- Email: [email protected]
The role of postoperative irradiation in the treatment of oligodendroglioma
Int. J. Radiation
Oncology
Pergamon
Biol. Phys., Vol. 30, No. 3, pp. 567-573, 1994 Copyright 0 1994 Elsevier Science Ltd Printed in the USA. All rights reserved 0360-3016/94 $6.00 + .OO
0360-3016(94)00281-9
??Clinical Original Contribution THE ROLE OF POSTOPERATIVE IRRADIATION TREATMENT OF OLIGODENDROGLIOMA DAVE E. GANNETT,
IN THE
M.D.,* WILLIAM M. WISBECK, M.D.,* DANIEL L. SILBERGELD, AND MITCHELL S. BERGER, M.D.+
M.D.+
Departments of *Radiation Oncology and +Neurosurgery, University of Washington, Seattle, WA Purpose: Controversy regarding the role of adjuvant radiation therapy for the treatment of oligodendroglioma continues to exist. To better define the utility of postoperative irradiation for this tumor, the experience at the Universitv of Washineton was retrospectivelv examined. Methods-and Materials: The histologic samples of 63 patients given the diagnosis of oligodendroglioma were reviewed by a panel of neuropathologists and 41 were classified as pure oligodendroglioma. The two treatment groups included 14 patients treated with surgery only and 27 who received surgery and postoperative radiation and were analyzed using univariate and multivariate analysis with respect to prognostic factors, freedom from relapse, and survival. Results: Univariate statistical analysis of 14 clinical variables showed that a poorer prognosis was associated in patients with high cell density (p = .008), necrosis (p = .017), hemiparesis (p = .026), and papilledema (p = .091), while patients presenting with seizures had a better prognosis (p = .0096). Multivariate analysis showed necrosis (p = .OOl) and hemiparesis (p = .02) to be associated with decreased survival. Multivariate and univariate analysis of the treatment groups found them to be homogenous with respect to prognostic factors. Survival times were significantly longer in the group treated with postoperative irradiation (median survival time 84 vs. 47 months, .032). The 5 and 10 year survival rates were 83% and 46%, respectively, for the irradiated patients compared with 51% and 36% for those treated with surgery alone. Freedom from tumor recurrence times were also longer in irradiated patients (median relapse free time 79 vs. 42 months, .Ol). Conclusion: Based on the results of this study, we recommend continuing the practice of treating oligodendroglioma with postresection irradiation until a prospective multicenter clinical trial is conducted to thoroughly evaluate the role of postoperative irradiation in the treatment of this tumor.
p
=
p=
Brain neoplasm, Oligodendroglioma,
Outcome, Radiation therapy.
INTRODUCTION
variability in radiation dose and technique, and lack of histologic confirmation of the cases. The benefit of postoperative x-ray therapy in the management of patients with cerebral oligodendroglioma has not, therefore, been adequately determined. In this study, we have evaluated the effects of postoperative radiation therapy on 41 histologically confirmed cases of oligodendroglioma treated at the University of Washington between 1956 and 1984.
Oligodendrogliomas comprise about 5% of all primary brain tumors (2, 6, 9, 21, 25). Since the clinical and pathologic characteristics of oligodendroglioma were first described in the 1920s (2,3), many series have been published which attempted to define the optimal treatment modalities for patients with this tumor. Although there is nearly unanimous agreement that surgery enhances survival (10, 11, 23, 34), controversy remains regarding the effect of radiation therapy, with several studies confirming (4, 7, 14, 15, 26, 28, 29, 3 1, 33) and discounting (5, 20, 30, 32) improved survival following surgery and subsequent radiation treatment. Interpretation of these studies is difficult due largely to limited numbers of patients,
METHODS
AND
MATERIALS
Patient population
Sixty-three patients with a diagnosis of oligodendroglioma who were treated at the University of Washington between 1956 and 1984 were evaluated. Forty-one of these
Reprint requests to: Dave E. Gannett, M.D., University of Arizona Health Sciences Center, Department of Radiation Oncology, 1501 N. Campbell Ave., Tucson, AZ 85724. Acknowledgements-We would like to thank Doctors E. Alvord, S. Lofton, and Cheng-Mei Shaw for their detailed review of the histologic sections and Doctors B. Stea and J. R. Cassady for
their review of the manuscript. This work was supported in part by National Institutes of Health, Grants NS07144 and KOSNSO1253-01. Dr. M. S. Berger is an American Cancer Society Professor of Clinical Oncology. Accepted for publication 13 May 1994. 567
I. J. Radiation Oncology 0 Biology 0 Physics
568
patients had sufficient histologic criteria for confirmation of the diagnosis and were included in this study. Two patients with anaplastic oligodendroglioma and two patients who died of postoperative complications within 30 days of surgery were omitted. The decision whether or not to give postoperative radiation therapy varied with the opinions of the surgeon and the consulting radiation oncologist. All patients had a Karnofsky performance status of at least 70. Twenty-seven patients were irradiated postoperatively. The fractionation ranged from 1.5-2.0 Gy per day and the treatment regimen was completed within 3 months of the craniotomy. Radiation was usually administered -2 weeks after surgery when the incision of the scalp was well healed and local subcutaneous fluid collections or edema had regressed. Doses ranged from 40.79-66.00 Gy, (mean 57.03 Gy, median 59.07 Gy; only one patient received < 50 Gy). Five patients received irradiation to the whole brain as a portion of their therapy, while 22 were irradiated to wide local fields. The followup period for survivors ranged from 5-22 years with a mean follow-up of 8.7 years. Two patients were lost to follow-up 4 and 13 years following surgery and were censored in calculation of survival times. Duration of survival and freedom from relapse were calculated from the date of initial resection. Histological evaluation Histologic sections of 63 patients previously coded in the pathology report as “oligodendroglioma” or “probable oligodendroglioma” were replated in 1989 and reviewed by three neuropathologists who were not aware of the clinical history. The diagnosis of oligodendroglioma depended on a uniform population of cells with perinuclear halos, a distinctive network of delicate parenchymal vessels, round to slightly oval nuclei, and a lack of fibrillarity with a pronounced tendency to infiltrate the cerebral cortex (24). Since virtually all oligodendrogliomas contain a component of either neoplastic or reactive astrocytes ( 1, 9, 19) tumors with an astrocytic composition of I 25% were defined as pure oligodendrogliomas ( 16) for the purposes of this study. Twenty cases had a significant fraction of tumor composed of another cell type accounting for greater than 25% of the entire tumor and were classified as mixed tumors. Two cases exhibiting significant necrosis or high cell density were classified as anaplastic. Statistical analysis The hospital records were analyzed with respect to the age at diagnosis, gender, race, location of tumor, extent
Table 1. Presenting
of resection, time from date of surgery to relapse and death, presenting signs and symptoms, and postoperative course. Histologic sections were reviewed for the presence of necrosis, high cell density, and microcysts. Freedom from relapse and survival rates were calculated using the method of Kaplan and Meier ( 13). Associations between categorical variables were evaluated by chi-square test on contingency tables or Fisher’s exact test (two-tailed) as appropriate. Intergroup differences were tested by MannWhitney U-test or t-test as appropriate. Intergroup differences in censored survival times was tested by the Wilcoxon test. Multivariate analysis of prognostic factors used the Cox proportional hazard regression model. Differences in censored survival based on computer-generated curves between groups were tested by the Mantel-Cox test. Differences with p-values I 0.05 were considered statistically significant. RESULTS
Patient population Of the 4 1 patients, 27 were male and 14 were female, a 1.9: 1 ratio in agreement with previous reports ( 12, 15). Thirty-nine of the patients were white and two were black. Patient ages at the time of surgery ranged from 3-78 years (median 45 years). The median age of the 14 patients who did not receive postoperative x-ray therapy was 36 years. The median age of the 27 patients who did receive postoperative irradiation was 42 years. The age distribution of irradiated and nonirradiated patients was similar to previous reports (4, 12, 15,26) with a bimodal distribution favoring ages 3-10 and 45-60. Three patients were < 18 years old. Twenty-five patients (6 1%) presented with some type of seizure disorder. Other presenting symptoms included mental and personality changes, nausea, vomiting, and headaches. The time from onset of symptoms to diagnosis varied from one month to 22 years. Common physical findings included hemiparesis, focal neurologic deficits including weakness, speech changes, and paresthesias, cranial nerve deficits, sensory deficits, and visual deficits. Presenting symptoms and signs are detailed in Table 1. Surgical findings Forty tumors (98%) were located in the cerebral hemispheres and one tumor was located almost exclusively in the thalamus. Of these, the majority (66%) were located either partially or completely in the frontal lobe. Nine tumors infiltrated the temporal lobe, five infiltrated the
signs, and location
for 4 1 patients
with oligodendroglioma Location
Signs
Symptoms Seizures Altered mental status Headaches Nausea/vomiting
symptoms,
Volume 30, Number 3, 1994
25 9 14 3
(61%) (22%) (34%) (7%)
Focal deficit Papilledema Hemiparesis Cranial nerve deficit
8 14 9 4
(20%) (34%) (22%) (10%)
Frontal Parietal Temporal Occipital
27 (66%) 6 (15%) 9 (22%) 2 (5%)
Treatment of oligodendroglioma 0 D. E.
GANNETT
569
et al
Freedom From Progression
Fig. 1. Kaplan-Meier
survival and freedom
from progression
lobe, and two tumors were found in the occipital lobe, in rough proportion to the volume of each lobe (Table 1). According to the operative report, 37 patients had a subtotal resection of their tumor, and, in four cases, the surgeon stated that all of the tumor tissue had been removed. patietal
Survival analysis Complete follow-up was available for 39 of the 41 patients in this study as of July 1989. Thirteen of the 41 patients were alive at the end of the study. The age at death varied from 12 to 82 years and the time from diagnosis to death ranged from 3 months to 11 years. The 5 and 10 year survival rates for the 41 patients with oligodendroglioma were 74% and 41%, respectively (Fig. 1). Freedom from relapse was 6 1% and 18% at 5 and 10 years. Univariate analysis of possible prognostic factors was performed to identify which factors, if any, influenced the time to death of the patients (Table 2). Time to tumor recurrence was not evaluated because of the relatively inexact nature of determining this variable retrospectively. The length of time from the manifestation of the symptom to diagnosis did not significantly affect survival. Age, gender, and race had no significant effect on survival. Seizure as the initial symptom was associated with a longer survival time QJ = .0096), while headaches, nausea and vomiting, and personality change had no significant effect. Of the physical findings, papilledema (p = .091) and hemiparesis (I, = .026) were associated with a decrease in survival time. Analysis of tumor histologic features showed necrosis (p = .017) and high cell density 07 = .008) to be associated with a significant reduction in survival time. Multivariate analysis of prognostic variables revealed necrosis @ = .OO1) and hemiparesis (p = .02) to be associated with decreased survival. Multivariate and univariate analysis of the distribution of prognostic variables between the surgery only and the surgery and x-ray therapy groups
curves for 4 1 patients
with oligodendroglioma.
showed no apparent heterogeneity between the groups (Table 3). A significant increase in survival time was evident in patients treated with postoperative irradiation. Five and 1O-year survival rates in the irradiated patients were 83% and 46%, respectively, and 5 1% and 36% in the nonirradiated patients (p = .032, Fig. 2). The median survival time of the irradiated patients was 7.0 years, compared to 3.9 years for those treated only with surgery. Time to tumor recurrence was also longer in the irradiated patients. Five and IO-year freedom from relapse rates were 74% and 17%, respectively, in the irradiated group, compared to 37% and 19% in the surgery only group @ = .Ol, Fig. 3). The median relapse free times were 79 months in the irradiated patients and 42 months in the surgery only group.
Table 2. Univariate patients
analysis of factors evaluated with oligodendroglioma
for 41
Variable
Univariate p value
Gender Age Race Headache Mental changes Seizures at presentation Nausea/vomiting Focal deficits Hemiparesis Papilledema Cranial nerve palsies Necrosis High cell density Microcysts
NS 0.18 NS NS NS 0.0096 NS 0.13 0.026 0.09 1 NS 0.017 0.008 NS
NS = Not significant.
570
I. J. Radiation Oncology 0 Biology 0 Physics
Table 3. Comparison of prognostic variables between treated with or without adjuvant radiotherapy Variable Number of patients Gender (male/female) Mean age (years) Race (black/white) Symptoms Seizure Mental change Headache Nausea/vomiting Signs Focal deficit Papilledema Hemiparesis Cranial nerve palsy Histopathology Necrosis High cell density Microcysts Extent of resection Subtotal Total Status (alive/dead)
Irradiated
Nonirradiated
patients
p-value
27 1819 39.2 l/26
14 915 40.0 l/13
NS NS NS NS
14 7 10 2
11 2 4 1
NS NS NS NS
5 9 6 3
3 5 3 1
NS NS NS NS
2 8 12
2 5 7
NS NS NS
25 2 10/17
12 2 3/11
NS NS 0.03
NS = Not significant.
Of the four cases in which the surgeon stated that the tumor had been completely removed, two patients are still alive at 13 and 22.6 years after surgery without evidence of recurrence, and the other two patients died 7.6 and 3.8 years postoperatively. The patients living 13+ and 7.6 years postoperatively had both received adjuvant radiation. All 28 patients who died during the follow-up period did so from complications related to the recurrent tumor. Histologic sections were available from eight patients who
I
Fig. 2. Survival (XRT).
rates for patients
Volume 30, Number 3, 1994
underwent a second surgical resection for recurrent tumor. Only three of these cases showed pure oligodendroglioma at the time of recurrence. In the remaining five cases, two were diagnosed as glioblastoma multiforme, two showed mixed gliomas, and one showed an moderately anaplastic astrocytoma without evidence of oligodendroglioma. Five patients received a second course of radiation therapy as part of their treatment for recurrent tumor; two died as a direct result of the tumor within a year of retreatment and three are alive without evidence of recurrence 2.2, 5.4, and 6.5 years after second treatment. The differences in survival between the two groups were not significantly altered by either censoring or excluding the patients who underwent a second surgical resection and or received additional radiation. Toxicity One patient developed a complication of treatment, unrelated to the tumor. She had a subtotal resection of a left frontal oligodendroglioma at 30 years of age. Postoperatively, she was irradiated with 45 Gy to the whole brain with a 15 Gy boost to the tumor bed. Two years later, a computed tomography (CT) scan showed evidence of recurrent tumor at the site of prior resection, and this was treated with an additional course of x-ray therapy consisting of 45 Gy to the tumor area. Approximately 5 years after the second course of irradiation, she was diagnosed with a Grade 2 (RTOG/EORTC late radiation morbidity scoring scheme) radiation induced proliferative retinopathy. A CT scan at that time showed no evidence of tumor recurrence to account for the visual loss and there was no evidence of diabetes. The retinopathy had worsened to Grade 3 at last follow-up, eight years after initial therapy. Three patients received some form of chemotherapy as part of their treatment. One died from recurrence of the
..~:‘..r’......_____
_____________.__._..--~’
NoXRT
with oligodendroglioma
treated
with and without
postoperative
x-ray therapy
Treatment of oligodendroglioma 0 D. E.
GANNETT
571
et al.
1
1
,
1
I-
O
5
10
15
20
YeUS
Fig. 3. Freedom x-ray therapy.
from progression
rates for patients with oligodendroglioma
tumor 24 months following therapy, one had a recurrence 7 years postoperatively, and the other patient is still alive without evidence of recurrence 9.2 years following surgery.
primary
DISCUSSION Since oligodendrogliomas, like other gliomas, tend to have an infiltrative growth pattern, complete resection is only rarely achieved. As with other gliomas, the tendency of oligodendroglioma to remain contained in the cerebrospinal axis suggests that irradiation as a regional therapy could potentially control these tumors. Consequently, recent literature (31, 35, 36) has recommended radiotherapy as adjuvant therapy to achieve local control and our results support the efficacy of this mode of treatment. In this series of 41 patients, the median survival time was significantly longer when radiation therapy was given postoperatively (p = .032). This group also had better 5 and lo-year survival rates (83% and 46% vs. 5 1% and 36%) when compared with the patients treated with surgery alone. Freedom from relapse times were also longer in the irradiated patients (p = .Ol). Given that the two patient groups were homogenous with respect to prognostic variables, this data suggests a benefit for postoperative radiation for patients with this tumor histology. Previous reports discussing the effects of radiation treatment on the course of oligodendroglioma have been hampered by small numbers of cases, lack of pathologic confirmation of the diagnosis, inadequate statistical analysis, and poor control of the population differences. In a review of the Cleveland Clinic experience in 1983, Reedy et al. (20) found that radiotherapy did not have a significant effect on the survival of 48 patients evaluated. However, radiation doses varied from 22-65 Gy, indicating significant heterogeneity in therapy with some patients
treated with and without postoperative
obviously undertreated. Furthermore, there was no pathologic review or clear evaluation of the distribution of histologic variables. The typical radiation dose used for malignant intrinsic brain tumors is usually greater than 40 Gy, with typical fractionation schemes delivering 1.82.0 Gy 5 days per week. Deviation from this type of treatment regimen may have a significant impact on tumor response. The comprehensive review by Bullard (5) which uses a rigid and thorough statistical analysis raises similar concerns. They too, concluded that there was no survival benefit for patients given postoperative irradiation following resection. However, 13 patients received < 40 Gy and 18 patients received anywhere from 40-60 Gy. Additionally, 17 patients were treated with orthovoltage technique. The histologic criteria for inclusion into the study were somewhat questionable as only one pathologist reviewed the slides with exclusion criteria for “lesions clearly containing neoplastic fibrillary astrocytes” as mixed gliomas. While limited by the constraints of a retrospective format, this study is strengthened by a number of factors. The replating and reevaluation of the pathology specimens by a panel of three experienced neuropathologists for all previously diagnosed oligodendrogliomas at our institution gives us reasonable confidence in the histologic diagnosis. University and departmental records were available for all confirmed cases and only two patients were lost to follow-up. In addition, an elaborate statistical analysis of the treatment groups show them to be homogenous with respect to prognostic factors. However, our study is limited by the relatively small size of the patient population, resulting in limited power to demonstrate a causal effect. In addition, a wide variety of field sizes, radiation equipment, doses and diagnostic strategies were used over the 28-year period of this study. The only way of circum-
572
I. J. Radiation Oncology 0 Biology 0 Physics
Volume 30, Number 3, 1994
Table 4. Review of previous series evaluating the impact of radiotherapy on oligodendroglioma Author, year
n
Survival rt
Survival no rt
Current study Wallner (33) 1988 Bullard (5) 1987 Lindegaard (14) 1987 Reedy (20) 1983 Pailas ( 18) 1982 Chin (7) 1980 Sheline (29) 1977 Weir (34) 1968 Roberts (22) 1966 Shenkin (30) 1964 Richmond (2 1) 1959 Horrax (12) 1951
41 42 71 170 48 63 35 36 63 33 11 22 22
10 years, 46% 10 years, 56% 5 years, 54% 5 years, 36% 5 years, 5 1% 92 month median 5 years, 100% 5 years, 85% 6.4 years mean 6.8 years mean 65 month median 5 years, 53% 87 month median
10 years, 36% 10 years, 18% 5 years, 48% 5 years, 27% 5 years, 54% 36 month median 5 years, 82% 5 years, 3 1% 4 years mean 5.2 years mean 13 month median (No patients) (No patients)
Dose GY) 40-66 > 45 < 40-62 20-60 22-65 > 30 53-70 > 20 NA > 60 30-65 NA NA
n = Number of patients; rt = Radiation therapy; NA = Not available.
venting the weaknesses of this and other retrospective studies on cerebral oligodendrogliomas, is to conduct a randomized prospective trial in which surgical resection with or without adjuvant radiation is compared. Contributing to the controversy regarding the role of radiation in the treatment of oligodendroglioma is the fact that the tumor pathology is not always clearly defined. Nearly all oligodendrogliomas contain areas of other glial cell differentiation. Most often this is limited to populations of reactive astrocytes and oligodendroglia, but may sometimes include neoplastic ependymal cells ( 1, 9, 19). Rubenstein (24) has estimated that half of all oligodendrogliomas contain some component of another tumor type. This creates confusion in pathologic diagnosis since there are no accepted criteria as to how much of a second neoplastic cell type is required to change the diagnosis from oligodendroglioma, and specific cell markers such as galactocerebroside are not applicable in paraffin fixed sections. In this study, we used the criteria proposed by Mork ( 16) in classifying pure oligodendrogliomas as those with < 25% astrocytic component. Recent research has provided insight to the etiology of the heterogeneity of oligodendrogliomas. Distinguishing the mature oligodendrocyte from an astrocyte is the presence of intracellular microtubules rather than glial filaments, therefore oligodendrocytes are negative for glial fibrillary acidic protein (GFAP). However, it has been noted that some immature human oligodendroglia and neoplastic oligodendroglia
may transiently express GFAP (17). This, coupled with evidence demonstrating that fetal radial glia give rise to cells of both astroglial and oligodendroglial lineage (8) suggest the possibility that the glial precursor cell is the target of the initial event preceding the phenotypic expression of neoplastic transformation. This could explain the high prevalence of astrocytic cell elements found in oligodendrogliomas, as well as the histologic progression of this tumor observed in this and other studies (12, 27).
CONCLUSION Based on this and other reported studies (Table 4), we offer the following recommendations regarding the indications for radiation therapy in the treatment of supratentorial oligodendroglioma: (a) Postoperative irradiation is suggested following subtotal resection of the tumor. A recent meta-analysis of the cancer literature (3 1) also suggests a benefit for patients with oligodendrogliomas who have undergone a subtotal resection. (b) Because of the possible change in histology documented in some recurrences and the similar appearance of radiation necrosis on imaging studies, it is advisable to obtain a biopsy prior to consideration of further radiation and or chemotherapy. (c) A randomized prospective trial will be necessary to definitively establish the role of adjuvant radiotherapy in the management of oligodendroglioma.
REFERENCES Aparicio, S. G.; Lumsden, C. E.; Jennings, M. Electron microscopy of &al reactions observed in CNS cultures. Brain Res. 7:268; 1968. Bailey, P.; Bucy, P. C. Oligodendrogliomas of the brain. J. Pathol. Bacterial. 32:735-751; 1929. Bailey, P.; Cushing, H. A classification of the tumors of the glioma group on a histogenetic basis with a correlated study of prognosis. Philadelphia: JB Lippincott; 1926: 175- 176. Bouchard, J. Effects of irradiation in treatment of intracranial gliomas-Treatment results by histologic groups. In:
Radiation therapy of tumors and diseases of the nervous system. Philadelphia: Lea & Febiger; 1966:78- 118. Bullard, D. E.; Rawlings, C. E., III; Phillips, B.; Cox, E. B.; Schold, S. C.; Burger, P. C.; Halperin, E. C. Oligodendroglioma-An analysis of the value of radiation therapy. Cancer 60:2179-2188; 1987. Burger, P. C.; Vogel, F. S. Surgical pathology of the nervous system and its coverings, 2nd ed. New York: John Wiley and Sons; 1982:335-349. Chin, H. W.; Hazel, J. J.; Kim, T. H.; Webster, J. H. Oli-
Treatment of oligodendroglioma 0 D. E. GANNETTet al.
8.
9. 10.
Il. 12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
godendrogliomas-A clinical study of cerebral oligodendrogliomas. Cancer 45:1458-1466; 1980. Choi, B. H.; Kim, R. C.; Lapham, L. W. Do radial glia give rise to both astroglial and oligodendroglial cells? Devel. Brain Res. 8:119; 1983. Escalona-Zapata, J. Uncommon oligodendrogliomas. Acta Neuropathol. 7(Suppl.):94; 198 1. Grant, F. C. A study of the results of surgical treatment in 2326 consecutive patients with brain tumors. J. Neurosurg. 13:479; 1956. Hart, W.; Petito, M.; Earle, W. Mixed ghomas. Cancer 33: 134; 1974. Horrax, G.; Wu, W. Q. Postoperative survival of patients with intracranial oligodendroglioma with special reference to radical tumor removal; A study of 26 patients. J. Neurosurg. 8:473-479; 195 1. Kaplan, E. L.; Meier, P. Nonparametric estimation from incomplete observations. J Am. Stat. Assoc. 53:457-481; 1958. Lindegaard, K. F.; Mork, S. J.; Eide, G. E.; Halvorsen, T. B.; Hatlevoli, R.; Solgaard, T.; Dahl, 0.; Ganz, J. Statistical analysis of clinicopathological features, radiotherapy, and survival in 170 cases of oligodendroglioma. J. Neurosurg. 67:224-230; 1987. Mork, S. J.; Lindegaard, K. F.; Halvorsen, T. B.; Lehman, E. H.; Solgaard, T.; Hatlevoli, R.; Harvel, S.; Ganz, J. Oligodendroghoma: Incidence and biological behavior in a defined population. J. Neurosurg. 63:881-889; 1985. Mork, S. J.; Halvorsen, T. B.; Lindegaard, K. F.; Eide, G. E. Oligodendroglioma: Histologic evaluation and prognosis. J. Neuropathol. Exp. Neurol. 45:65-78; 1986. Ogawa, H.; Satoi, Y.; Tateishi, K.; Kitamura, K. Transient expression of glial fibrillary acidic protein in developing oligodendrocytes in vitro. Devel. Brain Res. 18: 133; 1985. Pailas, J. E.; Combalbert, A.; Bernard-Badier, M.; Bille, J.; Frank, R. Etude sur l’evolutiondes oligodendrogliomes de l’encephale. A propos d’une serie operatoire de 34 cas. Acta Neural. Psych. Belg. 64:537-55 I; 1964. Ravens, J. R.; Adamkiewicz, L. L.; Groff, R. A. Cytology and cellular pathology of oligodendrogliomas of the brain. J. Neuropathol. Exp. Neurol. 14:142; 1955. Reedy, D. P.; Bay, J. W.; Hahn, J. F. Role of radiation therapy in the treatment of cerebral oligodendroglioma: An analysis of 57 cases and a literature review. Neurosurgery 13:499-503; 1983. Richmond, J. J. Malignant tumours of the central nervous system. In: Raven, R. W., ed. Cancer 5. London: Butterworth; 1959:375-389.
573
22. Roberts, M.; German, W. J. A long term study of patients with oligodendrogliomas. Follow-up of 50 cases, including Dr. Harvey Cushing’s series. J. Neurosurg. 24:697-700; 1966. 23. Rossi, G. R.; Feoli, F.; Fernandez, E. The role of surgery in the treatment of supratentorial brain gliomas. In: Paoletti, P., Walker, M. D., Butti, G., et al. eds. Multidisciplinary aspects of brain tumor therapy. Amsterdam: Elsevier North Holland; 1979:155-163. 24. Rubenstein, L. J. Tumors of the central nervous system. Atlas of tumor pathology, 2nd series, fascicle 6. Washington, DC: Armed Forces Institute of Pathology. 25. Russel, D. S.; Rubenstein, L. J. Pathology of tumors of the nervous system, 4th ed. London: Edward Arnold; 1977: 195203. of intracranial 26. Scanlon, P. W.; Taylor, W. F. Radiotherapy astrocytomas-analysis of 4 17 cases treated from 1960- 1969. Neurosurg. 5:301-308; 1979. R.; Ansuini, G.; Pisani, A. E.; 27. Scarpelli, M.; Monitroni, Rychlicki, F.; Papo, I.; Mariuzzi, G. M. The value of morphometry in prediction of survival in oligodendrogliomas. Clin. Neuropathol. 6:155; 1987. 28. Shaw, E. G.; Scheithauer, B. W.; O’Fallon, J. R.; Tazelaar, H. D.; Davis, D. H. Oligodendrogliomas: The Mayo Clinic experience. J. Neurosurg. 76:428-434; 1992. 29. Sheline, G. E.; Boldrey, E.; Karlsberg, P.; Phillips, T. L. Therapeutic considerations in tumors affecting the central nervous system; Oligodendrogliomas. Radiology 82:84-89; 1964. therapy on oli30. Shenkin, H. A. The effect of roentgen-ray godendrogliomas of the brain. J. Neurosurg. 22:57-59; 1965. 31. Shimizu, K. T.; Tran, L. M.; Mark, R. J.; Selch, M. T. Management of oligodendrogliomas. Radiology 186:569572; 1993. 32. Uihlein, A.; Colby, M. Y.; Layton, D. D.; Parsons, W. R.; Carter, T. L. Comparison of surgery and surgery and irradiation in the treatment of supratentorial gliomas. Acta. Radiol. Ther. 5:67-78; 1968. 33. Wallner, K. E.; Gonzales, M.; Sheline, G. E. Treatment of oligodendrogliomas with or without postoperative irradiation. J. Neurosurg. 68:684-688; 1988. An analysis of 63 cases. J. 34. Weir, B. E. Oligodendrogliomas: Neurosurg. 29:500-505; 1968. 35. Wilkins, R. H.; Renagachary, S. S., eds. Neurosurgery. New York: McGraw-Hill; 1985. 36. Youmans, J. R., ed. Neurological surgery, 2nd ed. Philadelphia: WB Saunders; 1982.
Oncology
Pergamon
Biol. Phys., Vol. 30, No. 3, pp. 567-573, 1994 Copyright 0 1994 Elsevier Science Ltd Printed in the USA. All rights reserved 0360-3016/94 $6.00 + .OO
0360-3016(94)00281-9
??Clinical Original Contribution THE ROLE OF POSTOPERATIVE IRRADIATION TREATMENT OF OLIGODENDROGLIOMA DAVE E. GANNETT,
IN THE
M.D.,* WILLIAM M. WISBECK, M.D.,* DANIEL L. SILBERGELD, AND MITCHELL S. BERGER, M.D.+
M.D.+
Departments of *Radiation Oncology and +Neurosurgery, University of Washington, Seattle, WA Purpose: Controversy regarding the role of adjuvant radiation therapy for the treatment of oligodendroglioma continues to exist. To better define the utility of postoperative irradiation for this tumor, the experience at the Universitv of Washineton was retrospectivelv examined. Methods-and Materials: The histologic samples of 63 patients given the diagnosis of oligodendroglioma were reviewed by a panel of neuropathologists and 41 were classified as pure oligodendroglioma. The two treatment groups included 14 patients treated with surgery only and 27 who received surgery and postoperative radiation and were analyzed using univariate and multivariate analysis with respect to prognostic factors, freedom from relapse, and survival. Results: Univariate statistical analysis of 14 clinical variables showed that a poorer prognosis was associated in patients with high cell density (p = .008), necrosis (p = .017), hemiparesis (p = .026), and papilledema (p = .091), while patients presenting with seizures had a better prognosis (p = .0096). Multivariate analysis showed necrosis (p = .OOl) and hemiparesis (p = .02) to be associated with decreased survival. Multivariate and univariate analysis of the treatment groups found them to be homogenous with respect to prognostic factors. Survival times were significantly longer in the group treated with postoperative irradiation (median survival time 84 vs. 47 months, .032). The 5 and 10 year survival rates were 83% and 46%, respectively, for the irradiated patients compared with 51% and 36% for those treated with surgery alone. Freedom from tumor recurrence times were also longer in irradiated patients (median relapse free time 79 vs. 42 months, .Ol). Conclusion: Based on the results of this study, we recommend continuing the practice of treating oligodendroglioma with postresection irradiation until a prospective multicenter clinical trial is conducted to thoroughly evaluate the role of postoperative irradiation in the treatment of this tumor.
p
=
p=
Brain neoplasm, Oligodendroglioma,
Outcome, Radiation therapy.
INTRODUCTION
variability in radiation dose and technique, and lack of histologic confirmation of the cases. The benefit of postoperative x-ray therapy in the management of patients with cerebral oligodendroglioma has not, therefore, been adequately determined. In this study, we have evaluated the effects of postoperative radiation therapy on 41 histologically confirmed cases of oligodendroglioma treated at the University of Washington between 1956 and 1984.
Oligodendrogliomas comprise about 5% of all primary brain tumors (2, 6, 9, 21, 25). Since the clinical and pathologic characteristics of oligodendroglioma were first described in the 1920s (2,3), many series have been published which attempted to define the optimal treatment modalities for patients with this tumor. Although there is nearly unanimous agreement that surgery enhances survival (10, 11, 23, 34), controversy remains regarding the effect of radiation therapy, with several studies confirming (4, 7, 14, 15, 26, 28, 29, 3 1, 33) and discounting (5, 20, 30, 32) improved survival following surgery and subsequent radiation treatment. Interpretation of these studies is difficult due largely to limited numbers of patients,
METHODS
AND
MATERIALS
Patient population
Sixty-three patients with a diagnosis of oligodendroglioma who were treated at the University of Washington between 1956 and 1984 were evaluated. Forty-one of these
Reprint requests to: Dave E. Gannett, M.D., University of Arizona Health Sciences Center, Department of Radiation Oncology, 1501 N. Campbell Ave., Tucson, AZ 85724. Acknowledgements-We would like to thank Doctors E. Alvord, S. Lofton, and Cheng-Mei Shaw for their detailed review of the histologic sections and Doctors B. Stea and J. R. Cassady for
their review of the manuscript. This work was supported in part by National Institutes of Health, Grants NS07144 and KOSNSO1253-01. Dr. M. S. Berger is an American Cancer Society Professor of Clinical Oncology. Accepted for publication 13 May 1994. 567
I. J. Radiation Oncology 0 Biology 0 Physics
568
patients had sufficient histologic criteria for confirmation of the diagnosis and were included in this study. Two patients with anaplastic oligodendroglioma and two patients who died of postoperative complications within 30 days of surgery were omitted. The decision whether or not to give postoperative radiation therapy varied with the opinions of the surgeon and the consulting radiation oncologist. All patients had a Karnofsky performance status of at least 70. Twenty-seven patients were irradiated postoperatively. The fractionation ranged from 1.5-2.0 Gy per day and the treatment regimen was completed within 3 months of the craniotomy. Radiation was usually administered -2 weeks after surgery when the incision of the scalp was well healed and local subcutaneous fluid collections or edema had regressed. Doses ranged from 40.79-66.00 Gy, (mean 57.03 Gy, median 59.07 Gy; only one patient received < 50 Gy). Five patients received irradiation to the whole brain as a portion of their therapy, while 22 were irradiated to wide local fields. The followup period for survivors ranged from 5-22 years with a mean follow-up of 8.7 years. Two patients were lost to follow-up 4 and 13 years following surgery and were censored in calculation of survival times. Duration of survival and freedom from relapse were calculated from the date of initial resection. Histological evaluation Histologic sections of 63 patients previously coded in the pathology report as “oligodendroglioma” or “probable oligodendroglioma” were replated in 1989 and reviewed by three neuropathologists who were not aware of the clinical history. The diagnosis of oligodendroglioma depended on a uniform population of cells with perinuclear halos, a distinctive network of delicate parenchymal vessels, round to slightly oval nuclei, and a lack of fibrillarity with a pronounced tendency to infiltrate the cerebral cortex (24). Since virtually all oligodendrogliomas contain a component of either neoplastic or reactive astrocytes ( 1, 9, 19) tumors with an astrocytic composition of I 25% were defined as pure oligodendrogliomas ( 16) for the purposes of this study. Twenty cases had a significant fraction of tumor composed of another cell type accounting for greater than 25% of the entire tumor and were classified as mixed tumors. Two cases exhibiting significant necrosis or high cell density were classified as anaplastic. Statistical analysis The hospital records were analyzed with respect to the age at diagnosis, gender, race, location of tumor, extent
Table 1. Presenting
of resection, time from date of surgery to relapse and death, presenting signs and symptoms, and postoperative course. Histologic sections were reviewed for the presence of necrosis, high cell density, and microcysts. Freedom from relapse and survival rates were calculated using the method of Kaplan and Meier ( 13). Associations between categorical variables were evaluated by chi-square test on contingency tables or Fisher’s exact test (two-tailed) as appropriate. Intergroup differences were tested by MannWhitney U-test or t-test as appropriate. Intergroup differences in censored survival times was tested by the Wilcoxon test. Multivariate analysis of prognostic factors used the Cox proportional hazard regression model. Differences in censored survival based on computer-generated curves between groups were tested by the Mantel-Cox test. Differences with p-values I 0.05 were considered statistically significant. RESULTS
Patient population Of the 4 1 patients, 27 were male and 14 were female, a 1.9: 1 ratio in agreement with previous reports ( 12, 15). Thirty-nine of the patients were white and two were black. Patient ages at the time of surgery ranged from 3-78 years (median 45 years). The median age of the 14 patients who did not receive postoperative x-ray therapy was 36 years. The median age of the 27 patients who did receive postoperative irradiation was 42 years. The age distribution of irradiated and nonirradiated patients was similar to previous reports (4, 12, 15,26) with a bimodal distribution favoring ages 3-10 and 45-60. Three patients were < 18 years old. Twenty-five patients (6 1%) presented with some type of seizure disorder. Other presenting symptoms included mental and personality changes, nausea, vomiting, and headaches. The time from onset of symptoms to diagnosis varied from one month to 22 years. Common physical findings included hemiparesis, focal neurologic deficits including weakness, speech changes, and paresthesias, cranial nerve deficits, sensory deficits, and visual deficits. Presenting symptoms and signs are detailed in Table 1. Surgical findings Forty tumors (98%) were located in the cerebral hemispheres and one tumor was located almost exclusively in the thalamus. Of these, the majority (66%) were located either partially or completely in the frontal lobe. Nine tumors infiltrated the temporal lobe, five infiltrated the
signs, and location
for 4 1 patients
with oligodendroglioma Location
Signs
Symptoms Seizures Altered mental status Headaches Nausea/vomiting
symptoms,
Volume 30, Number 3, 1994
25 9 14 3
(61%) (22%) (34%) (7%)
Focal deficit Papilledema Hemiparesis Cranial nerve deficit
8 14 9 4
(20%) (34%) (22%) (10%)
Frontal Parietal Temporal Occipital
27 (66%) 6 (15%) 9 (22%) 2 (5%)
Treatment of oligodendroglioma 0 D. E.
GANNETT
569
et al
Freedom From Progression
Fig. 1. Kaplan-Meier
survival and freedom
from progression
lobe, and two tumors were found in the occipital lobe, in rough proportion to the volume of each lobe (Table 1). According to the operative report, 37 patients had a subtotal resection of their tumor, and, in four cases, the surgeon stated that all of the tumor tissue had been removed. patietal
Survival analysis Complete follow-up was available for 39 of the 41 patients in this study as of July 1989. Thirteen of the 41 patients were alive at the end of the study. The age at death varied from 12 to 82 years and the time from diagnosis to death ranged from 3 months to 11 years. The 5 and 10 year survival rates for the 41 patients with oligodendroglioma were 74% and 41%, respectively (Fig. 1). Freedom from relapse was 6 1% and 18% at 5 and 10 years. Univariate analysis of possible prognostic factors was performed to identify which factors, if any, influenced the time to death of the patients (Table 2). Time to tumor recurrence was not evaluated because of the relatively inexact nature of determining this variable retrospectively. The length of time from the manifestation of the symptom to diagnosis did not significantly affect survival. Age, gender, and race had no significant effect on survival. Seizure as the initial symptom was associated with a longer survival time QJ = .0096), while headaches, nausea and vomiting, and personality change had no significant effect. Of the physical findings, papilledema (p = .091) and hemiparesis (I, = .026) were associated with a decrease in survival time. Analysis of tumor histologic features showed necrosis (p = .017) and high cell density 07 = .008) to be associated with a significant reduction in survival time. Multivariate analysis of prognostic variables revealed necrosis @ = .OO1) and hemiparesis (p = .02) to be associated with decreased survival. Multivariate and univariate analysis of the distribution of prognostic variables between the surgery only and the surgery and x-ray therapy groups
curves for 4 1 patients
with oligodendroglioma.
showed no apparent heterogeneity between the groups (Table 3). A significant increase in survival time was evident in patients treated with postoperative irradiation. Five and 1O-year survival rates in the irradiated patients were 83% and 46%, respectively, and 5 1% and 36% in the nonirradiated patients (p = .032, Fig. 2). The median survival time of the irradiated patients was 7.0 years, compared to 3.9 years for those treated only with surgery. Time to tumor recurrence was also longer in the irradiated patients. Five and IO-year freedom from relapse rates were 74% and 17%, respectively, in the irradiated group, compared to 37% and 19% in the surgery only group @ = .Ol, Fig. 3). The median relapse free times were 79 months in the irradiated patients and 42 months in the surgery only group.
Table 2. Univariate patients
analysis of factors evaluated with oligodendroglioma
for 41
Variable
Univariate p value
Gender Age Race Headache Mental changes Seizures at presentation Nausea/vomiting Focal deficits Hemiparesis Papilledema Cranial nerve palsies Necrosis High cell density Microcysts
NS 0.18 NS NS NS 0.0096 NS 0.13 0.026 0.09 1 NS 0.017 0.008 NS
NS = Not significant.
570
I. J. Radiation Oncology 0 Biology 0 Physics
Table 3. Comparison of prognostic variables between treated with or without adjuvant radiotherapy Variable Number of patients Gender (male/female) Mean age (years) Race (black/white) Symptoms Seizure Mental change Headache Nausea/vomiting Signs Focal deficit Papilledema Hemiparesis Cranial nerve palsy Histopathology Necrosis High cell density Microcysts Extent of resection Subtotal Total Status (alive/dead)
Irradiated
Nonirradiated
patients
p-value
27 1819 39.2 l/26
14 915 40.0 l/13
NS NS NS NS
14 7 10 2
11 2 4 1
NS NS NS NS
5 9 6 3
3 5 3 1
NS NS NS NS
2 8 12
2 5 7
NS NS NS
25 2 10/17
12 2 3/11
NS NS 0.03
NS = Not significant.
Of the four cases in which the surgeon stated that the tumor had been completely removed, two patients are still alive at 13 and 22.6 years after surgery without evidence of recurrence, and the other two patients died 7.6 and 3.8 years postoperatively. The patients living 13+ and 7.6 years postoperatively had both received adjuvant radiation. All 28 patients who died during the follow-up period did so from complications related to the recurrent tumor. Histologic sections were available from eight patients who
I
Fig. 2. Survival (XRT).
rates for patients
Volume 30, Number 3, 1994
underwent a second surgical resection for recurrent tumor. Only three of these cases showed pure oligodendroglioma at the time of recurrence. In the remaining five cases, two were diagnosed as glioblastoma multiforme, two showed mixed gliomas, and one showed an moderately anaplastic astrocytoma without evidence of oligodendroglioma. Five patients received a second course of radiation therapy as part of their treatment for recurrent tumor; two died as a direct result of the tumor within a year of retreatment and three are alive without evidence of recurrence 2.2, 5.4, and 6.5 years after second treatment. The differences in survival between the two groups were not significantly altered by either censoring or excluding the patients who underwent a second surgical resection and or received additional radiation. Toxicity One patient developed a complication of treatment, unrelated to the tumor. She had a subtotal resection of a left frontal oligodendroglioma at 30 years of age. Postoperatively, she was irradiated with 45 Gy to the whole brain with a 15 Gy boost to the tumor bed. Two years later, a computed tomography (CT) scan showed evidence of recurrent tumor at the site of prior resection, and this was treated with an additional course of x-ray therapy consisting of 45 Gy to the tumor area. Approximately 5 years after the second course of irradiation, she was diagnosed with a Grade 2 (RTOG/EORTC late radiation morbidity scoring scheme) radiation induced proliferative retinopathy. A CT scan at that time showed no evidence of tumor recurrence to account for the visual loss and there was no evidence of diabetes. The retinopathy had worsened to Grade 3 at last follow-up, eight years after initial therapy. Three patients received some form of chemotherapy as part of their treatment. One died from recurrence of the
..~:‘..r’......_____
_____________.__._..--~’
NoXRT
with oligodendroglioma
treated
with and without
postoperative
x-ray therapy
Treatment of oligodendroglioma 0 D. E.
GANNETT
571
et al.
1
1
,
1
I-
O
5
10
15
20
YeUS
Fig. 3. Freedom x-ray therapy.
from progression
rates for patients with oligodendroglioma
tumor 24 months following therapy, one had a recurrence 7 years postoperatively, and the other patient is still alive without evidence of recurrence 9.2 years following surgery.
primary
DISCUSSION Since oligodendrogliomas, like other gliomas, tend to have an infiltrative growth pattern, complete resection is only rarely achieved. As with other gliomas, the tendency of oligodendroglioma to remain contained in the cerebrospinal axis suggests that irradiation as a regional therapy could potentially control these tumors. Consequently, recent literature (31, 35, 36) has recommended radiotherapy as adjuvant therapy to achieve local control and our results support the efficacy of this mode of treatment. In this series of 41 patients, the median survival time was significantly longer when radiation therapy was given postoperatively (p = .032). This group also had better 5 and lo-year survival rates (83% and 46% vs. 5 1% and 36%) when compared with the patients treated with surgery alone. Freedom from relapse times were also longer in the irradiated patients (p = .Ol). Given that the two patient groups were homogenous with respect to prognostic variables, this data suggests a benefit for postoperative radiation for patients with this tumor histology. Previous reports discussing the effects of radiation treatment on the course of oligodendroglioma have been hampered by small numbers of cases, lack of pathologic confirmation of the diagnosis, inadequate statistical analysis, and poor control of the population differences. In a review of the Cleveland Clinic experience in 1983, Reedy et al. (20) found that radiotherapy did not have a significant effect on the survival of 48 patients evaluated. However, radiation doses varied from 22-65 Gy, indicating significant heterogeneity in therapy with some patients
treated with and without postoperative
obviously undertreated. Furthermore, there was no pathologic review or clear evaluation of the distribution of histologic variables. The typical radiation dose used for malignant intrinsic brain tumors is usually greater than 40 Gy, with typical fractionation schemes delivering 1.82.0 Gy 5 days per week. Deviation from this type of treatment regimen may have a significant impact on tumor response. The comprehensive review by Bullard (5) which uses a rigid and thorough statistical analysis raises similar concerns. They too, concluded that there was no survival benefit for patients given postoperative irradiation following resection. However, 13 patients received < 40 Gy and 18 patients received anywhere from 40-60 Gy. Additionally, 17 patients were treated with orthovoltage technique. The histologic criteria for inclusion into the study were somewhat questionable as only one pathologist reviewed the slides with exclusion criteria for “lesions clearly containing neoplastic fibrillary astrocytes” as mixed gliomas. While limited by the constraints of a retrospective format, this study is strengthened by a number of factors. The replating and reevaluation of the pathology specimens by a panel of three experienced neuropathologists for all previously diagnosed oligodendrogliomas at our institution gives us reasonable confidence in the histologic diagnosis. University and departmental records were available for all confirmed cases and only two patients were lost to follow-up. In addition, an elaborate statistical analysis of the treatment groups show them to be homogenous with respect to prognostic factors. However, our study is limited by the relatively small size of the patient population, resulting in limited power to demonstrate a causal effect. In addition, a wide variety of field sizes, radiation equipment, doses and diagnostic strategies were used over the 28-year period of this study. The only way of circum-
572
I. J. Radiation Oncology 0 Biology 0 Physics
Volume 30, Number 3, 1994
Table 4. Review of previous series evaluating the impact of radiotherapy on oligodendroglioma Author, year
n
Survival rt
Survival no rt
Current study Wallner (33) 1988 Bullard (5) 1987 Lindegaard (14) 1987 Reedy (20) 1983 Pailas ( 18) 1982 Chin (7) 1980 Sheline (29) 1977 Weir (34) 1968 Roberts (22) 1966 Shenkin (30) 1964 Richmond (2 1) 1959 Horrax (12) 1951
41 42 71 170 48 63 35 36 63 33 11 22 22
10 years, 46% 10 years, 56% 5 years, 54% 5 years, 36% 5 years, 5 1% 92 month median 5 years, 100% 5 years, 85% 6.4 years mean 6.8 years mean 65 month median 5 years, 53% 87 month median
10 years, 36% 10 years, 18% 5 years, 48% 5 years, 27% 5 years, 54% 36 month median 5 years, 82% 5 years, 3 1% 4 years mean 5.2 years mean 13 month median (No patients) (No patients)
Dose GY) 40-66 > 45 < 40-62 20-60 22-65 > 30 53-70 > 20 NA > 60 30-65 NA NA
n = Number of patients; rt = Radiation therapy; NA = Not available.
venting the weaknesses of this and other retrospective studies on cerebral oligodendrogliomas, is to conduct a randomized prospective trial in which surgical resection with or without adjuvant radiation is compared. Contributing to the controversy regarding the role of radiation in the treatment of oligodendroglioma is the fact that the tumor pathology is not always clearly defined. Nearly all oligodendrogliomas contain areas of other glial cell differentiation. Most often this is limited to populations of reactive astrocytes and oligodendroglia, but may sometimes include neoplastic ependymal cells ( 1, 9, 19). Rubenstein (24) has estimated that half of all oligodendrogliomas contain some component of another tumor type. This creates confusion in pathologic diagnosis since there are no accepted criteria as to how much of a second neoplastic cell type is required to change the diagnosis from oligodendroglioma, and specific cell markers such as galactocerebroside are not applicable in paraffin fixed sections. In this study, we used the criteria proposed by Mork ( 16) in classifying pure oligodendrogliomas as those with < 25% astrocytic component. Recent research has provided insight to the etiology of the heterogeneity of oligodendrogliomas. Distinguishing the mature oligodendrocyte from an astrocyte is the presence of intracellular microtubules rather than glial filaments, therefore oligodendrocytes are negative for glial fibrillary acidic protein (GFAP). However, it has been noted that some immature human oligodendroglia and neoplastic oligodendroglia
may transiently express GFAP (17). This, coupled with evidence demonstrating that fetal radial glia give rise to cells of both astroglial and oligodendroglial lineage (8) suggest the possibility that the glial precursor cell is the target of the initial event preceding the phenotypic expression of neoplastic transformation. This could explain the high prevalence of astrocytic cell elements found in oligodendrogliomas, as well as the histologic progression of this tumor observed in this and other studies (12, 27).
CONCLUSION Based on this and other reported studies (Table 4), we offer the following recommendations regarding the indications for radiation therapy in the treatment of supratentorial oligodendroglioma: (a) Postoperative irradiation is suggested following subtotal resection of the tumor. A recent meta-analysis of the cancer literature (3 1) also suggests a benefit for patients with oligodendrogliomas who have undergone a subtotal resection. (b) Because of the possible change in histology documented in some recurrences and the similar appearance of radiation necrosis on imaging studies, it is advisable to obtain a biopsy prior to consideration of further radiation and or chemotherapy. (c) A randomized prospective trial will be necessary to definitively establish the role of adjuvant radiotherapy in the management of oligodendroglioma.
REFERENCES Aparicio, S. G.; Lumsden, C. E.; Jennings, M. Electron microscopy of &al reactions observed in CNS cultures. Brain Res. 7:268; 1968. Bailey, P.; Bucy, P. C. Oligodendrogliomas of the brain. J. Pathol. Bacterial. 32:735-751; 1929. Bailey, P.; Cushing, H. A classification of the tumors of the glioma group on a histogenetic basis with a correlated study of prognosis. Philadelphia: JB Lippincott; 1926: 175- 176. Bouchard, J. Effects of irradiation in treatment of intracranial gliomas-Treatment results by histologic groups. In:
Radiation therapy of tumors and diseases of the nervous system. Philadelphia: Lea & Febiger; 1966:78- 118. Bullard, D. E.; Rawlings, C. E., III; Phillips, B.; Cox, E. B.; Schold, S. C.; Burger, P. C.; Halperin, E. C. Oligodendroglioma-An analysis of the value of radiation therapy. Cancer 60:2179-2188; 1987. Burger, P. C.; Vogel, F. S. Surgical pathology of the nervous system and its coverings, 2nd ed. New York: John Wiley and Sons; 1982:335-349. Chin, H. W.; Hazel, J. J.; Kim, T. H.; Webster, J. H. Oli-
Treatment of oligodendroglioma 0 D. E. GANNETTet al.
8.
9. 10.
Il. 12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
godendrogliomas-A clinical study of cerebral oligodendrogliomas. Cancer 45:1458-1466; 1980. Choi, B. H.; Kim, R. C.; Lapham, L. W. Do radial glia give rise to both astroglial and oligodendroglial cells? Devel. Brain Res. 8:119; 1983. Escalona-Zapata, J. Uncommon oligodendrogliomas. Acta Neuropathol. 7(Suppl.):94; 198 1. Grant, F. C. A study of the results of surgical treatment in 2326 consecutive patients with brain tumors. J. Neurosurg. 13:479; 1956. Hart, W.; Petito, M.; Earle, W. Mixed ghomas. Cancer 33: 134; 1974. Horrax, G.; Wu, W. Q. Postoperative survival of patients with intracranial oligodendroglioma with special reference to radical tumor removal; A study of 26 patients. J. Neurosurg. 8:473-479; 195 1. Kaplan, E. L.; Meier, P. Nonparametric estimation from incomplete observations. J Am. Stat. Assoc. 53:457-481; 1958. Lindegaard, K. F.; Mork, S. J.; Eide, G. E.; Halvorsen, T. B.; Hatlevoli, R.; Solgaard, T.; Dahl, 0.; Ganz, J. Statistical analysis of clinicopathological features, radiotherapy, and survival in 170 cases of oligodendroglioma. J. Neurosurg. 67:224-230; 1987. Mork, S. J.; Lindegaard, K. F.; Halvorsen, T. B.; Lehman, E. H.; Solgaard, T.; Hatlevoli, R.; Harvel, S.; Ganz, J. Oligodendroghoma: Incidence and biological behavior in a defined population. J. Neurosurg. 63:881-889; 1985. Mork, S. J.; Halvorsen, T. B.; Lindegaard, K. F.; Eide, G. E. Oligodendroglioma: Histologic evaluation and prognosis. J. Neuropathol. Exp. Neurol. 45:65-78; 1986. Ogawa, H.; Satoi, Y.; Tateishi, K.; Kitamura, K. Transient expression of glial fibrillary acidic protein in developing oligodendrocytes in vitro. Devel. Brain Res. 18: 133; 1985. Pailas, J. E.; Combalbert, A.; Bernard-Badier, M.; Bille, J.; Frank, R. Etude sur l’evolutiondes oligodendrogliomes de l’encephale. A propos d’une serie operatoire de 34 cas. Acta Neural. Psych. Belg. 64:537-55 I; 1964. Ravens, J. R.; Adamkiewicz, L. L.; Groff, R. A. Cytology and cellular pathology of oligodendrogliomas of the brain. J. Neuropathol. Exp. Neurol. 14:142; 1955. Reedy, D. P.; Bay, J. W.; Hahn, J. F. Role of radiation therapy in the treatment of cerebral oligodendroglioma: An analysis of 57 cases and a literature review. Neurosurgery 13:499-503; 1983. Richmond, J. J. Malignant tumours of the central nervous system. In: Raven, R. W., ed. Cancer 5. London: Butterworth; 1959:375-389.
573
22. Roberts, M.; German, W. J. A long term study of patients with oligodendrogliomas. Follow-up of 50 cases, including Dr. Harvey Cushing’s series. J. Neurosurg. 24:697-700; 1966. 23. Rossi, G. R.; Feoli, F.; Fernandez, E. The role of surgery in the treatment of supratentorial brain gliomas. In: Paoletti, P., Walker, M. D., Butti, G., et al. eds. Multidisciplinary aspects of brain tumor therapy. Amsterdam: Elsevier North Holland; 1979:155-163. 24. Rubenstein, L. J. Tumors of the central nervous system. Atlas of tumor pathology, 2nd series, fascicle 6. Washington, DC: Armed Forces Institute of Pathology. 25. Russel, D. S.; Rubenstein, L. J. Pathology of tumors of the nervous system, 4th ed. London: Edward Arnold; 1977: 195203. of intracranial 26. Scanlon, P. W.; Taylor, W. F. Radiotherapy astrocytomas-analysis of 4 17 cases treated from 1960- 1969. Neurosurg. 5:301-308; 1979. R.; Ansuini, G.; Pisani, A. E.; 27. Scarpelli, M.; Monitroni, Rychlicki, F.; Papo, I.; Mariuzzi, G. M. The value of morphometry in prediction of survival in oligodendrogliomas. Clin. Neuropathol. 6:155; 1987. 28. Shaw, E. G.; Scheithauer, B. W.; O’Fallon, J. R.; Tazelaar, H. D.; Davis, D. H. Oligodendrogliomas: The Mayo Clinic experience. J. Neurosurg. 76:428-434; 1992. 29. Sheline, G. E.; Boldrey, E.; Karlsberg, P.; Phillips, T. L. Therapeutic considerations in tumors affecting the central nervous system; Oligodendrogliomas. Radiology 82:84-89; 1964. therapy on oli30. Shenkin, H. A. The effect of roentgen-ray godendrogliomas of the brain. J. Neurosurg. 22:57-59; 1965. 31. Shimizu, K. T.; Tran, L. M.; Mark, R. J.; Selch, M. T. Management of oligodendrogliomas. Radiology 186:569572; 1993. 32. Uihlein, A.; Colby, M. Y.; Layton, D. D.; Parsons, W. R.; Carter, T. L. Comparison of surgery and surgery and irradiation in the treatment of supratentorial gliomas. Acta. Radiol. Ther. 5:67-78; 1968. 33. Wallner, K. E.; Gonzales, M.; Sheline, G. E. Treatment of oligodendrogliomas with or without postoperative irradiation. J. Neurosurg. 68:684-688; 1988. An analysis of 63 cases. J. 34. Weir, B. E. Oligodendrogliomas: Neurosurg. 29:500-505; 1968. 35. Wilkins, R. H.; Renagachary, S. S., eds. Neurosurgery. New York: McGraw-Hill; 1985. 36. Youmans, J. R., ed. Neurological surgery, 2nd ed. Philadelphia: WB Saunders; 1982.