[1] 21w Standard (60 Gy) or Short-Course (40 Gy) Irradiation Plus Concomitant and Adjuvant Temozolomide for Elderly Patients With Glioblastoma: A Propensity-Matched Analysis

[1] 245w Standard chemoradiotherapy has been associated with a survival benefit in older patients with glioblastoma, although the toxicity of aggressive treatments is of concern in this patient population. Our study shows that the combination of concomitant and adjuvant temozolomide with an abbreviated course of radiation therapy (40 Gy) is a reasonable option for treating older patients with glioblastoma, enabling survival benefit Purpose: To evaluate 2 specific radiation schedules, each combined with temozolomide (TMZ), assessing their efficacy and safety in patients aged 65 years with newly diagnosed glioblastoma (GBM). Methods and Materials: Patients aged 65 years with Karnofsky performance status (KPS) 60 who received either standard (60 Gy) or short-course (40 Gy) radiation therapy (RT) with concomitant and adjuvant TMZ between June 2004 and October 2013 were retrospectively analyzed. A propensity score analysis was executed for a balanced comparison of treatment outcomes. Results: A total of 127 patients received standard RT-TMZ, whereas 116 patients underwent short-course RT-TMZ. Median overall survival and progression-free survival times were similar: 12 months and 5.6 months for the standard RT-TMZ group and 12.5 months and 6.7 months for the short-course RT-TMZ group, respectively. Radiation schedule was associated with similar survival outcomes in either unadjusted or adjusted analysis. O 6 -methylguanine-DNA methyltransferase promoter methylation was the most favorable prognostic factor (PZ.0001). Standard RT-TMZ therapy was associated with a significant rise in grade 2 and 3 neurologic toxicity (PZ.01), lowering of KPS scores during the study (PZ.01), and higher posttreatment dosing of corticosteroid (PZ.02).

[1] 133w Glioblastoma (GBM) is increasing among the elderly, and approximately one-third of all patients are aged >65 years (1). Radiation therapy (RT) is generally administered in this population, conferring a median survival from diagnosis of approximately 4 to 10 months (2)(3)(4)(5). Chemotherapy with the alkylating agent temozolomide (TMZ) or hypofractionated courses of RT are treatment options associated with similar survival benefits (6)(7)(8)(9)(10)(11)(12). In a randomized study of 100 patients aged 60 years with GBM who received either radical RT (60 Gy) or short-course RT (40 Gy), Roa et al (9) observed a median survival time of 5.1 months and 5.6 months for standard and short-term RT groups, respectively. A median survival time of 6 to 8 months has been reported with doses of 30-37.5 Gy in 6-15 daily fractions in some prospective studies (6)(7)(8).

[2] 159w Radical RT plus TMZ is standard treatment for adult patients with GBM and has been advocated as an effective approach in older patients with a good performance status. Although a few studies have shown promising survival times of 10 to 14 months with this regimen (13)(14)(15)(16), the likely toxicity is of concern, given the limited life expectancy of this patient population. Thus, many physicians may opt for less-aggressive treatment. Using a radiation schedule of 40 Gy in 15 fractions in combination with concomitant and adjuvant TMZ, we previously reported a median overall survival (OS) of 12.4 months and a median progression-free survival (PFS) of 6 months in 70 older patients aged 70 years with newly diagnosed GBM (17). This approach was associated with improved quality of life and performance status (18). Owing to results of the above prospective study, this particular regimen is often used at our institution as an alternative to standard radiochemotherapy for older patients with GBM.

[3] 67w To date there are no studies that have evaluated the clinical outcomes and toxicity of different RT schedules in combination with TMZ for older patients with GBM. With this in mind, we compared data from older patients aged 65 years with GBM who received standard or shortcourse RT in conjunction with TMZ. A propensity score analysis was done to adjust for between-group disparities of clinically relevant covariates.

[1] 76w A total of 243 consecutive patients aged 65 years with newly diagnosed GBM received chemoradiotherapy between June 2004 and October 2013. Patient characteristics are shown in Table 1. Subjects received either standard (nZ127) or short-course (nZ116) RT. There were no statistically significant differences between groups in terms of gender, KPS scores, site of tumor, PTV, and MGMT promoter methylation status. However, patients given standard RT-TMZ were more likely to be younger and to undergo total/subtotal resection.

[2] 86w Median follow-up time for the standard and short-course RT treatment subsets were 24.0 months and 22.5 months, respectively. At the time of analysis (April 2014), 41 patients were still alive (standard RT-TMZ, 27; short-course RT-TMZ, 14). In the standard and short-course RT-TMZ treatment groups, respectively 49% (58 of 117) and 39% (43 of 109) of patients who experienced disease progression were given salvage therapy (chemotherapy only, 51; surgery only, 8; stereotactic RT only, 10; surgery or RT plus chemotherapy, 32). Treatments administered were similar between groups.

[3] 102w Median group OS was comparable: standard RT-TMZ, 12.0 months; short-course RT-TMZ, 12.5 months (PZ.5) (Fig. 1). Likewise, 12-month and 24-month survival probabilities were similar: standard RT-TMZ, 50% (95% confidence interval [CI] 39-62%) and 18% (95% CI 5-31%); short-course RT-TMZ, 54% (95% CI 41-77%) and 12% (95% CI 3-21%). Progression-free survival did not differ significantly by group. Median PFS times were 5.6 months and 6.7 months for standard and short-term RT-TMZ, respectively (PZ.5), and corresponding 12-month PFS rates were 23% (95% CI 8-38%) and 20% (95% CI 6-34%). Complete and partial response rates of 13% and 24%, respectively, were similar as well (PZ.2).

[4] 105w In the multivariate Cox proportional hazards regression model, total/subtotal resection (PZ.03), KPS >70 (PZ.01), and the MGMT promoter methylation (PZ.0001) emerged as significant indices of prolonged OS (Table 2). O 6 -methylguanine-DNA methyltransferase promoter methylation was the only factor associated with a significantly better PFS (PZ.0001). Median PFS of patients with methylated tumors was 9.5 months (1-year PFS, 35%), as opposed to 5.5 months (1-year PFS, 6%) for those with unmethylated tumors (PZ.0001). The impact of KPS, extent of resection, and MGMT methylation status on survival outcomes of each group was similar, with MGMT promoter methylation being the most significant predictor of OS (Fig. 2).

[5] 68w Propensity score matching resulted in 90 matched pairs. The overall c 2 test for balance (PZ.996) and the L1 index (0.84) suggested that the treatment groups were well balanced across all covariates (Table E1; available online at www.redjournal.com). No significant differences in OS and PFS between groups were observed (Table 3). Stratified analysis of prognostic factors, including MGMT methylation status, KPS, and extent of resection yielded similar results.

[6] 149w Ultimately, 8.6% of patients (11 of 127) subjected to standard RT (6-week course) and 1.7% of patients (2 of 116) on short-course RT did not complete treatment. In all instances, RT was discontinued owing to clinical signs of deterioration. Twenty-eight patients undergoing standard RT-TMZ and 11 subjected to short-course RT-TMZ experienced worsening of neurologic status early after concomitant RT and TMZ, although symptoms were reversed in 65% of these patients by steroid administration. Twelve more patients in the standard RT-TMZ group and 3 patients in the short-course RT-TMZ group experienced late (>6 months) neurologic deterioration (grade 2/3 cognitive disability) after RT, with no evidence of local recurrence. Overall, 40% of patients (51 of 127) who received standard RT-TMZ and 14% of patients (16 of 116) receiving short-course RT-TMZ experienced grade 2/3 neurologic Abbreviation: CI Z confidence interval. Other abbreviations as in Table 1. E2; available online at www. redjournal.com).

[7] 163w Temozolomide was discontinued early during RT in 4 patients (standard RT-TMZ, 3; short-course RT-TMZ, 1) owing to grade 3/4 hematologic toxicities. During adjuvant TMZ therapy, a total of 1416 cycles (698 and 718 cycles for standard and short-course RT-TMZ, respectively) were administered, with a median of 5 cycles for each patient. Grade 3/4 thrombocytopenia and lymphocytopenia were encountered in 24 patients (12%) and 51 patients (24%), respectively. Grade 3 neutropenia developed in 14 patients (6.5%), and 10 patients (5%) displayed grade 3 anemia. Myelosuppressive effects were similar between groups (PZ.2) (Table E2; available online at www.redjournal. com). Chemotherapy was stopped in 19 patients (9%) and reduced or delayed in 38 patients (18%). The most common nonhematologic adverse event of moderate-to-severe fatigue was registered by 76 patients (standard RT-TMZ, 49; short-course RT-TMZ, 27; PZ.04). Grade 1/2 nausea occurred in 26 patients. Twelve patients experienced deep vein thrombosis, and 5 suffered pulmonary embolism. Seven patients had severe lung infections (pneumonia) that resolved with medical therapy.

[8] 51w Transient radiologic imaging abnormalities (ie pseudoprogression) were seen in 28 patients (standard RT-TMZ, 21; short-course RT-TMZ, 7; PZ.01). Of these patients, 7 underwent surgery for deteriorating neurologic conditions, whereas all others continued TMZ therapy. Pseudoprogression was documented in 7 patients with unmethylated tumors and in 21 patients with methylated tumors (PZ.004).

[9] 104w Karnofsky performance status scores monitored during treatment and until the point of disease progression are shown in Figure 3. In patients given standard RT-TMZ, KPS scores worsened significantly over time, although with marked variability among patients. The largest declines were observed 1 month (PZ.001) and 3 months (PZ.005) after cessation of RT. In contrast, changes in KPS scores over time were not substantial for the shortcourse RT-TMZ group. Overall, KPS scores of standard and short-course RT-TMZ, respectively (assessed in at least 2 consecutive evaluations) improved in 14% and 23%, worsened in 44% and 23% (PZ.01), and remained stable in 42% and 54% of patients.

[10] 33w Ninety-eight patients required an increase of dexamethasone 4 mg during and/or early after RT. Daily dexamethasone dosage was increased in 61 patients given standard RT-TMZ and in 37 patients given short-course RT-TMZ (PZ.02).

[1] 456w Clinical outcomes of this study, in which either standard or short-course RT was delivered with concomitant and Fig. 2. Kaplan-Meier analysis of overall survival according to O 6 -methylguanine-DNA methyltransferase (MGMT) promoter methylation status in patients receiving standard or short-course radiation therapy plus temozolomide (RT-TMZ). There was a significantly longer survival in MGMT methylated patients (PZ.0001 by the logerank test for both groups), with a risk reduction of 70% (hazard ratio for death, 0.30; 95% confidence interval 0.21-0.42) compared with unmethylated patients. adjuvant TMZ to patients aged 65 years with GBM, were similar in terms of 1-year survival, PFS, and OS. Hence, a chemoradiotherapeutic regimen of short-course versus standard RT plus TMZ does not seem to compromise patient survival in this context. The above findings are strengthened by propensity score analysis, which addresses potential bias when retrospective data of 2 nonrandomized groups are compared. The median OS of 12.5 months observed here with use of short-course RT-TMZ is consistent with that seen in prior published series of standard chemoradiotherapy in elderly patients with GBM (13-16, 24, 25) and compares favorably to survival achieved through standard or hypofractionated RT (2)(3)(4)(5)(6)(7)(8)(9). In a retrospective review of 293 elderly patients (aged 65 years), Behm et al (25) reported significant survival benefit from use of chemoradiation, as opposed to RT alone. Median OS in patients undergoing RT alone or in combination with TMZ was 3.6 and 8.7 months, respectively (P<.0001). Survival advantages of chemoradiation over RT alone were observed in all patients, including advanced age groups >70 years and >75 years (P<.0001). Barker et al (24) conducted a retrospective analysis at Memorial Sloan-Kettering Cancer Center, assessing 291 patients aged 65 years with GBM who received standard RT, with or without concurrent TMZ. In patients with good performance status given RT plus TMZ, median OS increased from 12 to 21 months in patients aged 65-70 years and from 10 to 13 months in patients aged >70 years. A smaller prospective study of 32 elderly patients aged 70 years with histologically confirmed GBM reported a median OS of 10.6 months after standard postoperative chemoradiotherapy at XXXX (13), and similar findings have been published by a few other investigators (15,16). Overall, our study indicates that an abbreviated course of RT plus TMZ is an effective treatment for elderly patients with GBM, resulting in OS that is comparable to a radical course of RT plus TMZ. The recently closed intergroup European Organization for Research and Treatment of Cancer 26,062-22,061-National Cancer Institute of Canada Clinical Trials Group CE6 randomized trial comparing hypofractionated RT (40 Gy/15 fractions) with or without concomitant and adjuvant TMZ for patients aged >65 years with newly diagnosed GBM may help to better clarify the potential advantages of radiochemotherapy over RT alone.

[2] 122w The primary drawback of aggressive irradiation (ie high radiation doses to large brain volumes) for GBM in older adults is the potential for neurocognitive deterioration and declining quality of life. Moreover, the ramifications of long-term treatment with alkylating agents or chemoradiation in this population are unclear. Sijben et al (14) registered serious treatment-related toxic reactions in 42% of 39 patients aged 70 years with GBM, including asthenia, electrolyte disturbances, infections, and myelosuppression. Brandes et al (15) also reported neurologic and neurocognitive deterioration in 56% of 58 patients aged 65 years after standard chemoradiation, and similar results have been recorded in a previous study (26). Moreover, older patients treated with a radical course of RT alone may suffer significant late neurologic toxicity (2,9).

[3] 162w In this study, 40% of patients given a standard 6-week course of RT and 14% of patients receiving short-course RT experienced grade 2 or 3 neurologic adverse events during or after irradiation. Patients who received shorter RT and TMZ had better KPS scores over time and required lesser increments in posttreatment daily dexamethasone dosages than those treated with standard chemoradiotherapy. These findings are noteworthy, primarily by offering support for use of less-intensive RT schedules in adults of older age, especially those with lower performance status. Although neurocognitive and quality of life assessments were not formally integrated into this study, temporal fluctuations in KPS, levels of corticosteroid use, and displays of neurologic toxicity all reliably reflect clinical status and functional capacity, attesting to a comparatively better safety profile for short-course (vs standard) RT. Furthermore, addition of TMZ to short-course RT proved to be well tolerated and relatively safe, with no more toxicity than others have documented in older patients receiving TMZ alone (10)(11)(12).

[4] 263w Both unadjusted and adjusted Cox proportional hazards regression analysis showed that MGMT promoter methylation status was the most important prognostic index of extended survival in both groups (PZ.0001). The role of MGMT as a predictive biomarker of improved response to TMZ in older patients has been recently addressed in 2 large, randomized studies comparing differing regimens of RT and TMZ as monotherapy (11,12). In a northern European (Nordic) study of patients aged >60 years with newly diagnosed GBM randomly assigned to standard TMZ for up to 6 cycles plus hypofractionated RT or radical RT, MGMT promoter methylation status was a strong predictor of OS; however, this was true only for patients who received TMZ (11). Median survival times were 9.7 and 6.8 months, respectively, for patients with methylated and unmethylated tumors who received TMZ (PZ.02); and for patients subjected to radical RT, respective times were 8.2 and 7.0 months. In the German Neuro-oncology Working Group randomized phase 3 trial (NOA-08) of patients aged >65 years with malignant gliomas, median and event-free survival times were 8.6 and 3.3 months, respectively, for patients treated with dose-dense TMZ and 9.6 and 4.7 months, respectively, for those undergoing RT (12). As for the Nordic study, MGMT promoter methylation status was associated with better clinical outcomes only in the TMZ treatment group. Although these results suggest that TMZ monotherapy may suffice in treating patients with MGMT promoter methylated tumors, future studies comparing TMZ alone and in conjunction with RT are needed to determine the optimal treatment choices for these patients in terms of survival and quality of life.

[5] 23w Our study has several limitations, owing to its retrospective nature. Choice of treatment was determined by physician and patient preferences, thus introducing bias.

[6] 49w Moreover, it was vulnerable to selection bias from unmeasured baseline characteristics, such as presence of comorbidities and radiation techniques. However, considering the dearth of randomized chemoradiotherapeutic trials designed to compare various schedules of RT, the present study provides evidence that clinical outcomes of standard and short-course RT-TMZ are comparable.

[7] 92w In conclusion, the combination of concomitant and adjuvant TMZ with an abbreviated course of RT seems to be a reasonable option for treating older patients with GBM, enabling survival benefit similar to those with radical RT while lowering toxicity. Prospective studies comparing the impact of various treatment strategies on survival and neurocognitive function, including chemotherapy alone or in conjunction with RT, and stratified according to MGMT methylation status, are needed to define optimal management of these complex patients. In future trials, assessing quality of life and neurocognitive function is a research priority.

[1] 148w Data collected from 329 patients aged 65 years with newly diagnosed, histologically confirmed GBM and Karnofsky performance status (KPS) 60 who received standard or short-course RT plus concomitant and adjuvant TMZ between June 2004 and October 2013 were evaluated in this tri-center, retrospective study. For patients treated at Sant'Andrea Hospital and Neuromed Institute all clinically relevant radiographic, surgical, and pathologic information was drawn from a prospectively maintained database of patients with brain tumors treated with RT. Data on patients treated at Regina Elena Hospital were collected from hospital charts and radiologic studies. After excluding 86 patients, owing to ineligible chemoradiotherapeutic regimens or insufficient pathology and clinical information, a total of 243 patients remained in the final analysis. The O 6 -methylguanine-DNA methyltransferase (MGMT) promoter methylation status was determined via a methylation-specific polymerase chain reaction method (19). Institutional review boards of Sapienza University and Sant'Andrea Hospital approved the study.

[2] 276w Patients received 3-dimensional conformal RT plus concomitant daily TMZ, followed by adjuvant TMZ. Radiation therapy consisted of standard (59.4-60 Gy, 30-33 fractions of 1.8-2 Gy) or short-course (40 Gy, 15 fractions of 2.66 Gy) focal irradiation, commencing within 4 to 6 weeks after surgery. In each patient, treatment volumes were delineated by fusing postoperative magnetic resonance imaging (MRI) with planning computed tomography images. The gross tumor volume included the resection cavity and any residual tumor seen on contrast-enhancing T1 postoperative MRI. The clinical target volume was the gross tumor volume plus a 2-cm margin modified to account for specific anatomical barriers to tumor spread (dura, ventricles, falx, and tentorium cerebelli). The planning tumor volume (PTV) was generated by adding a 3-to 4-mm margin to the clinical target volume. The prescribed dose was normalized to 100% at isocenter, ensuring at least 95% isodose surface coverage of PTV. Concomitant chemotherapy consisted of TMZ 75 mg/m 2 , given 7 days per week from the first day of RT. Adjuvant TMZ was started 4 weeks after cessation of RT and delivered 5 days every 28 days for 6-12 cycles. The first-cycle dose of 150 mg/m 2 was boosted to 200 mg/m 2 for subsequent cycles, with reduction or suspension for development of grade 3 or 4 toxicity (defined by National Cancer Institute Common Terminology Criteria for Adverse Events version 3.0; http://ctep.cancer.gov) or for disease progression. Neurologic status and tolerance to treatment were assessed weekly during RT and monthly thereafter. Typically, MRI scans were done before RT, before the first cycle of adjuvant TMZ, every 2 months during adjuvant chemotherapy, and then every 2 to 3 months or as appropriate.

[3] 81w As stipulated by the Response Assessment in Neuro-Oncology Working Group (20), changes in magnitude of all measurable enhancing and nonenhancing tumorous lesions on contrast-enhanced T1-weighted and fluid-attenuated inversion recovery MRI sequences were indicative of radiographic response. The same neuroradiologist (A.B.) interpreted all responses. Transient radiologic abnormalities, such as incremental edema and contrast enhancement on MRI, occurring within 12 weeks of RT cessation and appearing stable or absent on subsequent clinical and radiographic assessments (without change in therapy), were recorded as pseudoprogression.

[4] 77w Overall survival and PFS estimates were calculated from the time of surgery by the Kaplan-Meier method and compared using the logerank test. Chi-square and nonparametric Mann-Whitney tests were used to examine between-group covariate differences, and the Cox proportional hazards model was used for univariate and multivariate analysis, to assess effects of radiation schedule, patient characteristics, and other prognostic factors on OS and PFS. Variables at significance levels of P<.1 on univariate analysis were included in multivariate analysis.

[5] 162w In a confirmatory analysis, comparing standard and shortcourse RT-TMZ, a propensity score was calculated to achieve balanced distribution of baseline characteristics in treatment groups. Propensity score adjustment is particularly advantageous in nonrandomized comparisons of treatment subsets to avoid the effects of confounding variables on outcome, thus eliminating bias (21). Propensity scores were derived from logistic regression, with radiation technique as the dependent variable (control condition, 60 Gy). Independent variables included age at diagnosis, gender, KPS score, extent of resection, and MGMT methylation status. Patients and controls were matched one-to-one by the nearest-neighbor method, using a caliper distance 15% of the standard deviation of estimated propensity score logit to ensure good matches. Overall survival and PFS served as outcome measures. The adjusted treatment subsets were assessed for balance, using the overall c 2 balance test (22) and the relative multivariate imbalance measure (denoted as L1), as proposed by Iacus et al (23). Standard software was used for statistical analysis (SPSS v21.0, Chicago, IL).