Briefing Document
`16 February 2000
`
`Confidential
`
`Page 1
`
`2/16/2000
`
`I TRIAL EXHIBIT
`
`TX 76
`
`CONFIDENTIAL
`ELAP00013763
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1084-0001
`
`

`
`PURPOSE OF THE MEETING
`
`RATIONALE FOR PROGRAMMATIC INTERVENTION
`
`4
`
`5
`
`Compound Overview and Link to Folate Metabolism
`
`Folinic Acid versus Folic Acid
`
`Summary of Data on the Relationship Between Folio Acid and the Toxicity or Activity of Antifolates6
`
`Cancer - Prec]inical Data
`
`Cancer - Clinical Data
`
`Rheumatoid Arthritis
`
`Biochemical Relevance of Homocysteine to Folate Metabolism
`
`Safety Analysis and Rationale for Programmatic Intervention
`
`Synopsis of Safety Analysis Findings
`
`Clinical Relevance of Homocysteine to LY231514 Studies
`
`IMCH Mortality and Safety Interventions
`
`6
`
`7
`
`9
`
`10
`
`10
`
`10
`
`11
`
`I1
`
`Question Ia Does the FDA agree that toxicity and morality data sttppom a programmatic intervention to
`
`improve patient safety in LY231514 trials and that daily low dose foIic acid supplementation appropriately
`
`serves this purpose ? 12
`
`RECENT CONCERNS THAT A RANDOMIZED TRIAL OF LY231514 WITH
`FOUC ACID VS LY231514 WITHOUT FOLIC ACiD IS NO LONGER FEASIBLE
`13
`
`Ethi~l Considerations
`
`Logistical Considerations
`
`13
`
`13
`
`Question lb Does the FDA agree that a randomized trial comparing patients receiving LF231514 with and
`
`without vitamins is no longer feasible or advisable given the demonstrated toxicity risks to LT’231314
`
`patients?
`
`14
`
`STATISTICAL IMPLICATIONS FOR PROPOSED ANALYSIS OF JMCH DATA
`15
`
`Efficacy Analysis
`
`Salety Analysis
`
`15
`
`16
`
`Confidential
`
`Page 2
`
`2/16/2000
`
`CONFIDENTIAL
`ELAP00013764
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1084-0002
`
`

`
`Question2 Do the proposed analyses of efficacy and saJ’ety described here for Study JMClf sufficiently
`
`address the impact of the foli¢ acid supplementation intervention an the results of this trial such that the
`
`trial will still qualify as a randomized, well-controlled trial for the mesothelioma and NSCLC indications?
`
`PROPOSED CHANGES TO THE SECOND-LINE NSCLC STUDY JMBQ
`
`17
`
`18
`
`Question 3a Does the agency support the replacement of vinoreibine with docetaxel as th¢ comparator in
`
`the JMBQ study?
`
`Question 3b Does the agen~ agree ~haf these modifications wilt allow Study JMBQ ~o comlnue to serve
`
`the role of a randomized, well-controlled trial in support of the mesothelioma and second-line NSCLC
`
`indications, as previously discussed in the End-of Phase H meeting in June of 19997
`
`REFERENCES
`
`AI-rACHMENT 1 LIST OF COMMUNICATIONS BETWEEN THE FDA AND
`LILLY CONCERNING LY231514 FOR MESOTHELIOMA AND NON-SMALL
`CELL LUNG CANCER
`
`A’rTACHMENT 2 LILLY’S RESPONSE TO FDA COMMUNICATION OF 14
`OCTOBER 1999
`
`19
`
`20
`
`22
`
`30
`
`Confidential
`
`Page 3
`
`2/16/2000
`
`CONFIDENTIAL
`ELAP00013765
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1084-0003
`
`

`
`Purpose of the Meeting
`
`The ongoing registration trial, H3E-MC-JMCH (]~(cid:128)ICH for brevity), has been modified
`through addition of folic acid for the purpose of promoting patient safety. The Medical
`Review Officer has stated that he does not support the addition of vitamins to an ongoing
`registration trial (FDA communications to Lilly on December 21, 1999 and January 6,
`2000). The sponsor has sought a face-to-face meeting to come to agre~nent as to the
`implications of the action of adding folic acid to the pivotal registration trial and for
`supporting trials.
`
`In addition, Li]ly would like to discuss proposed modifications to our second-line NSCLC
`trial supporting the mesothetioma registration trial. This document includes the issues
`that we are seeking guidance on and background iafonnation related to these issues.
`
`Confidential
`
`Page 4
`
`2/16/2000
`
`CONFIDENTIAL
`ELAP00013766
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1084-0004
`
`

`
`Rationale for Programmatic Intervention
`
`Compound Overview and Link to Folate Metabolism
`
`The antituraor activity of LY231514, a multit~rgeted antifolate, is derived from
`simultaneous and multiple inhibition of several key folate-requiring enzymes of the
`thymidine and purine biosynthetic pathways. LY231514 has been found to inhibit cell
`growth by interfering with the action of the enzymes thymidylate synthase (TS),
`dihy&ofolate reductas¢ (DI-IFR), and glycinar~de ribonucleotide formyltransferase
`(GARFT) by competing with the reduced folates 5,10-methylene tetrahydrofolate,
`dihydrofolate, and 10-forrayl tetrahydrofolate for binding sites on the respective enzymes
`(Shih et al. 1997). Thus, the mechanism of action of LY231514 and other folio acid
`analogues such as methotrexate and lometrexol is critically linked to intracellular folate
`metabolism. The effects of antifolates are also significantly modulated by the formation
`of intracellular polyglutamates. Polyglut&mates are retained within the cell for long
`periods thus increasiug the potency of the antifolate. In addition the polyglutamate
`derivatives of LY231514 are significantly more potent inhibitors of TS and GARFT than
`the parent compound and may thus serve to enhance the action of the drug on these
`targets.
`
`Preclinical and clinical studies evaluating the impact of dietary folic acid on the toxicity
`or efficacy of antifolates such as LY231514 and lometrexol have been reported. Because
`tumor tissue and normal tissue, such as bone marrow, presumably have different folate
`requirements, it is possible to decrease the toxicity to healthy tissue while maintaining
`antitumor effect through careful adjustment of folio acid intake. This has been shown in
`experimental systems for LY231514 and another antifolate, lometrexol (Worzalla et al.
`1998; Alati et al. 1996) and in clinical u’ials with lometrexol (Young et al. 1992;
`Laohavinij et al. 1996). In addition, it has been clinically observed that the efficacy of low
`dose methotrexate used in the treatment of rheumatoid arthritis is not negatively affected
`by folic acid supplementation, while an improvement in toxicity is seen (Morgan et al.
`1998).
`
`Folinic Acid versus Folic Acid
`
`If a patient receiving an antifolate experiences severe, life-threatening toxicity, stand~d
`medical treatment includes rescue with high doses of the reduced folate leucovorin
`(folinic acid). Because folates are not efficiently stored in the body, depletion and
`repletion can occur relatively quicldy with supplementation. For example, megaloblastic
`bematopoeisis reverts to normal hematopoeisis within 12 to 48 boars of folinic acid
`supplementation (Antony 1991). Reversal of methotrexate toxicity by folinie a~id is due
`
`Confidential
`
`Page 5
`
`2/16/2000
`
`CONFIDENTIAL
`ELAP00013767
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1084-0005
`
`

`
`to replenishment of the tetrahydrofolate pool by folinic acid and is therefore non-
`competitive with respect to methotrexate. In the case of LY2315 t4, reversal may be
`achieved by (i) competition of folinic acid for enzymatic binding sites and (ii)
`competition for the formation of the more potent polyglutamate forms of the drug.
`
`There are a number of masons why leucovorin is used as a rescue agent and folic acid is
`not. Folic acid is not used by the body as is, but must undergo a transformation to its fully
`reduced form before it can be utilized as a cofactor in the one-carbon transfer reactions
`critical in the synthesis of thymidine and purines. Folie acid is also a very poor substrate
`for an enzyme that catalyzes this reduction, dihydrofolate reductase, which is why this
`reduction is probably the rate-limiting step for folic acid utilization in mammalian
`systems. Folinic acid, on the other hand, is already fully reduced and is available for use
`as a cofactor in thymidine and pudne synthesis as quickly as it can be transported into the
`cell.
`
`In biological systems, the reduced folate carrier (RFC) is the predominant mechanism of
`folate transport into cells. While folinic acid is very efficiently transported by RFC, folic
`acid is not. This limits the amount of folic acid that can enter the cell via this mechanism.
`
`Folinic acid is preferred over folio acid as a rescue agent when ceils at risk for toxicity
`require reduced folates quickly. By contrast,, folio acid is a normal dietary supplement and
`can replenish folate pools efficiently, albeit more slowly, at low oral doses. Thus dietary
`folio acid may be useful in modulating the toxicity of antifolate chemotherapy agents.
`
`Summary of Data on the Relationship Between Folic Acid and
`the Toxicity or Activity of Antifolates
`
`Cancer- Preclinical Data
`
`Worzalla and coworkers have studied the effects of folic acid on the toxicity and
`antitumor activity of LY231514 in the in vitro and in vivo settings. In a number of
`human tumor cell lines, folie acid protected cells from cytotoxicity at concentrations 100-
`to 1000-fold higher than those required for folinie acid protection, indicating that the
`action of LY231514 is less sensitive in vitro to folic acid than it is to folinic acid. They
`also found that in mice fed a low folate diet (LFD), tumor growth inhibition was complete
`at LY231514 doses of 0.9 to 3.0 rag/m2, with 100% lethality occurring at LY231514-
`doses of 9.0 rag/m2 or higher. Mice receiving the same LFD who were supplemented
`with high doses of folio acid at 15 mg/kg/day (a dose approximately 10-fold greater than
`that in the normal diet) experienced complete tumor growth inhibition at LY231514 doses
`of 90 to 3000 mg/m2 without any lethality. Mice on the standard diet (approximately one
`tenth of the folio acid given to the supplemented mice) saw a virtually identical dose
`
`Confidential
`
`Page 6
`
`2/16/2000
`
`CONFIDENTIAL
`ELAP00013768
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1084-0006
`
`

`
`t
`
`response, but ~eater lethality, with I00% lethality occurring at 2400 rag/m2 (Worzalla ¢t
`a]. 1998).
`
`Doses of LY231514 for Maximum Antitumor Activity and kethalit
`
`’in Mice
`
`Diet
`
`Standard Diet
`(Dai|y,folic acid intake = 4,5 m~m~)
`LP’D ÷ 15 mg/kg Folic Acid
`(Daily folio acid intake = 45 m~m~)
`
`=Doses in m~m2/day
`
`Doses~ of
`LY’231514 Where
`
`Antitumor Activity
`is Observed
`From 90 to 1200
`
`Frvm’ 90 to 3000
`
`LY231514 Where
`Lethality is
`Observed
`2400 (100% IethaIity)
`
`(No lethality seen up
`to 3000)
`
`These data show that antitumor activity is virtually identical in mice receiving a standard
`diet to that in mice reeeivit~g a 10-fold increase in daily folio acid. Mice recei’~ing the
`extra folio acid also showed a decreased lethality at higher doses of LY231514. These
`data support the hypothesis that folie acid supplementation can protect healthy tissue
`from the toxic effects of LY231514 with retention of antitumor activity.
`
`The effect of folate status on the efficacy and toxicity of chemotherapy was also
`investigated in weanling Fischer 344 rats maintained on diets of varying folate content or
`supplemented with daily injections of folic acid, 50 mg/kg, for 6 to 7 weeks. Results
`showed ~at correction of folate deficiency approximately doubled the efficacy of
`cyclophosphamide in rats with much tess host toxicity, and folate repletion improved
`survival in 5-FU-treated animals. This indicates that nutritional folate status has an
`important influence on the efficacy and toxicily of some commonly used
`chemotherapeutic drugs (Branda et al. 1998).
`
`Cancer - Clinical Data
`
`A study of LY231514 in gastric cancer began in late 1996. After 3 of the first 6 patiems
`treated died, enrollment into this study was suspended. At that time, the protocol was
`modified to include folio acid supplementation at 5 mg per day 2 days raefore, the day of,
`and 2 days after the administration of 500 rag/m2 LY231514. This level of folic acid
`supplementatiort has been shown to be effeeti,ce in lo~¢ering homocysteine levels in a
`phase I study of LY231514 and folio acid. In the 7 patients treated since the study re-
`started, there have been no drag-related deaths, and responses have been reported in three
`patents (2 CR, 1 PR).
`
`Confidential
`
`Page 7
`
`2/16t2000
`
`CONFIDENTIAL
`ELAP00013769
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1084-0007
`
`

`
`L¥231514 in Gastric Cancer - Trial Results
`
`Patient
`Number
`
`LY231514/folic acid # Cycles Outcome
`received
`
`1
`2
`
`4
`
`6
`7
`8
`9
`10
`
`Patients without Folic Acid Supplementation
`6
`Stable disease
`500 rag/m2/none
`Discontinued due to neutropenia,
`500 mg/m2/none
`death 3 weeks later of pneumonia
`Discontinued due to an adverse event
`Death from drug-rdated anemia
`Pro~e.ssive disease
`Death from drug-related toxicity
`(vomiting and diarrhea leading to
`renal failure)
`
`500 mg/m2/none
`5.00 m~/m2/none
`500 mg/m~-/none
`500 mg/mZ/none
`
`1
`1
`4
`2
`
`Patients with Folic Acid Sup
`500 me/m2/5 mg x 5
`6
`500 me/m2/5 mg x 5
`6
`500 m~/m2/5 m~x 5
`6
`500 m~m2/5 mg x 5
`6
`500 m~m2/5 mg × 5
`5
`500 moffm2/5 mg x 5
`2
`500 me/m2/5 mg x 5
`6
`
`flementation
`Stable disease
`Pa, rti~, response - confirmed
`Stable disease
`Stable disease
`Compl©te response - confirmed
`Pro~essive dis.case,
`Complete response- post-baseline
`measurement by ¢chocardiography
`
`Although these data are preliminary, they point out that folic acid supplementation of
`patients receiving LY231514 as treatment for cancer may:
`
`1) Lead to a reduction in severe toxicity and drag-related death
`
`2) Allow patients to receive therapy for more cycles, resulting in
`
`3) Retention of antitumor activity (three responses in seven patients).
`
`In a dose-finding phase I study of lometrexol, a GARFT inhibitor, 53 patients were
`entered and received lometrexol, with 19 also receiving oral folk acid at 1 mg/m2 daily.
`In patients who did not receive folic acid, the maximally tolerated dose was 2.7
`twice weekly, and cumulative toxicity was observed. In patients receiving folic acid, the
`maximally tolerated dose was 4 to 5 mg/m2 twice weekly, with no cumulative toxicity
`observed. In 9 of these patients, andtumor effects were observed, with one patient
`experiencing a complete response of 18 months in oropharyngeal cancer (Young et al.
`1992).
`
`Confidential
`
`Page 8
`
`2/16/2000
`
`CONFIDENTIAL
`ELAP00013770
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1084-0008
`
`

`
`In another phase I dose-finding study of lometrexol, 43 patients received 5 mg oral folio
`acid daily for 7 days prior to and 7 clays after each lometrexol dose. Doses ranged from I2
`mg/m2 every 4 weeks to 170 mg/m2 every 3 weeks. Plasma folate concentrations
`increased from 9 to 20 n~mL in the 7 days of folic acid supplementation prior to course
`1. Pharmacokdnetic s~ndies showed a linear relationship between lometrexol plasma
`levels and doses at all doses for which data were available (12 - 130 m~m2). The plasma
`levels cor~lated well with those obtained from previous studies in which patients did not
`receive folio acid (data available for 12 - 45 ms/m2). Major toxicities were
`thrombocytopenia and mucositis; however maximum tolerated dose was not reached. One
`partial response and one minor response were observed; 4 additional patients experienced
`disease stabilization for a raedian duration of 3+ months (Laohavinij et al. !996).
`
`Rheumatoid Arthritis
`
`Several t_rials have been conducted to evaluate the effects of o~al folic acid on the toxicity
`associated with long-term, low-dose methotrexate in the treatment of rheumatoid arthritis.
`Two of note are summarized here.
`
`To determine the effect of either 5 mg or 27.5 mg of folic acid weekly on the toxicity and
`efficacy of low dose oral methotrexate therapy for rheumatoid arthritis, a randomized,
`double blind, placebo-controlled study was .conducted in 79 patients (Morgan et al. 1994).
`Folio acid supplementation at eider dose did not affect the effic.acy of methotrexate
`therapy. Patients given folic acid supplements had lower toxicity scores than did the
`placebo group. Low blood folate levels and increased mean corpuscular volumes were
`associated with substantial methotrexate toxicity, it was concluded that folie acid is safe
`in a broad range of doses and protects patients with rheumatoid atttuStls from
`methotrexate toxicity while preserving efficacy.
`
`Thirty-two patients with rheumatoid arthritis completed a 24-week, placebo-controlled,
`double-blind trial of folie acid (FA) supplementation during low-dose methotrexate
`(MTX) therapy (Morgan et al. 1990). Administration of the daily FA supplement
`significantly lowered toxicity scores without affecting efficacy, as measured by joint
`counts, joint indices, and physician evaluation of disease activity. Fifteen patients
`experienced some sort of toxicity; 67% in the placebo group, and 33% in the FA
`supplement group. Four patients in the placebo group had toxicity levels serious enough
`to require discontinuation of the MTX, while no patients in the FA supplement group
`discontinued MTX because of toxicity. It was concluded that a daily supplement of 1 mg
`of FA during low-dose MTX therapy is useful in lessening toxicity without altering
`efficacy during the first 6 months of treatment.
`
`Confidential
`
`Page 9
`
`2/16/2000
`
`CONFIDENTIAL
`ELAP00013771
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1084-0009
`
`

`
`Biochemical Relevance of Homocysteine to Folate Metabolism
`
`One of the principal routes of homocysteine metabolism is the folate-dependent
`mechanism, Through this route, homocysteine is converted to methionine by the enzyme
`methionine synthase, which is dependent on vitamin B 12 and incorporates a methyl group
`from 5-methyltelxahydtofolate into homocysteine, gNing metb2onine. Therefore, a folate
`deficiency will result in lowered methionine synthase activity and lead to an elevation of
`plasma homocysteine levels. Indeed, homocysteine has been found to be a sensitive
`marker for folic acid as well as B 12 deficiency (Morgan et al. 1991; Selhub and Miller
`1991).
`
`Recent studies in cardiovascular patients have suggested that folate supplementation with
`or without supplementation with B 19. and B6 can significantly reduce homocysteine
`levels (Homocysteine Lowering Triatists’ Collaboration 1998; Malinow et al. 1998).
`Folio acid supplementation of 400 ~tg daily in elderly patients with elevated homocysteine
`has been shown to substantia!ly reduce plasma hornocysteine within 2 weeks. The level
`continues to drop slightly for another 2 weeks, and then plateaus (Bronstrup et al. 1999).
`Brouwer and coworkers showed that low dose folic acid (250 ~tg - 500 ~tg) intervention
`significantly decreases homocysteine levels. An 8-week washout period was not sufficient
`for blood folate and plasma homocysteine Ievels to return to baseline (Brouwer et al.
`1999). Niyikiza et al have shown that in patients who are not supplemented,
`homocysteine levels do not change over the course of treatment with LY231514 (Niyikiza
`et al. 1998).
`
`Additionally it has been zhowrt that mice on low folate diet also have high plasma
`homoeysteine levels (Worzalla et al. 1996) and that supplementation with dietary folic
`acid reduced the plasma homoeysteine levels to near normal levels.
`
`On the other hand, TS, DHFK and GARFT are not known to have any binding sites for
`B12 or B6 and are not known in may way to be affected by these two vitamins. Therefore,
`increasing the ceIlular concentrations of B 12 and B6 is not expected to have an impact on
`the growth inhibitory effects of LY231514-.
`
`Safety Analysis and Rationale for Programmatic Intervention
`
`Synopsis of Safety Analysis Findings
`
`Multivariate and multiple logistic regression analyses carried out by Niyikiza and
`coworkers have shown that a patient’s pre-treatment serum homocysteine is a statistically
`significant predictor for his or her risk of developing serious toxicity (Grade 4
`neutropenia + Grade 3/4 infection, Grade 4 thrombocytopenia, or Grade 3/4 diarrhea,
`p <0. 00001) during the course of treatment with LY231514. This analysis was performed
`
`Confidential
`
`Page 10
`
`2/16/2000
`
`CONFIDENTIAL
`ELAP00013772
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1084-0010
`
`

`
`on patients from several trials in multiple minor types. (Safety analysis communicated to
`FDA on 03 December 1999.) Specifically, a homocysteine level above 12 lamol/L puts a
`patient at a much greater risk of developing severe toxicity than a homocysteine level
`be2ow this threshold level. Multivariate modeling also showed that toxicity such as Grade
`4 neutropenia coupled with Grade 3/4 infection is significantly correlated with death (p
`<0. 00001). Although the sample size of patiet~ts who died from drug-related death ~d
`who also had homocysteine levels measured is too small (n = I 1) to provide any direct
`correlation between elevated homocysteine levels and death, this relationship may be
`inferred.
`
`Clinical Relevance of Homocysteine to LY231514 Studies
`
`These data clearly point out the clinical relevance of homocysteine to LY231514 studies.
`While hematologic toxicity does not always result in ¢lirdcal symptoms, we have found
`that patients who experience Grade 4 neutropenia coupled with Grade 3 or 4 infection are
`at a higher risk of death. Because a patiem’s pre-treatment hornocysteine level is an
`extremely sensitive predictor for severe hematologic toxicity, we are collecting this
`information on each patient.
`
`JMCH Mortality and Safety Interventions
`
`In the ongoing mesothelioma registration trial in which nearly 40% of patients have
`elevated (defined as greater than 12 runnel/L) homoeysteine, 3 treatment related deaths
`occurred in the fLrst 42 patients enrolled into the experimental m-rn (LY231514 and
`cisplatin). This information, coupled with the results of the multivariate analyses of
`safety, led Lilly to explore intervention options, as well as to seek guidance from external
`experts. The consensus resulting from this series of discussions was ihat a 7% rate of
`death in a registration trial is unacceptable and that an intervention should be taken
`immediately. Lilly decided to give folio acid and B i2 to a/l patients for the following
`
`reasons:
`
`Given the strong body of data discussed above, we feel that dally supplementation of
`all LY231514 patients with a daily low-dose of folio acid and periodic BI2 injections
`will allow the greatest number of patients with mesothelioma to benefit from
`treatment with tl’fis novel antifoIate while also providing for greater patient safety in
`all LY231514 trials.
`
`Parents whose homocysteine is less than 12 vxno!/L may also experience a benefit
`from low dose relic acid supplementation. That is, a reduction in homocysteine may
`Ix~ seen even in patients with homocysteine lower than 12 Fvaol/L, and thus, these
`patients may also benefit from supplementation.
`
`Confidential
`
`Page 11
`
`2/16f2000
`
`CONFIDENTIAL
`ELAP00013773
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1084-0011
`
`

`
`In our database, 15% of patients have elevated methyl malonie acid, a marker for
`vitamin B12 deficiency, and there is a risk that some of these patients may not see a
`
`decrease in homocysteine with folio acid supplementation alone. Therefore, we are
`also recommending 1312 injections at regular intervals for all patients.
`
`The level of supplementation that we have implemented (--450 ~g/day) is not high, and in
`fact is similar to the US recommended dietary allowance and the amount recommended to
`pregnant women to reduce the risk of neural tube defects, and also to coronary heart
`patients to reduce the risk of cardiovascular events (Berry et a!. 1999; Czeizel 1993;
`Rimm et al. 1998; Graham 1999; Bronstrup et al. 1999; Brouwer et el. !999). We
`strongly feel that the balance between the risk of severe toxicity to patients and the
`concern of a detrimental effect on antitumor activity is now weighing in favor of reducing
`toxicity.
`
`Question la
`Does the FDA agree that toxicity and mortality data support a
`programmatic intervention to improve patient safety in LY231514
`tria/s and that daffy/ow dose folic acid supp/ementation appropriate/y
`serves this purpose?
`
`Confidential
`
`Page 12
`
`2/16/2000
`
`CONFIDENTIAL
`ELAP00013774
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1084-0012
`
`

`
`Recent Concerns that a Randomized Trial of LY231514 with Folic
`Acid vs LY231514 without Folic Acid is no Longer Feasible
`
`Ethical Considerations
`
`Given the strong body of data discussed above, Lilly now feels that conducting a
`clinical trial in which patients would be randomized to receive or not receive
`vitamin supplements while receiving LY231514 would be ill-advised. Exposing
`patients in the control arm to LY231514, particularly those with elevated homocysteine
`levels, without the potential benefit of’~itamin supplement support would b~ ill-advised,
`particularly when all other patients in LY231514 trials are receiving the supplements. For
`similar reasons, Lilly feels that investigators would be reluctant to enroll patients to such
`a study and patients would be reluctant to consent to risk not receiving vitamin
`supplements, making the trial rto longer feasible. Also, external consultants have said
`Ethical Review Boards would b~ reluctant to approve a trial such as this.
`
`Logistical Considerations
`
`We have concerns about the ability of patients to remain compliant on both arms of
`such a trial and also that the difficulty in.controlling for dietary habits in the patient
`population is a significant confounding factor. The amount of folio acid that will be
`used to supplement LY231514 patients is extremely small (-450 ~tg/.day). For those who
`are assigned to receive vitamin supplements, we are concerned that, given the easy
`accessibility of multivitamin and health food store preparations, patients may knowingly
`or inadvertently take additional folic acid supplements while on the trial. Similarly, those
`on the control arm not receiving vitamin supplements mn a risk of taking folic acid in
`some other preparation while on study. Again, the ease by which such folic acid could b~
`taken without knowing about it is substantial. Furthermore, it would be very difficult to
`control dietary intake such that patents on the control arm would not be taking amounts
`of folic acid through commercial cereal product supplementation or other dietary sources
`of folio acid and risk confounding the results of the trial.
`
`WbJle choosing a population at particular risk for nutritional deficiency might be best to
`answer the question most clearly, this population also takes nutritional supplements (eg,
`Ensure) regularly and again this may confound the results. Choosing a more general
`population would potentially make the result more generalizable but it would be very
`difficult to estimate the sample size if efficacy is a primary outcome, since effic~zy data
`varies by tumor type. It will also be difficult to get investigators to agree to treat patients
`in the front line setting with single agent MTA where the response rates are high enough
`to consider a feasible sample size for a randomized vitamin versus no vitamin trial. If we
`
`Confidential
`
`Page 13
`
`2/16/2000
`
`CONFIDENTIAL
`ELAP00013775
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1084-0013
`
`

`
`choose a second or third line population where single agent MTA may be feasible and
`acceptable, the response rate will be so low that the feasibilily of doing an equivalency
`trial based on the huge number of patients required to show a difference would be
`seriously questioned.
`
`Question l b
`Does the FDA agree that a randomized trial comparing patients
`receiving LY231514 with and without vitamins is no longer feasible or
`advisable given the demonstrated toxicity risks to LY231514
`patients?
`
`Confidential
`
`Page 14
`
`2/16/2000
`
`CONFIDENTIAL
`ELAP00013776
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1084-0014
`
`

`
`Statistical Implications for Proposed Analysis of JMCH Data
`
`Efficacy Analysis
`
`Lilly believes, as previously stated, that the addition of low amounts of folic acid will
`not adversely affect the primary endpoint of survival in the ongoing mesothelioma
`registration trial, JMCIt. The same intervention of vitamin supplementation is applied
`at the same time to both arms for the purpose of maintaining the integrity of the blinding,
`and therefore, no systematic bias is introduced. Since the trial was sized based on the
`survival endpoint and since the initial statistical hypotheses underlying the original
`sample size calculations remain unchanged with the vitamin supplementation, the trial
`sample size of 280 qualified patients should not be altered. Hence we propose the
`following for the efficacy analysis of this trial:
`
`Primary endpoint:
`
`At the completion of the current 280 patient trial, the primary endpoint of survival will be
`analyzed as currently stated in the protocol, Section 4 of ~IMCH(c). Furthermore, vitamitt
`supplementation will be incorporated as a prognostic factor in addition to previously
`planned prognostic factors (eg, age, sex, etc) i~a the Cox regression models to evaluate its
`possible impact o~ survival.
`
`Secondary endpoints:
`
`Secondary time-to-event endpoints such as time to progression, time to treatment failure
`(per South West On¢ology Group (SWOG) definition), and duratiorr of response will also
`be analyzed, as ctmently stated in the protocol, Section 4 of JMCH(c). Additionally,
`vitamin supplementation will be incorporated as a prognostic factor in addition to
`previously planned prognostic factors (eg, age, sex, etc) in the Cox regression model to
`evaluate its possible impact on those time-to-event variables.
`
`Fisher Exact Tests for differences in tumor and clinical benefit response rates will be
`performed between treatment arms as described in the protocol. Additionally, a
`dichotomous covmiate of vitamin supplementation (Yes versus No) will be used in
`multivariate models to assess the possible impact of vitamin supplementation on tumor
`and clinical benefit response rates. Similar analyses wiIl be conducted for quality of life
`scores analysis.
`
`Other additional explorato~ analyses will be conducted as deemed appropriate.
`
`Confidential
`
`Page 15
`
`2/16!2000
`
`CONFIDENTIAL
`ELAP00013777
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1084-0015
`
`

`
`Safety Analysis
`
`All prospectively planned endpoin~s will be analyzed as currently stated in the protocol.
`
`To address the affect of folic acid on safety, the following analyses are proposed. Let:
`
`¯
`
`¯
`
`"Arm A" represent the LY231514 plus cisplatin arm,
`
`"Arm B" represent the cisplatin alone arm.
`
`Arm A
`Treatment l
`
`Group A.vits
`
`LY231514 + cisplatin
`
`without vitamins
`(earLy patients)
`
`Group A+v~ts
`LY231514 + cisplatin
`
`with vitamins
`flare palfents)
`
`Treatment
`Arm B
`
`Group B.vits
`
`cisplatin alone
`
`without vitamins
`Ie~dy p~tients)
`
`Group B+vits
`cisplatin alone
`
`with vitamins
`,~late patientsI ,,
`
`Dec 2, 1999
`
`To assess the impact of folic acid supplementation on safety on a per patient basis, after
`280 qualified patients have completed the protocol, subgroup safety analyses will be
`performed on:
`
`those patients who had no vitamin supplementation during their participation in
`JMCH (Group A, -vits versus Group B -vim) and
`
`ii.
`
`those patients who were supplemented the entire time that they participated in
`JMCH (Group A +vi~ versus Group B ÷vits).
`
`Because there are a number of patients who began to receive folic acid supplementation
`after one or more cycles of treatment without supplemention, Lilly will also assess the
`impact of folic acid on safety on a per cycle basis. After 280 qualified patients have
`completed the protocol, safety data will be analyzed for:
`
`Confidential
`
`Page 16
`
`2/16/2000
`
`CONFIDENTIAL
`ELAP00013778
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1084-0016
`
`

`
`those cycles in which patients did not receive vitamin supplementation (Group A
`versus Group B) and
`
`ii.
`
`those cycles in which patients clid receive vitamin supplementation (Group A
`versus Group B)
`
`This per cycle analysis has been previously described in protocol amendment (c) (serial
`no. 200 on December 22, 1999 and no. 201 on December 22, !999).
`
`In addition, in order to assess the impact of folio acid supplementation intervention on the
`toxicity of LY231514 plt~s cisplatin in JMCH, safety data from this study will be
`aua!yzed in the following way:
`
`those patients on the experimenlal treatment arm
`(Group A_vits versus Group A ÷vits)
`
`Lilly proposes to also explore the impac