Survival trends in elderly patients with glioblastoma multiforme: Resective surgery, radiation, and chemotherapy

Neoplasm

Survival Trends in Elderly Patients With Glioblastoma Multiforme: Resective Surgery, Radiation, and Chemotherapy Ravish V. Patwardhan, M.D.,* Cedric Shorter, B.S.,* Brian K. Willis, M.D., FACS,* Praveen Reddy, M.D.,* Donald Smith, M.D.,* Gloria C. Caldito, Ph.D.,† and Anil Nanda, M.D., FACS* *Department of Neurosurgery,†Department of Biostatistics, Louisiana State University Health Sciences Center in Shreveport, Shreveport, Louisiana

Patwardhan RV, Shorter C, Willis BK, Reddy P, Smith D, Caldito GC, Nanda A. Survival trends in elderly patients with glioblastoma multiforme: resective surgery, radiation, and chemotherapy. Surg Neurol 2004;62:207–15.

tively). No significant difference was found in KPS, comparing all individual groups versus each other (p ⱖ 0.06). Remarkably, 6 patients survived over 14 months (range, 14.1–22.7 months), all of which received RS⫹RT⫹C.

BACKGROUND

CONCLUSIONS

It is appropriate to investigate and to determine survival trends following glioblastoma multiforme treatment using resective surgery, radiation therapy, and/or chemotherapy in patients aged 59 years and higher. METHODS

We retrospectively reviewed 30 elderly patients (ⱖ59 years old) who were treated for histopathologically confirmed glioblastoma multiforme at our tertiary care institution from 1990 through 2002. All patients were treated with steroids. In addition, 22 patients underwent resective surgery (RS), 17 patients underwent radiation therapy (RT), and 10 patients underwent chemotherapy (C). Many patients underwent these treatments in various combinations: 6 underwent biopsy only, 7 RS only, 6 RS⫹RT only, and 9 RS⫹RT⫹C. For each case, pretreatment Karnofsky performance scores (KPS), tumor location, presenting symptoms and signs, associated surgical morbidity, and pre-existing medical conditions were also recorded. Patients were categorized into one of four treatment subgroups: Biopsy only, RS only, RS⫹RT, and RT⫹RS⫹C. For each of these subgroups, pretreatment KPS and post-treatment survival were compared. RESULTS

Post-treatment survival following biopsy only was 3.2 ⫾ 0.8 months (mean ⫾ SE); RS 2.2 ⫾ 0.5; RS⫹RT 5.5 ⫾ 1.2; RS⫹RT⫹C 13.6 ⫾ 2.1. A longer survival trend was noted for the RS⫹RT versus RS group (two-tailed unpaired t test, p ⫽ 0.02;), as well as the RS⫹RT⫹C group, which showed consistently higher survival in comparison to most of the other groups (p ⫽ 0.0021, 0.00039, 0.013 vs. the biopsy only, RS only, and RS⫹RT groups, respecAddress reprint requests to: Anil Nanda, M.D., FACS, Professor and Chairman, Department of Neurosurgery, Louisiana State University Health Sciences Center in Shreveport, 1501 Kings Highway, PO Box 33932, Shreveport, Louisiana 71130 –3932. Received March 4, 2003; accepted November 4, 2003. © 2004 Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010 –1710

This study suggests a significant improvement in elderly patients treated with the combination of resective surgery, radiation therapy, and chemotherapy, rather than either treatment alone or other combination. This significant improvement does not appear to be biased by pretreatment KPS, as mean KPS values did not significantly differ between any of these groups. However, a greater number of patients in each group must be considered to achieve the power to make more definitive treatment guidelines. © 2004 Elsevier Inc. All rights reserved. KEY WORDS

Glioblastoma multiforme, geriatric, survival, resective surgery, radiation, chemotherapy.

oorer prognosis in older patients suffering from glioblastoma multiforme (GBM) has been well documented [1,8,13,20,21]. The role of resective surgery and radiation therapy, the mainstay of treatment for most younger patients, has been questioned in older patients [8,7]. Proponents of more conservative therapy argue that a biopsy to establish diagnosis may be all that is surgically warranted since these elder patients have a shorter quality-of-life span, which is further compromised by recovery from resective surgery and/or radiation therapy [8]. Others argue that elderly patients are being unfairly undertreated, and hence should be offered the standard of care established for younger patients [3,7]. The goal of the present study is to determine whether the treatment modalities of solely biopsy, resective surgery, radia-

P

0090-3019/03/$–see front matter doi:10.1016/j.surneu.2003.11.016

208 Surg Neurol 2004;62:207–15

tion therapy, chemotherapy, or a combination thereof, significantly improved the survival of elderly patients suffering from GBM.

Materials and Methods PATIENT POPULATION Patients with histopathologically proven glioblastoma multiforme between 1990 and 2002 were retrospectively identified from the Department of Neurosurgery’s database at the Louisiana State University Health Sciences Center in Shreveport (LSUHSC-S). Only patients who were age 59 years or higher at the time of initial diagnosis were included in this study. Patient descriptive data, such as age, sex, and race were recorded, along with medical data such as pre-existing conditions, symptoms, signs, and Karnofsky Performance Scores (KPS) at initial presentation. Discovery of the intracranial mass was by computed tomography (CT) or magnetic resonance imaging (MRI). The date of diagnosis was recorded as the date at initial admission during which the histopathological sample was identified as GBM. Patients were excluded if they did not have pathologic confirmation, were not within the designated age of 59 years or higher, or had a solely infratentorial mass. Survival data (i.e., dates of death) were acquired from the Louisiana State Department Office of Public Health, following appropriate guidelines of the Institutional Review Board (IRB) at LSUHSC-S. SURGICAL MANAGEMENT Patients underwent either biopsy for diagnostic purposes or resective surgery for reduction of mass effect. As it has been well established that GBM in older patients is incurable by resection alone [1,3,14,17], all patients were categorized from the surgical standpoint into either the “biopsy only” or “resective surgery only” subgroups. All patients underwent at least a biopsy or resective surgery to obtain tissue diagnosis. The extent of surgical resection was limited to “noneloquent areas of brain” since tumors that extended into speech, motor, thalamus/hypothalamus, or brainstem regions were not aggressively resected. Tumors in more superficial, less eloquent cortex underwent a more extensive resection. All patients were treated with steroid therapy, which was individualized according to evident edema, mass effect, and a patient’s clinical status. RADIATION THERAPY Radiation therapy (RT) was offered to all patients as an option. Of the 30 patients, 15 underwent radia-

Patwardhan et al

tion therapy, with 48 – 64 Gy being delivered to the tumor (in 30 daily fractions, 5 days a week, over a 6 week course); 3 patients underwent gamma knife (GK) radiosurgery in addition to RT, with 12–14 Gy at the 50% isodose line. Two patients who underwent gamma knife alone (in the absence of other radiation) were excluded from this study. Patient choice, given the options, led to the decision of which therapy was instituted; 15 patients did not undergo radiation therapy by choice. CHEMOTHERAPY Adjuvant chemotherapy was comprised of the alkylating agents 1,3-bis (2-chloroethyl)-1-nitrosourea (BCNU), and temazolamide. Generally, BCNU was the first line of therapy, and if patients had treatment failures or toleration difficulty, temazolamide was substituted. The dose for BCNU was 200 mg per square meter times 4 doses. In addition, 1 patient had Gliadel® BCNU-containing wafers (Guilford Pharmaceuticals, Inc., Baltimore, Maryland) placed within the tumor cavity. At least 7 of 9 patients (78%) received BCNU as initial therapy, and temazolamide (350 mg po qd times 5 days, for 4 cycles) was added to 1 patient’s regimen following thrombocytopenia while taking BCNU. The remaining 2 patients received their adjuvant treatments at other centers. While these patients had received chemotherapy per records, the exact agents and doses were unobtainable. PATIENT SUBGROUPS As one of the goals of this study was to determine individually the influence of resective surgery (RS), RT, or chemotherapy (C) in prolonging survival, each subgroup was analyzed separately. This was not possible, however, for the “chemotherapy only” or “radiation therapy only” subgroups, as nearly all patients who had undergone chemotherapy or radiation therapy had concomitant radiation therapy or resective surgery. The following four subgroups were hence designated as follows: biopsy only, RS only, RS⫹RT, RS⫹RT⫹C. STATISTICAL ANALYSES For each of the designated subgroups, a t test (unpaired, 2-tailed) was carried out to determine statistical significance. Values of p ⬍ 0.017 were considered statistically significant (since 4 total groups were compared, therefore p of 0.05 divided by 3); values where 0.017 ⬍ p ⱕ 0.033 represented a statistically significant trend. T-test analysis was performed upon post-treatment survival (days), as well as pretreatment KPS (to determine whether KPS introduced a bias).

Survival of Elderly Patients with GBM

1

Surg Neurol 209 2004;62:207–15

Descriptive Composition of Entire Study Population

Sex Race KPS (pretreatment) Presenting symptoms

14 female, 16 male 19 White, 10 African-American, 1 Other Mean 67.9 ⫾ 2.8 (standard error) Focal weakness (35.5%) Headache (25.8%) Seizures (19.4%) Confusion (16.1%) Aphasia (16.1%) Aphasic speech–expressive (16.1%) Focal weakness (48.4%) Aphasia (19.4%) Confusion (9.7%) Frontal lobe (33.3%) Parietal lobe (33.3%) Temporal lobe (26.7%) Occipital lobe (23.3%) Thalamus (3.3%) Left (63.3%) Right (33.3%) Bilateral (3.3%) Biopsy only (20%) RS (73%) Conventional Radiation (50%) Conventional ⫹ Gamma Knife Radiation (10%) C (30%)

Presenting signs Brain region predominantly involved (some tumors were predominant in more than one region)

Brain hemisphere predominantly involved at diagnosis Types of procedures performed (% of patients)

Results The descriptive composition of the patient population as a whole is shown in Table 1. Individual subgroup analyses, as performed according to the scheme designated in the Materials and Methods section above, appears in Table 2. Survival curves for each of the 4 subgroups are shown in Figure 1 A–D. Among complications of treatment, one patient developed thrombocytopenia (from which the patient was asymptomatic) while being treated with BCNU, and his regimen was changed to temazolamide. Another patient developed a postoperative hematoma at the surgical site, which required re-

2

operation for evacuation; he recovered fully from this postoperative complication.

Discussion Little improvement in prognosis for patients diagnosed with GBM has been realized over the past half-century [21], and the uniformly poor prognosis has been emphasized by numerous studies [1,8,13,20]. Present therapeutic options include steroids, resective surgery, radiation therapy (administered via external beam or stereotactically), and chemotherapy. Overall 5-year survival for children suffering from GBM was 4 to 26% (with extent of

Subgroup Analyses: p-Values for Two-Tailed Unpaired t-Test Survival in Months, [KPS]

PROCEDURE: AVERAGE SURVIVAL ⴞ SE IN MONTHS [AVERAGE KPS ⴞ SE]

BIOPSY ONLY: P-SURVIVAL [P-KPS]

RS ONLY: P-SURVIVAL [P-KPS]

RS ⴙ RT ONLY P-SURVIVAL [P-KPS]

RS ⴙ RT ⴙ C P-SURVIVAL [P-KPS]

Biopsy only: 3.2 ⫾ 0.8 [75.0 ⫾ 5.6] RS only: 2.2 ⫾ 0.5 [60.0 ⫾ 4.5] RS ⫹ RT only: 5.5 ⫾ 1.2 [70.0 ⫾ 6.3] RS ⫹ RT ⫹ C: 13.6 ⫾ 2.1 [68.9 ⫾ 5.6]

– – – –

0.29 [0.058] – – –

0.15 [0.42] 0.023 [0.47] – –

0.0021 [0.30] 0.00039 [0.25] 0.0126 [0.82] –

Boldface indicates statistically significant trend; boldface underline indicates statistical significance. (p ⬍ 0.017 is statistically significant, whereas 0.017 ⬍ p ⱕ 0.033 indicates a statistically significant trend for this study). RS ⫽ resective surgery; RT ⫽ radiation therapy; C ⫽ chemotherapy.

210 Surg Neurol 2004;62:207–15

1

Patwardhan et al

(A–E): Kaplan-Meier survival curves for patients treated with steroids and (A) entire study population, (B) biopsy only, (C) RS only, (D) RS⫹RT only, (E) RS⫹RT⫹C.

Survival of Elderly Patients with GBM

resection leading to higher survival rates) [22], whereas in elderly adults the 5-year survival is practically universally zero [14]. This is particularly significant since malignant gliomas (predominantly GBM) constitute 13.3% of all primary intracranial tumors in patients age 70 years or higher [12,20]. The average survival in elderly patients is 4 to 8 months, according to most studies [1,2,10,14,21]. It is even poorer in elderly adults falling into a “poor functional status” [9,19]. The significantly diminished survival, compounded by poor functional status in many cases, calls into question the decision to administer “aggressive” therapy (such as resective surgery, radiation therapy, and chemotherapy) following histopathological diagnosis of GBM following biopsy. Advocates of “aggressive therapy” via one or more of the postbiopsy treatment modalities argue that an older patient should not be denied the standard of care which may prolong survival [3,7]; others argue that “aggressive, heroic treatment” does not necessarily result in significant meaningful improvement, and only worsens the quality of life for the remaining survival period [7,8]. The goal of the present study was to compare the survival among each of the following subgroups: Biopsy only, RS only, RS⫹RT only, and RS⫹RT⫹C. TREATMENT OPTIONS Steroids are widely used in the treatment of GBM, and some form of steroids (usually dexamethasone) was administered to all of our patients (with dose tailored according to edema and clinical status as indicated). All patients also underwent tissue sampling, whether via stereotactic or open biopsy, or during resective surgery. Following diagnostic histopathological confirmation (i.e., malignant glioma displaying associated necrosis), patients (or respective caretakers, if patients were not mentally capable of making a rational decision) were given options of RT or RT⫹C. Based upon individual case decisions, patients underwent either treatment or no further treatment. In cases of a tumor in a superficial, less eloquent cortex causing significant mass effect, an attempt at safe, decompressive surgery was made. Each of the patient treatment subgroups is considered here; note that we have not grouped GBM and anaplastic astrocytoma (AA) as other “malignant glioma” studies have, because of the strong survival bias introduced by the proportion of AA patients. The definitive diagnosis of GBM is histopathological. In 22 (66%) of our patients his-

BIOPSY ONLY.

Surg Neurol 211 2004;62:207–15

tology was obtained by biopsy alone, without further resective surgery; 7 of these patients had neither RS, RT, or C. The mean survival was 3.2 ⫾ 0.8 months for the group undergoing biopsy only. This did not differ significantly from the RS only, RT only, or RS⫹RT only subgroups; however, this was significantly shorter than the RS⫹RT⫹C group (13.6 ⫾ 2.1 months, p ⫽ 0.0005). In an analysis of patients aged 60 years or higher diagnosed with supratentorial glioma in the United Kingdom, Whittle et al noted a mean survival of 7 weeks following treatment with biopsy and steroids [21]. Kelly and Hunt [11] noted a postbiopsy survival of 15.4 weeks in 88 patients age 65 years or older who were histopathologically diagnosed with Grade 4 malignant gliomas. The data from the present study (3.2 months) falls well within the 1.75 to 4 month range supported by these other studies. Resection extent of GBM is correlated positively by some with survival, although this is more remarkable in pediatric cases [22]. With increasing age, the effect of resection extent becomes less dramatic; the previously noted study by Kelly et al [11] noted that a “modest” (i.e., 27 weeks vs. 18 weeks with gross total resection vs. stereotactic biopsy) survival improvement was attributable to resective surgery. While the role of any resective surgery in GBM cases is debated, some authors feel that a small benefit may arise by reducing mass effect; whether the cytoreductive effect has any bearing upon chemotherapy is questionable [21]. For supratentorial glioblastoma in patients age 70 years or greater, Mohan et al [14] have noted a median patient survival of 7.2 months for patients undergoing subtotal resection and radiation, versus 17.3 months for patients undergoing gross total resection and radiation. In the study by Whittle et al involving supratentorial gliomas, median survival was 16 weeks for the biopsy and RT-treated group, versus 23 weeks for the “cytoreduction ⫹ RT” treated group [21]. Our study did not show a significant improvement when comparing the patients who underwent biopsy only versus those undergoing resective surgery only (3.2 months vs. 2.2 months, respectively, 2-tailed t test, p ⫽ 0.58).

RS ONLY.

Several recent articles have addressed the issue of possible benefit from radiation. Hoegler et al, in a prospective study of patients in “malignant glioma” (with 23 of 25 patients diagnosed with GBM and 2 with AA), found an overall median survival of 8.0 months following 37.5 Gy administration in 15 daily fractions; however, since this study grouped the patients with biopsy only (40% of patients) and resective surgery, an independent analysis of RT

RSⴙRT.

212 Surg Neurol 2004;62:207–15

effect was not performed [10]. Thomas et al. [19] found that hypofractionated radiotherapy, 30 Gy in 6 fractions over 2 weeks, in poor prognosis patients (defined as elderly and/or disabled) with “highgrade glioma” was associated with median survival of 6 months. Bauman et al [2] found that RT for GBM at 30 Gy/10 fractions/2 weeks in patients aged ⱖ65 years or with KPS ⱕ50 lived a median of 6 months. Other studies which report on varying RT dose regimens are difficult to interpret, containing various combinations of GBM and RT, and combining groups which undergo surgery, biopsy, and/or chemotherapy into their analysis and results [17]; some studies have designated patients as having “malignant glioma” without histology, relying upon “clinical course and typical radiographic appearance,” [13] stating these results were better than historical controls. The study by Whittle et al [21] found a median survival of 16 weeks with RT (47– 60 Gy in 30 fractions) in patients age 60 years with a “supratentorial glioma.” In looking at specifically GBM patients, Mohan et al found a 7.3 month median survival following radiation of 55 Gy or more in patients age 70 years and older. Their study found a small difference between the groups undergoing subtotal RS⫹RT versus biopsy⫹RT (7.2 vs. 5.1 months mean, respectively); this difference increased substantially when considering the “gross total RS⫹RT” group with median survival 17.3 months. While higher, these “gross total RS⫹RT” group figures may be biased by the tumors location and size which would make them more amenable to undergoing gross total resection versus subtotal resection, for the 7 patients reported in that study [14]. In yet another study, working with a lower than “standard” dose of RT (30 Gy in 10 fractions), Newall et al [15] found a median survival of 44 weeks in patients 63 to 75 years of age, all of which were diagnosed with GBM. The combination of RS and RT, as noted, yielded a mean survival of 5.5 months in the present study. This was not significantly improved versus the biopsy only or RS only groups independently; KPS scores were also similar among all groups. One patient required reoperation for a postoperative hematoma evacuation, from which full recovery occurred. RSⴙRTⴙC. Patients undergoing RS⫹RT⫹C had the best response of all groups in the present study. Patients were predominantly treated with BCNU (200 mg/sq meter times 4 doses). This regimen is higher than Gilbert et al’s study [5] which used a combination of BCNU (40 mg/sq m/day times 3 days), along with RT (60.2 Gy) and surgery (RS in 35%, biopsy in 65%); they determined a median

Patwardhan et al

survival of 51.1 weeks in 17 patients with mean age 71 years old. However, an opposing viewpoint was presented by Grant et al [6], who found shorter median survival time following chemotherapy (with BCNU at 250 mg/sq meter q 6 to 8 weeks, which was a higher dose than that given in the present study). They noted a higher myelosuppressive rate with advancing age as well. In the present study, BCNU dose was slightly reduced in comparison to their study, hence promoting survival with fewer myelosuppressive complications—1 patient did develop thrombocytopenia while on BCNU, and his regimen was changed to temazolamide. Other study designs were influenced by KPS scores. For example, Pierga et al [17] found a 58week median survival; although this was biased by better KPS in the chemotherapy group, it is still substantial. As noted, the present study did not show a significant difference between the RS⫹RT⫹C group compared individually to others (mean KPS 68.9, p ⱖ 0.058 vs. each other group). Because patients undergoing RS⫹RT⫹C did significantly better than other groups, the question of selection bias (i.e., whether only longer-term survivors were started on chemotherapy) was raised. Upon analysis, however, it was noted that the mean time between diagnosis and initiation of chemotherapy was only 29 days. Hence, the effects of such selection bias were minimized. Decision-making regarding which patients received adjuvant therapy was analyzed as well. Interestingly, there was a statistically significant difference in age—more advanced age was correlated with less likelihood of undergoing adjuvant therapy (median 70 years versus 64 years old for receiving adjuvant radiation (t test, p ⫽ 0.027), and median 70 years versus 61 years old for receiving adjuvant chemotherapy (t test, p ⫽ 0.030). However, there was not a significant difference in performance status between those patients who received adjuvant therapy and those who didn’t, as evidenced by lack of difference of Karnofsky performance score between these two groups (66.9 vs. 68.2 average KPS for the nonadjuvant therapy receiving group and the adjuvant therapy-receiving groups, respectively (p ⫽ 0.82). This study did not specifically evaluate socioeconomic status as an independent variable; although it is the practice of this state-serving institution to equally refer patients for adjuvant therapy independent of socioeconomic status, actual receipt of adjuvant therapy correlating with socioeconomic status was not specifically analyzed and may be a factor. Additionally, there was no significant difference with respect to accompanying co-

Survival of Elderly Patients with GBM

morbidities for the nonadjuvantly treated versus adjuvantly treated groups [average number of medical problems 1.7 versus 1.8, respectively (t test, p ⫽ 0.82)]. Three patients in our study underwent gamma knife radiation. No significant survival difference was apparent in comparing those receiving gamma knife versus others in respective subgroups. There is insufficient data presently in the literature as to the efficacy of GK in older patients diagnosed with GBM; occasional papers that address GBM are not specific to the elderly age group as studied in this paper [4,16,18]. GAMMA KNIFE RADIOSURGERY.

Conclusion The findings in this study are remarkable for the six patients with histologically diagnosed GBM who survived over 14 months; all underwent RS⫹RT⫹C. While the RS⫹RT⫹C group shows significantly improved survival versus most other groups (without apparently being influenced by KPS at admission), it is important to note that this study does not have the power to make more definitive statements regarding this regimen. Nonetheless, the trend calls for further studies with accrual of greater numbers of patients to achieve the statistical power needed to more firmly establish treatment guidelines. One of the most important issues, which needs to be assessed on an individual patient basis, is whether the gain of 10 months of a deteriorating life is worth surgery, radiation therapy, and chemotherapy. We echo the sentiment of others [7,8] who emphasize the quality of life issues in a very limited life.

Surg Neurol 213 2004;62:207–15

6.

7.

8. 9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

The authors thank Jeanie McGee for her editorial assistance and Felicia Jackson for her assistance in obtaining medical records.

REFERENCES 1. Ampil F, Fowler M, Kim Kookmin. Intracranial astrocytoma in elderly patients. J Neuro-Oncol 1992;12: 125–30. 2. Bauman GS, Gaspar LE, Fisher BJ, Halperin EC, MacDonald DR, Cairncross JG. A prospective study of short-course radiotherapy in poor prognosis glioblastoma multiforme. Int J Radiation Oncol Biol Phys 1994;29:835–9. 3. Brandes A, Fiorentino M. Treatment of high-grade gliomas in the elderly. Oncology 1998;55:1–6. 4. Coffey, RJ. Boost gamma knife radiosurgery in the treatment of primary glial tumors. Stereotact Funct Neurosurg 1993;61(Suppl 1):59 – 64. 5. Gilbert MR, Bozik M, Armstrong TS, et al. Aggressive chemotherapy treatment of brain tumors in elderly:

19.

20.

21.

22.

results from phase II study (abstract). ASCO Proc 1996;15:A271. Grant R, Liang BC, Page MA. Age influences chemotherapy response in astrocytomas. Neurology 1995; 45:929 –33. Grau JJ, Verger E, Brandes AA, Rigon A, Monfardini S, DeAngelis LM. Radiotherapy of the brain in elderly patients. Eur J Cancer 2000;36:443–52. Halperin EC. Malignant gliomas in older adults with poor prognostic signs. Oncology 1995;9:229 –38. Hernandez JC, Maruyama Y, Yaes R. Accelerated fractionation radiotherapy for hospitalized glioblastoma multiforme patients with poor prognostic factors. J Neurooncol 1990;9:41–5. Hoegler DB, Davey P. A prospective study of short course radiotherapy in elderly patients with malignant glioma. J Neuro-Oncol 1997;33:201–4. Kelly PJ, Hunt C. The limited value of cytoreductive surgery in elderly patients with malignant gliomas. Neurosurgery 1994;34:62–7. Kuratsu J, Ushio Y. Epidemiological study of primary intracranial tumours in elderly people. J Neurol Neurosurg Psychiatry 1997;63:116 –8. Meckling S, Dold O, Forsyth PAJ, Brasher P, Hagen NA. Malignant supratentorial glioma in the elderly: is radiotherapy useful? Neurology 1996;47:901–5. Mohan DS, Suh JH, Phan JL, Kupelian PA, Cohen BH, Barnett GH. Outcome in elderly patients undergoing definitive surgery and radiation therapy for supratentorial glioblastoma multiforme at a tertiary care institution. Int J Radiat Oncol Biol Phys 1998;42:981–7. Newall J, Ransohoff J, Kaplan B. Glioblastoma in the older patient. How long a course of radiotherapy is necessary? J Neurooncol 1988;6:325–7. Nwokedi EC, DiBiase SJ, Jabbour S, Herman J, Amin P, Chin LS. Gamma knife stereotactic radiosurgery for patients with glioblastoma multiforme. Neurosurgery 2002;50:41–7. Pierga J-Y, Hoang-Xuan K, Feuvret L, et al. Treatment of malignant gliomas in the elderly. J Neuro-Oncol 1999;43:187–93. Regine, WF, Patchell, RA, Strottman, JM, Meigooni, A, Sanders, M, Young, B. Combined stereotactic splitcourse fractionated gamma knife radiosurgery and conventional radiation therapy for unfavorable gliomas: a phase I study. J Neurosurg 2000;93(Suppl 3): 37– 41. Thomas R, James N, Guerrero D, Ashley S, Gregor A, Brada M. Hypofractionated radiotherapy as palliative treatment in poor prognosis patients with high grade gliioma. Radiother Oncol 1994;33:113–6. Villa` S, Vin ˜ olas N, Verger E, et al. Efficacy of radiotherapy for malignant gliomas in elderly patients. Int J Radiat Biol Phys. 1998;42:977–80. Whittle IR, Denholm SW, Gregor A. Management of patients aged over 60 years with supratentorial glioma: lessons from an audit. Surg Neurol 1991;36:106 – 11. Wisoff JH, Boyett JM, Berger MS, et al. Current neurosurgical management and the impact of the extent of resection in the treatment of malignant gliomas of childhood: a report of the Children’s Cancer Group trial no. CCG-945. J Neurosurg 1999;90:1147–8.