High Grade Neuroepithelial Tumors (Gliomas)


Glioblastoma Multiforme (GBM) WHO Grade IV

Neuroepithelial tumors comprise a group of tumors including GBM, anaplastic astrocytoma and low grade astrocytomas. Other neuroepithelial tumors include Oligodendroglial tumors including Oligodendrogliomas and anaplastic oligodendroglomas. There are also mixed gliomas in the pathologic classification including Oligoastrocytomas and anaplastic oligoastrocytomas. These diseases have differing clinical behaviors and progrnoses and are best each considered in their own right.

Glioblastoma Multiforme is a primary brain tumor with an aggressive course.They account for about half of all adult primary brain tumors and 75% of all high grade gliomas. The majority of adult GBMs do not form a capsule, but rather grow in a rapid, aggressive, infiltrative pattern. they infiltrate brain parenchyma but nearly never metastasize outside the CNS. WHO Grade III and IV are classified as malignant gliomas, but the diseases are distinct with different behaviors, response to treatment and prognosis. There is frequent vasogenic edema associated with the lesion which can create mass effect and the presenting symptoms. Presenting patterns are generally focal with headaches in faster growning lesions. Seizures are more common in low grade tumors. Glioblastomas generally present with focal deficits, headache, seizures, unilateral weakness or menstal status changes. They appear on MRI and contrast CT as a hypodense central lesion surrounded by ring enhancement and associated edema.

Epidemiology and Natural History

GBMs tend to run a rapid course. Untreated, time from diagnosis to death is rapid, frequently within months. Better performance status, younger age result in slower progression, but the disease is almost univerally fatal within 20 months of diagnosis. Headaches associated with GBM are due to mass effect and increased intracranial pressure on the dura and vessels. The histopathologic features include diffuse infiltraiton involving large portions of the brain, nuclear atypia, mitotic activity, vascular proliferation and ring enhancement areound areas of central necrosis.

The prognosis for GBM is universally poor with median survivals around 14 months in highly selected patients with aggressive treatment. A recursive partition analysis study identified age as the most important factor in prognosis followed by performance status and normal mental status. Age < 50 and KPS ≥ 70 were favorable prognostic indicators. The RPA divided GBM into six classification groups I - VI for malignant gliomas.

More recently these groups have been collapsed into three sets: I-II Anaplastic astrocytoma, age ≤ 50 and normal mental status or age > 50 and KPS > 70 and symptoms > 3 months. This group has a median survival of 40-60 months.

The next RPA set is III-IV with two distinct groups, one anaplastic astrocytoma (WHO Grade III), the other glioblastoma multiforme (WHO Grade IV). The first group consists of AA with age ≤ 50 and abnormal mental status or age > 50 and symptom duration < 3 months. The second group are GBMs with age < 50 or age > 50 and KPS ≥ 70. This group has a median survival expected in the range of 11 - 18 months.

The final set is RPA class V-VI which includes only GBM for age > 50 and KPS < 70 or abnormal mental status. This set has the poorest prognosis at 5 - 9 months median survival.

RPA For Survival pre-Temozolomide
RPACharacteristicsMed Survival (mo.)OS1OS2
I Anaplastic Astrocytoma, Age < 50 normal mental status 59 mo.90%76%
II Anaplastic Astrocytoma, age ≥ 50, KPS ≥ 70, Symptoms > 3 months 3785%68%
III GBM, Age <50, KPS 90-100, or Class I but abnormal MS 1885%35%
IV GBM, Age < 50, KPS < 90 or Class II but Symptoms < 3 mo. or Class V but able to work 1145%15%
V GBM, Age ≥ 50, KPS ≥ 70, or KPS < 70 and normal MS 930%6%
VI GBM, age ≥ 50, KPS < 70, normal mental status 520%4%

RPA For Survival with Temozolomide
RPACharacteristicsMed Surv.
TMZ+RT
(Control)
OS2p
III GBM, Age <50, KPS 90-100, or Class I but abnormal MS 21 (15) 43% (20%)< 0.0001
IV GBM, Age < 50, KPS < 90 or Class II but Symptoms < 3 mo. or Class V but able to work 16 (13)28% (11%)< 0.01
V GBM, Age ≥ 50, KPS ≥ 70, or KPS < 70 and normal MS 10 (9)17% (6%)=0.05

In patients who have difficulty tolerating temozolomide, the MGMT marker may discriminate who would benefit from temozolomide and those who receive relatively lower benefit. There are data which suggest that methylated PTEN gene in low grade glioma transformations may be linked to the malignant transformation to GBM. Secondary GBMS have a mutation in teh IDH2 gene and possibly p53 as well.

GBMs have a rapid doubling time and rapid progression. Doubling times are on the average, 2-3 months. Infiltrating disease is several centimeters beyond a GTR. Surrounding edema contains tumor and a GTR is nearly impossible. The median survival is 12-14 months with the standard of care being Radiation with temozolomide.

Clinical Workup and Staging

GBMs are not staged. The workup consists of detailed MRI imaging with contrast, T2, T2/FLAIR, and T1 without and with gadolinium. For patients who are not able to obtain an MRI, then a CT with contrast is eseential. Once imaging has demonstrated a probable high grade glioma, a tissue diagnosis is mandatory. Referal to neurosurgery for craniotomy and maximum safe resection followed by definitive radiation therapy with concurrent temozolomide followed by maintenance temozolomide is the current standard of care.

Treatment Recommendations

The ACR does not have appropriateness criteria for high grade gliomas. Radiation therapy is generally considred appropriate and recommended. Dose escalation studies have failed to demonstrate an advantage for higher doses and do increase the risk of radionecrosis. Recurrences locally generall happen within 3 cm of the original tumor, therefore most centers recommend treating the pre-operative edema plus 2 cm. The dose to the initial volume has been reported variously between 46 Gy - 54 Gy, with the most common doses with a boost to 60 Gy in 33 fractions total (59.4 Gy at 1.8 Gy/fraction). Several studies over the years add insight to appropriate dosing and volumes. The Brain Tumur Cooperative Group looked at doses ranging from < 45 Gy to 60 Gy at 1.7 - 2.0 Gy /fraction. Median survival improved from 28 weeks to 42 weeks when radiation dose was between 50 and 60 Gy. The MRC-UK also showed significant survival advantage with doses at 60 Gy over those receiving 45 Gy (12 v. 9 months).

External beam radiotherapy is the standard of care with 60 Gy at 2 Gy/fraction to the GTV + 2-3 cm margin. This should be treated with multiple fields, using 3dCRT. Median survival is 3-6 months without radiation and 9-12 months with radiation. 60 gy is better than 45 Gy, but 70 is not better than 60 Gy. UMich demonstrated that 90 Gy is no better with 90% of failures in field. A brachytherapy boost study was likewise a negative trial.

For patients with poor performance status, advanced age, the prognosis is much worse. For patients with limited expected survival who are not able to tolerate conventional treatment a short palliative course may be beneficial. Older patients (> 65), with poor performance status have been shown to experience limited post-treatment survival or rapid neurologic deterioration following standard radiation therapy. A french study demonstrated MS of 3.9 months (no treatment), 6.7 months with 50.4 Gy in patients ≥ 70 and KPS ≥ 70. A Canadian study demonstrated equipvalence of 60 Gy, 40 Gy and nearly the same as the French study with doses of 60 Gy (5.1 mo.), 40 Gy (5.6 mo.), 50.4 Gy (6.7 mo.)

Dose Escalation

Dose escalation studies > 60 Gy has not been shown to be beneficial. RTOG and ECOG randomized patients to 60 Gy v. 60 Gy + 10 Gy boost for high grade gliomas treated with conformal 3D or IMRT radiation. No benefit to dose escalation above 60 Gy was seen. Median survival was 9.3 months (60 Gy group) and 8.2 months (70 Gy Group). Hyperfractionation or accelerated treatment has likewise demonstrated no benefit. One study demonstrated a benefit to hyperfractionation, but these results have not been confirmed in prospective randomized trials. No survival difference was found.

SRS/FSRT

Improved median survival in worse prognosis groups was reported with SRS in a 3 institution study with 2 year actuarial survival for all patients at 45% and median survival of 96 weeks. This study was 115 patients, while a much larger RTOG study of 1587 patients showed similar results, however, the study selection criteria limited which patients recieved the boost. A prospective randomized trial compared 60 Gy (conventional) plus BCNU ± SRS in 203 patients. At 62 months, no significant improvement in median survival, no difference in patterns of failure and no difference in quality of life and cognitive decline were seen. An RTOG study (RTOG 0023) looked at integrated FSRS with patients treated to 50 Gy with interspersed SRS boosts of 5 - 7 Gy to a cummulative dose of 70 - 78 Gy in 29 treatments over 6 weeks. Significant toxicity including Grade 4 toxicity (neurologic, constitutional, metabolic), and Grade 3 (necrosis) was seen. Median surival was 12.5 months and no advantage was seen when compared with historical data.

Given the various dose fraction schemes, the lack of a formal "appropriateness" criteria, perhaps the best source of current recommendations for radiation dose/fraction schemes in malignant gliomas comes from the RTOG protocols. A variety of recent protocols exploring concommitant agents have been opened. The radiation section of these protocols give insight into present radiation protocols. Most of these protocols propose a dose of 46 Gy at 2 Gy/fraction 5 days/week given to the edema as seen on pre-operative MRI plus a 2 cm margin, corrected for anatomical barriers. This is followed by a boost to the T1/Gadolinium enhancing area or resection cavity plus a 2 cm margin to a total of 60 Gy at 2 Gy/fraction 5 days/week. Radiation is to commence between 3 and 5 weeks of surgery.

Treatment recommendations for GBM after maximum safe resection are for radiation therapy with temozolomide. Radiation fields are described as follows:

  • To the pre-operative tumor volume as defined by edema + 2 cm margin: 48.6 Gy at 1.8 Gy/fraction
    • If the volume is smaller then 2 Gy to 46 is ok, and there may not be any difference between schemes.
  • This should be followed by a boost to the T1/Gadolinium enhancing region or the resection cavity plus a 2 cm margin with an additional dose of 10.8 Gy at 1.8 Gy/fraction in 6 fractions for a total dose of 59.4 Gy.
  • Radiation should commence between 3 weeks and 5 weeks post-operatively.
  • Temozolomide should be given with radiation and continued adjuvantly post radiation. MGMT methylation may indicate a reduced effectiveness of temozolomide, but it is generally well tolerated.

RTOG trials are now recommending this scheme in its standard arm radiation therapy protocols or in all arms where radiation dose/volume/fraction is not being compared. A somewhat older trial for Grade III anaplastic gliomas (9402) uses 50.4 Gy to MR/T2+2 cm and boosts MR/T1-Gad+1 cm to 59.4 Gy1

Treatment should be delivered with 6 MV photons or energies appropriate for optimizing dosimetry. IMRT or 3D CRT techniques may be used depending on the need to protect critical structures, including the brain stem, chiasm, orbits, optic nerves. Simulation should take place in the supine position using a thermoplastic mask sufficient to full immobilize the patient. Simulation should take place using CT and MRI with fusion of images for treatment planning. Generally the MRI imaging should be as close as possible to the CT geometry to minimize fusion errors.

Initial target volume will be as follows:

CTV-Initial/GTV1
The initial GTV1 consists of the T2 or FLAIR abnormality on the post-operative MRI. This volume must include all the post-operative enhancement and the surgical cavity.
From the GTV1, the CTV1 will be generated by adding a 2.5 cm margin. The CTV1 may be reduced to 0.5 cm at anatomical barriers to spread (skull, ventricles) and to allow sparing of critical structures.
The initial PTV will be the CTV1 plus 3-5 mm expansion, depending on immobilization methods and reproducibility.
This volume will be treated to 45-46 Gy at 1.8 - 2.0 Gy/fraction 5 days/week, once per day.
CTV-Boost/GTV2
The boost GTV2 target volume will be defined by the contrast enhanced T1 MRI fused to the planning CT. This GTV is defined as the post-operative MRI scan abnormality and must include the entire surgical defect cavity margins.
The CTV2 boost volume will be the GTV2 plus a 2 cm margin. This margin can be reduced to 5 mm if required to avoid critical structures and anatomical barriers to spread.
The PTV2 will be a 5 mm expansion on the CTV2 depending on local setup uncertainties. The PTV2 should not be adjusted for OAR structures.
Dose to the PTV2 should be a cummulative dose of 59.4 - 60 Gy at 1.8 - 2.0 Gy/fraction, while respecting OAR dose limits
OAR/PRV
Each organ at risk can be used to generate a planning risk volume. The PRV will be the OAR volume plus a 3 mm margin.

Critical structure maximum dose constraints are defined by a maximum point dose to a volume greater than 0.03 cc:

  • Lens: 7 Gy
  • Retina: 50 Gy
  • Optic Nerves: 55 Gy
  • Optic Chiasm: 56 Gy
  • Brainstem: 60 Gy

Planning goals include 95% of PTV2 gets 60 Gy, 99% of PTV2 gets 54 Gy. Treatment breaks should be kept to less than 5 days total, but 6 - 7 days are acceptable.

Anaplastic Gliomas — WHO Grade III High Grade Gliomas

These tumors include anaplastic astrocytomas, anaplastic oligoastrocytoma, and anaplastic oligodendrogliomas. They account for about 25% of all high grade gliomas in adults, occurring during young to middle adulthood. They have increased cellularity but without necrosis. They show similar MRI imaging characteristics to GBM: T1-Gad ring enhancing lesion and T2/FLAIR edema. Up to 1/3 of these tumors may not enhance. Grade III disease prognosis is highly influenced by molecular markers. The median survival is around 3 years but the population is heterogeneous.

Pathology and Natural History

The prognosis of Grade III tumors is heavily influenced by molecular markers. Loss of 1p19q has been identified in 63% of anaplastic oligodendroglioma, 52% of mixed anaplastic astrocytomas. 1p19q del is associated with improved survival, greater chemosensitivity and greater radiosensitivity. RTOG examined the role of various treatments and found no influence on the type of treatment as prognostic, but did find the role of 1p19q deletion as prognostic with loss of heterozygosity (deleted) a favorable influence. The RTOG reported in 9402, a survival benefit from PCV chemotherapy only in 1p19q co-deleted tumors. In this populations median survival improved from 2.7 to 8.7 years. An 11 year update demonstrated an overall survival benefit from adding PCV to radiation with a median survival of 14.7 years v. 7.3 years without PCV (radiation alone).

IDH1 mutations have been shown to be a strong prognostic factor in the EORTC 26951 trial, independent of 1p19q co-del. IDH1 mutations are frequently present in secondary Grade IV glomas that arise from lower grade gliomas. Prospective trials of anaplastic gliomas have demonstrated prognostic significance in terms of both progression-free survival and overall survival. A German study demonstrated IDH1 mutations to be a stronger prognostic factor than either 1p19q co-deletion or MGMT methylation. Both MGMT methylation and IDH1 mutation are correlated with 1p19q loss.

Other tumor markers include MGMT methylation, which is not as predictive in Grade III as it is in Grade IV tumors. MGMT has not shown a similar predictive value for PCV (procarbazine, CCNU and vincristine) chemotherapy as it has for temozolomide in Grade IV tumors. It is not predictive of the efficacy of PCV chemotherapy.

Treatment Recommendations

Surgery → Radiation Therapy

The present standard of care for Grade III anaplastic gliomas is maximum safe resection followed by post-operative radiation therapy. The radiotherapy targets are the same as for GBM. RTOG 9402 used 50.4 Gy to T2+2 cm (edema) and boosted the tumor to 59.4 Gy to the T1-Gad + 1 cm. Perez states (6th Ed., Anaplastic Glioma) that the treatment fields and volumes are the same as for

Chemotherapy — Anaplastic Astrocytoma

RTOG found PCV was associated with an improved outcome in anaplastic astrocytomas. A retrospective review failed to confirm this. The MRC-UK randomized 674 patients, 117 had anaplastic astrocytoma after surgery to RT ± PCV (procarbazine, CCNU, Vincristine). There was no advantated to PCV in any subgroup. The median survival in this study was substantially lower than other studies at 13-15 months, compared with 2 - 3 years. Temozolomide has been used in patient with recurrent anaplastic astrocytoma at first relapse with PFS6 months of 46% and OS of 13.6 months.

Chemotherapy — Oligodendroglioma/oligoastrocytoma

These entities are thought to be chemosensitive, based on response rates to PCV in several studies. With 1p19q co-deleted disease, there was a near doubling in median survival, thus establishing chemotherapy as the standard of care in the oligodendroglioma/oligoastrocytoma subset. PCV has significant toxicity and many have substituted temozolomide, which is better tolerated. TMZ has produced high response rates in patients with anaplastic oligodenroglioma and has been shown to be effective in 43% of those who had previously failed PCV. Up front TMZ in elderly patients (age > 70) with anaplastic oligodendroglioma/oligoastrocytoma improved outcomes in patients with a PFS of 6.9 months and MS of 12.1 months in patients with MGMT methylation. Response to TMZ has also been associated with 1p del in a small retrospective study. In cases without 1p19q del, combination chemotherapy-radiotherapy was observed to be associated with a longer time to progression.

In cases with 1p19q co-deletion, combination chemotherapy and radiation was associated with longer median time to progression, but no survival advantage was demonstrated. Longer time to progression was observed with PCV compared with TMZ in co-deleted 1p19q. Medical oncologists have been using TMZ in lieu of PCV due to the increased toxicity, but data is weak and possibly leads to a worse outcome than using PCV.

Evidence for treatment recommendations comes from the above referenced study data and NCCN reviews and recommendations.

  1. Maximum safe resection is generally associated with more favorable outcomes and is recommended when possible.

  2. Post-operative radiation therapy has been shown to provide a survival advantage in several clinical trials involving Grade III/IV tumors. There has been no clinical trial testing radiation in Grade III only patients. The standard of care is the same for GBM in terms of RT target volumes, doses and fractionation schemes:
    1. CTV1=T2 or FLAIR MRI + 2.5 cm to 46-50.5 Gy at 1.8 or 2 Gy/fraction
    2. CTV2(boost)=T1-Gad MRI + 1 cm to 59.4 - 60 Gy at 1.8-2 Gy/fraction in 6 weeks
    3. IMRT SIB has not been tested or reported.

  3. The role of chemotherapy is uncertain for non-codeleted anaplastic gliomas. BCNU and PCV have no demonstrated survival benefit and do have toxicities. Temozolomide is active in recurrent anaplastic astrocytomas and is being tested in up-front setting. Radiation is considered category 1 with prospective evidence, however chemotherapy is category 2 by the NCCN: the panels generally agree with no preference for TMZ, BCNU or PCV chemotherapy.

  4. Co-deleted 1p19q have a more favorable prognosis and respond better to both chemotherapy and radiotherapy. In this subset RTOG and MRC studies have demonstrated a proven benefit of PCV over radiation alone. The NCCN panel assigns a category 1 recommendation to the use of PCV+radiation in this setting, with prospective evidence supporting its use. The role of TMZ is not yet clear, but the NCCN assigns a level 2 recommendation (no significant disagreement, but no prospective evidence either). Perez, 6th Ed. is less enthusiastic than the NCCN.