RECORD OF TELEPHONE CONTACT
`
`Date of Contact:
`
`October 30, 1995
`
`Reference:
`
`From:
`
`Subject:
`
`Participants:
`
`FDA:
`
`BB-IND 4517 rhuMAb HER2
`Roxanne Bale~.S
`
`FDA Telecon on H0648g Protocol Amendment
`
`Susan Jerian, OTRR/DCTDA
`Peter A. (Tony) Lachenbruch, DBE
`Terry Nieman, DBE
`
`Genentech: Roxanne Bales, Regulatory Affairs
`Jim Green, Project Team Leader
`Sue Hellman, Medical Affairs
`Gracie Lieberman, Biostatistics
`Corsee Sanders, Biostatistics
`
`SUMMARY
`On October 30, 1995, we held a conference call with FDA to discuss the amendment
`to our pivotal trial protocol to allow patients treated with doxorubicin in the
`adjuvant setting to be treated with taxol in combination with rhuMAb HER2 or
`placebo. The Agency is in basic agreement with our proposed changes to the
`
`protocol and said they would wait to receive the protocol amendment and fully
`
`analyze it to give their final review. The issues we had asked them and a summary
`of their answers are as follows:
`
`1/ 4517-024 roe
`ss
`PROTECTIVE ORDER MATERIAL
`
`GENENTECH_0000026
`
`
`
`1 of 8
`
`Celltrion, Inc. 1058
`Celltrion v. Genentech
`IPR2017-01122
`
`

`

`BB-IND 4517
`October 30, 1995
`Page2
`
`Does the Agency agree with the proposal to use paditaxel off-label in the pivotal
`comparator trial H0648g?
`
`Dr. Susan Jerian cautioned us saying that they would not stop us from doing the
`
`trial, but they felt that it adds enormous complexity to the trial, that we have no
`
`Phase II data on any of the combination therapies, and they are concerned about our
`proposal to use only pooled data. In our discussions with them we assured them
`
`that we had considered the risks of the trial design and that we still intended for the
`endpoint to be as stated in the protocol, i.e., a comparison of time-to-progression in
`patients receiving rhuMAb HER2 and those on placebo. At the conclusion of the
`
`conversation Dr. Jerian reiterated the cautionary statement, saying that a protocol
`with two different combination chemotherapies could cause a host of problems, but
`
`she thinks we have considered them.
`
`Does the Agency agree with the proposed changes of the analysis plan for Protocol
`H0648g?
`
`We spent most of the teleconference reviewing the analysis plan with particular
`focus on the interim analysis for safety and confirmation of the TTP in the placebo
`
`arm. The two statisticians understood the nature of the interim analysis and that it
`would affect the time of follow-up, not the number of patients or the initial length
`
`of follow-up for efficacy. There was also some discussion of the DSMB, its charter,
`and the communications between the DSMB and GNE.
`
`2/4517-024 roe
`ss
`PROTECTIVE ORDER MATERIAL
`
`GENENTECH_0000027
`
`
`
`2 of 8
`
`Celltrion, Inc. 1058
`Celltrion v. Genentech
`IPR2017-01122
`
`

`

`BB-IND 4517
`October 30, 1995
`Page3
`
`Does the agency agree that the patients currently enrolled in Protocol H0648g will be
`included as part of the sample size and will be included in the final analysis?
`
`The patients already enrolled can remain in the analysis.
`
`ACTION ITEMS:
`We need to send the protocol amendment, the DSMB charter and CVs of its
`members to the Agency, and the statistical analysis plan for Protocol H0648g as soon
`
`as they are available.
`
`DETAILED MINUTES:
`After introductions around both sides of the teleconference, we began by addressing
`
`the questions posed in the October 16, 1995 letter to FDA.
`RB: Let's start with the first question in the letter: "Does the Agency agree with
`the proposal to use paclitaxel in the pivotal comparator trial?"
`SJ: We have presented this at a couple of internal meetings and the consensus is
`that we would caution you about this change. We wouldn't stop you from this, but
`
`we feel that it complicates the trial by having multiple combinations with
`
`chemotherapies for which you have no Phase II data. You say you are pooling the
`data for the efficacy analysis. Have you considered doing a separate trial with just
`
`taxol?
`SH: We discussed this, but at this point patient accrual is our main concern. We
`
`want our trial to address the best standards of care and it appears that everyone is
`now using doxorubicin as an adjuvant. We did discuss doing separate trials, but it
`
`would be two trials with the same n and would not solve our accrual problem.
`SJ:
`Previously you did not allow dexamethasone treatment. What are you doing
`about this in the revised protocol?
`
`3/4517-024 roe
`ss
`PROTECTIVE ORDER MATERIAL
`
`GENENTECH_0000028
`
`
`
`3 of 8
`
`Celltrion, Inc. 1058
`Celltrion v. Genentech
`IPR2017-01122
`
`

`

`BB-IND 4517
`October 30, 1995
`Page4
`
`-~
`
`SH: For the patients in the taxol arm, dexamethasone can be used per the package
`
`insert recommendations.
`
`SJ:
`
`But you are still not allowing any in the CA arm?
`
`SH: We changed it for the doxorubicin/ cyclophosphamide patients so they could
`be on a stable low dose of corticosteroids. Our goal was to standardize it in the taxol
`
`per the PI.
`TN: We have questions on the power of the trial: you want to see a significant
`result in the pooled sample, so if you see statistical significance in the pooled results
`and positive in both arms, will you want approval for both cotherapies?
`
`GL: Right
`TN: What if you see significance in only one arm?
`GL: We probably won't get that--to see significance, a single arm would have to
`
`have a much larger increase in time to disease progression than 50%.
`
`TN: Would you ask for approval for just one arm?
`SH: We'd look at the impact of rhuMAb HER2 on other parameters besides the
`primary endpoint of time to progression.
`TL:
`Say if in one arm you had p=0.04 and in the other p=0.3 and overall you had
`p=0.2, we would think that the trial would not support licensure.
`
`GL: We understand that, but otherwise we couldn't enroll the trial.
`
`SH: We understand that that is the risk.
`
`GL: We think we will see an increased TTP, maybe 50% in each arm, but we won't
`
`reach the statistical significance level without pooling the arms for the final
`
`analysis.
`CS:
`In the analysis plan we specify the primary endpoint and that is what we
`
`intend.
`
`TL: You are saying you will focus on the primary endpoint?
`GL: Yes, that is the primary patient group we will be looking at.
`
`4/4517·024 roe
`ss
`PROTECTIVE ORDER MATERIAL
`
`GENENTECH_0000029
`
`
`
`4 of 8
`
`Celltrion, Inc. 1058
`Celltrion v. Genentech
`IPR2017-01122
`
`

`

`BB-IND 4517
`October 30, 1995
`Pages
`
`TN: We also have questions on the interim analysis, but since we only got your
`fax with the analysis section on Thursday we don't yet have a complete opinion.
`
`GL: This interim analysis was developed before the taxol amendment was
`written, but was on the recommendation of the DSMB. We want to look at the
`
`lower boundaries of the TTP for safety reasons.
`TN: You said you want to see if the TTP for placebo is 8 months. How will this be
`determined by this analysis? How much longer will those patients be followed?
`
`GL:
`
`Some will be 6 months and some will be 12 months and then the power will
`
`be adjusted.
`TN: What does that mean? How will it be adjusted?
`
`GL: By adjusting the length of follow-up.
`In your study you are looking for 12 months follow-up. If you run it for 3
`TL:
`years, will you follow-up the patients for 13 months or a full 24 months?
`GL: We will follow-up for the full 24.
`
`SH:
`TL:
`
`[on follow-up period to maintain the power of the study]
`Time used in the protocol for the initial cutoff is 18 months, so your critical
`
`value will be based on 0.75 of the information for a 24-month follow-up. So that is
`one cutoff. But if you go to 36 months then the interim at 18 months would have
`
`0.50, so it changes the critical value for your interim analysis.
`CS: At 18 months we will assess to see if the assumption of TTP of 8 months in
`
`the placebo group is correct. If not, we will recommend to continue to follow-up.
`
`TL:
`
`The concern is that the critical value for the log rank test is based on the
`
`relative fraction of information and the greater the relative fraction, the stricter the
`critical value. So being at 0.25 vs. 0.50 vs. 0.75 in the trial affects the critical value. If
`
`you extend the study you will have to deal with different critical values.
`
`5/4517-024 roe
`ss
`PROTECTIVE ORDER MATERIAL
`
`GENENTECH_0000030
`
`
`
`5 of 8
`
`Celltrion, Inc. 1058
`Celltrion v. Genentech
`IPR2017-01122
`
`

`

`BB-IND 4517
`October 30, 1995
`Page6
`
`CS: The original plan does not have the option to stop for efficacy. If we look at
`the data and if our assumption is valid we would take a look at the 12-month
`completion for each patient as was originally planned. If the assumption is not
`valid, we will extend the follow-up. We will use a 5% level for efficacy for the full
`
`data in both cases. No stopping for efficacy at the interim analysis is planned. The
`
`DSMB will consider only at safety for stopping. We are not going to stop the trial
`
`and declare efficacy.
`TN: So you will extend the follow-up time only if the TTP in the placebo group is
`greater than 8 months. There was concern about looking too many times. If you are
`concerned about accrual, have you considered a patient-related rather than
`time-related interim timepoint such as when 50% of the patients have progressed?
`CS: We thought about that, but if we plan to wait until half the patients progress,
`and if our assumption is wrong, then we would have to wait for an unknown time
`to have the interim timepoint. For example, if the median TTP is 8 months, then
`
`we would expect 50% of the patients to have progressed within 8 months and the
`interim time point would occur at 8 months. But if the median TIP is 13 months,
`then the interim timepoint would be at 13 months which is already beyond our
`
`current 12-month follow-up for each patient.
`If safety is the consideration, then it would be a good thing if patients aren't
`TN:
`
`progressing.
`If very few patients are progressing at the time of the interim analysis, then
`CS:
`that means the treatments are effective, but then it may mean the placebo group is
`
`not progressing either so we would have 2 concerns: safety analysis by the DSMB
`and validating the efficacy assumptions. We are not saying to stop the trial and
`
`declare efficacy.
`If your accrual isn't as expected, how many analyses will you do?
`SJ:
`
`6/4517-024 roe
`ss
`PROTECTIVE ORDER MATERIAL
`
`GENENTECH_0000031
`
`
`
`6 of 8
`
`Celltrion, Inc. 1058
`Celltrion v. Genentech
`IPR2017-01122
`
`

`

`BB-IND 4517
`October 30, 1995
`Page7
`
`GL: We are still planning just this one interim analysis and then the final one.
`There would be the possibility of stopping if the trial is not going anywhere.
`
`So you might stop altogether?
`SJ:
`GL: We would stop the trial for safety, if the placebo group shows an increase in
`TIP as compared to the treatment group.
`TL: We have no problem with interim analysis for safety. My concern is efficacy.
`When you don't stop the trial, the inference is that it is not highly positive, but also
`not a safety problem. Because you still continue the trial, the enthusiasm of the
`investigators could wane.
`GL: At the interim analysis we have very conservative boundaries for the efficacy
`endpoint. In the interim analysis we only want to confirm the assumptions.
`TL: One option might be to look at TTP only for the placebo group.
`GL: That's what we originally planned to do, but the DSMB was concerned that
`the treatment arm might have a shorter TIP and wanted us to look at both arms.
`SJ: We have a question on the DSMB. Who comprises the board?
`GL: William Stewart of Canada is the chairman; he is an oncologist. Susy Scholl
`from Europe is another oncologist. Scot Emerson is the statistician and Kay
`Dickersin, an epidemiologist from the University of Maryland is a representative of
`the advocacy groups.
`SJ: What is their charter? Their purpose?
`
`GL: We can submit the charter if you like. The DSMB is to review safety, since we
`have no data on the combination therapies they are reviewing the Adverse Events
`as we go along. They will look at the interim analysis for safety and will tell us if the
`TTP assumption in the placebo group is correct.
`SJ: What are their communications with the company?
`
`7/4517-024 roe
`ss
`PROTECTIVE ORDER MATERIAL
`
`GENENTECH_0000032
`
`
`
`7 of 8
`
`Celltrion, Inc. 1058
`Celltrion v. Genentech
`IPR2017-01122
`
`

`

`BB-IND 4517
`October 30, 1995
`Pages
`
`GL: We have a CRO who is providing all the data to the DSMB. They will
`
`provide reports as requested by the DSMB. All closed meetings will be held between
`
`the CRO and DSMB.
`CS: We'll have the charter finalized shortly and we can send it in.
`
`SJ: What members are connected to GNE or other companies?
`
`CS:
`
`Part of the charter is financial disclosure. If you want more information we
`
`can send the DSMB members' CV s.
`
`SJ: We will complete our analysis after we get the protocol amendment and the
`
`DSMB charter. We caution you that having two chemotherapies could cause a host
`of problems, but it appears that you have considered them.
`
`SH: Add to what if get different subgroup results pooled analysis is our endpoint
`
`and we accept the risk. We do have other trials (2 Phase Ils and other Phase Ills) and
`
`would look at all the data.
`SJ:
`But you have no Phase Ils with these chemotherapies.
`
`SH: Correct.
`If that is everything, let's address the last question from the letter "Does the
`RB:
`
`Agency agree that the patients currently enrolled in Protocol H0648g will be
`
`included as part of the sample size and will be included in the final analysis?"
`
`SJ:
`
`No problem.
`
`8/4517-024 roe
`ss
`PROTECTIVE ORDER MATERIAL
`
`GENENTECH_0000033
`
`
`
`8 of 8
`
`Celltrion, Inc. 1058
`Celltrion v. Genentech
`IPR2017-01122
`
`