CLINICAL REVIEW
`
`Clinical Review Section
`
`Cisplatin Exposure in JMCH
`
`In the pre--NDA meeting Briefing Document (scheduled for the January 30, 2002), the proposed
`Indication for malignant pleural mesothelioma stipulated: '
`.__.
`
`'
`Also, the proposed
`Dosage and Administration section of the package insert outlined three regimens:
`,‘_.__/
`The draft Protocol for\_
`treatment JMFE (submitted April 3, 2002, serial #399) initially contained these regimens
`
`The rationale for the inclusion ofthe
`V
`_
`was based on patients who could not tolerate cisplatin. FDA disagreed with the inclusion of
`two ofth'e three regimens in the label and in the expanded access program. This was
`because the combination of alimta + cisplatin was reported to increase survival in IMCH
`and there was no data that showed an increase in survival with alimta alone or the
`
`combination of
`Thus, the FDA
`di d not believe it \xas appropriate to offer expanded access to alimta alone or the
`
`A
`combination of _____
`
`’
`
`Later, in an amendment to JWE (submitted 12/16/2002,), it was stipulated that patients would
`receive alimta + cisplatin who have been previously treated with cisplatin-based regimen and
`responded for six months, and who did not have medical contra-indiCations to receivingmore
`cisplatin, i.e., renal insufficiency, significant neuropathy, ototoxicity and very low left ventricular
`ejection fraction. Again, all of these reasons did not appear appropriate to exclude cisplatin. First,
`patients, who have renal insufficiency and cannot have more cisplatin, cannot receive alimta--a
`drug excreted renally. Second, patients who have a very low left ventricular ejection fraction,
`which contraindicated cisplatin, may not tolerate-three days of potent corticosteroidsua part of
`the alimta regimen. Third, patients who have a non-response to prior ciSplatin can have cisplatin
`+ alimta in view ofthe claimed synergy between cisplatin and alimta in an in vitro model.
`
`
`
`may ha\e been derived from
`How ever, the promotion of
`safety concerns or investigator preferences in JMCH. Review of the doseintensity tables
`prmided by Lilly in the IMCH study report suggested that overall planned ciSplatin dose—
`intensity was the same as planned alimta dose-intensity (table below). Based on this analysis, it
`did not appear that alimta was given without cisplatin to a significant extent.
`
`173
`
`

`

`
`
`Clinical Review Section
`
`Ta ble JMCH.12.4.
`
`Dose Intensity (DI)
`RT Population
`H3E-MC-JMCH
`
`Statistics
`Number of Patients
`
`Planned Menu r Patient (mgfmlfweek)
`Delivered Meanf Patiml
`Percent of planned Dl (delivered/Named)
`
`92.0%
`
`| 66.7
`[53.4
`
`Cis ulatin
`Cis nlatin
`222
`
`25
`21M
`96.4%
`
`Table JMCH.12.5.
`
`Dose Intensity (DI)
`RT Po'pUlatlon by Supplementation Stews
`H3E-MC-JMCH
`
`
`
`Number of Patients
`
`
`
`
`l 66. 7
`
`
`1 54.6
`23.4
`22.6
`93.6% 90.4%
`
`
`92.7%
`
`
`
`Planned Mean/ Patient (nig/inZ’week)
`Delivered Mean 1- Patient
`lanned Di delivered/t
`Percent of
`
`lame-
`
`In Appendix 16.1.10, Listing of Patients Receiving Test Drug(s) or
`Investigational Product(s) by Lot or Batch Number (p. 1763-1874), of the JMCH study
`report, it appeared that there were several patients who did not have cisplatin lot or batch
`numbers recorded at baseline and/or at some time during the study. Non~recording of the
`cisplatin lot number may have been because the. site did not record it or the cisplatin lot number
`was not recorded because cisplatin was not given to the patient. Below is the portion of the CRF
`where the information was to be recorded.
`
`174
`
`

`

`
`
`Clinical Review Section
`
`«9% Cfmical Report Form
`A Single-blind Randomized Phase 3 Trial of MTA plus
`Cisplatin versus Cisplatin in Patients with Malignant
`Pleu'a'Mm‘helbm
`' HSEMC-JMCH
`_
`
`I
`
`'
`
`
`
`Normal Saline & Clsplatln (2de (Visit)
`Study Drug Packet
`
`1
`
`I
`
`lnillals
`
`STUDY DRUG CT NUMBER 2 CISPLATIN
`
`F
`
`"Mo or more vials with the same Lot [amber are used for the infusion; record the Lot number only I
`once.
`If there are only one, No, or three Lot numbers to record. leave other spaces blank.
`
`
`
`DaleVerified
`
`lnitlale
`
`Below is a table of patients on the alimta + cisplatin arm, who the cisplatin lot number was not
`reported at baseline and throughout the treatment.
`
`INVESTIGATOR
`SITE
`
`'
`
`PATIENT # # OF CYCLES dose delayed or reduced
`cycle#—reason for dose
`dela or reduced
`
`1072
`
`clearance
`
`2,3,6-cisplatin & alimta
`delayed, creatinine
`clearance; 5-cisplatin &
`alimta delayed,
`neutrophil; S—alimta
`reduced, stomatitis
`
`4-cisplatin & alimta
`delayed, creatinine
`
`175
`
`

`

`CLINICAL REVIEW
`
`Clinical Review Section
`
`‘
`
`PATIENT # # OF CYCLES dose delayed or reduced
`
`INVESTIGATOR
`
`
`I-_dela ' or reduced
`SITE
`
`
`
`
`
`2
`
`2-cisplatin & alimta
`delayed. anemia
`
`
`
`_-
`
`
`
`
`A sample from Appendix 16.1.10 is patient #130-1261 (also, included is patient #130-1 196 who
`had the cisplatin lot numbers recorded).
`
`130 u" nuts, 1
`2
`)
`I
`‘
`£
`
`no
`
`11::
`
`In/Cllp
`
`III“
`:al!:l
`MN}
`:uhl
`nu:
`ulhl
`“I“
`:JIZIJ
`1|“:
`un"
`inn
`xatsl7
`allix
`14.;2
`140;:
`140;:
`11"!
`alas:
`
`In
`txsrna:xn
`m
`Gnu":-
`In
`cx nun:
`m
`cxtrLaax-
`In
`trunn-
`In
`CISYLAYXI
`
`ME!”
`
`Below is a table ofpatients on the alimta + cisplatin arm, who the cisplatin lot number was not
`reported at baseline and the cisplatin lot number was reponed in later cycle(s).
`
`
`PATIENT #
`# OF CYCLES
`TOTAL #
`
`INVESTIGATOR
`
`
`dose delayed or
`reduced
`SITE
`CISPLATIN LOT
`' CYCLES
`
`
`
`NUMBER NOT
`cycle‘ri-reason for
`
`
`
`REPORTED
`dose delay or
`
`
`
`
`reduced
`
`
`
`
`130
`
`c cle
`
`c cle v
`
`1631
`
`1450
`
`2550
`
`131
`
`136
`
`140
`
`251
`
`510
`
`554,
`
`lst 3 cycles, 5th-
`12th cycle; only 4th
`cycle with cisplatin
`
`12
`
`
`
`
`
`
`
`
`
`
`4-cisplatin reduced,
`deafness; 5-12-
`cisplatin omitted,
`deafness; 9-alimta
`delayed, URI
`
`lst 2 cycles
`
`'
`
`3
`
`5103
`
`lst cycle
`
`5516
`
`lst cvcle
`
`2-cisplatin & alimta
`reduced, nausea
`
`2-cisplatin & alimta
`reduced, platelet
`count reduced
`
`5-cisplatin & alimta
`reduced,
`deh dration
`
`3-cisolatin & alimta
`
`176
`
`

`

`5 TNVESTIGATOR PATIENT #
`I
`
`CLINICAL REVIEW
`
`Clinical Review Section
`
`# OF CYCLES
`CISPLATIN LOT
`NUMBER NOT
`REPORTED
`
`TOTAL #
`CYCLES
`
`-_-_
`
`dose delayed or
`reduced
`cycle#-reason for
`dose delay or
`reduced
`
`delayed, creatinine
`clearance
`
`‘ A sample from Appendix i]6.1.10 is patient #136-1631.
`13‘
`at): m/nup 1
`nu:
`2
`110;}
`I
`um
`A
`nun
`sun”
`6
`It“?
`um
`6
`1
`1H“
`I
`1n“
`9
`1‘ I52
`10
`1"“
`.u
`1"“
`12
`um
`
`In
`m
`In
`In
`sunny
`In
`In
`m
`n:
`In
`In
`”A
`m
`
`Below is a table ofpatients on the alimta + cisplatin am, who the cisplatin lot number was
`
`re 011ed at baseline and the cisplatin lot number was not re orted in later cvcle(s).
`TOTAL #
`TOTAL # OF
`dose delayed or
`CYCLES MTA+CISPLAT
`reduced
`
`CISPLATTNV
`
`LOT NUMBER
`NOT
`REPORTED
`
`PTS. @ SITE
`
`cycle#-reason for
`dose delay or
`reduced
`
` 2-cisplatin & alirnta
`
`6 were not
`
`reported for both
`cis-latin +alimta.
`
`
`
`delayed, creatinine
`clearance; 3-cisplatin
`& alirnta delayed,
`white blood count
`
`2, 4, 5-cisplatin &
`alimta delayed,
`creatinine clearance;
`2-cisplatin & alimta
`reduced, serum
`creatinine increased
`4-cislatin & alimta
`
`177
`
`

`

` CLINICAL REVIEW
`
`Clinical Review Section
`
`
`
`TOTAL # 0E
`
`
`
`
`TMESTJGATO {PATIENT #
`: OE CYCLES
`ClSPLATIN
`MTA+CISPLAT
`
`
`
`LOT NUMBER
`PTS. @ SITE
`
`
`NOT
`
`
`
`REPORTED
`
`
`reduced,neutrophil
`count reduced
`
`3-cisplatin & alimta
`
`delayed and reduced,
`vomiting; 4-cisplatin
`
`omitted. vomittin -~
`
`
`1
`
`dose delayed or
`reduced
`cycle#-reason for
`dose delay or
`reduced
`
`
`
`Samples from Appendix 16.1.10 are patients #136-1633 and #720-7200.
`11‘ m)
`Inter-p 7
`run
`In
`mm cunum
`mm cuunn
`I
`non
`In
`an,“
`In
`9
`
`720
`
`72" mitt-p 1
`3
`3
`I
`
`"mm m
`PDI'7I‘I
`“$1381“
`“3110”
`in
`mm)
`(xsnnn
`"min In
`’mn)!
`unnu-
`”31105)
`In
`
`The tables suggested that several patients might not have received cisplatin at baseline and/or at
`some time during the JMCH study. In response to FDA concern about this, Lilly stated that only
`two patients--#136-l631 and #720-7200 had cisplatin omitted (response dated 9/19/2003). For
`patient #136-163, Lilly acknowledged that cisplatin was omitted cycles 5 -l2. Appendix 16.1.10
`indicated that the cisplatin lot number was also not reported for cycles 1-3. By using this
`appendix, there was no way to tell the difference between cycles that cisplatin was omitted and
`cycles that the cisplatin lot number was not recorded. Also, Lilly stated that no patients on the
`. alimta/cisplatin arm of study JMCl-l received --———s
`at baseline or at any time during the
`study and that there were no patients on the alimta/cisplatin am of study JMCH who had alimta
`omitted and received only cisplatin at baseline or at any tinie during the study.
`
`In their response submitted 11/6/2003, Lilly stated, "on inspection of Appendix 16.1.10 in the
`JMCH study report, it might appear that some patients received Alimta but not cisplatin."
`Additionally, Lilly stated that the cisplatin 10t numbers were not collected for these patients and
`that only two patients had cisplatin omitted in the alirnta/cisplatin arm of study JMCH.
`
`In conclusion, the requests for inclusion of regimens oft.
`the first proposed package insert and Protocol for Treatment were not based on information
`generatedin the pivotal trial, JMCH. Except for the two patients acknowledged by Lilly, Lilly
`stated that all patients on the alirnta + cisplatin arm received both alimta + cisplatin while they
`_ were on the JMCl-l study.
`
`t in
`
`
`
`178
`
`

`

`
`
`Clinical Review Section
`
`Survival: The Primary Endpoint
`
`No source documents were provided or reviewed. The FDA statistician used datasets submitted
`by Lilly on December 6, 2002. The datasets were located in the Electronic Document Room
`(EDR) of CDER of FDA under the Letter Date “24-OCT-2002” and “6-DEC-2002”,
`respectively. The major data set for the efficacy analysis was “SURVLOCK” which defines the
`survival time and events.
`
`Below are the results of the FDA statistician's survival analysis of study JMCH.
`
`Survival Analysis of Randomized and Treated Patients
`
`Table 1.
`
`, Patients dead“
`
`Primary Endpoint: Survival for 'RT Population SFDA Analysis.)
`RT Population
`FS Population
`PS+NS Population
`(N=448
`’N=33l
`N=l 1
`LY/cis
`Cisplatin
`LY/Cis
`Cisplatin
`LY/cis
`Cisplatin
`(N=226)
`('N=222)
`(N=l68)
`('N=163)
`(N=58)
`(N=59)
`n (%
`n (%)
`n (%)
`n (%)
`n (‘34:)
`n (%
`145 (64)
`159 (72)
`95 (57 )
`103 (63)
`50 (86)
`56 (95)
`
`-
`
`_
`
`WW
`-Meditin
`
`12.1
`
`9.3
`
`13.3
`
`10.0
`
`9.5
`
`7.2
`
`(95% C1)
`p-vglue"
`Long-rank
`Wileoxort
`
`>
`
`(100.144)
`
`(7.8, 10.7)
`
`(114,149)
`
`(8.4. 11.9)
`
`(8.1, 10.8)
`
`(6.5. 9.9)
`
`0.021
`0.028
`
`0.051
`0.039
`
`0.253
`0.440
`
`0.798
`0.758
`0766
`Hazard Ratio”
`(0.54, l.l7
`(0.57 1.0)
`0.61 0.96
`95% CI for Hazard Ratio‘
`
`Statistical rcviewer’s results based on the analysis data sets provided by the sponsor.
`" Patients were died for different reasons: study disease related. study toxicity, and other causes.
`1’ P-value is based on the test results for the two treatment groups.
`‘ Hazard Ratio is bused on the proportional—hazards model with the monent as single independent variable.
`
`. In the randomized and treated (RT) (n=448), the median survivals for alimta/cisplatin and
`cisplatin alone were 12.1 and 9.3 months, respectively (log-rank, p=0.021); this was a
`statistiCally significant increase in median survival of 2.8 months. In the subgroup149 of the fiilly
`folic acid and vitamin B12 supplemented patients (n=331), the median survivals for
`alimta/cisplatin and cisplatin alone were 13.3 and 10 months, respectively (log-rank, p=0.051);
`this was a marginally statistically significant increase in median survival of 3.3 months.
`In the
`underpowered subgroup of partially folic acid and vitamin B12 supplemented plus never
`
`”9, Lilly tested three models in the prognostic evaluation of survival the optimal parameterization was found to be
`Model FS+PS versus NS. A comparison of Model FS versus PS+NS (defined in the statistical analysis plan) had
`less prognostic power than the alternative parameterization (FS+PS versus NS). This finding was based on the fact
`that Model FS+PS versus NS had a smaller p-value for the supplemenatation group factor and a larger log-likelihood
`value. These results suggested that, with respect to survival, PS patients were more like FS patients than NS patients.
`
`179
`
`

`

`CLINICAL REVIEW
`
`Clinical Review Section
`
`supplemented patients, the median survivals for alimta/cisplatin and cisplatin alone were 9.5 and
`7.2 months, respectively (log-rank, p=0.253); although this was a 2.3 month increase in survival,
`it was not statistically significant. The hazard ratios of 0.766, 0.758, and 0.798, for the
`respective survival analyses were consistent with regard to a survival benefit in the
`alimta-’cisplatin arm compared to the cisplatin alone arm.
`
`,There were 8 patients (2 alimta/cisplatin, 6 cisplatin alone) who were randomized and not
`included in the survival analysis. With 456 randomized patients (304 events, 152 censored), i.e.,
`448 + 8 patients, the results'of the FDA survival analysis were:
`
`"lntent-to-Treat" Analysis of Survival
`
`'F—NTENT-TO—TREAT ALIMTA/‘CISPLATIN CISPLATIN
`
`p-value
`
`(N=153) Survival, median
`
`0.0205
`9.3 months
`12 months
`
`(10, 14.4)
`(7.8, 10.7)
`(95% Cl)
`
`In the intent-to-treat population (n=456), the median survivals for alimta/cisplatin and cisplatin
`alone were 12 and 9.3 months, respectively (log-rank, p=0.0205); this was a statistically
`significant increase in median survival of 2.7 months.
`
`The intent-to-treat analysis (with the inclusion of the 8 patients, i.e., n=456) was comparable to
`the randomized and treated analysis (n=448) of survival.
`
`Confirmed Pathological Diagnosis of Mesothelioma
`
`In the past, expert panels have been set up to review suspected malignant pleural mesothelioma
`cases. One editonalist wrote about the need for a panel of experts to review pathological
`material to guarantee the accuracy of diagnosis. ”0 The reason for thisIS three-fold. First,
`epithelial cell type has been associated with a more favorable prognosis in most large series; the
`fibrosarcomatous type carries the worst prognosis, and the mixed type is intermediate. Second,
`it is important to differentiate mesothelioma from adenocarcinoma-~tumors with histologic
`similarities-since it may influence the treatment and the natural history. Adenocarcinomas from
`primary lung, breast, ovary, stomach, kidney, or prostate cancer frequently metastasize to the
`pleura and can be extremely difficult to distinguish from epithelial mesothelioma cytologically or
`histologically. Metastatic adenocarcinoma with extensive pleural involvement may grossly
`resemble mesothelioma and has been called pseudomesothelioma. Third, sarcomatous
`mesotheliomas must be distinguished from fibrosarcoma, malignant fibrous histiocytoma,
`malignant schwannoma, and hemangiopericytoma. Synovial sarcoma and carcinosarcomas,
`
`150.lett JR. Malignant pleural mesothelioma. A proposed new staging system. Chest. 1995;] 08:895-897)
`
`180
`
`

`

`'
`
`CLINICAL REVIEW
`
`'
`
`Clinical Review Section
`
`which may also have mixed sarcomatous and epithelial components, usually present as a
`localized mass in the lung.
`
`1 In genera‘ mesothelioma15 difficult to diagnose, even by expert pathologists. Initial
`misdiagnosis15 common.
`
`In a FDA comment faxed to Lilly on 8/31/2000,151 the importance of independent pathology
`review was stated:
`
`Although all patients may not have sufficient tissue for an independent review of
`histopathology, the slides should be available for review by an independent pathologist.
`The rigor of the study, regarding confidence in the histopathological diagnosis, will be
`decreased without independent review of all cases. In view that only one randomized
`trial in mesothelioma will be accepted for this indication, the one study in mesothelioma
`must be strictly performed.
`
`The following were amendments made to the JMCH protocol, regarding pathology and its
`independent review:
`
`19 June 2000 (~323 out of 574 patients entered on study IMCH at this time)]52:
`
`3.4.2.1. Inclusion Criteria - Not all patients have sufficient tissue for an
`independent review, but will still be allowed in our analysis. (p. 1141 of study
`report JMCH)
`
`Patients may be entered and randomized based on local pathology; however,
`independent centralized pathology review will be carried out on all patients i_f
`feasible. In case of a discrepancy between the assessment of the independent
`reviewer and the investigator, the assessment of the independent reviewer will
`take precedence. (p. 1145)
`
`24 January 2001 (~518 out of 574 patients entered on study IMCH at this time):
`
`Patients may be entered and randomized based on local pathology; however,
`independent centralized pathology review will be carried out on all patients if
`feasible. In—easeetla-daserepaaerbemamssmeneefthe—mdepeadem
`
` W 153(p. 1166)
`
`151 This was in response to submission serial #242, dated 7/12/2000).
`'53 Lilly met with the FDA on 6/21/2000, This was a follow-up to EOPZ re. mesothelioma indication. One ofissues
`forddiscussion \\as whether FDA would accept an interim analysis of secondary endpoints from the mesothelioma
`trial.
`
`153 The strikeouts were pan ofthe citation.
`
`181
`
`

`

`CLINICAL REVIEW
`
`Clinical Review Section
`
`The protocol submitted in the JMCl—l study report stated:
`
`Histologically proven diagnosis of mesothelioma ofthe pleura in patients not candidates
`for curative surgery. Patients will be clinically staged using the IMlG TNM staging
`criteria (see Protocol Attachment JMCHJ). Patients may be entered and randomized
`based on local pathology; however, independent centralized pathology review will be
`carried out on all patients if feasible.154
`
`On page 959 of the JMCH study report, it was stated that: "\ A will assay the blood
`'chemistries, homocysteine, and calculated creatinine clearance (CrCl) and will manage the
`centralized independent pathology review and pharmacokinetic samples."
`
`However, the ENTRY PROCEDURES AND CRITERIA FOR ENROLLMENT
`forrn’ssindicated that independent centralized pathology review was to be carried out on all
`patients.
`
`WORKSPEET
`HJEMZW
`
`gar,
`
`
`'
`ENTRY PROCEDURES AND CRITERM FOR ENROLLMENT
`COMRACEPTK‘IN is ensured through meta-c):
`
`Vat o
`Pa): 1 ol 5
`
`~
`
`D Sterilization tstroial mradtationnmndy
`U Postmenopausal
`D Oral mneept'wes‘
`U Diaphragm
`D Sponge' DI spemic'rde'
`D Comm 316 spen'nidae'
`
`u Mia-meme devioetIJD)
`U Cmtraoen’m indam' or DepuPrwen?‘
`E] 9nd abstinence
`D Solitary part-er who is vasectomlzed
`D Na unruly ad'we
`U Na amicable
`taut-as tn male pawns or mammal tangles
`‘Ena mam‘bzam name as tie (maxim shim page ducal-‘6 bet‘md a sewer: m;
`
`Yes No
`
`lncnsion Criteria: The amass ler tern; 140 must he YES to qualiylor study.
`
`D U 1. Histologirzly prawn dis-31m; ol nesathzliam M he plan it patients not ms
`for wratve sunry. Patient: will be diriully staged ushg the MG TNMstagtr-a attefa
`(see Prdccot Alumni JucHJi Patients may be entered and randomized based on
`bza! mind-am: renewal independent cmaized patholcqy review ‘1'! be can‘ed cu m
`at patients.
`In as: at a dismount-y between he awessmem ol the independert
`"Mum and I: 'nvestigakx. the asasmu‘tt at the ndependent twine! nil take
`pendent:
`
`For pathological diagnosis, the case report form (CRF) provided for checking-off of the box.
`There was no indication on whether the pathological and subtype diagnoses were from the local
`site or from independent centralized pathology review.
`
`15" Page 932 of the .lMCH study report
`”5 Page 1179 ofthe .lMCl-l study report
`
`182
`
`

`

`
`
`Clinical Review Section
`
`AShghUr-d hidwlhzdfluSTrialdMTA
`% u‘m‘d Rap“ Fonn
`5:th weighti- in Wetsuit.
`humPhudbmtalin-u
`I msmm
`DIAGNO$IS :
`INITIAL PATHOLOGICAL chnosns
`
`mm” o
`
`I
`
`
`
`@ "mpmd Du. Slur-Inuit! Mammalian
`
`Dmd in‘tinl mac aw:
`panama—W
`
`Gall: d hmdqinld'qnli
`[Churn—Q
`
`i lv'e-z-qu w
`
`In response to FDA query, Lilly responded with (dated 1/10/2003): "One of the entry
`requirements for study JMCH was to have local pathologic confirmation of malignant pleural
`mesothelioma. This requirement was validated by independent (independent from the site)
`monitors who were fluent in the local language. In addition, local pathology could be validated
`by the FDA during site audits."
`
`In response to FDA query, Lilly responded with (dated 2/13/2003): "Regarding DODP's request
`for pathological confirmation documentation for the patients entered on JMCH, the monitors
`(independent from the site) verified that the diagnosis of mesothelioma on the Case Report Form
`(CRF's) matches the diagnosis shown on the local pathology report."
`
`Although the published report of the JMCH study did not mention central review of pathology
`specimens,156 the accompanying editorial stated that "Central review of all CT scans and all
`pathology specimens was performed. This rigorous approach to analysis lends credibility to the
`study results, especially in a disease for which correct pathologic diagnosis can still be difficult,
`and for which there has been little uniformity in measuring response to treatment."157
`
`'56 Vogelzang NJ, Rusthoven JJ, Symanowski J, et a1: Phase III study of pemetrexed in combination with cisplatin
`versus cisplatin alone in patients with malignant pleural mesothelioma. J Clin Onc012122636-
`2644, 2003
`‘57 Rusch VW. Pemetrexed and Cisplatin for Malignant Pleural Mesothelioma: A New Standard of Care? Journal of
`Clinical Oncology, 21:2629-2630, 2003
`
`183
`
`

`

`CLINICAL REVIEW
`
`Clinical Review Section
`
`The information below, regarding independent central pathology review, was requested from
`Lilly on 9/2/2003 and the response received by FDA on 9/22/2003.
`
`INDEPENDENT
`
`RANDOMIZED AND -
`
`confirm pathology of
`malignant mesothelioma
`
`Documented as tissue
`
`°/0)
`
`(2193%)
`
`
`
`CENTRAL PATHOLOGY
`TREATED, N=448
`I REVIEW CATEGORIES '
`(%)
`
`
`302
`1 Independent review
`
`
`
`(67%)
`1 confirmed pathology of
`
`
`‘ malignant mesothelioma
`
`16
`Independent review
`
`
`(3.6“s)
`suggestive/consistent of
`
`1 malignant mesothelioma
`
`
`Independent review did not
`(630
`
`
`unsatisfactory to confirm
`
`
`l‘athOlO
`
`
`
`Not feasible to send in
`samples for independent
`
`? .atholo -
`review
`
`
`
`
`
`l l l l
`
`87
`
`(19. 4%)
`
`, 67% of the randomized and treated patients had the diagnosis of mesothelioma confirmed by
`independent review; 3.6% of the randomized and treated patients‘ pathology was suggestive
`oficonsistent with malignant mesothelioma. 6.7% of the patients did not have the diagnosis of
`mesothelioma confirmed. 22.3% ofthe patients‘ either had tissue that was unsatisfactory to
`' confirm pathology or it was not feasible to send samples for independent pathology review.
`‘ view that only one randomized trial in mesothelioma will be accepted for this indication, the
`~ lMCl—I study in mesothelioma was not strictly performed.
`
`In
`
`.
`
`Lilly stated that "no adjudication took placein cases where there was discrepancy between local
`and centralized pathology reviews "'58
`
`The information provided on independent pathology review did not take into account the
`histological subtypes ofmesothelioma , i.e., epithelial, sarcomatoid, and mixed. As stated in
`- FDA‘s BACKGROUND ON MESOTHELIOMA section in this review, the histological subtype
`, ofmesotlzeliomana baseline stratificationfactor in study JMCH--can have impact on prognosis
`and an imbalance would aflect the results ofa survival analysis. FDA requested this
`
`1st
`" Response received from Lilly dated 9/22/2003.
`
`184
`
`

`

`
`TLINICAL REVI’
`
`Clinical Review Section
`
`information, as well as, the charter of the independent pathology review and what responsibilities
`were charged to the review ‘59
`
`Lilly sent FDA a flow sheet, illustrating the Independent Pathology Review on 12/ 16/2003. Note
`the date on the sheet is "27Sep02"--about a month prior to when Rolling submission of NBA
`began and conflicts with pn'or amendments and correspondences from Lilly.
`
`Nanci-63m PM Review Process flow
`(meta/amm- 27890!
`
`-wmm “Nimww- Au"
`-Wtflpmhuiuupunm nan-a
`«mil-1mm
`"amwMNMIM
`MI—W
`
`
`
`
`
` BESTPOSSIBLECOPY
`
`~9£quwmsmvmmhmm
`Knmumnmwmw
`inn-W. comm-I‘m"
`. w-mmmhwmmm
`mammal-m2-
`-Manum1m:-ua
`nun-ymuaqu-wum—anwr
`
`Summary of the Independent Pathology Review process:
`
`0 Local investigator site: slides or blocks, and local pathology report were sent to
`
`.__.
`
`0 At - ———-
`
`4-: pathologist interprets slide and enters diagnosis into a blinded database--Pathologist
`1
`
`I DSP staff enters local diagnosis, subtype, differentiation into a blinded database--
`Pathologist 2
`
`IF DiagnosisPamoyogisu = DiagnosisPathologisa -) results entered
`
`_
`'59 From the JMCH study report (p. 77)
`tissue samples for pathological determination (transported and reported via v
`
`
`
`’zAnalysis of tumor-
`
`185
`
`

`

`CLINICAL REVIEW
`
`Clinical Review Section
`
`1F DiagnosisPamologisu == DiagnosisPamotogisa -) Pathologist 3 reads slides-)
`DiagnOSiSpamoto'gisg '9 FINAL _
`
`1n Lilly's response (dated 9/22/2003) to FDA query, a statement was made that "no adjudication
`took place in cases where there was discrepancy between local and centralized pathology
`reviews". According to the Independent Pathology Review Process FIOW outlined above, it
`appears that the determination by Pathologist 3 was the final diagnosis if there was a discrepancy
`between local and Pathologist 1 (review pathologist).
`
`Below is the analysis of mesothelioma subtype derived from independent pathology review
`submitted by Lilly on 12/16/2003. This analysis is on patients whose diagnosis of mesothelioma
`was confirmed and the mesothelioma subtype was confirmed or determined after independent
`review. 21% of the 302 confirmed mesothelioma patients (alimta/cisplatin: 24%, 3 7 out of J 53
`confirmed; cisplatin alone: 18%, 27 out 149 confirmed) had their subope changedfrom the
`designation determined at the investigators’ site.
`
`153 patients on the alimta/cisplatin amt had the diagnosis of mesothelioma confirmed by
`independent pathology review; 149 patients on the cisplatin alone arm had the diagnosis
`confirmed.
`
`Folic acid and vitamin B12 supplement statuses were balanced on both arms in confirmed
`mesothelioma pathology patients (table below).
`
`i rouc ACID/fiTAMIN 1312 ALIMTA/CISPLATIN CISPLATIN Amt?
`SUPP_L_EMENT STATUS
`
`'
`
`186
`
`

`

`i
`
`.
`
`‘~
`
`CLINICAL REVIE‘V
`
`Clinical Review Section
`
`I
`
`Gender was balanced on both arms in confirmed mesothelioma pathology patients (table below).
`
`iGENDER ALIMTA/CISPLATIN ClSPLATlN ALONE '
`
`i
`
`26(17%)
`127 (83%)
`
`25(17%
`124 (83%)
`
`Confirmed Pathological Diagnosis of Mesothelioma Subtypes
`
`The table below illustrates the list of pathological diagnoses entered from the investigators' site
`from patients with confirmed mesothelioma. The independent review consolidated the varied
`mesothelioma diagnoses to subtypes of epithelial, mixed, and sarcomatoid.
`
`ll
`
`:
`
`INDEPENDENT
`REVIEW
`alimta’cisplatin
`Cisplatin
`alone
`127
`
`Cisplatin
`alone
`-
`E-ithelial Pleur Meso —- 130
`
`Alimta’cisplatin
`
`187
`
`Bihasical Fleur. Meso
`Meso Fibrosum Cellular _=——‘
`
`Carcinoma
`Tubulo-P21111211, S-indle Cell—-——
`—!1---——_
`Other -___
`-—-——-
`_-_——
`
`l il il
`
`l‘P-ATl-IOLOGIC DIAGNOSIS
`
`INVESTIGATOR'S
`
`
`1
`
`1—
`9
`
`——i
`11119.11 Cell 12111111 Meso
`Sarcomaroid Pieur. Meso n==9
`Neo M Meso —_—=i
`Pa-illarPleur Meso —-__1
`l_—-——i
`—i-II--’
`
`

`

`
`
`Clinical Review Section
`
`None of the results of the independent pathology subtype review and diagnoses were recorded in
`the DIAGDATA database (the CRF page is below) and there was no "blank" to record the
`information on the CRF.
`
`FMZLZJQSQZQ. ,.
`
`AW?' mp‘mun'udP—ISTHMHTA
`filCiflflthiflfithfish-ih
`
`V mug-ammun-
`mm -
`
`DEAGNOSIS : INITKLPATHOIDGICAL DMNDSIS
`
`
`
`Nod hiflq'flhdqiul hail
`
`0-1:deth
`paw—W
`
`
`
`BESTPOSSIBLEcepv
`
`37 alimta/cisplatin patients had their mesothelioma subtype changed or determined after
`independent pathology review; 27 cisplatin alone had the subtype changed or determined.
`
`The table below illustrates the pattern of change in or determination of subtype diagnoses from
`the investigator to the independent review for the alimta/cisplatin arm.
`
`CHANGE IN PATHOLOGY FROM INVESTIGATOR TO INDEPENDENT REVIEW
`
`independent review
`I atholo 3
`Malign. Meso, Epithelial
`T ‘0 e, Pleur.
`V
`
`alimta/cisplatin {
`i
`
`17
`
`Other
`
`Sarcomatoid Pleur. Meso
`
`Spindle and Epitheloid
`
`188
`
`

`

`l
`
`CLINICAL REVIEW
`
`Clinical Review Section
`
`i
`
`
`
`
`‘mT
`
`l
`‘
`
`uln-
`I‘alhOIO ’
`oe, Pleur.
`Malign. Meso, Mixed
`T\ oe, Pleur.
`
`Malign. Meso, Mixed -(
`T e, Pleur.
`
`i l i
`
`Malign. Meso,
`Sarcomatoid T oe, Pleur.
`
`Malign. Meso,
`Sarcomatoid T p e, Pleur.
`
`
`
` investigator's pathology
`
`Epithelial Fleur. Meso
`
`Meso Malignum
`
`Sareomatoid Pleur. Meso
`
`Epithelial Pleur. Meso
`
`Mixed Cell Pleur. Meso
`
`The table below illustrates the pattern of change in or determination of subtype diagnoses from
`the investigatorto the independent review for the cisplatin alone arm.
`
`
`
`
`CHANGE IN PATHOLOGY.FROM INVESTIGATOR TO INDEPENDENT REVIEW
`
`independent review
`cisplatin
`alone
`atholog
`
`1
`
`12
`
`2
`
`l
`
`2
`
`Malign. Meso, Epithelial
`T pe, Pleur.
`
`Malign. Meso; Epithelial
`. T ne, Pleur.
`
`Malign. Meso, Epithelial
`‘I e, PICUI.
`
`Malign. Meso, Epithelial
`T ‘ne, Pleur.
`
`Malign. Meso, Epithelial
`T re, Pleur.
`
`.
`
`Malign. Meso, Epithelial
`T oe, Pleur.
`
`Malign. Meso, Epithelial
`T re, Pleur.
`
`Malign. Meso, Epithelial
`Tx oe, Pleur.
`
`Malign. Meso, Epithelial
`- T e, Pleur.
`
`Malign. Meso, Mixed
`T p e, Pleur.
`
`Malign. Meso, Mixed
`T pe, Pleur.
`Malign. Meso, Mixed
`
`
`
`
`
`
`189
`
`
`
`
`
`
`
`Iinvestigator's pathology
`Biphasical Pleur. Meso
`
`Mixed Cell Pleur. Meso
`
`Neop M, Meso
`
`Neop M, NOS
`
`Papillar Pleur. Meso
`
`Pleur. Meso
`
`Poorly Differentiated Carcinoma
`
`Sarcomatoid Pleur. Meso
`
`Tubulo-Papillar, Spindle Cell
`
`Biphasieal Pleur. Meso
`
`Epithelial Pleur. Meso
`
`
`
`
`
`Meso Fibrosum Cellular
`
`

`

`CLINICAL REVIEW
`
`Clinical Review Section
`
`
`
`-'!_—
` Mixed Cell Pleur. Meso
`
`Malign. Meso,
`
`
`i
`Sarcomatoid T‘ne, Pleur.
`
`In both treatment arms; independent pathology review shifted more patients to the epithelial
`' mesothelioma subtypes or good prognosis subtype. There was a moderate decrease in the mixed
`_ subtype or intermediate prognosis subtype There was minimal change in the sarcomatoid
`subtype or poor prognosis subtype.
`
`The two tables below illustrate the effect on prognosis due to the change in mesothelioma
`subtype from the investigators's site diagnosis to the independent pathology review diagnosis.
`Although there is an overall improvement in subtype prognosis, the changes appear balanced
`, with respect to both treatment arms.
`
`v investigator‘s pathology
`
`
`
`Independent review
`pathology
`
`’
`
`Malign. Meso
`Eoithelial T p e, Pleur
`Malign. Meso,
`E-ithelial T oe, Pleur.
`
`prognosis
`determination
`
`17
`
`5
`
`intermediate '
`
`'good
`
`good
`
`poor' good
`-
`
`'—
`
`3
`
`Eithelial p e, Pleur.
`
`Malign. Meso,
`Eithelial T e, Pleur.
`
`Malign. Meso,
`
`Mixed Cell Pleur. Meso
`
`i
`
`Neop M, Meso
`
`Sarcomatoid Pleur. Meso
`-
`
`Spindle and Epitheloid
`
`l
`i
`
`Eithelial Tpe Pleur.
`,
`IL_'__
`l‘
`Biphasica] Pleur. Meso . _- unchanged
`
`
`_
`
`-
`
`‘
`
`'
`
`'
`
`good
`
`.
`
`1 1
`
`‘
`Epithelial Pleur. Meso
`
`- Meso Malignum
`‘
`
`.
`
`T 1 e, Pleur.
`
`T 1' e, Pleur.
`Malign. Meso, Mixed
`. T e, Pleur.
`
`T e, Pleur.
`
`good ' intermediate
`
`intermediate
`
`
`
`Other
`
`Malign. Meso,
`
`
`
`Malign. Meso, --
`. EpithelialPleur. Meso
`}
`Sarcomatoid Type,
`
`
`
`Pleur.
`Malign. Meso,
`2
`intermediate 0 poor
`' Mixed Cell Pleur. Meso
`
`
`Pleur.
`
`Sarcomatoid Type,
`
`I90
`
`,
`
`-
`
`.-
`I
`
`,
`
` ..___,.;__'r'
`
`

`

`l
`
`V
`
`CLINICAL REVIEW ‘
`
`Clinical Review Section
`
`!
`
`
`
`lnvestigator‘s pathology
`
`Independent review cisplatin
`pathology
`
`
`
`change in prognosis
`or prognosis
`determination
`
`intermediate . 'good
`
`12
`
`intennediate' ‘good
`
`good
`
`good
`
`unchanged
`
`good
`
`good
`
`Malign. Meso,
`
`Eoithelial T\. oe, Pleur.
`Malign. Meso,
`Eoithelial T\ oe, Pleur.
`
`Malign. Meso,
`Eoithelial Tx oe, Pleur.
`Malign. Meso,‘
`Eoithelial T ‘oe, Pleur.
`
`Malign. Meso,
`
`Eoithelial Toe, Pleur.
`
`Malign. Meso,
`Eoithelial Toe, Pleur.
`
`Malign. Meso,
`Eoithelial ‘o e, Pleur.
`Malign. Meso,
`E oithelia] ‘0 e, Pleur.
`
`Malign. Meso,
`Eoithelial T oe, Pleur.
`Malign._Meso, Mixed
`
`Biphasical Ple‘iir. Meso
`
`Mixed Cell Pleur. Meso
`
`. N
`
`eop M, Meso
`
`Neop M, NOS
`
`. Papillar Pleur. Meso
`
`Pleur. Meso
`
`gl
`
`.
`I
`
`.
`
`|
`I
`
`‘
`
`|
`
`.
`
`Poorly Differentiated
`Carcinoma
`Sarcomatoid Pleur. Meso
`
`Tubulo—Papillar, Spindle Cell
`
`poor' ‘good
`
`unChanged
`
`intermediate ° good
`
`Biphasical Pleur. Meso
`
`1—-
`T oe, Pleur.
`
`Meso Fibrosum Cellular
`
`1
`
`intermediate
`
`L
`'
`'
`
`Mixed Cell Pleur. Meso
`
`T ‘o e, Pleur.
`Malign. Meso,
`Sarcomatoid Type,
`
`intermediate 0 p001.
`
`i
`{
`'
`
`l
`
`19]
`
`

`

`CLINICAL REVIEW
`
`Clinical Review Section
`
`Survival Analyses of Confirmed Mesosthelioma Pathology
`
`On page 962 of the JMfiCl—l study report was the following statement:
`
`"Because there may be a discrepancy between the pathological diagnosis assessment of
`the independent reviewer and the investigator, data analysis will also be performed on all
`patients whose diagnoses were confirmed by the independent reviewer."
`
`This analysis was not in the JMCH study report. Belowuis that analysis:
`
`In the 9/22/2003 Lilly response, the following directions were provided in orde