|
Correction
to
Esteller et al.,
|
| |||||||||||||||||||||||||||||||||||||
Second, although the difference in age between the group of patients with unmethylated tumors and the group with methylated tumors was not statistically different, there was a trend toward an older age in the group with unmethylated tumors. We are not given the distribution of ages above or below 50 years. Since the prognosis worsens with each additional decade of age, such information would be useful in assessing the balance in age between the two groups.2,3
Third, the median period of survival was approximately 20 months for the patients with unmethylated tumors and approximately 30 months for those with methylated tumors, as compared with an average of about 12 months in other series.2,3 Moreover, since no deaths occurred before 12 months, it is unlikely that the patients chosen for this study were representative of most patients with the disease. Finally, the small numbers of deaths in the two groups of patients make the results of a statistical comparison questionable.
Jan C. Buckner, M.D.
Timothy J. Moynihan, M.D.
Mayo Clinic
Rochester, MN 55905
References
David I. Quinn, M.B., B.S., Ph.D.
University of Southern California
Los Angeles, CA 90033-0804
References
Uwe Schlegel, M.D.
University Hospital Bonn
D-53105 Bonn, Germany
References
First, contrary to the statement by Esteller et al., the most frequent site of DNA base alkylation by monofunctional and bifunctional nitrosoureas and related alkylating agents, such as temozolomide and procarbazine, is not the O6-position of guanine but rather the N7 position of guanine and the N3 position of adenine. Second, MGMT does not repair the DNA interstrand cross-links resulting from the O6-chloroethylguanine adducts of nitrosoureas, as suggested in Figure 1 of the article by Esteller et al. Furthermore, in that figure, the DNA interstrand cross-link produced by carmustine is incorrectly depicted as a diguanyl cross-link. It is actually an N1-deoxyguanosinyl-N3-deoxycytidyl cross-link.1 Its correct structure is shown in Figure 1 here.
|
Francis Ali-Osman, D.Sc.
Kalkunte Srivenugopal, Ph.D.
Raymond Sawaya, M.D.
University of Texas M.D. Anderson Cancer Center
Houston, TX 77030
References
To the Editor: Dr. Quinn correctly points out that improvements in the response to treatment and survival cannot be definitively attributed to carmustine without a comparison with no treatment. The study Quinn cites showed no relation between MGMT enzyme activity and survival among patients not given chemotherapy, but it also failed to show such a relation among patients receiving chemotherapy.1 Previous studies demonstrated a direct relation between MGMT expression in glioma cell lines and a response to the alkylating agent nimustine (ACNU), but not other chemotherapeutic agents.2 Other work we have done suggests that MGMT inactivation predicts prolonged survival in patients with lymphoma who are treated with an alkylating agent (unpublished data) but not in patients with colorectal cancer who do not receive an alkylating agent (unpublished data).
Drs. Buckner and Moynihan raise several questions. Although there was a slight imbalance between the two groups in the number of patients who were over 50 years old, the age distribution did not differ statistically. In a univariate analysis, age was minimally associated with progression-free survival (hazard ratio for the risk of progression, 0.99) and overall survival (hazard ratio for the risk of death, 0.92); the associations were not statistically significant. Most important, the association of MGMT methylation with overall and progression-free survival was independent of age, as indicated in the legend to Figure 3 of our article. Differences in survival between our study and others may be due to differences in treatment regimens, performance status, and tumor grade (with a higher prevalence of grade 3 tumors in our study); however, these differences do not change the conclusions of our study. Our statistical analysis took into account the size of the sample.
We regret that in our article we did not clearly state that we obtained MRI scans for all patients after surgery in order to provide a base line for evaluating the response to treatment, and we thank Dr. Schlegel for allowing us to make this clarification. We thank Ali-Osman et al. for clarifying issues related to the chemistry of alkylating agents. Although other sites of DNA base alkylation may be more frequent, the O6 position appears to be most important for sensitivity to alkylating agents and the adduct most closely related to MGMT expression.3 Although MGMT does not repair cross-links, it prevents their formation by the removal of alkyl groups. Finally, translational and post-translational changes in MGMT that were not determined by examination of promoter-region methylation would be relevant only in the tumors with unmethylated MGMT promoters, which transcribe the gene. Such changes, if they had been present, would not have led to the observed association.
Manel Esteller, M.D., Ph.D.
James G. Herman, M.D.
Johns Hopkins Oncology Center
Baltimore, MD 21231
References
To the Editor: Dr. Quinn is correct when he points out that the value of the findings reported by Esteller et al. would be enhanced by comparison with data from a group of patients with gliomas who were not treated with carmustine (or other alkylating agents). If methylation of the MGMT promoter region in tumor samples from such patients were not correlated with improved overall and disease-free survival, that finding would strongly support the hypothesis of a causal relation between the activity of carmustine and the methylation. Going one step further, a survey of the methylation status of other promoter regions in the glioma samples analyzed by Esteller et al. would indicate whether the putative relation with methylation was specific to the MGMT promoter. Nonetheless, the authors' principal conclusion that "methylation of the MGMT promoter in gliomas is a useful predictor of the responsiveness of the tumors to alkylating agents" stands without such additional studies. One could argue that the term "alkylating agents" is too broad, since the data are only for the nitrosourea carmustine, but the data do identify a "useful predictor" if one assumes (on the basis of the clinical course and timing) that the observed responses were actually due to the treatment with carmustine. Dr. Quinn's critique illustrates the difficulty of establishing pharmacogenomic causality, especially in the clinical setting.
John N. Weinstein, M.D., Ph.D.
National Cancer Institute
Bethesda, MD 20892
| |||||||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
