`
`JUL
`
`3 2014
`
`Timothy P. Walbert
`Chainnan, President, and CEO
`Horizon Pharrna, Inc.
`520 Lake Cook Road, Suite 520
`Deerfield, IL 60015
`
`Food and Drug Administration
`10903 New Hampshire Avenue
`Building #51
`Silver Spring, MD 20993
`
`Re: Docket No. FDA-20 14-P-0209
`
`Dear Mr. Walbert:
`
`This letter responds to the citizen petition submitted to the Food and Drug Administration
`(FDA) by Horizon Pharma, Inc. (Horizon), dated February 3, 2014 (Petition). In the Petition,
`Horizon asks FDA to take certain actions with regard to any abbreviated new drug
`application (ANDA) for which Vimovo (naproxen and esomeprazole magnesium) Delayed(cid:173)
`Release Tablets (new drug application (NDA) 022511, held by Horizon) is the reference
`listed drug (RLD).
`
`Specifically, Horizon requests that FDA:
`
`1. Require that any application listing Vimovo as the RLD, which seeks approval for a
`generic product that employs less than a complete enteric coating around the
`naproxen component, be supported by either
`
`a. pharmacokinetic data sufficient to show that the timing of release of naproxen
`in the generic product is equivalent to that of Vimovo, or
`
`b. data from clinical trials demonstrating that the proposed generic product does
`not cause more frequent or more severe gastrointestinal adverse events than
`does Vimovo;
`
`2.
`
`If additional clinical trials described above are necessary to support safety of the
`proposed generic product, require that any such application be filed as a 505(b)(2)
`application, I rather than as an ANDA; and
`
`3. Refuse to approve any currently pending ANDA for such a product, and require the
`applicant to withdraw its ANDA and resubmit it as a 505(b)(2) application,
`accompanied by the additional clinical testing described above.
`
`Petition at 2.
`
`I An~pplication submitted under section 505(b)(2) ofthe Federal Food, Drug, and Cosmetic Act (FD&C Act)
`(2 I U.S.c. 355(b)(2)).
`
`I
`
`HZ167631
`
`Page 1
`
`
`
`Docket No. FDA-2014-P-0209
`
`We have carefully considered the information submitted in the Petition. For the reasons
`stated below, the Petition is denied.
`
`I.
`
`BACKGROUND
`
`A.
`
`Vimovo
`
`On April 30, 2010, FDA approved an NDA for Vimovo (naproxen and esomeprazole
`magnesium) Delayed-Release Tablets (NDA 022511, held by Horizon). Vimovo is currently
`approved for relief of signs and symptoms of osteoarthritis, rheumatoid arthritis, and
`ankylosing spondylitis, and to decrease the risk of developing gastric ulcers in patients at risk
`of developing non-steroidal anti-inflammatory drug (NSAID)-associated gastric ulcers. 2
`Vimovo is available in two strengths: 375 milligrams (mg) ofnaproxen with 20 mg of
`esomeprazole, and 500 mg of naproxen with 20 mg of esomeprazole. The 500-mg
`naproxen/20-mg esomeprazole strength is the RLD.
`
`The active ingredients of Vimovo are naproxen, which is an NSAID with analgesic and
`antipyretic properties, and esomeprazole magnesium, which is a proton pump inhibitor (PPI)
`that suppresses gastric acid secretion. Vimovo tablets consist of an enteric-coated naproxen
`core surrounded by an immediate-release esomeprazole magnesium layer. The enteric
`coating prevents naproxen from being released when gastric pH is under 5.5.3
`
`The Vimovo application referenced two previously approved NDAs:
`
`•
`
`EC-Naprosyn (naproxen) Delayed-Release Tablets (NDA 020067, held by Roche
`Palo), and
`• Nexium (esomeprazole magnesium) Delayed-Release Capsules (NDA 021153, held
`by AstraZeneca).
`
`EC-Naprosyn is available as enteric-coated tablets containing 375 or 500 mg ofnaproxen for
`oral administration. The dissolution of the enteric-coated naproxen tablet is pH-dependent
`with rapid dissolution above pH level 6. There is no dissolution below pH level 4. Nexium
`is available as delayed-release capsules containing 20 or 40 mg of esomeprazole (present as
`22.3 mg or 44.5 mg of esomeprazole magnesium trihydrate) in the form of enteric-coated
`granules5 that protect the esomeprazole from being degraded in the stomach's acidic
`environment.
`
`2 See labeling for Vimovo approved on March 27, 2014, available at
`http://www.accessdata.fda.gov/drugsatfda docs/labell20 14/02251 IsO 10lbl.pdf.
`
`3 Id. at section 12.1 (Clinical Pharmacology: Mechanism of Action).
`
`4 EC-Naprosyn label approved on March 22, 2013, available at
`http://www.accessdata.fda.gov/drugsatfda docs/label/20 13/017581 s111,0181648061,0189655020,020067s0181
`bl.pdf.
`
`5 Nexiwn label approved on March 27, 2014, available at
`http://www.accessdata.fda.gov/drugsatfda docs/labell20 14/021153s046,021957sO 15,022101sO 12lbl.pdf.
`
`2
`
`HZ167632
`
`Page 2
`
`
`
`Docket No. FDA-2014-P-0209
`
`The Vimovo application was also supported by two clinical studies that were submitted to
`support the efficacy ofVimovo in reducing the risk of developing NSAID-associated gastric
`ulcers. The SOO-mg naproxen dose of the fixed-combination Vimovo was compared with
`enteric-coated naproxen in both studies. Both studies showed that Vimovo, given as a 500(cid:173)
`mg naproxen/20-mg esomeprazole tablet twice daily, showed a statistically significant
`reduction in the 6-month cumulative incidence of gastric ulcers, as compared to enteric(cid:173)
`coated naproxen given as a 500 mg tablet twice daily.6 The approval of the other dose level
`ofVimovo (375-mg naproxen/20-mg esomeprazole) and approval ofthe indications for use
`in treatment of rheumatoid arthritis and ankylosing spondylitis were based on the approval of
`EC-Naprosyn and bioequivalence (BE) studies between Vimovo and EC-Naprosyn.
`
`B.
`
`Statutory and Regulatory Basis for Approving 505(b)(2) Applications
`andANDAs
`
`Section 505(b) of the FD&C Act establishes the approval requirements for NDAs. To be
`approved, an application submitted under 505(b) must, among other things, be supported by
`well-controlled investigations showing the drug product to be safe and effective.s One
`pathway under section 505(b) provides for approval ofNDAs that are supported entirely by
`investigations either conducted by the applicant or to which the applicant has a right of
`reference (stand-alone NDA). The 1984 Drug Price Competition and Patent Term
`Restoration Act (Hatch-Waxman Amendments) provided an alternate pathway under section
`50S(b)(2) for approval of an NDA for which some or all of the safety and efficacy
`investigations relied upon for approval were not conducted by or for the applicant and for
`which the applicant has not obtained a right of reference or use.9 Like a stand-alone NDA, a
`505(b)(2) application is approved under section 505(c) of the FD&C Act.
`
`The Hatch-Waxman Amendments also provide for submission of ANDAs for approval of
`generic versions of listed drugs.! 0 A listed drug is a drug product with an effective approval
`under section 505(c).I! The ANDA process shortens the time and effort needed for approval
`by, among other things, allowing an ANDA applicant to rely on FDA's previous finding of
`safety and effectiveness for a listed drug rather than requiring the ANDA applicant to repeat
`the studies conducted to support approval of the listed drug. To rely on such a finding, the
`ANDA applicant must show that, among other things, its proposed drug product is the same
`
`6 Vimovo label approved on March 27, 2014 (supra note 2) at section 14 (Clinical Studies). See also Goldstein
`JLl, Hochberg MC, Fort JG, Zhang Y, Hwang C, Sostek M. Clinical trial: the incidence ofNSAID-associated
`endoscopic gastric ulcers in patients treated with PN 400 (naproxen plus esomeprazole magnesium) vs. enteric~
`coated naproxen alone. Aliment Pharmacol Ther. 2010 Aug; 32(3):401-413.
`7 In addition, two clinical studies were conducted to evaluate the efficacy ofthe 500 mg naproxen component of
`Vimovo for treating the signs and symptoms of osteoarthritis. The results of these two studies provided
`sufficient evidence to support the efficacy of Vimovo for the indication of treatment of signs and symptoms of
`osteoarthritis.
`8 Section 505(b)(l) of the FD&C Act.
`
`9 The Vimovo NDA was itself submitted under section 505(b)(2).
`10 Section 5050) of the FD&C Act.
`11 21 CFR 314.3(b).
`
`3
`
`HZ167633
`
`Page 3
`
`
`
`Docket No. FDA-2014-P-0209
`
`as the listed drug with respect to its active ingredient, dosage fOlID, strength, route of
`administration, and, with certain exceptions, labeling, and that the proposed product is
`bioequivalent to the listed drug. 12
`
`C.
`
`Bioequivalence and ANDAs
`
`The basic assumption underlying the Hatch-Waxman Amendments is that bioequivalent drug
`products that meet the FD&C Act's criteria are therapeutically equivalent and may be
`substituted for each other. A generic drug product is bioequivalent to the RLD "if the rate
`and extent of absorption of the drug do not show a significant difference from the rate and
`extent of absorption of the listed drug when administered at the same molar dose of the
`therapeutic ingredient" (section 5050)(8)(B)(i) of the FD&C Act). FDA regulations at 21
`CFR 320.1 (e) specify that
`
`[Bjioequivalence means the absence of a significant difference in the rate and
`extent to which the active ingredient or active moiety in pharmaceutical equivalents
`or pharmaceutical alternatives becomes available at the site of drug action when
`administered at the same molar dose under similar conditions in an appropriately
`designed study.13
`
`FDA's regulations at 21 CFR 314.94(a)(7) set forth the BE requirements for an ANDA, and
`the regulations in part 320 (21 CFR Part 320) set forth the procedures for detennining BE.
`The regulations discuss the various methods of establishing BE in general descending order
`of accuracy, sensitivity, and reproducibility. These methods include pharmacokinetic (PK)
`studies, pharmacodynamic (PD) studies, comparative clinical trials, and in vitro studies (21
`CFR 320.24). In addition, as under section 505(j)(8)(C) of the FD&C Act, section
`320.24(b)(6) ofthe regulations states that FDA has the flexibility to use "[a]ny other
`approach deemed adequate by FDA to ... establish [BE]." It is well-accepted that FDA has
`considerable discretion in determining how the BE requirement is met. FDA's discretion
`need only be based on a "reasonable and scientifically supported criterion, whether [the
`Agency] chooses to do so on a case-by-case basis or through more general inferences about a
`category of drugs....,,14 Courts have expressly upheld FDA's regulatory implementation of
`the FD&C Act's BE requirements. IS
`
`Standard BE PK studies are conducted using a two-treatment crossover study design. Single
`oral doses of the test and reference drugs are administered, and blood or plasma levels of the
`drug are measured over time. The rate and extent of drug absorption are statistically
`measured. The relevant PK parameters calculated from these data include the area under the
`
`12 Section 5050)(2) of the FD&C Act.
`13 See also 21 CFR 320.23(b).
`
`14 Bristol-Myers Squibb Co. v. Shalala, 923 F. Supp 212, 218 (D.D.C. 1996) (quoting Schering Corp. v.
`Sullivan, 782 F. Supp 645, 651 (D.D.C. 1992), vacated as moot sub nom, Schering Corp. v. Shalala, 995 F.2d
`1103 (D.C. Cir. 1993).
`15 See, e.g., Schering Corp. v. FDA, 51 F.3d 390,397-400 (3d Cir. 1995); Fisons CO/po v. Shalala, 860 F. Supp.
`859,863-67 (D.D.C. 1994).
`
`4
`
`HZ167634
`
`Page 4
`
`
`
`Docket No. FDA-2014-P-0209
`
`plasma concentration versus time curve (AVC), calculated to the last measured concentration
`time (AVCo_I), and AVC extrapolated to infinity (AUCoo). These parameters represent the
`extent of absorption. Another relevant PK parameter is the maximum or peak drug
`concentration (Cmax). Cmax is used to reflect the rate of absorption.
`
`To establish BE, the calculated 90 percent confidence interval for the log transformed ratio
`of geometric means for AVC and Cmax values between the generic product and the RLD
`should fall entirely within an 80 percent to 125 percent acceptance interval (0.8 to 1.25).16
`The use ofan 80 to 125 percent acceptance interval to compare two products with the same
`active ingredient, dosage form, route of administration, and strength is a scientific judgment
`about the best statistical practices for BE determinations. It reflects decades of scientific data
`on the variability of product characteristics within and between batches, as well as biological
`variability in patients. Because the point estimate of the test/reference ratio lies in the center
`of the 90 percent confidence interval, the mean of the data is usually close to 100 percent (a
`test/reference ratio of 1).17 The PK parameter time to reach maximum concentration (Tmax) is
`also a rough indicator of the rate of drug absorption and serves as supportive data in
`evaluating BE.
`
`For modified-release drug products, FDA may consider an additional PK parameter, lag
`time. 18 Lag time means the time from administration of a drug product until release and
`absorption of one or more active ingredients of that product (as reflected by quantifiable
`plasma concentrations). 19 For example, for an oral delayed-release tablet, the effect of an
`enteric coating intended to slow release of the active ingredient would be reflected in PK
`data as the lag time.
`
`D.
`
`Product-Specific Bioequivalence Guidance
`
`Under a process set forth in the guidance for industry on Bioequivalence Recommendations
`for Specific Products (BE Recommendation Process Guidance), FDA develops
`
`16 See FDA's Approved Drug Products with Therapeutic Equivalence Evaluations (Orange Book), Chapter 1.3,
`"Statistical Criteria for Bioequivalence," available at
`http://www.fda.gov/downloadslDrugslDevelopmentApprovaIProcess/UCM071436.pdf; See also FDA
`guidance for industry on Bioavailability and Bioequivalence Studies for Orally Administered Drug Products(cid:173)
`General Considerations (Revision I), available at
`http://www.fda.gov/Drugs/GuidanceComplianceRegulatofyln.formation/Guidances/default.htm.
`
`17 In general, for the 90 percent BE confidence interval to fall between 0.8 and 1.25, the point estimate should
`fall within .90 and 1.11.
`
`18 See, e.g., FDA guidance for industry on Food-Effect Bioavailability and Fed Bioequivalence Studies,
`available at http://www.fda.gov/downloads/regulatoryinfonnation/guidances/ucm126833.pdf
`at 6"7.
`
`19 See, e.g, description of lag time with regard to delayed-release products in FDA guidance for industry on
`Bioavailability and Bioequivalence Studies for Orally Administered Drug Products - General Considerations
`(Revision I), supra note 16, at 14 (stating "[a]s defined in the USP, delayed-release drug products are dosage
`forms that release the drugs at a time later than immediately after administration (i.e., these drug products
`exhibit a lag time in quantifiable plasma concentrations)").
`
`5
`
`HZ167635
`
`Page 5
`
`
`
`Docket No. FDA-2014-P-0209
`
`recommendations for demonstrating BE of specific products.2o These recommendations are
`posted on the FDA Drugs guidance Web page21 in draft and announced periodically by
`publication of a notice of availability (NOA) in the Federal Register. At any time following
`the posting of a draft recommendation on the FDA's Web site, a member of the public may
`submit comments on that draft recommendation by following the instructions listed on the
`Web site. Following the close of the comment period that is announced in the NOA, FDA
`considers all comments received and issues either final recommendations or revised draft
`recommendations for further comment.
`
`In March 2011, FDA issued a draft BE guidance regarding naproxenJesomeprazole
`magnesium delayed release tablets (the Naproxen/Esomeprazole Draft Guidance)?2 This
`draft guidance provides detailed recommendations for BE studies to applicants submitting
`ANDAs for which Vimovo is the RLD. The draft guidance recommends conducting two
`single-dose, two-way, crossover in vivo BE studies - one fasting and one fed - with PK
`endpoints in healthy subjects using 500 mg/EQ 20 mg base strength naproxenlesomeprazole
`magnesium tablets. In vivo studies ofthe lower strength tablets may be waived if certain
`conditions are met.
`
`II.
`
`DISCUSSION
`
`The Petition requests that any ANDA for a product that lists Vimovo as the RLD and
`includes a portion of naproxen outside of the enteric-coated core include PK or clinical data
`beyond that currently recommended in the NaproxenlEsomeprazole Draft Guidance (Petition
`at 2 and 8). In particular, the Petition asks FDA to require either (l) PK data sufficient to
`show that the timing of release of the naproxen from such a product is equivalent to that of
`Vimovo, or (2) clinical trial data demonstrating that the proposed product does not result in
`more frequent, or more severe, gastrointestinal adverse events than Vimovo (Petition at 2 and
`8). Finally, the Petition states that an application supported by such clinical trial data should
`be required to be submitted under section 505(b)(2) rather than section 5050) of the FD&C
`Act.
`
`FDA does not agree that this additional testing should be required. As discussed below, we
`do not agree with the Petition's claims regarding the clinical significance of the relative
`timing of the release of esomeprazole and naproxen. Accordingly, we do not believe any
`testing beyond that recommended in the existing draft BE guidance is necessary either to
`evaluate BE or to ensure comparable safety. Given that FDA does not agree that such
`additional testing is necessary, we need not consider whether, if clinical studies were
`
`20 The BE Recommendation Process Guidance is available at
`http;/Iwww.fda.goY/Drugs/GuidanceComplianceRegulatoryInfonnation/Guidances/ucrnO?5207.htm.
`
`21 Available at
`http://www.fda.gov/Drugs/GuidanceComplianceRegulatoryinformation/Guidanccs/ucm075207.htm.
`
`22 Draft Guidance for Esomeprazole Magnesium; Naproxen, available at
`http://www.fda.goY/downloads/Drugs/GuiQ.am:eComplianceRegulatoryInformation/Guidances/lJCM249220.pd
`f.
`
`6
`
`HZ167636
`
`Page 6
`
`
`
`Docket No. FDA-2014-P-0209
`
`necessary, an application containing data from such studies would need to be submitted
`under section 505(b)(2) rather than 5050) of the FD&C Act.
`
`A.
`
`The Petition Fails to Establish that the Release of Esomeprazole
`Before Naproxen Contributes Materially to the Gastroprotective
`Effect of Vimovo.
`
`The Petition states that Vimovo has been "specifically formulated to allow esomeprazole (a
`proton pump inhibitor) to achieve its gastroprotective impact before naproxen is released into
`the system" (Petition at 5). According to the Petition, "by the time Vimovo begins to release
`its naproxen, Vimovo's esomeprazole has already begun to impact gastric pH. This provides
`a less favorable environment for the development of gastric ulcers" (Petition at 7).
`In contrast, according to the Petition, "a generic product without all of the naproxen
`enterically coated would release at least some portion of its naproxen in a less
`gastroprotective environment, before gastric pH has begun to rise" (Petition at 7). According
`to the Petition, "a reduction in the enteric coating surrounding the naproxen core could
`subject patients to significantly increased risk of potentially fatal side-effects" (Petition at 7).
`
`The Petition provides no data (and FDA is not aware of any data) showing that the
`gastroprotective effect of the esomeprazole is related to or materially enhanced by the fact
`that in the Vimovo formulation, esomeprazole is released before naproxen. The two clinical
`trials described in section LA. of this response do not support the petitioner's claim that the
`relative timing of release of esomeprazole and naproxen is clinically meaningful. Rather,
`they demonstrate that the combination of esomeprazole and naproxen in Vimovo results in a
`reduced risk of gastric ulcers, compared to naproxen alone. Moreover, FDA's approval of
`Vimovo does not reflect a finding as to the clinical significance, if any, of the sequence of
`release of the active ingredients.23
`
`In addition, we note two properties of esomeprazole that are inconsistent with the Petition's
`hypothesis that Vimovo's release of esomeprazole before naproxen has a material beneficial
`effect on the product's safety profile. First, it takes several days for the esomeprazole in
`Vimovo to achieve its full PD effect on gastric pH,24 after which this effect plateaus,zs Once
`
`23 See Vimovo labeling, supra note 2; see also Summary Review, NDA No. 022511, available at
`http://www.accessdata.fda.gov/drugsatfda docs/nda/20 10/022511 Orig1sOOOSumR.pdf.
`
`2~ See, e.g., Miner P, Jr., et a!., Clinical trial: evaluation ofgastric acid suppression with three doses of
`immediate-release esomeprazole in thejixed-dose combination ofPN 400 (naproxenlesomeprazole magnesium)
`compared with naproxen 500 mg and enteric-coated esomeprazole 20 mg: a randomized, open-label Phase I
`study in healthy volunteers, Aliment Pharmacol Ther 2010; 32: 414-424 (study data cited in the Petition at 6
`and in the Vimovo NDA). In this study ofnaproxen/esomeprazole magnesium, the antisecretory effect of
`esomeprazole was assessed on days I and 9 and the antisecretory effect on day 9 was significantly greater than
`on day 1. The mean percent time intragastric pH was greater than 4.0 ranged from 13% to 28% on day 1 (20.5%
`for Vimovo) and from 41.1% to 76.8% on Day 9 (71.4% for Vimovo).
`
`25 By "plateau" we mean that if the PD effect is compared at the same time each day, the effect reaches a
`maximum after 4-5 days of dosing. See, e.g., Prilosec labeling approved on March 27,2014, at p. 21
`(Pharmacodynamics Section), available at http://www.accessdata.fda.gov/drugsatfda docs/label/
`2014101981Os098,022056s0.l41hl.p(.lf("[t]he inhibitory effect of omeprazole on acid secretion increases with
`repeated once-daily dosing, reaching a plateau after four days").
`
`7
`
`HZ167637
`
`Page 7
`
`
`
`Docket No. FDA-2014-P-0209
`
`gastric pH has reached its plateau, it appears that the release of esomeprazole before
`naproxen would be immaterial. Accordingly, even if there were a safety benefit associated
`with the release of esomeprazole before naproxen, this advantage would likely apply only to
`the first few days of administration. But the Petition identifies (at 2) a concern with
`significant GI adverse events associated with chronic NSAID use, not just the first few days
`ofNSAID use. Second, the non-enteric coated esomeprazole in Vimovo must be
`bioavailable in order to have a beneficial effect on gastric pH. Given that esomeprazole is
`acid-labile, meaning that it is not stable in acidic environments, its bioavailability is
`estimated to be only about 50% on the first day of administration for immediate-release
`esomeprazole formulations (such as Vimovo) as compared to enteric-coated delayed-release
`esomeprazolc products (such as Nexium), which release esomeprazole in a less acidic
`environment.2 This makes it less plausible that the early release of the esomeprazole in
`Vimovo somehow enhances the esomeprazole's GI-protective effects even in the first few
`days of administration.
`
`To the extent that the petitioner claims that the enteric coating of naproxen is responsible for
`Vimovo's effectiveness,27 we note that there has been considerable debate within the
`literature regarding whether surrounding NSAIDs in an enteric coating reduces gastric injury
`or ulceration. Early evidence in the literature suggesting a benefit28 does not seem to be
`supported by further research, and, as discussed below, subsequent data and analysis suggest
`that enteric-coated NSAIDs may have similar levels of risk for gastrointestinal injury as do
`uncoated NSAIDs.
`
`For example, there is evidence that enteric-coated, low-dose aspirin (also an NSAID) has no
`benefit over uncoated aspirin. One study concluded that "[t]he coating ofthe active principle
`in order to spare the stomach does not reduce the risk of upper gastrointestinal complications,
`neither for the stomach nor for the duodenum.,,29 Similarly, a meta-analysis of enteric(cid:173)
`coated aspirin studies30 concluded that "rnlo clinical benefits in terms ofreduction of
`gastrointestinal bleeding or ulceration with enteric coating have, therefore, been successfully
`demonstrated, although the endoscopic studies show that potentially these benefits could
`exist." Studies submitted during the approval of EC-Naprosyn (one of the drug products
`
`26 See, e.g., Clinical Pharmacology and Biophamaceutics Review(s), NDA No. 022511, available at
`http://www.accessdata.fda.gov/drugsatfda docs/ndaJ20 10/022511 Origl sOOOClinPharmR.pdf.
`
`27 For example, after discussing the multifactorial etiology ofNSAID-induced damage to the gastric mucosa
`and the role oflocally-present NSAIDs (Petition at 3 and Figure 1), the Petition goes on to state that "[t]he
`enteric coating of the NSAID in [Vimovo] prevents that exposure," and that "drugs that are not completely
`enteric coated might lead to ... increased damage by exposure of the drug in the stomach lumen" (Petition at
`3).
`
`28 See, e.g., Hawthorne, AB, Mahida, YR, Cole, AT, and Hawkey, CJ. Aspirin-induced gastric mucosal
`damage: prevention by enteric coating and relation to prostaglandin synthesis. Br. J. CHn. Pharmacal. (1991),
`32,77-83.
`
`29 De Abajo, FJ and Garcia Rodriguez, LA. Risk of upper gastrointestinal bleeding and perforation associated
`with low-dose aspirin as plain and enteric-coated fonnulations. BMC elin Pharmacal. (2001) 1: I. PMC32172.
`
`30 Walker, J, Robinson, J, Stewart, J, and Jacob, S. Does enteric-coated aspirin result in a lower incidence of
`gastrointestinal complications compared to normal aspirin? Interact CardioVasc Thorac Surg (2007) 6 (4): 519~
`522.
`
`8
`
`HZ167638
`
`Page 8
`
`
`
`Docket No. FDA-2014-P-0209
`
`cited in Vimovo's NDA) and discussed on the EC-Naprosyn label include a head-to-head
`comparison of enteric-coated naproxen vs. non-enteric-coated naproxen. The results of this
`comparison indicated that EC-Naprosyn and non-enteric-coated Naprosyn showed no
`significant differences in efficacy or safety and had a similar prevalence of minor 01
`complaints?1 In addition, in long-term open-label trials involving EC-Naprosyn, the rates for
`clinically-diagnosed peptic ulcers and GI bleeds were similar to what has been historically
`reported for long-term NSAID use.32 Another researcher notes that "[a]ttempts have been
`made to produce NSAIDs with reduced gastric irritancy through formulation in slow release
`of enteric-coated tablets or as prodrugs that require hepatic metabolism to be active.
`However, the incidence of gastroduodenal ulceration with these drugs is comparable to what
`is observed with standard NSAIDS.,,33
`
`In summary, although FDA agrees that Vimovo has demonstrated a gastroprotective effect,
`we do not agree that the available data demonstrate that the release of esomeprazole prior to
`the release of naproxen is material to that effect.
`
`B.
`
`The Petition Fails to Establish that the Presence of Any Naproxen
`Outside an Enteric Coating in ANDA Products Increases the Risk of
`Adverse Gastrointestinal Effects.
`
`The Petition states that several generic manufacturers have filed ANDAs referencing
`Vimovo as the RLD, and that Horizon has received notice of one ANDA for a proposed
`generic product that "contains a measurable amount ofnaproxen outside ofthe enteric
`coating" (Petition at I). In particular, the Petition states that Horizon received a Paragraph
`IV notice letter34 from a sponsor ofa proposed generic version ofVimovo that described the
`proposed product as follows:
`"A measurable portion of the entire amount of the naproxen (NSAID) in
`[the ANDA applicant's} proposed product is not surrounded by a coating
`that prevents release at a pH of less than 3.5 but instead is coated in a
`polymer that completely dissolves at a much lower pH, and is released at
`that pH "
`
`Petition at 1-2, footnote 3.
`
`31 See EC-Naprosyn label approved on March 22, 2013, at p. 7 (Clinical Studies Section) available at
`http://www.accessdata.fda.gov/drugsatfda docs/label/20 ] 3/017581 s] 11,018] 64s061 ,0 I 8965s020,020067sO 181
`bl.pdf (discussing the results ofthree 6-week, double-blind, multicenter studies with EC-Naprosyn (naproxen)
`(375 or 500 mg bid, n=385) and Naprosyn (375 or 500 mg bid, n=279) that were conducted comparing the two
`products and included 355 rheumatoid arthritis and osteoarthritis patients who had a recent history ofNSAID~
`related GI symptoms).
`
`12 l d. (discussing the results of long-tenn open-label trials involving 553 patients who received EC-Naprosyn
`(mean length oftreatment was 159 days».
`
`33 Wallace, J. Nonsteroidal Anti-inflammatory Drugs and Gastroenteropathy: The Second Hundred Years
`Gastroenterology (1997); 112: 1000-1016.
`
`34 Section SOSG)(2)(B) of the Act and 21 CFR 3 14.94(a)(12)(i)(A)(4) require that the generic drug applicant
`provide notice to the patent holder and NDA sponsor of any certification made in the ANDA that the patent in
`question is invalid or will not be infringed by the generic product (known as a "paragraph IV certification").
`
`9
`
`HZ167639
`
`Page 9
`
`
`
`Docket No. FDA-2014-P-0209
`
`The Petition states that such an ANDA product "'specifically obviates [Vimovo's] careful
`design" and would impair the gastroprotective benefits of esomeprazole by "allowing release
`ofa measurable amount of naproxen before the gastroprotective benefits of esomeprazole
`can take effect" (Petition at 1-2 and 5). The Petition claims that "a reduction in the enteric
`coating surrounding the naproxen core could subject patients to significantly increased risk
`of potentially fatal side-effects" (Petition at 7).
`
`FDA does not generally comment on pending ANDAs, including whether a given ANDA has
`in fact been submitted, and we will not do so in this case.35
`
`Also, for the reasons discussed above in section ILA, even if a proposed generic product
`referencing Vimovo were to release some amount of naproxen before all of the esomeprazole
`release had occurred,36 we do not agree with the Petition that there is reason to be concerned,
`in the abstract, that this could result in materially different gastrointestinal adverse event
`profiles.
`
`Accordingly, the available data do not lead us to conclude that an ANDA product with any
`amount of naproxen outside of the enteric-coated core would, for this reason alone, be
`associated with increased gastrointestinal injury risk compared to Vimovo.
`
`C.
`
`Additional Data or Testing Are Not Necessary for Approval of
`Certain ANDA Products
`
`The Petition states that additional data or testing should be required for any ANDA product
`that cites Vimovo as the RLD if the product employs less than a complete enteric coating
`around the naproxen component, and thus may include a portion ofnaproxen outside the
`enteric-coated core (Petition at 2 and 8).
`
`In particular, the Petition requests that applications for such products must include either:
`
`•
`
`•
`
`PK data sufficient to show that the timing of release of naproxen in the generic
`product is equivalent to that of Vimovo, or
`
`data from clinical trials demonstrating that the proposed generic product does not
`cause more frequent or more severe gastrointestinal adverse events than does
`Vimovo.
`
`We disagree. For the reasons discussed in section ILA of this response, we do not agree with
`the Petition's claims regarding the clinical significance of the release of esomeprazole before
`
`35 See 21 CFR 314.430(b).
`
`36 We note that if dissolution testing (also discussed in note 37 of this response) shows or predicts early
`naproxen release from a proposed generic formulation ofVimovo, such a product might not be considered a
`delayed-release product. In that case, the product would not be approvable as an ANDA to Vimovo because it
`would not be the same dosage form as Vimovo (unless it was submitted pursuant to an approved suitability
`petition for a different dosage form). See sections 505U)(2)(A)(iii) and 505G)(4)(D)(i) ofthe FD&C Act and 21
`CFR 314.94(a)(6) and 3 14.127(a)(4)(i).
`
`10
`
`HZ167640
`
`Page 10
`
`
`
`Docket No. FDA-2014-P-0209
`
`naproxen. For the reasons discussed in section ILB of this response, we do not agree with
`the Petition that the presence, in a proposed generic version ofVimovo, of some portion of
`the naproxen outside the enteric core would result in increased risk of gastrointestinal injury.
`
`Accordingly, were FDA to receive an ANDA for a generic version ofVimovo that, as the
`Petition puts it, "includes a portion ofnaproxen outside of the enteric coated core" (Petition
`at 8), we would not, solely on that basis, require any testing beyond that recommended in the
`existing draft BE guidance either to evaluate BE or to ensure comparable safety.37 Given
`that we do not agree that additional testing would be required, we need not consider whether,
`if such testing took the fonn of clinical studies, an application containing data from such
`studies would need to be submitted under section 505(b)(2) rather than 5050) of the FD&C
`Act.
`
`We note that this response addresses the Petition's requests generally. It shoul