MINUTES OF ThE ANTIFOLATE ADVISORY PANEL MEETING,
`OCTOBER 28TH AND 29TL1 1997, WASHINGTON DC.
`
`Present:
`
`Panel:
`Dr. Dan Von Hoff (Chair)
`Prof. Hilary Calvert
`Prof. Herbie Newell
`Dr. Luis Paz Ares
`Prof. Hans Schmoll
`Dr. Carmen Allegra
`Dr. Peter Dannenberg
`Prof. Jean Louis Missett
`Dr. Robert Allen
`
`LY 309887 Investigators (afternoon 29th October)
`Dr. Eric Rowinaky
`Dr. Cheryl Aylesworth
`Dr. Sharyn Baker
`Prof. Herbie Newell
`Prof. Hilary Calvert
`
`Lilly:
`Diana Kelley
`Mark Miller
`Don Thornton
`Jerry Thompson
`Steve Hailiburger
`Jackie Walling
`Victor Chen
`Rick Schultz
`Barbara Sterner
`
`Agenda
`As previously published except item 7 (MTA investigator meeting discussion),
`which was deferred.
`
`CON FI DENTIAL
`
`AIrEXHIBIT 312_
`
`DeponentDRp.DEPOBOOK.COM
`
`b
`
`AV00008682
`
`TX 312
`
`Lilly Ex. 2112
`Sandoz v. Lilly IPR2016-00318
`
`

`
`MINUTES
`
`COLORECTAL CANCER.
`MTA data. Front-line RR% = 17%. Survival 14.5 months. Secondline: as yet
`no responses, TTP is 1.6 months. This compares with historical data for CPT
`11 in this setting of 2.6 months. However, details of patients prior 5 FU
`therapy was not available. Third line: no responses, TTP = 2.3 months.
`
`Future development: Third line: Probably should give up. Only possible
`future study would be randomized against BSC.
`
`Secondline. We need to understand prognostic factors of the patient
`population exposed and to check the TTP figures; If one were to proceed the
`study that would be needed would be a randomized study vs. OPT 11.
`However emerging data suggests that CPT11 would be very liard to beat in
`this setting. The response rates in 3 studies of CPT 11 were: 21, 17 and
`12.5%, with the latter two studies being conducted by community oncologists
`
`Front-line. We need to decide whether a study of monotherapy MTA is
`needed to demonstrate the value of MTA, or whether one should only proceed
`in combination. As monotherapy, one would need to randomize against either
`5FU or Tomudex. One would need to demonstrate either better survival or
`equivalent survival with better QOL. It would be preferable to randomize
`against the Mayo clinic schedule of 5 FU since this is more tox and less
`efficacious than the continuos infusion or DeGramont regimes, or at least is
`perceived to be so. Alternatively it might be possible to randOmize against
`the physicians choice of 5 FU regime. There was some divergent opinions
`about the ùeed to do this study. Some members thought that it ws the only
`possibility of demonstrating a survival benefit. Prof. Missett and Prof.
`Schmoll thought that the study should be done. Dr. Paz Ares and Prof.
`Mis sett thought that Tomudex was a useful drug. In particular the Q3 week
`schedule is attractive.
`
`In combination, the issue here is the large number of new players. This is an
`issue not only from a regulatory hurdle perspective, but also in terms of
`competition for clinical trial patients. For example the cómbination of 5FU
`and oxaliplatin is reported to have a response rate of 50%, with median
`duration of survival of 17 to 18 months. There is also an ongoing study of
`5FU and OPT 11 in front-line therapy and a study of CPT11 and 5FU is in
`the planning stages. One possible study design for a randomized trial
`(dependent obviously on phase I data being permissive) would be MTA
`CPT11 vs.. 5FU CPT11. Although it has proven difficult to combine MTA with
`5 FU, the panel felt that it was important to persevere with this combination,
`even ifa low dose of 5FU had to be taken into phase II clinical trial.
`
`CONFIDENTIAL
`
`AV00008683
`
`Lilly Ex. 2112
`Sandoz v. Lilly IPR2016-00318
`
`

`
`NB Buyse has published a useful overview showing that both TIP and RR%
`are poor predictors of eventual survival in colorectal cancer.
`
`Daily x 5.
`The panel felt that thisi schedule should only be examined further in other
`tumors if the colorectal data were positive (currently too early). Daily x 5
`topoteean is not acceptable in the palliative setting in France, because of the
`QOL factor.
`
`NSCLC
`MTA Data: there is a 20 to 25 % response rate in two multicentric phase II
`studies, one of which is complete (Canadian) and the other ongoing (South
`Africa and Australia) . The responses from the Canadian study have been
`independently validated.
`Although a number of responses were seen in stage Ilib disease, there was a
`17%. RR in stage W disease.
`
`Conclusions
`Need greater confidence in the phase II setting (i.e. more data)
`In a phase II study of MTA plus cisplatin, one would be looking for a 1 yr.
`survival of 50 to 60%, with a response rate of 40%
`
`JMBQ
`This is a study in a secondiline setting for patients with either stage 11Th or W
`disease, who have failed i prior regime containing platinum and a taxane.
`The study is a phase 11/111 design. In the phase II portion patients will be
`randomized to MTAvs. Navelbine, and the primary endpoint of this study
`will be response rate. Assuming sufficient activity for MTA the study will
`continue to a phase Ill, where survival will be the primary endpoint.
`Additional endpoInts will include TIP, response, clinica.i benefit? QOL (to be
`decided), and safety. Patients will be stratified for response on prior therapy
`(yes or no), chemotherapy free interval , performance status, and baseline
`homocysteine levels (see later discussion for information in respect of
`functional folate status, and a modification to this design, based on the need
`to evaluate the effect of vitamin supplementation on toxicity).
`
`Richard Schilsky liked this study design
`
`Further development: Parallel path the development of MTA in NSCLC in
`second and front-line settings. The preferred route to registration would
`obviously be in a seçondline setting (speed, fewer competing players). Even in
`a secondiline setting, response rate is not viewed as a useful endpoint.
`Taxotere was turned down on initial application to the FDA, which was based
`on response rate. RPR were told to do a randomized trial.
`
`CONFIDENTIAL
`
`AV00008684
`
`Lilly Ex. 2112
`Sandoz v. Lilly IPR2016-00318
`
`

`
`One possible study in. a front-line setting would be MTA plus Carboplatin vs.
`Carbopiatin Taxol. CÀLGB is currently doing a study of earboplatin Tax vs.
`Tax in patients with stage W disease.
`
`BREAST CANCER
`MTA data: An approximate 30% RR in a single phase II study in patients of
`whom some had received prior chemotherapy for metastatic disease
`Future development: If MTA has sufficient activity in patients that have
`failed anthracydines and taxanes (JMBT phase II), then Rick Schilsy felt
`that a randomized study would be needed in the US. A phase II with an
`approximate 18 to 20 RR% will not suffice. However the caveat to that is that
`a number of phase II studies involving a large number of patients, je at least
`150, with a consistent RR% of 18 to 20% MIGHT however suffice..
`Capecitabine has a response rate of 20 to 23 % in this patient population with
`good TTP.
`
`Potential study designs for a randomized study in these patients would
`include MTA plus Navelbine vs. Navelbine, 5 FU (this would also be a good
`study to differentiate MTA from a mechanistic point of view. Possible
`alternatives would included MTA vs. 5FU, MTA vs. Mitomycin C, or MTA vs.
`Navelbine. There is a study of 5FU and oxaliplatin just starting in this
`patient population (note: where is this study being done? In France?, and is
`this a phase II?). Thecombination of MTA plus oxaliplatin is possible in this
`setting, but obviously phase I data is needed.
`
`NB. Dan Von Hoff bet Hilary Calvert that the response rate of MTA would
`not be more than 30% in this patient population. Hilary accepted the bet!
`
`MTA might have a role in the adjuvant setting, when combined with an
`alkylating agent. In the second line setting after failure of anthracyclines, a
`randomized trial of MTA plus Taxotere vs. Taxotere would be appropriate.
`
`The panel felt that an anthracycine combination was of lower priority than.
`some of the other studies.
`
`BLADDER CANCER
`MTA phase II: This is a study in which patients have had no prior
`chemotherapy for metastatic disease. Currently 19/24 patients are
`evaluable. The response rate is approximately 27 %, and at least 2 of the
`patients have had decreases in tumor volume of at least 85%. There has been
`an excess of toxicity, particularly myelosuppression, and a number of patients
`have had reduced creatinine clearance. Action; look at creatinine Clearance
`criterion in the protocol, is it too high for repeat dosing, or should we be dose
`
`CONFIDENTIAL
`
`AV00008685
`
`Lilly Ex. 2112
`Sandoz v. Lilly IPR2016-00318
`
`

`
`reducing in the event of a Cr Clearance of say <z 60...this would need to be
`determined from inspection of the data
`
`Further development: This is difficult given the number of new agents
`being evaluated in bladder cancer, and the high response rates obtained
`(including gemcitabine)
`
`ECOG are currently doing a study of gemcitabine vs. piritrexim
`
`HEAD AND NECK CANCER
`MTA Data: This is a 2nd line study conducted in France. Currently there are
`4/7 partial responses in the study with a further patient reported to be
`"responding". No patients have yet had progressive disease. All patients
`have had 5FU and cisplatin front line. The chemotherapy free intervals in 3
`of 4 of the responders are 8, 10 and 12 months respectively
`
`Future studies: In patients that have failed 5FU and.cisplatin front-line, a
`randomized study companrig MTA to methotrexate. This is identical to the
`study design used by Aguoron in evaluating Thymitaq, where the response
`rate is reported to be 13% for Thymitaq. This type of study could be difficult
`to do in the US, where patients typically would receive radiotherapy and
`Taxol in this setting. However it might be possible to do this study elsewhere,
`Pakistan is worth considering for example.
`
`Another possible study design would therefore be to randomize against Taxol
`in a second line setting
`
`As a prelude to moving into a front-line setting, it would be necessary to
`conduct a phase II study with MTA plus cisplatin. Another option in an
`adjuvant setting might be MTA plus XRT vs. XRT. However both precinical
`and clinical data with this combination is required before this study could be
`done.
`
`CERVIX.
`MTA data. This is confounded by the loss of many (if not all) responders to
`follow up . Hence a replicate study is needed.
`
`Future data: It may be necessary to use a cooperative group in order to get
`sufficient patients , GOG might be a possibility. Mexico would be a good
`source of cervix patients. In front-line, an appropriate phase III randomized
`trial would be MTA vs. cisplatin.
`
`CONFIDENTIAL
`
`AV00008686
`
`Lilly Ex. 2112
`Sandoz v. Lilly IPR2016-00318
`
`

`
`OVARY
`There is no MTA data as yet. A phase II will start in Canada in early 1998
`
`Assuming MTA has sufficient activity to proceed then a suitable randomized
`study in a second line setting would be MTA vs. topotecan. Topotecan has a
`12% RR in patients who have failed front-line cisplatin and Taxol.
`
`SARCOMA
`Both methotrexate and edatrexate have some activity in this disease, and
`hence a study of MTA might be a good bet. Specifically the EORTC might be
`interested, once a study of Tomudex has completed. This has just started.
`
`CARCINOID
`Infusional 5FIJ has activity in this disease. Hence it might be useful to study
`MTA. Prof. Schmoll was interested in this study.
`
`MESO'rHILIOMA
`Given the observations of 4 PRS in 7 patients treated within the context of.
`the phase I study of M'FA plus cisplatin (JMAP), there is some interest in the
`activity of NTA as a single agent and in combination in this disease. The
`panel felt that due to the difficulties in radiological assessment of responses
`in mesothelionia, two phase II studies should be conducted, certainly with the
`combination, if not with the single agent before proceeding further.
`
`PRECLINICAL
`Given the interpatient variation in circulating plasma levels of thymidine
`(0.1 to luM), it would be helpful to look at the concentration dependence of
`end product reversal by thymidine (Note Joe Shih has this data, mm sure I've
`seen him present it). Similarly levels of FPGS expression are likely to vary
`between tumors and between patients, and hence this could control response
`and or toxicity. The panel asked if MTA formed a quarternary complex with
`5FU and TS. This is not known, however there is no reason to think that
`MTAFdUMP would be more stable (Herbie).
`
`CONFIDENTIAL
`
`AV0 0008687
`
`Lilly Ex. 2112
`Sandoz v. Lilly IPR2016-00318
`
`

`
`VITAMIN METABOLISM AND ITS CLiNICAL RELEVANCE
`Background: Elevated homocysteine is a very sensitive indicator of functional
`folate deficiency.
`There is published data showing that a cocktail of folie acid 0.65mg, B6
`(pyridoxine)12.2 mg, and B12 (cobalamin) 0.4 mg will lower homocysteine
`levels in patients with elevated levels (definition here?) by 50%. Folic acid
`supplementation alone will lower levels by 42%. It is not clear if this
`difference is statistically significant (i.e. not sure if this data was derived
`form a controlled randomized trial or if there were adequate number of
`patients studied to detect a difference of this magnitude). In patients who
`have raised methylmalonic acid ( a marker of B12 deficiency),
`supplementation with B12 will reduce levels of MMA. Oral B12 would appear
`to be as goodas parenteral.
`Contrary to popular belief, B12 is not stored in the liver, what is there is all
`bound. In the diet cobalamin is bound to protein and it must be cleaved in the
`stomach.
`Cystathjonine is the most sensitive indicator of B 6 deficiency. Patients with
`elevated cystathioriine often present with fatigue
`10% of the population is homozygous for a mutant ofmethioninesynthase.
`
`MULTWARIATE ANALYSIS OF MTA DATA
`The data suggests a very strong correlation of 03 and 4 neutropenia with
`homocysteine levels >10, suggesting that one might predict those patients are
`at risk of developing serious toxicity from MTA, that vitamin
`supplementation, particularly folie acid, might alleviate toxicity, and that
`supplementation might allow an increase in dose intensity. There is an
`ongoing phase I study with folie acid (5mg orally daily for 5 days: around the
`dose of MTA) which is currently at the 800mg/m2 dose level soit is still a
`little early to tell if an increase in dose intensity is possible. The panel asked
`if the relationship might be explained on the basis of renal function
`deterioration. However, there appears to be no correlation of markers of
`renal function with toxicity. There appears to be no relationship of
`homocysteine with response, suggesting that it might be possible to
`supplement with folic acid and preserve antittunor effect.
`
`Integration of this data into future trials: This data if confirmed would
`suggest that it would be important to stratiu for homocysteine levels in
`randomized trials with MTA.
`
`How much folic acid, B6 and B12 would you need to give and how long should
`they be given for?
`Robert Allen recommends using a cocktail (see combination data above) in
`order to reduce homocysteine levels, of 2mg B12, 25mg B6 and i mg folie
`acid. A weeks supplementation should normalize levels of homocysteine.
`
`CONFIDENTIAL
`
`AVO 0008688
`
`Lilly Ex. 2112
`Sandoz v. Lilly IPR2016-00318
`
`

`
`An important question to answer wouid be to asses the effect of
`supplementation on toxicity and also efficacy of MTA. It was felt that this
`question could be addressed within the study ¡MBQ (see above). Hence it was
`decided to make the study 3 arm in the first instance, adding an MTA plus
`vitamin arm (BG, B12, folie acid).
`
`LUNG TOXICITY.
`This is very, rare. It is difficult to asses if this is really drug related. At this
`stage it will be important to make sure that any toxicity is well characterized
`whether it is reversible etc. The panel did not feel it appropriate to add
`additional tests such as PFTs to the current protocols. High resolution CT
`scanning should be used if a patient develops lung toxicity
`
`PK
`Population Pk: Data is available in courses i to 3. It would be helpful to
`have data from later cycles, and also to understand the duration of
`neutropenia. Is there a pk interaction with decadron, it might be important
`to look. (In cell culture decadron inhibits cytotoxic induced apoptosis). In the
`MTA cispIatin study the panel suggested it might be important to get
`CrEDTA data from the MTA cisplatin sequence and then the reverse
`sequence to look at potential renal toxicity.
`
`CTC CRiTERIA
`The new criteria are' currently out for further testing. The panel thought that
`the approved version would be available in the April I May timeframe and
`that it would bé very important for Lilly to start using them as soon as
`possible once they have been approved.
`
`CON FI DENTIAL
`
`AyO 0 0 0 8 68 9
`
`Lilly Ex. 2112
`Sandoz v. Lilly IPR2016-00318