CENTER FOR DRUG EVALUATION AND
`RESEARCH
`
`
`APPLICATION NUMBER:
`
`205834Orig1s000
`
`SUMMARY REVIEW
`
`
`
`
`

`

`Decisional Review for NDA 205834
`
`
`Date
`September 23, 2014
`From
`Debra Birnkrant, M.D.
`mm- Division Director's Summa Review
`NDNBLA #
`NDA 205834/Original Submission
`Su .
`. #
`Proprietary]
`Established
`
`' [ledipasvir (LDV)/sofosbuvir (SOF)]
`
`Harvoni
`
`
`
`USAN names
`
`Dosage forms I
`strength
`Proposed
`Indication(s)
`
`Fixed-dose combination, 90 mg/400 mg tablets, once daily
`
`Indicated for the treatment of chronic hepatitis C (CHC) in
`adults with Genotype 1 infection
`
`[EE— Auroval
`
`1. Introduction to Review: This Division Director’s memorandum provides a
`topline summary of NDA 205834 for Gilead Sciences’ New Drug Application
`(NDA) for the fixed—dose combination of ledipasvir, an NS5A inhibitor and
`sofosbuvir, a hepatitis C virus (HCV) nucleotide analog NS5B polymerase
`inhibitor for treatment of adult patients with Genotype 1 (GT1) infection. This
`decisional review summarizes pertinent findings from the original NDA
`submission and FDA’s multidisciplinary reviews and product labeling.
`
`2. Background/Regulatory History/Previous Actions/Foreign
`Regulatory Actions/Division of Scientific Investigations (DSI) Status:
`Chronic hepatitis C viral infection is a public health burden. Based on NHANES
`data from 2003-2010, it is estimated that 2.7- 3.9 million people in the United
`States are infected with the virus. Most are unaware that they are infected and
`consequently treatment is not reaching those in need. With chronic infection, it is
`projected that there will be an increase in cases of hepatocellular carcinoma
`(HCC) with an estimated peak incidence in 2019 of 14,000 cases/year and
`decompensated cirrhosis with a projected peak incidence in 2020 of >145,000
`cases/year (Davis, et al., Gastroenterology 2010) because the epidemic began
`decades ago with 75% of those with HCV in the United States born between
`1945-1965 (CDC).
`
`Treatment regimens have improved over the years based on better tolerability
`and enhanced effectiveness. In 2013, two new direct—acting antivirals were
`approved from two different classes. SOF, an HCV nucleotide analog NS5B
`polymerase inhibitor was approved in combination with ribavirin, with and without
`pegylated interferon for treatment of chronic hepatitis C (CHC) GT 1- 4 infection.
`SVR12 rates increased to 89% with this new direct-acting antiviral in combination
`with pegylated interferon and ribavirin (P/R) for CHC GT 1 viral infection.
`Simeprevir (SIM), an NS3/4A protease inhibitor was also approved for use in
`
`Reference ID: 3632863
`
`

`

`combination with P/R in 2013. Overall SVR12 rates were 80% for GT1a and
`GT1b patients naïve to treatment. Off-label use of the combination of SOF and
`SIM in the COSMOS trial was recently published. Overall SVR rates from the
`COSMOS trial were 93% for the populations enrolled (Lawitz, et al., Lancet,
`2014).
`
`Treatment of CHC infection allows for a chance of virologic cure. Virologic cure
`as measured by sustained virologic response (SVR) 12 weeks after completion
`of treatment is associated with histologic benefit, a decrease in all-cause and
`liver-related mortality, and decreases in rates of HCC and hepatic
`decompensation (van der Meer, et al. JAMA 2012). Regarding treatment for
`CHC, current standards are outlined in the 2014 American Association for the
`Study of Liver Diseases (AASLD)/Infectious Diseases Society of America (IDSA)
`treatment guidelines that were recently revised as of August 11, 2014.
`
`The NDA for the FDC was submitted on February 8, 2014 and reviewed under
`the PDUFA V program. As previously mentioned, SOF was approved in 2013.
`LDV is the first drug in its class and the FDC is recommended for use in HCV
`GT1 patients based on Phase 3 clinical trials that enrolled both treatment-naïve
`and treatment-experienced patients. Duration of treatment is dependent on
`whether cirrhosis is present and other baseline factors. For example, the FDC is
`recommended for use for 12 weeks duration for GT1 patients who are treatment-
`naïve. A treatment duration of 8 weeks can be considered in treatment-naïve
`patients without cirrhosis who have pre-treatment HCV RNA less than 6 million
`IU/mL based on an analysis of baseline predictors of relapse. A 24-week
`treatment duration is recommended for treatment-experienced patients with
`cirrhosis; treatment-experienced is defined as those patients who have failed
`treatment with either P/R or an HCV protease inhibitor plus P/R. See Table 1
`below that is excerpted from product labeling.
`
` Table 1 Recommended Treatment Duration
`Treatment Duration
`Patient Population
`Genotype 1
`Treatment-naïve
`+/- cirrhosis
`Genotype 1
`Treatment-
`experienced**
`without cirrhosis
`Genotype 1
`Treatment-
`experienced**
`With cirrhosis
`
`12 weeks*
`
`12 weeks
`
`24 weeks
`
` The FDC for 8 weeks can be considered in treatment-naïve patients
` without cirrhosis who have pre-treatment HCV RNA < 6 million IU/mL
`
`Reference ID: 3632863
`
`2
`
`

`

`The safety and efficacy of LDV/SOF in pediatric patients have not been studied.
`
`The application was granted a priority review because CHC is a serious and life-
`threatening disease and the FDC appeared to provide improvement in safety and
`effectiveness. In addition, the FDC was designated as a Breakthrough Therapy
`under FDASIA,Title IX because preliminary clinical evidence indicated substantial
`improvement over available therapies in the treatment of CHC-infected adults
`including the first ribavirin-free regimen for GT1.
`
`The original NDA submission contained clinical data from three Phase 3 trials in
`GT1 CHC viral infection and Phase 2 trials as well as other studies that
`supported dose selection. ION-1 and ION-3 trials were conducted in treatment-
`naïve patients and ION-2 was conducted in patients who had previously failed a
`P/R regimen including subjects who may have failed a PI-based regimen.
`
`Per Dr. El-Hage, Office of Scientific Investigations (OSI), six domestic Phase 3
`clinical trial sites underwent inspection. Overall, the data submitted from these six
`sites are considered acceptable in support of the pending application.
`
`The application was not presented before the Antiviral Drugs Advisory
`Committee because a preliminary review of the NDA, including labeling did not
`reveal any significant clinical or safety issues that would benefit from an advisory
`committee discussion.
`
`3. Chemistry/Manufacturing/Controls (CMC): The CMC reviewers of the
`LDV/SOF NDA are: Drs. George Lunn, Sandra Suarez and Steven Donald. Dr.
`Rapti Madurawe supervised the CMC review with Dr. Stephen Miller serving as
`CMC-Lead. The CMC team reviewed data to assure the identity, strength, purity
`and quality of the FDC; complete SOF drug substance data was provided in
`approved NDA 204671 and was incorporated in this NDA by reference. The CMC
`team is in agreement with the Applicant that the FDC contains 90 mg of LDV and
`400 mg SOF and concluded the following:
`
`The composition, manufacturing process and specifications for the FDC tablets
`are appropriate. Specifically, for LDV, there are two isolated intermediates that
`have acceptable specifications. Per Dr. Lunn’s review, a reasonable specification
`that includes tests for appearance, identity, acetone content, water, assay,
`impurities, residual solvents, and elemental impurities is also provided.
`
`The stability data contained in the FDC NDA support an expiry date of 24 months
`when stored at or below 30 degrees centigrade. Further, the container-closure
`system and labeling are appropriate.
`
`An inspectional report has been completed and the outcome is satisfactory.
`
`Reference ID: 3632863
`
`3
`
`

`

`4. Pharmacology/Toxicology: Please see review of submitted nonclinical
`toxicology studies by Dr. Christopher Ellis, supervised by Dr. Hanan Ghantous.
`
`The nonclinical safety profile of SOF has been previously reviewed under NDA
`204671 with the exception of two-year carcinogenicity studies that were
`reviewed with this application. Per Dr. Ellis’ review, the nonclinical safety profile
`of LDV has been evaluated in: safety pharmacology studies in rats and dogs;
`repeat-dose toxicology studies in mice, rats and dogs for up to 1, 6 and 9 months
`duration, respectively; up to two-week repeat-dose toxicology studies to qualify
`impurities; phototoxicity studies in mice and rats; fertility and pre- and post-natal
`developmental studies in rats; embryo-fetal developmental studies in rats and
`rabbits; and genetic toxicology studies (Ames, in vitro chromosomal aberration
`and in vivo rat micronucleus assays). In addition, numerous in vitro and in vivo
`nonclinical pharmacokinetic studies evaluating the absorption, distribution,
`metabolism and excretion of LDV have been conducted; rat and mouse
`carcinogenicity studies with LDV are currently in progress.
`
`Myocardial inflammation and degeneration occurred in rats administered oral
`GS-9851 doses, a 1:1 mixture of SOF and its diasteriomer, of 2,000 mg/kg/day in
`a 7-day toxicology study. Cardiac toxicity was not observed in rats administered
`oral doses of SOF up to 750 mg/kg/day for approximately 20 months, or in dogs
`and mice administered SOF up to 500 and 1,000 mg/kg/day for 9 and 3 months
`respectively, with corresponding exposures approximately 9-fold (rat), 17-fold
`(dog) and 24-fold (mouse) that in humans at the recommended SOF dose of 400
`mg once daily. Nonetheless, cardiac effects seen in nonclinical studies were
`further examined in clinical trials and are summarized in Dr. Sarah Connelly’s
`clinical review.
`
`As described in Dr. Ellis’ review, no clear target organs of toxicity were identified
`in repeat-dose toxicology studies in mice, rats and dogs administered LDV doses
`of up to 300, 100 and 30 mg/kg/day for 1, 6 and 9 months, respectively.
`Therefore, no specific overlapping toxicity of potential significant clinical concern
`was identified in animals administered LDV or SOF alone. However, a potential
`LDV-related mild hepatobiliary toxicity signal (not considered adverse and not
`clearly dose-dependent) was noted, with slight increases in ALP and/or ALT
`associated with increased liver/gall bladder weight (high-dose males only)
`without correlating histopathology changes observed in mice following oral
`administration of LDV at up to 300 mg/kg/day (AUC0-24hr~164 & 271 g.h/ml for
`LDV in females and males, respectively). In addition, minimal-to-slight random
`foci of hepatocyte necrosis (males) and bile duct hyperplasia (males and
`females) were noted in rats following oral administration of LDV at up to 100
`mg/kg/day (AUC0-24hr~56 g.h/ml for LDV). These non-adverse hepatobiliary
`findings were observed at LDV AUC exposure ~8- and 30-fold higher, in rats and
`mice respectively, than in humans at the recommended LDV dose. In addition,
`slight increases in cholesterol and triglycerides were noted in rats at 100
`mg/kg/day. In dogs, no clear clinically relevant LDV-related findings were
`
`Reference ID: 3632863
`
`4
`
`

`

`observed following oral administration of LDV at up to 30 mg/kg/day (AUC0-
`24hr~41.3 & 80.3 g.h/ml for LDV in males and females, respectively), resulting in
`LDV AUC exposure ~9-fold higher than in humans at the recommended LDV
`dose.
`
`The FDC is pregnancy category B.
`Ledipasvir was not genotoxic in a battery of in vitro or in vivo assays.
`Carcinogenicity studies of ledipasvir in mice and rats are ongoing.
`Two-year carcinogenicity studies of SOF were conducted in mice and rats and
`reviewed under this NDA. No increase in the incidence of drug-related
`neoplasms were observed at the highest doses tested in mice and rats, resulting
`in AUC exposure to the predominant circulating metabolite GS-331007 of
`approximately 4- and 18-fold (in mice) and 8- and 10-fold (in rats), in males and
`females respectively, compared to the exposure in humans at the 400 mg dose.
`5. Clinical Pharmacology: The Office of Clinical Pharmacology reviewers were
`Drs. Jenny H. Zheng and Leslie Chinn with secondary review provided by Dr.
`Shirley Seo; Jeffry Florian and Yaning Wang provided primary and secondary
`pharmacometrics reviews of the NDA, respectively.
`
`Multiple clinical studies were conducted to characterize the pharmacokinetics
`(PK) of LDV and SOF when administered either as single agents or as the FDC.
`Both intensive and sparse plasma concentration data from 391 healthy subjects
`and 2147 HCV-infected subjects who received LDV/SOF FDC, SOF and LDV
`administered together as single agents, or LDV as a single agent from 14 clinical
`studies (9 Phase 1, 2 Phase 2, and 3 Phase 3 studies) were used for population
`PK evaluation of SOF, its predominant circulating metabolite GS-331007, and
`LDV.
`
`Dose selection rationale for SOF can be found in the review of NDA 204671.
`Dose selection rationale for LDV was based on an abbreviated monotherapy trial
`and a Phase 2 dose-ranging trial where the 90 mg dose of LDV had higher SVR
`rates and lower breakthrough rates than a 30 mg once daily dose.
`
`Below, key PK characteristics for LDV as a single agent are summarized from Dr.
`Zheng’s review:
` A high-fat meal reduced LDV AUC and Cmax by approximately 45%.
` LDV is >99.8% bound to human plasma proteins.
` The half-life of LDV is approximately 47 hours
` Following a single 90 mg oral dose of [14C]-LDV, mean total recovery of the
`[14C]-radioactivity in feces and urine was approximately 87%, with most of the
`radioactive dose recovered from feces (approximately 86%). Unchanged LDV
`excreted in feces accounted for a mean of 70% of the administered dose and
`the oxidative metabolite M19 accounted for 2.2% of the dose. The data
`
`Reference ID: 3632863
`
`5
`
`

`

`indicate that biliary excretion of unchanged LDV is a major route of
`elimination with renal excretion being a minor pathway.
` LDV exhibits dose linearity for AUCinf and Cmax over the 3- to 100-mg range.
` LDV PK is not affected by race (as determined by both popPK analysis of
`Phase 3 data as well as one dedicated Phase 1 study in Japanese subjects)
`or age (18-80 years)
` Relative to healthy subjects, LDV AUC0-24 and Cmax were 24% lower and 32%
`lower, respectively, in HCV patients.
` AUC and Cmax of LDV were 77% and 58% higher, respectively, in females
`than males. After correcting for body weight differences between genders,
`females still have approximately 40% higher exposure as compared to males.
`This observation has no clinical relevance because neither response rate nor
`rate or severity of adverse events was significantly different between genders.
` Renal impairment has no clinically significant effect on LDV PK and no dose
`adjustment of LDV is needed. The effect of hemodialysis on LDV PK was not
`evaluated; however, due to the high protein binding of LDV, hemodialysis is
`unlikely to have a significant impact.
` Hepatic impairment (including cirrhosis) has no clinically significant effect on
`LDV PK, and no dose adjustment of LDV is needed for any degree of hepatic
`impairment.
` LDV solubility decreases as pH increases. Drugs that increase gastric pH are
`expected to decrease systemic concentrations of LDV.
`o A substantial decrease in LDV plasma exposure (~ 42% to 48% lower
`AUC and Cmax) was observed upon administration of the proton pump
`inhibitor (PPI) omeprazole (20 mg) 2 hours prior to LDV administration.
`o LDV absorption was unaffected upon simultaneous or staggered (12
`hours) administration of the H2-receptor antagonist (H2RA) famotidine (20
`mg).
` LDV is a substrate of the drug transporters P-gp and BCRP.
` LDV is an inhibitor of P-gp and BCRP.
` LDV is not expected to inhibit OATP1B1, OATP1B3 and BSEP at
`concentrations achieved in vivo at the recommended dose
` LDV 30 mg once daily increased SIM Cmax and AUC by 161% and 169%,
`respectively, due to P-gp inhibition, which is a similar magnitude of the effect
`of DRV/RTV on (SIM) and DRV/RTV is not recommended to be
`coadministered with simeprevir. Thus, SIM is not recommended to be
`coadministered with LDV/SOF.
` At the supratherapeutic dose of 120 mg twice daily, LDV does not prolong
`QTc to a clinically relevant extent.
`
`The PK characteristics of SOF and its major metabolite GS-331007are detailed
`in NDA 204671.
`
`Below, key PK characteristics for SOF, GS-331007 and LDV following oral
`administration of LDV/SOF FDC are excerpted from Dr. Zheng’s review; drug-
`drug interactions are depicted in Tables 3, 4 and 5 of product labeling:
`
`6
`
`Reference ID: 3632863
`
`

`

` LDV/SOF FDC can be administered without regard to food (as instructed in
`Phase 3 studies).
` LDV/SOF FDC tablets have been studied with antiretrovirals as well as acid
`reducing agents. The following are the results from these studies:
`o No clinically significant effects on SOF and GS-331007 exposures
`were observed with any of the above agents, which was similar to what
`was observed for SOF administered as a single agent.
`o The effects of these drugs on LDV AUC or Cmax are in the range of a
`34% decrease by Atripla to ~100% increase by ritonavir-boosted
`atazanivir. Safety data to support these shifts in plasma concentrations
`come from clinical trials where P-gp inhibitors were allowed; efficacy
`data come from a co-infected trial where all participants achieved SVR
`four weeks after the end of treatment (n=16).
`o LDV solubility decreases as pH increases. Consequently, drugs that
`increase gastric pH are expected to decrease the concentration of
`LDV. Dosing of antacids containing aluminum and magnesium
`hydroxide with the FDC should be separated by 4 hours. Famotidine
`(40 mg single dose) was administered either simultaneously with or 12
`hours apart from LDV/SOF, while omeprazole (20 mg) was studied
`only with simultaneous administration with LDV/SOF. Since the onset
`of the antisecretory effect of H2RAs occurs within 1 hour, while the
`onset of antisecretory effect of PPIs may be delayed and prolonged,
`staggered administration of LDV/SOF with H2RAs or PPIs may result
`in lower LDV concentrations. As shown in Study GS-US-256-0110, a
`substantial decrease in LDV plasma exposure (~42% to 48% lower
`AUC and Cmax, respectively) was observed with administration of
`omeprazole 2 hours earlier than LDV as a single agent. Therefore,
`H2RAs should only be administered simultaneously or 12 hours apart
`with LDV/SOF, at doses comparable to famotidine 40 mg twice daily.
`PPIs should only be administered simultaneously with LDV/SOF under
`fasted conditions at doses comparable to omeprazole 20 mg once
`daily (or lower).
`o No effect on the pharmacokinetic parameters of the following
`coadministered drugs was observed with ledipasvir or sofosbuvir:
`abacavir, cyclosporine, efavirenz, emtricitabine, lamivudine,
`methadone, or rilpivirine.
`o No effect on the pharmacokinetic parameters of ledipasvir, sofosbuvir
`and GS-331007 was observed with abacavir/ lamivudine,
`emtricitabine/rilpivirine/ tenofovir disoproxil fumarate or raltegravir.
`
`Tenofovir concentrations are increased with concomitant use of a ritonavir-
`boosted protease inhibitor. There are no clinical safety data with LDV/SOF and
`regimens containing tenofovir plus ritonavir-boosted HIV protease inhibitors.
`Table 3 in product labeling recommends consideration of alternative HCV or
`antiretroviral therapy to avoid increases in tenofovir exposures. However, if
`coadministration is necessary, labeling recommends monitoring for tenofovir-
`
`7
`
`Reference ID: 3632863
`
`

`

`associated adverse reactions. There are also no clinical data with LDV/SOF and
`elvitegravir/cobicistat/emtricitabine/tenofovir DF (Stribild, STR) and
`coadministration with STR is not recommended. The rationale for the different
`recommendations in Table 3 for these respective drug interactions is based upon
`benefit-risk considerations. Patients on regimens containing tenofovir and HIV
`protease inhibitors boosted with ritonavir may not have alternative HIV
`antiretroviral options, whereas patients receiving STR are generally treatment-
`naïve and have other options. Thus, the wording in Table 3 is intended to
`communicate the risk of increased tenofovir exposure with LDV/SOF
`coadministered with regimens containing tenofovir plus an HIV protease
`inhibitor/ritonavir, allow provider flexibility to determine the preferred treatment
`regimen for their patient and convey the need for renal monitoring for patients
`who receive these medications concomitantly.
`
`Tenofovir concentrations are also increased with concomitant use of Atripla.
`Review of summary clinical trial safety data from approximately 175 subjects who
`received LDV/SOF plus Atripla is determined to be adequate to support labeling
`for LDV/SOF and Atripla coadministration. Wording in Table 3 conveys the need
`for renal monitoring for patients who receive these medications concomitantly.
`Obtaining additional safety data in subjects receiving concomitant LDV/SOF and
`Atripla (or its components) from the ongoing trial GS-US-337-0115 is a
`recommended postmarketing requirement.
`
`No dose adjustment of LDV/SOF is required for patients with mild, moderate or
`severe hepatic impairment (Child-Pugh Class A, B or C).
`
`No dose adjustment of LDV/SOF is required for patients with mild or moderate
`renal impairment. However, no dose recommendation can be given for patients
`with severe renal impairment (eGFR < 30 mL/min/1.73m2) or with end stage
`renal disease due to higher exposures of the predominant SOF metabolite.
`Further, the use of P-gp inducers is not recommended (e.g., rifampin) as they
`may lead to a reduced therapeutic effect of the FDC.
`
`6. Clinical Virology: Please see extensive reviews by Drs. Lisa Naeger and
`Eric Donaldson who conducted the review of virology and resistance data,
`including next generation sequencing (NGS) data with supervisory concurrence
`by Dr. Jules O’Rear. Our virology review staff concluded that this FDC NDA for
`LDV and SOF is approvable with respect to virology for the treatment of GT1
`HCV infection. They recommend the 12-week duration of LDV/SOF for
`treatment-naive and treatment-experienced patients without cirrhosis. For
`subjects with cirrhosis, a 24-week duration of LDV/SOF is recommended.
`Per Drs. Naeger and Donaldson’s reviews and product labeling, the following
`wording describes resistance in cell culture: Reduced susceptibility to LDV was
`associated with the primary NS5A amino acid substitution Y93H in both
`genotypes 1a and 1b. Additionally, a Q30E substitution emerged in genotype 1a
`replicons. Site-directed mutagenesis of the Y93H in both genotypes 1a and 1b,
`
`8
`
`Reference ID: 3632863
`
`

`

`as well as the Q30E substitution in genotype 1a, conferred high levels of reduced
`susceptibility to LDV (fold change in EC50 greater than 1000-fold).
`HCV replicons with reduced susceptibility to SOF have been selected in cell
`culture. Reduced susceptibility to SOF was associated with the primary NS5B
`substitution S282T in all replicon genotypes examined. Site-directed
`mutagenesis of the S282T substitution in replicons of 8 genotypes conferred 2- to
`18-fold reduced susceptibility to SOF.
`Virology reviews and resultant product labeling contain the following wording
`related to resistance in clinical trials: In
` a pooled analysis
`of subjects who received the FDC, 37 subjects (29 with genotype 1a and 8 with
`genotype 1b) qualified for resistance analysis due to virologic failure (35 with
`virologic relapse and 2 with breakthrough on-treatment due to documented non-
`adherence). Post-baseline NS5A and NS5B deep sequencing data (assay cutoff
`of 1%) were available for 37/37 and 36/37 subjects’ viruses, respectively.
`Of the 29 genotype 1a virologic failure subjects, 55% (16/29) of subjects had
`virus with emergent NS5A resistance-associated substitutions K24R, M28T/V,
`Q30R/H/K/L, L31M, or Y93H/N at failure. Five of these 16 subjects also had
`baseline NS5A polymorphisms at resistance-associated amino acid positions.
`The most common substitutions detected at failure were Q30R, Y93H or N, and
`L31M.
`Of the 8 genotype 1b virologic failure subjects, 88% (7/8) had virus with
`emergent NS5A resistance-associated substitutions L31V/M/I or Y93H at failure.
`Three of these 7 subjects also had baseline NS5A polymorphisms at resistance-
`associated positions. The most common substitution detected at failure was
`Y93H.
`At failure, 38% (14/37) of virologic failure subjects had 2 or more NS5A
`substitutions at resistance-associated positions.
`In phenotypic analyses, post-baseline isolates from subjects who harbored NS5A
`resistance-associated substitutions at failure showed 20- to >243-fold reduced
`susceptibility to LDV.
`Treatment-emergent NS5B substitutions L159 (n=1) and V321 (n=2) previously
`associated with SOF failure were detected in the Phase 3 trials. In addition,
`NS5B substitutions at highly conserved positions D61G (n=3), A112T (n=2),
`E237G (n=2), and S473T (n=1) were detected at low frequency by next
`generation sequencing in treatment failure subjects infected with HCV GT1a.
`The D61G substitution was previously described in subjects infected with HCV
`GT1a in a liver pre-transplant trial. The clinical significance of these substitutions
`is currently unknown.
`The SOF-associated resistance substitution S282T in NS5B was not detected in
`any failure isolate from the Phase 3 trials. NS5B substitutions S282T, L320V/I,
`and V321I in combination with NS5A substitutions L31M, Y93H, and Q30L were
`
`Reference ID: 3632863
`
`9
`
`(b) (4)
`
`

`

`detected in one subject at failure following 8 weeks of treatment in a Phase 2
`trial.
`Not surprisingly, LDV was fully active against the SOF resistance-associated
`substitution S282T in NS5B while all LDV resistance-associated substitutions in
`NS5A were fully susceptible to SOF. Both LDV and SOF were fully active against
`substitutions associated with resistance to other classes of direct-acting antivirals
`with different mechanisms of actions, such as NS5B non-nucleoside inhibitors
`and NS3 protease inhibitors. NS5A substitutions conferring resistance to LDV
`may reduce the antiviral activity of other NS5A inhibitors. Importantly, the efficacy
`of the FDC has not been established in patients who have previously failed
`treatment with other regimens that include an NS5A inhibitor.
`No data are available on the persistence of LDV or SOF resistance-associated
`substitutions. NS5A resistance-associated substitutions for other NS5A inhibitors
`have been found to persist for more than one year in some patients.
`Critical analyses were conducted to explore the association between pre-existing
`baseline NS5A polymorphisms at resistance-associated positions and relapse
`rates. In the pooled analysis of the Phase 3 trials, 23% (370/1589) of subjects’
`virus had baseline NS5A polymorphisms at resistance-associated positions (any
`change from reference at NS5A amino acid positions 24, 28, 30, 31, 58, 92 or 93
`identified by population or deep sequencing).
`In treatment-naïve patients whose virus had baseline NS5A polymorphisms at
`resistance-associated positions in Studies ION-1 and ION-3, relapse rates were
`low and comparable to subjects without baseline polymorphisms: 6% (3/48) after
`8 weeks and 1% (1/113) after 12 weeks of treatment with baseline
`polymorphisms versus 5% (8/167) after 8 weeks and 1% (3/306) after 12 weeks
`treatment without baseline polymorphisms.
`In treatment-experienced patients whose virus had baseline NS5A
`polymorphisms at resistance-associated positions, relapse rates were 22% (5/23)
`after 12 weeks and 0% (0/19) after 24 weeks of treatment with the FDC. Patients
`with multiple NS5A polymorphisms at resistance-associated positions appeared
`to have higher relapse rates. SVR was achieved in all 24 subjects (N=20 with
`L159F+C316N; N=1 with L159F; and N=3 with N142T) who had baseline
`polymorphisms associated with resistance to NS5B nucleoside inhibitors. The
`S282T SOF resistance-associated substitution was not detected in the baseline
`NS5B sequence of any subject in Phase 3 trials by population or deep
`sequencing.
`
`7. Efficacy and Safety: Clinical reviews were conducted by Dr. Sarah Connelly
`with secondary review provided by Dr. Kim Struble. The Biometrics review was
`conducted by Dr. Karen Qi with secondary review provided by Dr. Fraser Smith
`and supervisory review provided by Dr. Daphne Lin.
`
`Reference ID: 3632863
`
`10
`
`

`

`The Phase 3 program encompasses treatment-naïve and treatment experienced
`patient populations with GT1 who enrolled in three principal clinical trials, ION-1,
`ION-2 and ION-3. Results from ION-1 and ION-3 supported an indication for
`treatment-naïve patients with and without cirrhosis. ION-1 was a randomized,
`multicenter, four-arm, open-label trial that examined LDV/SOF for 12 or 24 weeks
`with and without ribavirin. The primary endpoint was SVR12 and the primary
`hypothesis was that each treatment arm had to be superior to a historical control
`rate of 60%, derived from the upper bound of the 95% confidence interval of the
`highest SVR rate for pegylated interferon and ribavirin from historical trials
`ADVANCE and SPRINT2. Per Dr. Struble’s CDTL review and Dr. Qi’s statistical
`review, efficacy from the 24 week arms were not included in the NDA because
`the Division agreed that if the two 12-week arms achieved an SVR12 rate of >
`90% in subjects with and without cirrhosis, separately, then the Applicant could
`submit an NDA and not wait for the 24 week arms to reach the SVR12 timepoint.
`Randomization was stratified by genotype subtype (1a or 1b) and presence or
`absence of cirrhosis. Of note, ION-1 was conducted in the United States and
`Europe.
`
`ION-3 evaluated LDV/SOF in three treatment groups: LDV/SOF +/- RBV for 8
`weeks and LDV/SOF for 12 weeks. Only non-cirrhotic subjects were enrolled and
`randomization was stratified by genotype subtype (1a or 1b). Of note, ION-3 was
`only conducted in the United States.
`
`Demographics and baseline characteristics were balanced across treatment
`groups in the individual trials. Results from ION-1 and ION-3 were robust. SVR12
`rates exceeded 90% for all treatment arms. The following table describing results
`from ION-1 appears in product labeling. As ribavirin did not alter treatment
`outcomes, only the LDV/SOF 12-week arm is displayed in Table 9 in product
`labeling.
`
`Table 9
`ION-1: SVR Rates for Selected Treatment Naïve Subgroups after 12 Weeks Treatment
`LDV/SOF 12 Weeks
`(N = 214)
`
`Genotypea
`Genotype 1a
`Genotype 1b
`Cirrhosisb
`99% (176/177)
`No
`94% (32/34)
`Yes
`a. One subject without a confirmed subtype for genotype 1 infection and one subject with genotype 4
`infection were excluded from this subgroup analysis.
`b. Subjects with missing cirrhosis status were excluded from this subgroup analysis.
`
`98% (142/145)
`100% (67/67)
`
`Reference ID: 3632863
`
`11
`
`

`

`Table 6 in product labeling presents the response rates for the FDC treatment
`groups in the ION-3 trial. Ribavirin was not shown to increase the response rates,
`therefore the 8-week FDC plus ribavirin arm is not presented in Table 6.
`Table 6
` ION-3: Response Rates in Treatment Naïve Patients after 8 and 12 Weeks Treatment
`LDV/SOF
`LDV/SOF
`8 Weeks
`12 Weeks
`(N = 215)
`(N = 216)
`94% (202/215)
`96% (208/216)
`
`SVR
`
`0/216
`1% (3/216)
`2% (5/216)
`
`0/215
`5% (11/215)
`<1% (2/215)
`
`Outcome for subjects without SVR
`On-Treatment Virologic Failure
`Relapsea
`Otherb
`Genotypec
`96% (165/172)
`93% (159/171)
`Genotype 1a
`98% (43/44)
`98% (42/43)
`Genotype 1b
`a. The denominator for relapse is the number of subjects with HCV RNA <LLOQ at their last on-treatment
`assessment.
`b. Other includes subjects who did not achieve SVR and did not meet virologic failure criteria (e.g., lost to
`follow-up).
`c. One subject without a confirmed subtype for genotype 1 infection was excluded from this subgroup
`analysis.
`The review team recommended 12 weeks of treatment with LDV/SOF based on
`relapse rates, 5% compared to 1% for 8 weeks versus 12 weeks. It was felt that
`to be able to give patients the best outcome with a first regimen it was
`appropriate to treat longer to attempt to overcome relapse. However, in those
`patients with more favorable baseline characteristics, a shorter treatment
`regimen could be considered. Relapse rates based on baseline viral load by
`treatment duration are depicted in Table 7 of product labeling.
`
`Table 7
`ION-3: Relapse Rates in Treatment Naïve Patients by Baseline Viral Load after 8 and 12
`Weeks of Treatment
`
`Number of responders at end of treatment
`Baseline HCV RNAa
`HCV RNA < 6 million IU/mL
`HCV RNA ≥ 6 million IU/mL
`
`LDV/SOF
`8 Weeks
`(N = 215)
`
`215
`
`2% (2/123)
`10% (9/92)
`
`LDV/SO