Author reply

Abstract

We read with interest Dr. Alliot's letter, which raised the issues of a potential patient selection bias, the response evaluation, and toxicity in our study of treatment with temozolomide without radiotherapy in elderly patients with glioblastoma multiforme (GBM).1 First, we agree that patient selection bias is always possible in such Phase II studies. However, it should be noted that, as previously described, although approximately half of the patients referred in our center (32 of 63 patients) were included in the study, the population of patients who were not included was rather heterogeneous and included not only 27% of patients with potentially poor prognostic characteristics (such as no prior biopsy or a Karnofsky performance status (KPS) < 60) (17 of 63 patients), but also 22% of patients with a potentially good prognostic profile (such as anaplastic histology or status that permitted radiotherapy with or without nitrosourea) (14 of 63 patients). Because 63 of 65 patients with malignant glioma were referred directly to our center because of our regional neurooncology organization, it is unlikely that our status as a referral center had an impact on patient selection. The median tumor dimension of 15.75 cm2 should be considered rather than a minimal tumor dimension, and it should be understood that patient inclusion was not restricted to those individuals with small tumors. Taken together, a median patient age of 75 years, a KPS of 60 (44% of patients), a median tumor size, and a percentage of macroscopic total resection of 3% do not reflect, in our opinion, a potentially favorable patient selection bias. Response evaluation is always critical in glioma patients because even well defined Macdonald response criteria have known potential limits. However, the reported response rate of 31% was confirmed by an external and independent expert (K.H.X.) as mentioned in our article. Response was evaluated in 29 patients, including 6 patients in whom an evaluable tumor remained after undergoing partial surgery whereas only 1 patient underwent macroscopic total resection with no evaluable residual tumor noted at the time of entry into the study. Moreover, the benefit of chemotherapy does not appear to be influenced by surgery as suggested by a meta-analysis performed by the Glioma Meta-analysis Trialists (GMT) Group,2 whereas treatment at the time of recurrence appears to have no impact on response or progression-free survival. We agree that steroid decrease per se is not a significant criteria for therapeutic benefit, but we consider that improvements in the KPS and/or MMS, combined with a decrease in steroid use, may be valuable in a palliative setting such as glioblastoma multiforme occurring in the elderly. Toxicity is in fact critical in these palliative situations and approximately 9% rate of Grade 3/4 nausea and emesis should be decreased through an improvement in supportive medication, but overall toxicity data and temozolomide compliance need to be considered behind other forms of treatment such as radiotherapy or other chemotherapies. We agree that in specific centers such as our hospital, the management of toxicity may be optimal, which raises two comments. First, the adequate control of oral chemotherapy with toxicities, even if limited, should still be performed; and second, the adequate management of symptoms associated with brain tumors, such as seizures or brain edema, also is of critical importance for a patient's quality of life, which emphasizes the importance of coordinating all medical support for patients with brain tumors. Finally, the purpose of our study1 was not to consider surgery and temozolomide as alternative treatments and we agree that the respective impact of surgery and temozolomide therapy cannot be addressed by this study. To our knowledge to date, surgical procedures mainly were discussed based on anatomic considerations and the patient's performance status, without any influence of the potential impact of other oncologic treatments noted. Independent of the surgical procedure performed, the first question raised with regard to elderly patients with glioblastoma multiforme concerns the role of oncologic complementary treatment. In that respect, the study conduct by the French group ANOCEF, which randomizes patients to receive either radiotherapy or best supportive care, and in which we are actively involved, may help to provide an answer. Based on our results, which are in accordance with the study published by Glantz et al.,2 the next step may be to compare the best treatment arm from the ANOCEF study with the use of temozolomide as exclusive treatment for elderly patients with glioblastoma multiforme. The results of our study1 may contribute further to the design of future appropriate comparative studies to determine the best standard of care for elderly patients with glioblastoma multiforme with respect to efficacy, treatment tolerance, convenience, and quality of life. Olivier-L Chinot M.D.*, * Unité de Neuro-Oncologie, Service de Neurochirurgie CHU Timone; Laboratoire de Cancérologie Expérimentale, Faculté de Médecine Secteur Nord Marseille, France.