GLIADEL (BCNU) WAFER PLUS CONCOMITANT TEMOZOLOMIDE THERAPY AFTER PRIMARY RESECTION OF GLIOBLASTOMA MULTIFORME

Background

• Glioblastoma multiforme (GBM), the most common malignant primary brain tumor in adults, is associated with a median survival of <15 months.
• Implantation of polifeprosan 20 with carmustine (BCNU) implants (Gliadel; MGI Pharma, Inc., Bloomington, MN), which provide local delivery of the chemotherapeutic polymer BCNU, is known to extend median survival by 2 to 4 months for patients with newly diagnosed and recurrent malignant astrocytoma.
• Adding the systemic chemotherapeutic agent temozolomide (TMZ) to standard radiation therapy has been reported to increase median survival by 2.5 months (vs radiotherapy alone), further extending median survival after primary resection.
• However, postoperative radiotherapy plus concomitant and adjuvant TMZ alone does not deliver therapeutic agent to the debulked tumor during the 3-week period between surgery and radiotherapy. The BCNU implant wafer essentially acts as a theoretical bridge during this nontherapeutic period, allowing continuous adjuvant therapy beginning immediately following tumor resection.
• Although trials are under way, no other published studies to date have examined whether combining BCNU implant and TMZ therapy is safe or further improves survival in patients with newly diagnosed GBM.

Objective

• The primary aim is to assess the overall survival and treatment-related morbidity for the radiotherapy (XRT) + BCNU implant + TMZ cohort (2004–2006). Overall survival of patients not receiving BCNU implant (XRT + TMZ) during this period is also assessed.
• The secondary aim is to compare overall survival and treatment-related morbidity between age-matched (18–70 years of age) XRT + BCNU implant + TMZ and XRT + BCNU implant alone cohorts.

Title/Citation

McGirt MJ, Than KD, Weingart JD, et al. Gliadel (BCNU) wafer plus concomitant temozolomide therapy after primary resection of glioblastoma multiforme. J Neurosurg. 2009;110:583-588.

Funding

One of the authors (Alessandro Olivi, MD) reported that he serves on the speakers’ bureau for MGI, Inc.

Trial design

Retrospective study in a single institution.

Enrollment

• 33 patients who underwent primary resection, implantation of BCNU implant wafers, adjuvant XRT, and concomitant TMZ for treatment of new-onset GBM between 2004 and 2006.
• 78 patients who underwent primary resection, implantation of BCNU implant wafers, adjuvant XRT, without TMZ for treatment of new-onset GBM between 1997 and 2004.
• 45 patients who underwent primary resection, adjuvant XRT, concomitant TMZ, without BCNU implant for treatment of new-onset GBM between 2004 and 2006.

Note: Prior to 2004, patients receiving BCNU implant wafer implantation and adjuvant XRT were not treated with additional adjuvant first-line chemotherapeutic agents. Therefore, all patients receiving BCNU implant wafer implantation prior to 2004 received adjuvant XRT and BCNU implant alone.

Inclusion criteria

Patients undergoing resection of malignant astrocytoma by 3 full-time faculty members at a single institution between 1997 and 2006 were reviewed.

Between 1997 and 2004, patients enrolled if undergoing all of the following:

• Primary resection for treatment of new-onset GBM
• Implantation of BCNU implant wafers
• Adjuvant XRT

Between 2004 and 2006, patients enrolled if undergoing primary resection for treatment of new-onset GBM, adjuvant XRT, and one or both of the following:

• Implantation of BCNU implant wafers
• Concomitant TMZ

Exclusion criteria

• Stereotactic needle biopsy
• Open diagnostic biopsy
• Surgery performed by part-time faculty members

Study Groups

• Primary resection + XRT + BCNU implant +TMZ cohort
• Primary resection + XRT + BCNU implant cohort
• Primary resection +XRT +TMZ cohort

End points

• Primary: median survival, 2-year survival, 6-month morbidity
• Secondary: median survival in patients ≤70 years of age

Statistical analyses

• Kaplan-Meier method for assessment of overall survival
• Log-rank test and multivariate proportional-hazards regression analysis (Cox model) for survival comparisons
• Parametric data expressed as means ± standard deviations and compared using Student t-test.
• Nonparametric data expressed as medians (and interquartile ranges) and compared using Mann-Whitney U test.
• Percentages compared using chi-square test or Fisher exact test

Baseline characteristics

• Mean age was 57 and 56 for XRT + BCNU implant + TMZ and XRT + BCNU implant cohorts, respectively.
• 60% of patients in XRT + BCNU implant + TMZ and XRT + BCNU implant cohorts were male.
• Agents used prior to 2004 differed from those used from 2004 to 2006.
• No significant differences between the 2 cohorts in the following:
• Preoperative age
• Degree of disability
• Extent of resection
• Incidence of subsequent adjuvant chemotherapy
• Incidence of revision resection

End points

XRT + BCNU Implant +TMZ Cohort:

• Primary end points: median survival 20.7 months, 2-year survival 36%
• Primary end point: 6-month morbidity: surgical site infection (3%), perioperative seizures (6%), deep vein thrombosis (3%), pulmonary embolism (9%), cerebral edema (3%). Seven patients (21%) required early termination of TMZ due to myelosuppression.
• Secondary end point: in patients ≤70 years of age median survival 21.3 months, 2-year survival 39%, no increase in perioperative morbidity compared with XRT + BCNU implant
• Progression-free survival

XRT + TMZ Cohort:

• Primary end points: median survival 14.7 months

XRT + BCNU Implant Cohort:

• Primary end points: median survival 12.4 months, 2-year survival 18%
• Secondary end point: in patients ≤70 years of age median survival 12.4 months, 2-year survival 18%

Conclusion

Concomitant TMZ therapy in addition to BCNU implant wafer implantation in patients undergoing primary resection and adjuvant radiotherapy for GBM was associated with a median survival of nearly 21 months, with an incidence of myelosuppression similar to that previously reported for TMZ treatment alone.

Study strengths/ limitations

Strengths:

• Other than the introduction of concomitant TMZ into practice during the 10-year period that the trial covered, patient characteristics (eg, demographics, radiological presentation of tumor) and practice patterns (eg, surgical strategies, follow-up care, adjuvant XRT) were similar between the cohorts.
• Demographic and clinical characteristics of patients in this trial are representative of those of most practices and previously reported cohorts.
• Patient selection did not underlie the favorable survival observed in the study (nearly 10% of patients were >70 years of age, 25% underwent subtotal resection).
• Although survival comparisons (between XRT + BCNU implant + TMZ, XRT + TMZ, and XRT + BCNU implant) were not part of the objective, they were assessed nevertheless (but with several limitations stated below).

Limitations:

• Given the retrospective nature of the trial, it remains unclear whether the examined multimodal therapy (XRT + TMZ + BCNU implant) or other unmeasured factors (eg, differences in MGMT-promoter methylation) contributed to the survival difference.
• Factors (eg, MGMT-promoter methylation positivity, differences in adjuvant chemotherapy used at time of tumor progression) other than addition of TMZ to BCNU implant might have contributed to the observed 8-month increase in survival.
• The 3 cohorts have disproportionate number of patients (78, 45, and 33).
• Although the aim of the study was to characterize survival and morbidity related to patients treated with XRT + BCNU implant + TMZ, there were considerable data collected for (and comparisons to) those given XRT + BCNU implant alone and XRT + TMZ alone. As such, there were 3 cohorts in total, however, they were not clearly defined throughout the study or presented consistently.
• The observed survival difference (6 months greater for XRT + BCNU implant + TMZ vs XRT + TMZ) may be attributed to the large discrepancy in extent of resection between the cohorts and did not reach statistical significance because the study was underpowered.
• Baseline characteristics for XRT + TMZ cohort not available.

Implications

• Combining immediately delivered local chemotherapy and more delayed systemic chemotherapy may result in complimentary effects, because TMZ is most effective in more vascular tumor regions whereas BCNU implant acts independently of vasculature and may provide a therapeutic bridge during the 3-week period between surgery and radiotherapy.
• Given the potential survival advantage of XRT + BCNU implant + TMZ, patients receiving BCNU implant wafers should be considered for adjuvant TMZ therapy.
• Prospective trials are needed to confirm the potential survival benefit observed with this multimodal therapy.