table 8.1 Definitions for Different MACE anal ses
`— “Broad MACE SMQ”
`
`
`
`
`
`_
`
`-
`
`
`
`
`
`-
`
` xxxxxxxxxxxxxx
`
`Silent mxocardial infarction
`Tro - onin I increased
`
`-
`
`’
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`—
`
`><
`.
`.
`A _nosia
`Amaurosis fu_ax .
`
`..—
`
`
`><><><><
`
`
`
`_
`
`
`
`Page 11 of2
`
`

`

`
`
`“Broad MACE SMQ”
`
`“FDA Custom MACE”
`
`
`
`Division Director Review
`
`Carotid arterial embolus
`Carotid arteriosclerosis
`Carotid arte
`ane
`sm
`
`b p ass
`Carotid arte
`disease
`Carotid arte
`dissection
`Carotid arte
`insufficienc
`Carotid arte
`occlusion
`Carotid arte
`stenosis
`Carotid arte
`stent insertion
`Carotid arte
`thrombosis
`Carotid arte
`Carotid endarterectom
`Central nain s
`drome
`Cerebellar arte
`occlusion
`Cerebellar arte
`thrombosis
`Cerebellar embolism
`Cerebellar hematoma
`
`Cerebellar hemorrha_e
`Cerebellar infarction
`Cerebellar ischemia
`
`.
`
`Cerebral aneu sm ru-tured s philitic
`Cerebral arteriosclerosis
`
`Cerebral arteriovenous malformation hemorrhaic
`Cerebral arte
`embolism
`Cerebral arte
`occlusion
`Cerebral arte
`stenosis
`Cerebral arte
`thrombosis
`
`
`
`
`
`
`
`Cerebral hematoma
`Cerebral hemorrha e -
`Cerebral hemorrha e fetal
`Cerebral hemorrha _e neonatal
`Cerebral infarction
`Cerebral infarction fetal
`Cerebral ischemia
`Cerebral thrombosis
`Cerebral vasoconstriction
`Cerebral venous thrombosis
`Cerebrovascular accident
`Cerebrovascular accident r0 h laxis
`Cerebrovascular disorder
`Cerebrovascular insufficienc
`Cerebrovascular s - asm
`Cerebrovascular stenosis
`Charcot-Bouchard microaneur sms
`Cranial nerve ualsies multi le
`Di 0 le ia
`D sarthria
`Embolic cerebral infarction
`Embolic stroke
`Facial .als
`Hematom elia
`Hemi aresis
`
`
`
`
`
`
`
`
`
`
`
`Page 12 of2
`
`
`
`
`
`
`
`
`
`
`

`

`
`
`
`
`
`
`__
`
`_
`
`
`
`
`x x
`
`X X X X X
`
`x x
`
`x x x x x
`
`xx
`
`x x
`
`x
`x
`
`
`
`Division Director Review
`
`Hemorrha_ic cerebral infarction
`Hemorrhaic stroke
`
`'
`
`Intracerebral ane
`
`sm 0 eration
`
`'
`
`‘
`
`
`
`Transient ischemic attack
`Vascular ence halo ath
`Vertebral aite
`occlusion
`Vertebral arte
`stenosis
`
`,
`
`‘
`
`
`
`Page 13 of2
`
`
` “Broad MACE SMQ”
`
`
`“FDA Custom MACE”
`
`
`
`
`
`x
`.
`x
`
`
`Intracerebral hematoma evacuation —
`
`
`
`
`lntracranial aneu sm
`
`
`
`
`Intracranial hematoma
`
`Intraventricular hemorrha e
`
`
`
`
`
`Intraventricular hemorrha - e neonatal
`
`
`_ X
`lschemic cerebral infarction
`-
`X
`Ischemic stroke' _— _
`
`_
`
`_m:—_
`
`_—
`
`
`—_—
`
`
`Monoparesis ,
`
`Monoplegia
`
`Moxamoxa disease
`
`
`Paral sis
`Para] sis flaccid
`
`—_
`Paraaresis
`' ——
`
`___
`
`I___
`
`x
`X
`Pos rocednral stroke ——
`
`
`
`Precerebral anemcclusion _
`
`
`Putamen hemorrhage
`
`
`
`
`Qua—driparesis
`
`ouadri ule- ia
`
`Reversible ischemic neuroloic deficit _
`
`Ru-tured cerebral aneur sm
`
`l_——
`
`
`l___x
`
`
`
`
`
`S -inal enidural hemorrhage
`'
`x —
`
`
`.
`S . inal hematoma
`x
`'
`~
`
`
`Stroke in evolution
`x
`'
`.
`
`
`
`
`'
`'
`Subarachnoid hemorrha e
`x
`
`
`
`
`Subarachnoid hemonrhae neonatal ——
`
`Subdural hemorrhae ——
`
`
`Subdural hemonhae neonatal _—
`Thalamic infarction
`'
`
`Thalamus hemonha-e
`
`
`Thrombotic cerebral infarction
`
`Thrombotic stroke
`
`
`
`
`
`' —
`’
`x
`x
`
`X X X X
`
`
`
`x x
`
`

`

`Division Director Review
`
`
`
`
`
`“Broad MACE SMQ” ’ “FDA Custom MACE”
`
`
`
`Vertebral artery thrombosis
`
`
`
`Vertebrobasilar insufficiency
`
`
`Visual midline shift 3
`drome
`
`
`Wallenber; s ndrome
`
`
`The following table is obtained directly from Ms. Mele’s review and was also presented at the
`April 1, 2009 advisory committee meeting.
`
`Table 3...-7 .1 Summau of MACE Results"
`Common Odds Ratio
`
`
`
`Saxagliptin
`Comparator
`
`(u=1251)
`(11:33:16)
`_ Stratified on Study
`
`
`
`
`(95% CI)
`
`
`
`
`
`CusstomMACE
`
`4 (0.1%)
`7 (0.6%)
`0.21 (0.04, 0.8)
`
`
`
`
`
`ST+LT
`0.52 (0.31.0
`
`
`
`SMQ MACE
`
`
`
`25 (2.0%)
`0.90 (0.6. 1.5)
`58 (1.8%)
`ST
`
`
`
`0.96 (0.7, 1.4)
`'
`100 (3.14%)
`41 (3.2%)
`ST+LT
`*The. ST+LT database for the. FDA analyses is the 120-day safety update database
`
`
`
`From these results the majority of the AC members cOncluded that saxagliptin has satisfied the
`requirements for approval with respect to excluding the 1.8 goal post. While one can also
`argue that a more definitive assessment of CV risk which excludes the l. 3 upper bound of the
`95% CI has also been satisfied by the applicant, the AC members unanimously voted for
`additional cardiovascular safety assessment in the post-marketing setting. In light of the low
`CV event rate (reflecting a low risk population studied thus far) and the absence of prospective
`adjudication of events, the FDA will require a postmarketing trial to provide'a more definitive
`CV risk assessment for saxagliptin.
`
`,
`8.2 Hypersensitivity Reactions
`Shortly after the approval of Januvia (sitagliptin), spontaneous postmarketing reports of
`allergic and hypersensitivity reactions were received resulting in labeling changes to the
`Warnings and Precautions section as follows:
`
`There have been postmarketing reports ofserious allergic and hypersensitivity reactions in
`patients treated with Januvia such as anaphylaxis, angioedema, and exfoliative skin
`conditions including Stevens-Johnson syndrome.
`
`Other DPP4-inhibitors have also had clinical and nonclinical findings of hypersensitivity—like
`reactions. Alogliptin had hypersensitivity—like reactions in its chronic dog studies and a higher
`rate of similarly-coded reactions was observed with alogliptin in the clinical trials.
`Noteworthy were two cases of angioedema including one which had a positive rechallenge and
`dechallenge.
`
`In this NDA, there was an imbalance in hypersensitivity reactions not favoring saxagliptin
`(2.4%; 50 caSes vs 0.6%; 5 cases) when evaluating the pooled Phase 3 monotherapy and add—
`
`Page 14 on
`
`

`

`Division Director Review
`
`on trials, including the lZO—day safety update. From Table 13 in Dr. Joffe’s review, the
`majority of these hypersensitivity reactions was coded only as hypersensitivity (saxagliptin
`n=1 8, placebo n=0) followed by urticaria (saxagliptin n=l6, placebo n= 2). Angioedema was
`reported in one saxagliptin-treated patient. Drs. Joffe and Lowy have provided selected
`narratives of s