Immediate surgical intervention and subsequent radiation therapy (RT) are advocated by some physicians. However, it remains unclear whether early intervention prolongs survival or changes the natural history of the disease, especially in patients who present with a seizure and a normal interictal examination.
Bailey and Cushing first proposed that gliomas were derived from transformation of normal glial cells at some point during their development. Astrocytomas were tumors with the appearance of astrocytes, while oligodendrogliomas had the appearance of oligodendrocytes. A grading system based upon histologic features was not incorporated into this classification.
Kernohan estimated the prognosis of glial type tumors based upon the extent of observed anaplastic features such as mitoses, endothelial proliferation, cellular atypia, and necrosis. Such a classification proved of prognostic value in these tumors, especially astrocytomas.
Both of these classification schemes persist in revised form. The first forms the foundation for the WHO classification schema, while the second continues to be used in modified form.
LGG is a frequently used term, but it is not clearly defined by either of these classifications. Although a formal definition is not available, LGG may be considered a hybrid term which describes a spectrum of primary brain tumors that are composed of cells which histologically resemble one or more differentiated types of macroglial cells (diffuse and fibrillary astrocytes, oligodendrocytes, ependymal cells) and which have no evidence of anaplasia (in the Kernohan scheme, this would encompass grades I and II).
These tumors have also been referred to as "benign" gliomas, but this is a misnomer. Although they have a more favorable prognosis than glioblastomas (GBMs), LGGs are only occasionally associated with prolonged (>10 years) survival and, over their course, frequently develop characteristics similar to more aggressive brain tumors.LGGs occur approximately one-fifth as frequently as the more malignant gliomas. They can develop in any central nervous system (CNS) location, although studies are usually directed to a particular location (ie, LGGs occurring in the cerebral hemispheres, optic pathways, brainstem, etc). Diffuse astrocytoma is the most common histologic subtype of LGG. All diffuse astrocytomas widely infiltrate surrounding neural tissues. They generally occur in adults with a peak age incidence in the late thirties, approximately twenty years younger than GBM. The most common location from which they arise is the cerebral hemispheres; the most common presenting symptom is seizure.
Diffuse astrocytomas can be subdivided further into three types: fibrillary, protoplasmic, and gemistocytic. The prognosis for those tumors varies greatly, independent of histology. Ten year survivals can occur in up to 20 percent of patients, while other tumors behave more aggressively with survivals indistinguishable from anaplastic astrocytomas.
Fibrillary — Fibrillary tumors typically arise from deeper white matter. They are firm and rubbery due to the neuroglial fibrils that they produce. Microscopically, fine and coarse neuroglial fibrils form a matrix between the cells.
Protoplasmic — Astrocytoma cells can resemble protoplasmic astrocytes; these tumors tend to arise in the cerebral cortex.
Gemistocytic — Gemistocytic cells appear exceptionally large and are densely packed. In many cases, distention of their cell bodies lends them a globoid appearance. Although it has been suggested that these tumors are better classified as anaplastic astrocytomas, the reported survival figures and prognostic determinants generally mirror those for diffuse fibrillary astrocytomas
These distinctions are not absolute. Among gemistocytic tumors, for example, both "pure" and "mixed" subtypes have been noted. The former has been defined as a glial tumor with more than 60 percent gemistocytes and a background of fibrillary astrocytes; the latter has been defined as containing 20 to 60 percent gemistocytes and a background of anaplastic astrocytes.
JPAs most frequently arise in the cerebellar hemispheres and around the third ventricle. However, they also are encountered in the cerebral hemispheres. They are frequently cystic and well-demarcated. Although almost always circumscribed on imaging studies, they enhance on CT and MR scanning because their blood vessels undergo chronic glomeruloid degenerative hyalinization. A different mechanism, endothelial proliferation, is responsible for enhancement in anaplastic astrocytomas and GBMs.
JPAs are distinguished from other LGGs because of their circumscribed nature, rare progression to GBM, and very favorable prognosis. Malignant transformation of these tumors to higher grades of astrocytoma occurs in less than 5 percent of cases. It may happen more frequently in patients who have been previously irradiated , raising a question of whether this modality should be avoided in these patients. Because it is potentially curable, the goal of surgery should be complete removal. However, multicentric spread may occur, especially with hypothalamic tumors.
Most oligodendrogliomas are slow-growing, hypodense, sometimes calcified tumors that are usually, but not always, nonenhancing with gadolinium. They constitute approximately 5 percent of intracranial adult gliomas, less if only those tumors with just an oligodendroglioma component are considered (ie, not a mixed tumor, see below). They are more likely to hemorrhage than other low-grade gliomas.
Although considered to have a somewhat more favorable prognosis than diffuse astrocytomas, the behavior of oligodendrogliomas is similar, in that younger patients, patients who undergo complete resections, and those with low grade (grade II) tumors, appear to have a better prognosis. Grade III (anaplastic) oligodendrogliomas have a worse outcome. In one series of 100 patients with oligodendroglioma diagnosed over a seven year period, the 2, 5, and 10-year survival rates for those treated for grade II lesions were 88, 88, and 85 percent, respectively. Results were worse for patients with initially low-grade tumors that underwent anaplastic transformation during the follow-up period (2, 5, and 10-year survival 79, 64, and 42 percent, respectively), and the corresponding values for those presenting with de novo grade III tumors were 43, 16, and 15 percent, respectively.
Anaplastic oligodendrogliomas and their treatment are considered elsewhere.
Although uncommonly encountered, they are important to recognize because of their generally good prognosis; further therapy beyond surgery, even when resections are incomplete, is usually not necessary.
Ependymomas, tend to occur in young patients and arise in the area of the fourth ventricle. However, they are generally not considered LGGs even though they are composed of cells that have the appearance of ependymal cells, glial cells that line the ventricular cavity.LGG is highly likely when a patient presents with a transient neurologic disturbance consistent with seizure, and imaging reveals an unenhancing hemispheric mass that produces little mass effect. Persistent symptoms generally reflect either the location of the tumor (eg, hemiparesis, ataxia) or the result of increased intracranial pressure (eg, headache, change in mental status). These relationships can be illustrated by the following observations:
Among 1028 patients with brain tumor, the prevalence of epilepsy was 49 percent in patients with GBM, 69 percent in anaplastic glioma, and 85 percent in LGG.
Among 831 consecutive patients with brain tumor, tumor contrast enhancement on CT was present in 96 percent of GBMs, 87 percent of high-grade gliomas, 57 percent of anaplastic gliomas, and 21 percent of LGGs. As noted above, JPAs represent an exception to this general rule for LGGs, since they typically show contrast enhancement.
Cranial imaging can be supplemented by the use of positron-emission tomography (PET) scanning. Because LGGs are characterized by glucose hypometabolism, PET images showing diffuse hypometabolism may support a decision to defer treatment, while the presence of hypermetabolic areas may indicated the presence of high-grade tumors, and signal the need for biopsy or treatment.
Since one can predict the subsequent clinical course with fair reliability from these parameters alone (ie, isolated transient symptoms and nonenhancement), physicians often delay making a histologic diagnosis until symptoms or imaging worsen. Whether such a delay in diagnosis and treatment significantly affects outcome is controversial and is still the subject of sometimes heated debate.
In order to decide whether delaying treatment is detrimental, it would be useful to have controlled studies in which to assess outcome. Unfortunately, studies which address immediate versus delayed surgery and radiation therapy (RT, which is the conventionally used adjuvant therapy) are not available. However, several retrospective studies have addressed outcome after histologic diagnosis is made. Not surprisingly, it is difficult to arrive at a consensus based upon retrospective data. Nevertheless, several general observations can be made from these reports concerning prognosis and progression.
The generally favorable prognosis in JPA can be illustrated by the following observations. In one series of 34 patients treated by surgical removal of the tumor, 82 percent had a good outcome at a mean follow-up of 17 years. In another report of 51 patients with a median follow-up of 15 years, 82 percent were alive at 10 years; 89 percent of these patients were fully active. Ten-year survival was higher in patients who underwent gross total or radical subtotal tumor resection (100 versus 74 percent in patients with subtotal tumor resection or biopsy alone.
Although certain less common LGGs such as gangliogliomas (see above) and pleomorphic xanthoastrocytomas might also have a very favorable prognosis, the prognosis of the other pathologic subtypes of LGGs is similar.
Among the more common LGGs, most patients eventually deteriorate. In one series of 53 adults with supratentorial astrocytomas, the majority of whom had either subtotal or gross total resection and postoperative RT, the median survival was 7.3 years with a five-year survival of 64 percent. Deaths were generally due to evolution into an anaplastic astrocytoma or GBM. The median time to recurrence was 4.5 years after the original surgery, and the survival from recurrence was only 12 months. In this and other studies, young age, good performance status, and the clinical presentation of an isolated seizure with an otherwise normal examination were generally good prognostic features. On the other hand, CT contrast enhancement of the original tumor appears to predict a high likelihood of progression to a malignant lesion
Mixed tumors have a similar prognosis to diffuse astrocytomas. In one series of 71 patients, for example, the five, ten, and fifteen year survival rates were 55, 29, and 17 percent, respectively; these values were significantly lower than age and sex matched controls. All patients were treated with surgery and 66 with postoperative RT. Tumor grade, using the Kernohan method, was the major prognostic feature. However, even patients with grade 1 and 2 tumors had a median survival of only 6.3 years. Progression was typically associated with anaplastic transformation. Other studies have noted lower survival rates although pathologic grade was still important.
More recent data suggests that patients with oligodendroglioma or mixed glioma have a prolonged natural history. In a retrospective review of 106 patients with low-grade oligodendroglioma or mixed glioma, the majority of whom had either biopsy alone or subtotal resection, the median time to progression was five years (range 0.5 to 14.2), and the median overall survival was 17 years. Some of this long survival may reflect the effect of lead time bias since the average time from symptoms to diagnosis was less than one month, which is shorter than in most other series. Tumor progression eventually developed in 68 percent of patients. Disease-free and overall survival did not differ with immediate or deferred therapy (see below).n one series of 25 such patients, the median survival was 8.2 years. None of the patients died from progressive low-grade disease; seven died as a consequence of their tumors dedifferentiating into a more malignant astrocytoma or GBM, with a median time of approximately five years after the diagnosis.
Similar findings have been noted in other studies, suggesting that one could expect a period, usually in the range of five to seven years after diagnosis, of radiographic and clinical stability after which tumors take on characteristics indistinguishable from more anaplastic astrocytomas and glioblastomas. For patients with diffuse astrocytomas or oligodendrogliomas, progression will occur in more than 50 percent of cases.
The seizures associated with hemispherically located LGGs can be a source of major morbidity and are more refractory to medical management than idiopathic epilepsy. In these instances, consideration should be given to performing surgery. Over one-half of patients with a short preoperative duration of seizures (less than one year) will have near complete or total resolution of seizures after tumor resection. Although it is difficult to definitively demonstrate that it is superior to a standard tumor resection, most clinicians advocate seizure-type surgery with intraoperative recording to localize the seizure focus in these instances.several studies suggest that the more extensive the tumor resection, the better the prognosis
However, specific conclusions cannot be drawn from these series as they all suffer from design-related limitations, particularly small size and the inclusion of histologic tumor types with varying natural histories. Furthermore, it is difficult to distinguish from these retrospective studies whether a polar tumor that presents as an isolated seizure or as a completely resectable lesion has a better prognosis irrespective of what treatments are offered. It therefore remains unclear whether attempting complete resection affords a better prognosis than a stereotactic biopsy, although it may be more effective at controlling morbidity.
In the absence of controlled prospective trials, there are several approaches that have been taken. The aggressive approach recommends extensive resection at the time of diagnosis, regardless of symptoms. This approach is based upon the opinion that complete resection may improve survival, and decrease the rate of malignant transformation. Intraoperative magnetic resonance imaging (MRI) has been used to help the surgeon identify small lesions and to improve the extent of tumor resection .
Subsequent to surgery, all patients with diffuse astrocytomas, oligodendrogliomas, or mixed tumors undergo RT with dosages to the local tumor bed exceeding 50 Gy.
Others, citing the lack of evidence that early intervention is beneficial, prefer to delay aggressive treatment until there is radiographic evidence of tumor growth, intractable seizures, progressive neurologic impairment, or malignant transformation .
One report retrospectively addressed the issue of deferring therapeutic intervention by comparing the clinical course in two groups of patients: 26 who presented with a transient neurologic event and radiographic evidence of LGG and received no treatment (except anticonvulsants); and 20 who presented similarly but underwent immediate intervention. Fifteen patients (58 percent) in the deferred group required surgery or RT at a median of 29 months. There was no difference in the rate of malignant transformation, survival, or quality of life between the two groups.
In a second retrospective report of 106 patients with oligodendroglioma or mixed glioma, 19 patients had gross total resection, 41 had subtotal resection, 28 biopsy only, and 18 had no surgical intervention. Following initial surgical management, 68 patients underwent observation alone, 20 had RT, 12 had chemotherapy and 6 had chemoradiotherapy. The median time to progression was 4.2 years in the observation group, and 3.9 years if the 18 patients without initial surgical diagnosis were excluded. The respective values for the patients treated with postoperative RT, initial chemotherapy, and combined chemoradiotherapy were 5.7, not reached after a median follow-up of 5.5 years, and 8.6 years. The authors concluded that there were no significant differences between the groups; a similar conclusion was reached for overall survival. Thus, deferral of treatment did not appear to compromise survival or outcome. However, treatment-related toxicity was prominent. Delayed cognitive impairment was noted in 15 of the 62 patients who received RT during the course of their illness, and radiation necrosis was diagnosed in nine patients at a median of 52 months following therapy. Overall, one-third of patients undergoing RT had significant late neurotoxicity and one-half of patients receiving chemotherapy had significant acute toxicity. retrospective studies assess the effect of postoperative RT in patients with LGGs involving the cerebral hemispheres. Two noted a small but statistically significant survival benefit, one found a significant survival benefit only in the subset of patients older than 40 years of age who underwent subtotal resection or biopsy and one found an increase in the median time to anaplastic transformation from 3.7 to 5.4 years in patients receiving RT
At least three randomized controlled trials have addressed the impact of RT on low grade astrocytomas. In one trial conducted by the European Organization for Research and Treatment of Cancer (EORTC), 311 patients with supratentorial LGG were randomly assigned following biopsy or subtotal excision to receive either RT (54 Gy in six weeks) or no therapy until progression. Gross total resection was accomplished in 62 percent of patients, and the remainder had biopsy only. With a median follow-up of five years, postoperative RT significantly prolonged progression-free survival (44 versus 37 percent), but had no impact on overall survival (63 versus 66 percent).
Two other randomized studies have failed to show a survival benefit from higher RT doses in patients with LGG
In the first trial, 340 patients were randomly assigned to 45 Gy or 59.4 Gy and then followed for a median of 74 months. There was no significant difference in survival (58 and 59 percent) or progression-free survival (47 and 50 percent) between the two groups.
The second report compared 50.4 Gy with 64.8 Gy in 203 patients with supratentorial LGG. There was a trend toward better two- and five-year survival in the lower dose group, and the incidence of severe radiation necrosis was twice as high in the high-dose arm (5 versus 2.5 percent).
In phase II studies, hyperfractionated radiation therapy is feasible for incompletely resected lesions with mild to moderate toxicity; however, improved survivals have not been observed. Postoperative fractionated stereotactic radiotherapy also offers an alternative to conventional RT, permitting sparing of irradiation to normal brain tissue.
Most authorities agree that radiation therapy is not indicated in the initial management of juvenile pilocytic astrocytoma. In other LGGs, postoperative RT has been recommended after complete resection by some, whereas others advise that it be withheld until there is evidence of tumor recurrence.
Another potential side effect is neuropsychologic symptoms including cognitive decline. It has been suggested that this complication is overrated, and that cognitive decline is a function not only of RT, but also the tumor itself and antiepileptic drug use. This issue was addressed in several studies:
A single institution report of 51 patients undergoing resection for cerebral LGGs compared 28 who received postsurgery RT (but no chemotherapy or secondary surgery), and 23 who received no further therapy, including RT. At a mean follow-up of seven years, the radiated group performed significantly worse on tests of cognitive function, and had more severe leukoencephalopathy, which correlated with poor memory.
A second EORTC study assessed the effects of RT dose on cognitive decline. An increased frequency of long-term adverse effects was noted in patients treated with a higher dose (59.4 Gy in 6.5 weeks versus 45 Gy in 5 weeks).
A low frequency of cognitive decline in patients without tumor progression was noted in a series of 203 adults with supratentorial LGGS who underwent focal RT and were periodically assessed with Folsom Mini Mental Status Exam (MMSE). In patients without tumor progression, significant deterioration from baseline occurred at years 1, 2, and 5 in 8.2, 4.6, and 5.3 percent of patients, respectively, while most patients with an abnormal baseline MMSE score (<27) experienced significant worsening. However, the lack of discrimination of this tool for neurocognitive assessment may have underestimated the number of affected patients.
RT itself may promote malignant transformation. Although early studies suggested that this possibility was remote, recent reports suggest a higher incidence of this development in patients with JPA who receive RTvariable results. One controlled trial in patients with incompletely resected LGGs that compared CCNU and radiotherapy to radiotherapy alone failed to show a benefit for chemotherapy in either response rate or median survival. Assessing benefit from chemotherapy can be difficult to judge using standard response criteria, since LGGs are typically nonenhancing, ill-defined, and difficult to measure. Unequivocal clinical improvement that is unaccompanied by a significant reduction in tumor size may indicate benefit, while in some studies, functional imaging (eg, with thallium-201 SPECT) appears to be a better predictor of tumor control and patient survival than CT or MRI. Nevertheless, the end points of most clinical trials are objective response rate and time to tumor progression.
In one study, for example, 52 patients with oligodendroglioma or mixed oligoastrocytoma were treated with up to six cycles of PCV Thirty-three patients responded, nine completely (objective response rate 63 percent). Response rates were higher in patients with necrosis in the histologic specimen (76 versus 45 percent). The median time to progression was 25 months in complete responders, 12 months in partial responders, and 7 months in nonresponders.
A somewhat lower response rate was noted in a second series, in which 28 patients with residual tumor on MRI scan following either biopsy or subtotal resection of low grade oligodendroglioma or oligoastrocytoma received up to six cycles of modified PCV (procarbazine 75 mg/m2 orally days 8 to 21, lomustine 130 mg/m2 orally on day 1, vincristine 1.4 mg/m2 days 8 and 29, every eight weeks) followed by RT within 10 weeks, or immediately in the absence of a chemotherapy response. Eight (29 percent) had objective tumor regression prior to RT.
PCV chemotherapy may provide a reasonable alternative to radiation for the initial treatment of large oligodendrogliomas; in such cases, deferred radiation may postpone or avoid the development of delayed leukoencephalopathy. However, randomized trials will be needed to confirm the efficacy of this approach.
LGGs are also responsive to second-line as well as first-line temozolomide
In one series of 46 patients (which included 20 with oligoastrocytoma and 16 with astrocytoma), only one of whom had received prior PCV, 11 had a complete response and 17 had a partial response (objective response rate 63 percent). Response rates for astrocytoma and oligodendroglioma were similar, and the one year progression-free survival rate for all patients was 76 percent.
A somewhat lower objective response rate was reported in a second report of 60 previously untreated patients with low-grade oligodendroglioma or oligoastrocytoma. Although clinical improvement was noted in 51 percent (particularly in those with uncontrolled epilepsy), only 17 percent had an objective partial response. Nevertheless, the one year progression-free survival rate was similar to the prior study, 73 percent.
Both grade II and grade III (anaplastic) oligodendrogliomas and oligoastrocytomas that harbor allelic loss of chromosome 1p or combined allelic losses of 1p and 19q are associated with increased sensitivity to chemotherapy, and longer recurrence-free survival and overall survival. These deletions can be detected by fluorescent in situ hybridization (FISH) analysis of paraffin-embedded tissues.
Was tissue diagnosis necessary prior to instituting treatment?
Was biopsy necessary if the patient were going to be observed?
Was observation an acceptable approach?
Was an attempt at total resection justified?
Was radiation therapy useful?
The results of this study are summarized in table. The only issue in which there was felt to be consensus was that there was no test except histologic verification that could ensure the diagnosis of LGG to a sufficient degree to justify radiation or chemotherapy. As for the other issues, some evidence was available that suggested a benefit of both extensive surgery and radiation, but no practice parameters could be defined based upon the available evidence and any recommendations could only be considered as optional. In particular, observation of a patient who presents with a seizure and imaging studies consistent with LGG could not be considered substandard care at this time.