B. Indication: For the treatment of adult patients with refractory anaplastic astrocytoma (AA), i.e., patients at first relapse who have experienced disease progression on a drug regimen containing a nitrosourea and procarbazine.

The TEMODAR^® clinical development program for malignant gliomas focused on GBM and AA from 1993 to 1998. The New Drug Application (NDA) for temozolomide capsules for the treatment of recurrent malignant GBM (glioblastoma multiforme) and AA was submitted to the FDA on August 12, 1998.

During the NDA review process, the FDA agreed that there was no standard of care for relapsed AA in recurrent disease, and agreed to review the AA indication under the accelerated approval regulations with a phase Phase IV 4 commitment to confirm clinical benefit in an AA population in a randomized controlled study to be approved by the Agency.

The clinical section of the AA accelerated approval NDA was based upon a multicenter 32 institution open-label Phase 2 study of temozolomide in the treatment of adult patients with AA at first relapse (Study C/I94-123). The primary efficacy endpoint of this study was Progression progression-free survival (PFS) at six months. Secondary efficacy endpoints were overall survival, objective response and HQL (Health-related Quality of Life). An analysis of event free survival was also included in the NDA.

The FDA clinical review for accelerated approval of temozolomide in AA was completed January 29, 1999. According to FDA analysis of the results of the study C/I94-123 (ITT population n=162), the response rate was 5% CR and 28% PR; median PFS was 6.18 months; PFS at 6 months was 51% (95% CI 43-59%); median OS was 13.59 6 months; and survival rate at 6 months was 75% (95% CI 68-82%); the safety profile was acceptable.

In AA patients who were refractory to both a nitrosourea and procarbazine, the FDA analysis was indicated that the overall response rate (CR + PR) was 22% with 9% complete response and the median duration of all response was 50 weeks (16-114 week range) with median duration of complete response of 64 weeks (range 52 to 114 weeks); the median PFS was 4.4 months and median OS was 15.9 months.

The Oncology Oncologic Drugs Advisory Committee (ODAC) in its 1/12/1999 meeting unanimously agreed (12-Yes, 0-No) that:

In eEarly February 1999, the sponsor submitted a Protocol protocol Concept concept Sheet sheet (PCS) for a proposed post-approval study in newly diagnosed AA patients to be conducted by the cooperative group Radiation Therapy Oncology Therapy Group (RTOG). At that time the proposal was for a two-arm study comparing radiation with BCNU to radiation with BCNU followed by temozolomide. There was no need for a Phase 1 safety program prior to this Phase 3 study.

On February 12, 1999, the FDA issued an approvableapprovable letter for TEMODAR^® for the indication "treatment of adult patients with refractory anaplastic astrocytoma, i.e., patients at first relapse with disease progression on a nitrosourea and procarbazine drug regimen".

The PCS of the RTOG study in newly diagnosed AA patients (Study 98-13) was reviewed by the FDA and initial comments were communicated to Schering Corporation on March 5, 1999. The protocol was revised in collaboration with RTOG to incorporate the FDA comments and was re-submitted to the Agency on June 24, 1999. The revised protocol included a third arm to study the doublet of BCNU/temozolomide. This revised protocol required a Phase 1-safety assessment of the doublet and the FDA required submission of the safety data prior to initiating the study with the BCNU/temozolomide combination. Additional FDA comments were received on July 8, July 30 and August 3, 1999.

On August 11, 1999, FDA approved TEMODAR^® capsules under 21 CFR §314 Subpart H for the treatment of adult patients with refractory anaplastic astrocytoma, i.e., patients at first relapse who have experienced disease progression on a drug regimen containing a nitrosourea and procarbazine.

The FDA approval was based on a surrogate endpoint (response rate) with a post-approval commitment to conduct a study entitled :entitled: "A phase Phase I/III1/3 randomized study of radiation therapy and temozolomide versus radiation therapy and BCNU versus radiation therapy and temozolomide and BCNU for anaplastic astrocytoma". The commitment further stipulated that safety data from a phase Phase I 1 segment of the study would be submitted to the FDA with agreement that initiation of the combination arm (BCNU/temozolomide) would be contingent on FDA approval to proceed. Furthermore, Schering Corporation committed to completing the two monotherapy arms of the study in the event that the combination arm was stopped for any reason.

Two GBM studies were submitted in the original NDA. The first study, C/I94-091, compared temozolomide (112 patients) to procarbazine (113 patients). The primary endpoint of this study was progression free survivalfree survival (PFS) with overall survival (OS) as a secondary endpoint.secondary endpoint. PFS at 6 months was 21% (95% CI 13-29%) for temozolomide versus 9% (95% CI 4-15%) for procarbazine (p=0.016 )0.016). Median overall survival in this study showed a trend favoring temozolomide (7.34 vs 5.82 months; p=0.067).

The second study (C/I94-122) was a single arm study of 138 GBM patients inpatients in which 6-month PFS was reported as 19%.

The GBM claim from the August 12, 1998 NDA was determined not to be eligible for standard or accelerated approval.

.

1. A phase Phase I/III1/3 randomized study of radiation therapy and temozolomide versus radiation therapy and BCNU versus radiation therapy and temozolomide and BCNU for anaplastic astrocytoma.

2. Other TEMODAR^® Commitments: On January 25, 2001, the FDA issued a Pediatric Written Request and SPRI submitted the required studies on September 12, 2002. As a result of this, pediatric exclusivity was granted on November 12, 2002 and this commitment is now completed.

The Radiation Therapy Oncology Group (RTOG) Protocol 98-13 has been reviewed by the FDA on several times occasions and in response to FDA guidance, has been amended three times. The date(s) entered between parentheses next to section headings refer to the dates of the protocol amendment.

The primary endpoint of RTOG 98-13 is survival. The standard arm is radiotherapy (RT) plus BCNU. The experimental arm is RT and temozolomide. Assuming that the Median Survival Time (MST) for RT+BCNU is 36 months and the RT and temozolomide arm has a MST of 54 months, then a sample size of 216 evaluable patients per arm will provide overall statistical power of 90% with a one-sided significance level of 0.05. Since it is expected that 5% of the patients will be ineligible, then a total of 454 randomized patients will be required. The primary analysis method will be Kaplan-Meier with a stratified log rank test.

According to Scott et al.⁵⁹ and Curran et al.⁹ [superscript numbers indicate references in the protocol] the recursive partitioning analysis (RPA) classes are prognostically important. Based upon eligibility criteria, patients may be in RPA classes I-IV which have decreasing estimated MST from 58.6 to 11.1. The distribution of patients by RPA class will affect the expected number of deaths during the study. It is assumed that 67% of the patients will be in class I, 25% in class III, and 8% in classes II and IV combined. If the percentage of RPA class II patients in the study sample is substantially higher than 25%, then the MST will be lower than 36 months, or if the percentage of RPA class I patients is higher than 75%, then the MST will be higher than 36 months. In either case the sample size may need to be adjusted. (8/17/01, 2/18/02)

Initially, fifteen patients were accrued to the RT+temozolomide+BCNU arm and the specified number of dose limiting toxicities was not exceeded (≥ 2 patients with grade ≥ 3 pulmonary toxicity or ≥ 5 patients with grade ≥ 4 thrombocytopenia or neutropenia). However, a sufficient number of patients had dose reductions resulting in protocol changes, notably a reduction in the BCNU dose from 200mg/m² to 150mg/m². In addition, the eligibility criteria were also changed (patients with oligodendrogial/astrocytic tumors are eligible if the oligodendrogial component is less than 25%; pre-study liver and renal function limits added; eligible DLCO increased from ≥ 60% to ≥ 70%).

An additional 14 patients were accrued to assess further the combination of the RT+temozolomide+BCNU. The dose of temozolomide was 150 mg/m² p.o. daily on days 1-5 with BCNU 150 mg/m² i.v. on day 5. Accrual was suspended for 3 months to assess safety. Based on the high proportion of dose reductions seen and the toxicities encountered, the outcome of this safety assessment was that the phase 3 component of this study will consist only of Arms 1 and 2 (see Treatment Schema).

An additional 454 patients are to be randomized to these two treatment arms discussed in Section 13.2 of the protocol.

Patients are to be randomized according to a permuted block design, balancing by institution within strata. The randomization is stratified by age (<50 vs >50), KPS (60-80 vs. 90-100), and prior surgery (resection vs. biopsy). These stratification factors ensure balance by RPA classes as well.

Three interim analyses of the primary study endpoint (survival) are scheduled as per the following table:

If a significance level is smaller than the H₀ values, then the null hypothesis will be rejected. These significance levels were calculated to ensure an overall significance level of 0.05. There will be two stochastic analyses: at 50% accrual and 75% accrual. If the stochastic analysis indicates less than 15% power to observe the alternative hypothesis, then the study will be recommended to be closed. The results of these interim analyses will only be reported in a blinded fashion to the RTOG Data Monitoring Committee (DMC). A report with recommendations will be given to the study chairman. Any problems or recommendations identified by the DMC, not results, will be reported to the Brain Committee, which is responsible for this study and, if necessary, the RTOG Executive Committee, so that corrective action can be taken.

This analysis is planned at the point where all patients have been followed for a minimum of 36 months, or a maximum of 251 deaths have occurred. The anticipated components of this analysis are:

b) reporting institutional accrual;

c) distribution of important prognostic baseline variables by treatment arms;

d) observed results with respect to the endpoints described in (need correct reference)

Survival is the primary endpoint. RT+BCNU will be compared to RT+temozolomide. A significance level of 0.0405 (one-sided) will be used, adjusting for prior analyses. Analyses within RPA classes, or other prognostic groups, may be performed if there are sufficient numbers of patients.

Time to tumor progression will be evaluated. Subgroup analyses within RPA classes, or other prognostic groups, selecting the best treatment, may be performed if there are sufficient numbers of patients.

Overall toxicity will be compared across treatments. The comparison will be performed using the Pearson chi-square test.

Pathologic samples will be analyzed for chromosomes 1p and 19q and CDKN2A. The distribution of the outcome of molecular analyses will be examined by treatment arm to identify any imbalance. If there is no imbalance then the treatment arms will be collapsed and survival and time to tumor progression will be compared by the groups identified by Cairncross et al. These groups will also be correlated with other pretreatment characteristics.

Following protocol review and approval within RTOG and NCI, the phase Phase 1 component of the study was initiated 16 June 2000June 16, 2000. Enrollment was suspended on 15 March 2001March 15, 2001 for evaluation of the safety profile of the combination arm in this cohort. This evaluation was completed by RTOG in June 2001 and it was decided that additional safety data for the combination of BCNU/temozolomide was needed before commencing the phase Phase 3 study. An additional cohort of 15 patients was added to the phase Phase 1 plan. The phase Phase 1 study re-opened on August 17, 2001 and closed to enrollment on January 25, 2002.

The safety results from the two cohorts of phase Phase 1 did not support starting phase Phase 3 portion of the study with the combination arm. Consequently, the third arm (BCNU/temozolomide doublet) was dropped, the amended protocol submitted to the FDA on October 8, 2002 and the phase Phase 3 study was submitted to IRBs beginning on 16 October 2002.October 16, 2002. The Phase III portion of the study was initiated in January, 2003.

The study completed accrual to initial cohort of the phase Phase 1 study (15 patients) on March 15, 2001 and the second cohort completed enrollment (14 patients) on January 25, 2002 whereupon enrollment was suspended. Following review of the safety data and the decision to drop the doublet arm, enrollment was opened to the Phase 3 portion of the study on January 10, 2003.

As of 30 January 2003January 30, 2003, 5 patients have been accrued in the Phase 3 study. The total accrual is targeted to be 454 patients. The patient accrual was originally projected to be 12 cases per month based upon accrual rates to study RTOG 94-04. At this rate, it will take approximately 38 months to reach the required total accrual of 454 cases. Several initiatives are being taken to increase the accrual rate in order to decrease the time needed to achieve the target study size.

At the time the phase Phase 3 study was designed, the anticipated patient accrual was 12 patients per month for 38 months. Based on the study status as of January 2003 the last patient would be enrolled in December 2006. If the accrual rate can be at least doubled through current initiatives, it is estimated that the last patient would be enrolled by the end of 2004. The primary endpoint of the study is survival and the presumed median survival time for the treated AA patient population is 36 months.

If the target of last patient enrollment by the end of 2004 is achieved, and the event rate occurs as predicted, an analysis of survival could be available by late 2007.

The final study protocol has scheduled interim analyses based on a prespecified number of events which may allow study results to be shared with the Agency in advanced of June 2007.

A report of the survival analysis, assuming the current initiatives to achieve accelerated enrollment are successful, may be available by late 2007.

Initiatives are being taken to increase accrual to the Phase 3 study to about twice the original estimates. Accordingly, it is estimated that the last patient could be enrolled by the end of 2004.of 2004. If the target of last patient enrollment by the end of 2004 is achieved, an analysis of survivalof issurvival is anticipated by Novemberby November of 2007. It would be possible to issue a summary report on survival by the end of 2007. The RTOG protocol has scheduled interim analysis, which may allow study results to be shared with the Agency prior to June of 2007.

The sponsor would like to explore with the Agency possible alternative means to meet the post-approval commitment (confirmation of clinical benefit), foremost among them submission of the results of an EORTC randomized study in first line GBM comparing radiotherapy with temozolomide to radiotherapy alone.