Controversy remains on the optimal treatment schedule of the alkylating agent temozolomide (TMZ) in malignant glioma and the value of methylguanine methyltransferase (MGMT) determination. This repair enzyme is key in reverting lethal DNA damage induced by alkylating agent chemotherapy, and in the absence of this enzyme, cells are more susceptible to TMZ. In this issue of the Journal of Clinical Oncology, Chinot et al report on a (too) small phase II study investigating a week-on and week-off dosedense TMZ administration regimen as neoadjuvant chemotherapy before definitive radiation in 29 eligible patients with unresectable glioblastoma. The aim of this alternative dose-dense administration schedule is not only to increase the dose-intensity by more than two-fold, but to overcome drug resistance by depleting the tumor cells of MGMT. In peripheral blood mononuclear cells, it has been shown that repeated and continuous exposure to TMZ will exhaust the cells of MGMT, but whether this also applies to tumor cells remains to be demonstrated. The results of this study were disappointing. Median survival was only 6 months (95% CI, 1 to 11), with only 28% of patients alive at 1 year. Even when considering that this is a poor prognosis group of patients, these results appear inferior to standard therapy, including concomitant chemoradiotherapy, where a median survival of 9 months was reported for patients with unresectable tumors. Chinot et al attempted to determine MGMT expression in 25 patients by immunohistochemistry, choosing an arbitrary cutoff based on the median percentage of cells staining positively. They found that 6 of 11 patients with low MGMT expression, but only one of the 14 patients with high MGMT expression responded. The study emphasizes again the importance of MGMT to predict a response to the alkylating agent therapy. The predictive value of MGMT for benefit from alkylating agent chemotherapy, whether this be a nitrosourea or temozolomide, has now been repeatedly shown. Methylation of the MGMT gene promoter has been found to be a predictive marker for benefit from TMZ treatment in patients treated with temozolomide chemotherapy. Analysis of a representative subgroup of patients in a randomized trial showed that the addition of TMZ only benefited patients with a methylated (and thus silenced) MGMT gene promoter, while overall survival and progression-free survival was comparable for TMZ-treated patients with a functional MGMT gene to patients having been treated with initial radiotherapy alone. Based on this data, one may want to conclude that patients without MGMT promoter methylation should not be treated with alkylating agent chemotherapy. In these patients, TMZ has marginal activity at best, and alternative strategies may offer better chances. However, outside clinical trials, alternative strategies are currently not available. Nevertheless, the treatment, although generally well tolerated, carries its risks for toxicity. Also, in countries without reimbursement for chemotherapy, some patients may choose to forego the expense for temozolomide in the absence of MGMT promoter methylation. Furthermore, this analysis was performed retrospectively, and although very provocative and suggestive, requires prospective validation. In an ongoing Intergroup trial by the Radiation Therapy Oncology Group and European Organisation for Research and Treatment of Cancer (RTOG 0525/EORTC 26052-22053), patients with newly diagnosed glioblastoma are stratified by MGMT methylation status before random assignment to a TMZ administration schedule (standard daily dose for 5 days or a dose-dense treatment for 3 of 4 weeks). To be eligible for this trial, patients must have undergone at least partial tumor resection, and tumor material must be available for centralized MGMT promoter methylation determination. Thus, not only efficient trial management is necessary, but also the goodwill and collaboration of the neurosurgeons and pathologists at many sites. Repeatedly, investigators advocate performing MGMT determination using the simpler immunohistochemistry staining, rather than a methylation-specific polymerase chain reaction assay (MSP) of the function of the MGMT gene. Although some associations may be found by immunohistochemical MGMT expression, this method has many shortcomings. Heterogeneous MGMT expression within tumors, and high levels of MGMT protein present in infiltrating lymphocytes, microglia, and blood vessels limit the reproducibility of the method. More important, MGMT is an inducible enzyme that may be upregulated after chemotherapy or radiotherapy, thus the expression of the protein at diagnosis might not reflect the functional protein during therapy. The MSP assay requires small quantities of DNA that can be extracted from paraffin-embedded tissue samples or, better yet, from cryopreserved tissue samples. In glioma, the presence of a methylated MGMT allele can be attributed solely to the neoplastic cells, irrespective of contamination of nontumor tissue. In our opinion, determination JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 25 NUMBER 12 APRIL 2
Read full abstract