J Neurooncol DOI 10.1007/s11060-015-1834-x
CLINICAL STUDY
Metronomic chemotherapy with daily low-dose temozolomide and celecoxib in elderly patients with newly diagnosed glioblastoma multiforme: a retrospective analysis Grit Welzel1 • Julian Gehweiler1 • Stefanie Brehmer2 • Jens-Uwe Appelt3 Andreas von Deimling4 • Marcel Seiz-Rosenhagen2 • Peter Schmiedek2 • Frederik Wenz1 • Frank A. Giordano1
•
Received: 5 November 2014 / Accepted: 30 May 2015 Ó Springer Science+Business Media New York 2015
Abstract Chemotherapy is often omitted in elderly patients with glioblastoma multiforme due to a fear of side effects. We applied metronomic chemotherapy with low-dose temozolomide and celecoxib (LD-TEM/CEL) during and after external beam radiotherapy (EBRT) and here report on how this regimen compares to standard temozolomide radiochemotherapy (SD-TEM) in elderly patients. We retrospectively analyzed records of 146 patients aged 65 years and older that underwent EBRT. Factors of interest were age, performance status, comorbidities, MGMT status, therapy (resection/biopsy, radiotherapy/dose, chemotherapy/regimen/dose), progression-free (PFS) and overall survival (OS) status. Irrespective of the regimen, addition of chemotherapy more than doubled
Grit Welzel and Julian Gehweiler have contributed equally to this work.
Electronic supplementary material The online version of this article (doi:10.1007/s11060-015-1834-x) contains supplementary material, which is available to authorized users.
median survival rates (EBRT only: 4.2 months; EBRT ? LDTEM/CEL: 8.5 months; EBRT ? SD-TEM: 10.8 months; p B 0.008). Although patients receiving metronomic LDTEM/CEL were significantly older (62 % were C75 years vs. 22 %; p \ 0.001), had significantly lower performance scores (50 % had a KPS \70 vs. 28 %; p = 0.049) and were significantly more comorbid (73 % had C4 comorbidities vs. 37 %; p = 0.002) than patients of the SD-TEM group, there were no significant differences in PFS and OS. Independent of other factors, omission of chemotherapy significantly impairs progression-free and overall survival. With all the limitations of a retrospective analysis, our data suggest that metronomic chemotherapy with LD-TEM/CEL may be equieffective and eventually better tolerated than SD-TEM. It may be offered to elderly patients that are not eligible for standard chemotherapy. Keywords Glioblastoma Elderly patients Chemotherapy
& Frank A. Giordano [email protected]
Introduction 1
Department of Radiation Oncology, Universita¨tsmedizin Mannheim, Medical Faculty Mannheim, Ruprecht-KarlsUniversity Heidelberg, Mannheim, Germany
2
Department of Neurosurgery, Universita¨tsmedizin Mannheim, Medical Faculty Mannheim, Ruprecht-KarlsUniversity Heidelberg, Mannheim, Germany
3
CLC Bio, Aarhus, Denmark
4
Department of Neuropathology, Ruprecht-Karls-University Heidelberg, and Clinical Cooperation Unit Neuropathology, and DKTK, DKFZ, Heidelberg, Germany
5
Department of Radiation Oncology, University Medical Center Mannheim, University of Heidelberg, TheodorKutzer-Ufer 1-3, 68167 Mannheim, Germany
Glioblastoma multiforme (GBM), the most aggressive and most lethal malignant brain tumor, is predominantly affecting elderly patients with the majority of patients aged 60 years and older [1]. After the pivotal EORTC-NCIC phase III trial, concomitant radiotherapy (RT) and temozolomide therapy followed by adjuvant temozolomide chemotherapy became the standard treatment [2]. Although the median age in the mentioned study was 56 years and patients older than 70 years were excluded from the trial (as in most clinical trials on GBM), the 5-year analysis of the data also suggested a mild therapeutic benefit in
123
J Neurooncol
patients aged 60–70 years, although this was an end point for which the study was not powered [3]. Despite compelling data from retrospective and prospective studies that concomitant and adjuvant temozolomide chemotherapy with normo- [4–6] and hypofractionated [7, 8] radiotherapy improves survival rates also in elderly patients, it is frequently omitted due to an increased incidence of toxicity in elderly patients. In the EORTCNCIC study on relatively young(er) patients, grade 3 or 4 haematotoxicity exclusively occurred in the temozolomide arm [2]. Compared to the incidence of 7 % under concomitant chemotherapy in the EORTC/NCIC study, the reported rates in elderly patients vary from 12 % in a cohort from MSKCC [5] over 37 % in a group of Japanese GBM patients [9] up to 42 % in a Dutch cohort [10]. Thus, novel regimens that avoid providing survival benefits at the expense of increased rates of toxicities are needed specifically for elderly patients with GBM. As an alternative to the standard ‘‘Stupp protocol’’, a locally established standard for elderly comorbid and frail patients is a daily low dose of temozolomide combined with the anti-inflammatory drug celecoxib during and after radiotherapy (LD-TEM/CEL) [11]. The metronomic regimen is based on pre-clinical data [11–14] and it was already seen safe and effective in a younger patient population with newly diagnosed GBM [15] and in patients with recurrent GBM [16–18]. Here, we present results of a retrospective analysis of 146 patients aged 65 years and older that were newly diagnosed with GBM and that underwent radiotherapy with and without concomitant and/or adjuvant chemotherapy in a standard (SD-TEM) or a metronomic (LD-TEM/CEL) manner. Our aim was to assess whether (i) chemotherapy influences outcome at all in elderly patients and (ii) how both regimens compare in terms of outcome and toxicity.
Patients and methods Patient population Following approval by the institutional review board (2013-832R-MA), we analyzed electronic health records (MOSAIQÒ, Elekta AB, Stockholm) of patients aged 65 years and older that were newly diagnosed with GBM between February 2006 and January 2013. For all patients, the diagnosis was established histologically after maximum safe resection or stereotactic biopsy. Patients with a history of low-grade gliomas were excluded from the study. For analysis purposes, we only included patients with followup (FU) visits in our department to ensure sufficient documentation (all follow-up visits included MRI scans and were scheduled quarterly). Factors gathered from the
123
records were age at the time of diagnosis, sex, postoperative Karnofsky-Performance-Status (KPS), comorbidities (cardiovascular, gastrointestinal, neurological, endocrinological or other comorbidities), surgery, O6-methylguanine-DNA methyltransferase (MGMT) promoter methylation status, radiotherapy (including dose and fractionation scheme), concomitant and adjuvant chemotherapy (regimen and duration), results from MRI scans and survival status. Radiotherapy All patients included in the analysis underwent external beam radiotherapy (EBRT) on average 3 weeks after resection or biopsy. The radiation doses and the fractionation schemes were either 60 Gy in 30 fractions of 2 Gy, 59.4 Gy in 33 fractions of 1.8 or 45 Gy in 15 fractions of 3 Gy. For all applied doses, we calculated the correlating equivalent (or isoeffective) dose of a 2 Gy fractionation scheme (EQD2) considering an a/b ratio of 3 Gy. All EQD2 \ 50 Gy (e.g. due to termination of radiotherapy before the envisaged total dose) were considered as therapeutically insufficient radiation doses. Three-dimensional conformal or IMRT-based treatment planning was performed on contrast enhanced MRI or CT imaging after patient positioning with thermoplastic mask immobilization. Radiotherapy was delivered to the (post-operative or post-biopsy) peritumoral edema plus a 2–3 cm margin. Salvage re-irradiation consisted of 36 Gy delivered in 3 Gy per fraction, but also included stereotactic radiosurgery (as boost or as single treatment, e.g. for satellite lesions). Concomitant and adjuvant chemotherapy The local standard regimen (SD-TEM) consists of 50 mg/m2/day temozolomide given only during radiotherapy (5 days a week) [19]. The local standard ‘‘low-dose’’ chemotherapy regimen (LD-TEM/CEL) during radiotherapy consists of 2 9 20 mg/day temozolomide and 200 mg/day celecoxib 7 days a week. If no contraindications (e.g., leuko- or thrombopenia) were present after radio(chemo)therapy, the patients either received cycling (28 days) adjuvant chemotherapy with 150–200 mg/m2/day temozolomide (SD-TEM) 4–5 weeks after the last radiotherapy fraction for 5 days, or the LD-TEM/CEL regimen (40 mg/day temozolomide and 200 mg/day celecoxib 7 days a week) was continued without interruption after radiotherapy. Assignment to either therapy was done by physicians (or tumor boards) on an individual basis taking into account age, clinical presentation, comorbidities and patient preferences. Salvage chemotherapies included switching from SDTEM to LD-TEM/CEL or vice versa or the application of
J Neurooncol
bevacizumab. Patients receiving chemotherapy underwent weekly complete blood counts (CBC) and medical examinations at the quarterly FU visits to assess toxicity. For this study, all blood workups generated within 6 months after initial diagnosis were analyzed. Toxicities were graded as per NCI Common Terminology Criteria for Adverse Events (version 4.0). Statistical analyses Progression-free survival (PFS) was measured from the date of diagnosis until the date of first progression in follow-up MRI scans. Overall survival (OS) was measured from the date of diagnosis until the date of death from any cause or date of last follow-up and was estimated by the Kaplan–Meier method. The log-rank and the generalized Wilcoxon tests were used for comparison of survival curves. The Cox proportional analysis was used to evaluate the significance of (risk) factors. All predictors with p values lower than 0.10 in univariate analysis were used in multivariate analysis (except variables with more than 25 % of missing data). The association between the treatment groups (SD-TEM, LD-TEM/CEL or EBRT alone) and other patient and disease characteristics were examined by the Chi square test and Fisher’s exact test, or the Mann– Whitney test, if required. For all analyses, a two-sided p value of less than 0.05 was considered statistically significant. SPSS software, version 20 (IBM Corp., Armonk, NY) was used for all statistical analyses.
Results Clinical features and treatment characteristics of both cohorts A total of 146 patients aged 65 years and older were included in the analysis. The median follow-up was 6.6 months (range: 0.9–43.4 months) for all patients. For patients alive at the time of analysis (including censored cases), the median FU was 7.5 months (1.4–43.4 months). An overview on the clinical characteristics is given in Table 1. Although higher age (C75) was associated with decreased OS in univariate analyses, multivariate analysis revealed no statistically significant influence, suggesting that age is no independent prognostic factor for survival in patients of 65 years and older. Sixty-seven patients (46 %) received EBRT and standard dose TMZ (SD-TEM), 26 patients (18 %) received EBRT and low dose TMZ, and 53 patients (36 %) received EBRT alone. The median age of all patients was 74 years (range: 65–94) with 71 years (range: 65–81) in the EBRT ? SD-TEM group, 75 years (range: 67–82) in the
EBRT ? LD-TEM/CEL group, and 76 years (range: 66–94) in the EBRT alone group. There were significantly more patients younger than 75 years, less patients with worse clinical performance, less patients with partial or complete resection and less patients with C4 comorbidities among the EBRT ? SDTEM group compared to the EBRT LD-TEM/CEL group. Of note, there were nearly twice as much endocrine disorders (e.g., diabetes, lipid disorders, etc.) in the LD(54 %) than in the SD-TEM group (28 %; p = 0.021). Except for resection status (complete/partial resection rate was 96 % in the LD-TEM/CEL group and 74 % in the EBRT only group), there was no significant difference in the composition of the EBRT alone and EBRT ? LDTEM/CEL cohorts. The median interval between surgery and EBRT was similar for all three cohorts (EBRT ? standard dose TMZ: 23 days, range: 13–251; EBRT ? LD-TEM: 22.5 days, range: 15–42; EBRT alone: 22 days, range: 11–62 days, data not shown). During EBRT, 15 patients (58 %) of the LD-cohort received LD-TEM/CEL concomitant to EBRT, 54 patients (81 %) of the SD-TEM cohort received concomitant temozolomide as per EORTC/NCIC protocol (supplemental Table 1). The median number of adjuvant cycles given in the SD-TEM cohort was 2 (range: 0–20), whereas nine patients (13.4 %) received six or more cycles and 19 patients (28 %) did not receive any adjuvant chemotherapy after EBRT ? SD-TEM. The median time on adjuvant metronomic chemotherapy given to the LD-TEM/CEL cohort was 3 months (range: 0–15), 5 patients (19.2 %) were on this regimen for 6 months and longer and two patients (8 %) did not receive any adjuvant chemotherapy after EBRT ? LD-TEM/CEL. Although statistically insignificant, adjuvant therapy was more often discontinued after EBRT ? SD-TEM (two out of 15/13 % in the LD-Tem/CEL group vs. 14 out of 54/26 % in the SD-Tem group; p = 0.492). Thirteen percent of LD-TEM/CEL patients switched to SD-TEM and 8 % switched from SD-TEM to LD-TEM/ CEL (supplemental Table 1). Salvage re-radiotherapy was more often applied in LD-TEM/CEL cohort (14 patients, 54 %) than in the SD-TEM cohort (three patients, 4 %). Addition of chemotherapy more than doubled OS in elderly Together with resection status, chemotherapy turned out to be a statistically significant independent prognostic factor for OS in this elderly cohort (Table 2). OS dramatically improved with chemotherapy (10.8 months for EBRT ? SD-TEM and 8.5 months for LD-TEM/CEL) as compared to EBRT alone (4.2 months) both in uni- and multivariate
123
J Neurooncol Table 1 Demographic and patient characteristics according to chemotherapy status Total
EBRT ? SDTEM
EBRT ? LDTEM/ CEL
n
%
n
%
n
%
Male
68
47
37
55
12
46
Female
78
53
30
45
14
54
Gender
p EBRT ? SDTEM versusEBRT ? LD-TEM/CEL
EBRT alone
%
0.035
0.378
–
19
36
–
–
–
34
64
–
–
CLINICAL STUDY
Metronomic chemotherapy with daily low-dose temozolomide and celecoxib in elderly patients with newly diagnosed glioblastoma multiforme: a retrospective analysis Grit Welzel1 • Julian Gehweiler1 • Stefanie Brehmer2 • Jens-Uwe Appelt3 Andreas von Deimling4 • Marcel Seiz-Rosenhagen2 • Peter Schmiedek2 • Frederik Wenz1 • Frank A. Giordano1
•
Received: 5 November 2014 / Accepted: 30 May 2015 Ó Springer Science+Business Media New York 2015
Abstract Chemotherapy is often omitted in elderly patients with glioblastoma multiforme due to a fear of side effects. We applied metronomic chemotherapy with low-dose temozolomide and celecoxib (LD-TEM/CEL) during and after external beam radiotherapy (EBRT) and here report on how this regimen compares to standard temozolomide radiochemotherapy (SD-TEM) in elderly patients. We retrospectively analyzed records of 146 patients aged 65 years and older that underwent EBRT. Factors of interest were age, performance status, comorbidities, MGMT status, therapy (resection/biopsy, radiotherapy/dose, chemotherapy/regimen/dose), progression-free (PFS) and overall survival (OS) status. Irrespective of the regimen, addition of chemotherapy more than doubled
Grit Welzel and Julian Gehweiler have contributed equally to this work.
Electronic supplementary material The online version of this article (doi:10.1007/s11060-015-1834-x) contains supplementary material, which is available to authorized users.
median survival rates (EBRT only: 4.2 months; EBRT ? LDTEM/CEL: 8.5 months; EBRT ? SD-TEM: 10.8 months; p B 0.008). Although patients receiving metronomic LDTEM/CEL were significantly older (62 % were C75 years vs. 22 %; p \ 0.001), had significantly lower performance scores (50 % had a KPS \70 vs. 28 %; p = 0.049) and were significantly more comorbid (73 % had C4 comorbidities vs. 37 %; p = 0.002) than patients of the SD-TEM group, there were no significant differences in PFS and OS. Independent of other factors, omission of chemotherapy significantly impairs progression-free and overall survival. With all the limitations of a retrospective analysis, our data suggest that metronomic chemotherapy with LD-TEM/CEL may be equieffective and eventually better tolerated than SD-TEM. It may be offered to elderly patients that are not eligible for standard chemotherapy. Keywords Glioblastoma Elderly patients Chemotherapy
& Frank A. Giordano [email protected]
Introduction 1
Department of Radiation Oncology, Universita¨tsmedizin Mannheim, Medical Faculty Mannheim, Ruprecht-KarlsUniversity Heidelberg, Mannheim, Germany
2
Department of Neurosurgery, Universita¨tsmedizin Mannheim, Medical Faculty Mannheim, Ruprecht-KarlsUniversity Heidelberg, Mannheim, Germany
3
CLC Bio, Aarhus, Denmark
4
Department of Neuropathology, Ruprecht-Karls-University Heidelberg, and Clinical Cooperation Unit Neuropathology, and DKTK, DKFZ, Heidelberg, Germany
5
Department of Radiation Oncology, University Medical Center Mannheim, University of Heidelberg, TheodorKutzer-Ufer 1-3, 68167 Mannheim, Germany
Glioblastoma multiforme (GBM), the most aggressive and most lethal malignant brain tumor, is predominantly affecting elderly patients with the majority of patients aged 60 years and older [1]. After the pivotal EORTC-NCIC phase III trial, concomitant radiotherapy (RT) and temozolomide therapy followed by adjuvant temozolomide chemotherapy became the standard treatment [2]. Although the median age in the mentioned study was 56 years and patients older than 70 years were excluded from the trial (as in most clinical trials on GBM), the 5-year analysis of the data also suggested a mild therapeutic benefit in
123
J Neurooncol
patients aged 60–70 years, although this was an end point for which the study was not powered [3]. Despite compelling data from retrospective and prospective studies that concomitant and adjuvant temozolomide chemotherapy with normo- [4–6] and hypofractionated [7, 8] radiotherapy improves survival rates also in elderly patients, it is frequently omitted due to an increased incidence of toxicity in elderly patients. In the EORTCNCIC study on relatively young(er) patients, grade 3 or 4 haematotoxicity exclusively occurred in the temozolomide arm [2]. Compared to the incidence of 7 % under concomitant chemotherapy in the EORTC/NCIC study, the reported rates in elderly patients vary from 12 % in a cohort from MSKCC [5] over 37 % in a group of Japanese GBM patients [9] up to 42 % in a Dutch cohort [10]. Thus, novel regimens that avoid providing survival benefits at the expense of increased rates of toxicities are needed specifically for elderly patients with GBM. As an alternative to the standard ‘‘Stupp protocol’’, a locally established standard for elderly comorbid and frail patients is a daily low dose of temozolomide combined with the anti-inflammatory drug celecoxib during and after radiotherapy (LD-TEM/CEL) [11]. The metronomic regimen is based on pre-clinical data [11–14] and it was already seen safe and effective in a younger patient population with newly diagnosed GBM [15] and in patients with recurrent GBM [16–18]. Here, we present results of a retrospective analysis of 146 patients aged 65 years and older that were newly diagnosed with GBM and that underwent radiotherapy with and without concomitant and/or adjuvant chemotherapy in a standard (SD-TEM) or a metronomic (LD-TEM/CEL) manner. Our aim was to assess whether (i) chemotherapy influences outcome at all in elderly patients and (ii) how both regimens compare in terms of outcome and toxicity.
Patients and methods Patient population Following approval by the institutional review board (2013-832R-MA), we analyzed electronic health records (MOSAIQÒ, Elekta AB, Stockholm) of patients aged 65 years and older that were newly diagnosed with GBM between February 2006 and January 2013. For all patients, the diagnosis was established histologically after maximum safe resection or stereotactic biopsy. Patients with a history of low-grade gliomas were excluded from the study. For analysis purposes, we only included patients with followup (FU) visits in our department to ensure sufficient documentation (all follow-up visits included MRI scans and were scheduled quarterly). Factors gathered from the
123
records were age at the time of diagnosis, sex, postoperative Karnofsky-Performance-Status (KPS), comorbidities (cardiovascular, gastrointestinal, neurological, endocrinological or other comorbidities), surgery, O6-methylguanine-DNA methyltransferase (MGMT) promoter methylation status, radiotherapy (including dose and fractionation scheme), concomitant and adjuvant chemotherapy (regimen and duration), results from MRI scans and survival status. Radiotherapy All patients included in the analysis underwent external beam radiotherapy (EBRT) on average 3 weeks after resection or biopsy. The radiation doses and the fractionation schemes were either 60 Gy in 30 fractions of 2 Gy, 59.4 Gy in 33 fractions of 1.8 or 45 Gy in 15 fractions of 3 Gy. For all applied doses, we calculated the correlating equivalent (or isoeffective) dose of a 2 Gy fractionation scheme (EQD2) considering an a/b ratio of 3 Gy. All EQD2 \ 50 Gy (e.g. due to termination of radiotherapy before the envisaged total dose) were considered as therapeutically insufficient radiation doses. Three-dimensional conformal or IMRT-based treatment planning was performed on contrast enhanced MRI or CT imaging after patient positioning with thermoplastic mask immobilization. Radiotherapy was delivered to the (post-operative or post-biopsy) peritumoral edema plus a 2–3 cm margin. Salvage re-irradiation consisted of 36 Gy delivered in 3 Gy per fraction, but also included stereotactic radiosurgery (as boost or as single treatment, e.g. for satellite lesions). Concomitant and adjuvant chemotherapy The local standard regimen (SD-TEM) consists of 50 mg/m2/day temozolomide given only during radiotherapy (5 days a week) [19]. The local standard ‘‘low-dose’’ chemotherapy regimen (LD-TEM/CEL) during radiotherapy consists of 2 9 20 mg/day temozolomide and 200 mg/day celecoxib 7 days a week. If no contraindications (e.g., leuko- or thrombopenia) were present after radio(chemo)therapy, the patients either received cycling (28 days) adjuvant chemotherapy with 150–200 mg/m2/day temozolomide (SD-TEM) 4–5 weeks after the last radiotherapy fraction for 5 days, or the LD-TEM/CEL regimen (40 mg/day temozolomide and 200 mg/day celecoxib 7 days a week) was continued without interruption after radiotherapy. Assignment to either therapy was done by physicians (or tumor boards) on an individual basis taking into account age, clinical presentation, comorbidities and patient preferences. Salvage chemotherapies included switching from SDTEM to LD-TEM/CEL or vice versa or the application of
J Neurooncol
bevacizumab. Patients receiving chemotherapy underwent weekly complete blood counts (CBC) and medical examinations at the quarterly FU visits to assess toxicity. For this study, all blood workups generated within 6 months after initial diagnosis were analyzed. Toxicities were graded as per NCI Common Terminology Criteria for Adverse Events (version 4.0). Statistical analyses Progression-free survival (PFS) was measured from the date of diagnosis until the date of first progression in follow-up MRI scans. Overall survival (OS) was measured from the date of diagnosis until the date of death from any cause or date of last follow-up and was estimated by the Kaplan–Meier method. The log-rank and the generalized Wilcoxon tests were used for comparison of survival curves. The Cox proportional analysis was used to evaluate the significance of (risk) factors. All predictors with p values lower than 0.10 in univariate analysis were used in multivariate analysis (except variables with more than 25 % of missing data). The association between the treatment groups (SD-TEM, LD-TEM/CEL or EBRT alone) and other patient and disease characteristics were examined by the Chi square test and Fisher’s exact test, or the Mann– Whitney test, if required. For all analyses, a two-sided p value of less than 0.05 was considered statistically significant. SPSS software, version 20 (IBM Corp., Armonk, NY) was used for all statistical analyses.
Results Clinical features and treatment characteristics of both cohorts A total of 146 patients aged 65 years and older were included in the analysis. The median follow-up was 6.6 months (range: 0.9–43.4 months) for all patients. For patients alive at the time of analysis (including censored cases), the median FU was 7.5 months (1.4–43.4 months). An overview on the clinical characteristics is given in Table 1. Although higher age (C75) was associated with decreased OS in univariate analyses, multivariate analysis revealed no statistically significant influence, suggesting that age is no independent prognostic factor for survival in patients of 65 years and older. Sixty-seven patients (46 %) received EBRT and standard dose TMZ (SD-TEM), 26 patients (18 %) received EBRT and low dose TMZ, and 53 patients (36 %) received EBRT alone. The median age of all patients was 74 years (range: 65–94) with 71 years (range: 65–81) in the EBRT ? SD-TEM group, 75 years (range: 67–82) in the
EBRT ? LD-TEM/CEL group, and 76 years (range: 66–94) in the EBRT alone group. There were significantly more patients younger than 75 years, less patients with worse clinical performance, less patients with partial or complete resection and less patients with C4 comorbidities among the EBRT ? SDTEM group compared to the EBRT LD-TEM/CEL group. Of note, there were nearly twice as much endocrine disorders (e.g., diabetes, lipid disorders, etc.) in the LD(54 %) than in the SD-TEM group (28 %; p = 0.021). Except for resection status (complete/partial resection rate was 96 % in the LD-TEM/CEL group and 74 % in the EBRT only group), there was no significant difference in the composition of the EBRT alone and EBRT ? LDTEM/CEL cohorts. The median interval between surgery and EBRT was similar for all three cohorts (EBRT ? standard dose TMZ: 23 days, range: 13–251; EBRT ? LD-TEM: 22.5 days, range: 15–42; EBRT alone: 22 days, range: 11–62 days, data not shown). During EBRT, 15 patients (58 %) of the LD-cohort received LD-TEM/CEL concomitant to EBRT, 54 patients (81 %) of the SD-TEM cohort received concomitant temozolomide as per EORTC/NCIC protocol (supplemental Table 1). The median number of adjuvant cycles given in the SD-TEM cohort was 2 (range: 0–20), whereas nine patients (13.4 %) received six or more cycles and 19 patients (28 %) did not receive any adjuvant chemotherapy after EBRT ? SD-TEM. The median time on adjuvant metronomic chemotherapy given to the LD-TEM/CEL cohort was 3 months (range: 0–15), 5 patients (19.2 %) were on this regimen for 6 months and longer and two patients (8 %) did not receive any adjuvant chemotherapy after EBRT ? LD-TEM/CEL. Although statistically insignificant, adjuvant therapy was more often discontinued after EBRT ? SD-TEM (two out of 15/13 % in the LD-Tem/CEL group vs. 14 out of 54/26 % in the SD-Tem group; p = 0.492). Thirteen percent of LD-TEM/CEL patients switched to SD-TEM and 8 % switched from SD-TEM to LD-TEM/ CEL (supplemental Table 1). Salvage re-radiotherapy was more often applied in LD-TEM/CEL cohort (14 patients, 54 %) than in the SD-TEM cohort (three patients, 4 %). Addition of chemotherapy more than doubled OS in elderly Together with resection status, chemotherapy turned out to be a statistically significant independent prognostic factor for OS in this elderly cohort (Table 2). OS dramatically improved with chemotherapy (10.8 months for EBRT ? SD-TEM and 8.5 months for LD-TEM/CEL) as compared to EBRT alone (4.2 months) both in uni- and multivariate
123
J Neurooncol Table 1 Demographic and patient characteristics according to chemotherapy status Total
EBRT ? SDTEM
EBRT ? LDTEM/ CEL
n
%
n
%
n
%
Male
68
47
37
55
12
46
Female
78
53
30
45
14
54
Gender
p EBRT ? SDTEM versusEBRT ? LD-TEM/CEL
EBRT alone
%
0.035
0.378
–
19
36
–
–
–
34
64
–
–
