Radiation Oncology Synopsis

High Grade Gliomas

General Background

40% of malignant primary brain tumors are malignant. The most common malignant CNS tumor in adults is the glioblastoma multiforme which accounts for 85% of all adult CNS neoplasms adn 20% of all primary tumors. The WHO high grade gliomas are WHO Grades III and IV. These are anaplastic astrocytoma, oligodendroglioma and oligoastrocytoma (Grade III) and glioblastoma multiforme. (Grade IV).

Less than 5% of all GBMs are multicentric. They tend to be localized.

The four defining characteristics of GBMs are:

Common genetic changes seen in malignant brain tumors include overexpression of EGFR seen in 50% and PTEN mutation seen in 30% - 40%.

For primary GBM, the genetic changes seen are EGFR overexpression/MDM2 amplification and LOH10/16p.

For secondary GBM the associated genetic changes p53 mutation → low grade glioma → LOH 19q/p16 loss → anaplastic astrocytoma → LOH 10, DCC → secondary GBM.

Cushing triad is a result of increased intracranial pressure. Cushings manifests as hypertension, bradycardia, respiratory irregularity.

High grade gliomas commonly present with morning headache (about 50%) seizures (20%) focal neurologic deficients and mental status changes. They appear on contrast CT as ring enhancing lesions and on MRI as gadolinium enhancing T1 lesions and T2/FLAIR with edema.

There are no staging criteria, other than histologic grading.

Treatment and Prognosis

Prognosis

Primary brain tumors are divided into high grade and low grade tumors. For low grade tumors, median survival is considerably longer than high grade gliomas:

RTOG Recursive Partition Analysis

The RTOG recursive partition analysis is a study of outcomes based on patient and tumor characteristics based on analysis of prognostic indicators in three RTOG trials. Age < 50 or age > 50 was the most significant prognostic indicator of survival with age > 50 worse. This was followed by performance status as the second most important prognostic indicator: worse do worse. Finally, degree of surgical resection and dose of radiation therapy.

The most important factors used in for the RTOG recursive partition analysis stratifications are:

  1. Age < 50; age > 50
  2. histology (GBM = ≥ III; AA Class I/II
  3. Karnoffsky performance status of 70 (watershed)
  4. Mental status changes and symptoms < 3 months (indicates rapid onset -- they do worse).

RPA classifications ≥ III are all associated exclusively with GBM, with exceptions. Class III anaplastic astrocytoma with poor mental status is also Class III. All Class iii patients are under age 50.

Class IV is defined as any histology with KPS < 70 and altered mental status.

GBMs fall under classes III-VI:

  1. Class III: age < 50, KPS 90-100
  2. Class IV: age < 50, KPS < 90 or age > 50 with good mental status
  3. Class V: age > 50, KPS < 70 and no change in mental status
  4. Class VI: KPS < 70 and mental status change

RPA describes median survival by class:

Current modified RPA classifications are based on outcomes with temozolomide. The OS-4 and median survival for the adapted RPA groups for malignant gliomas (Miraminoff, 2007) are:

Class OS-4 Median Survival
III (age < 50; PS 0) 28.4% (control: 6.4%) 21 mo. (control: 15 mo.)
IV (age < 50, PS 1-2) 11.3% (control: 3.3%) 16 mo. (control: 13 mo.)
V(age ≥ 50; mmse < 27) 6% (control 1%) 10 mo. v. 9 mo.

The European Nomogram GBM Calculator also considerede MGMT methylation status and the extent of resection. The only factors which were found to be prognostic were MGMT, Performance Status and Mental Status. Extent of resection was not found to be prognostic. (Lancet Oncology 2008). MGMT is a DNA repair enzyme that removes alkyl groups from the O-6 position of guanine. When methylated it leads to longer survival.

More recent data indicates that the extent of resection may be associated with improved outcomes.

Treatment

The general treatment paradigms are:

TMZ doses are oral tablets 7 days/week at 75 mg/m2 for 5 days, 30 minutes prior to radiation. → 1 month break,

NCCN

The NCCN guidelines 2011 discussion recommends maximum safe surgical resection. They recommend whenver possible the maximum safe tumor resection be achieved unless CNS lymphoma is suspected. If CNS lymphoma is suspected, then biopsy only is preferred. Gliadel (BCNU wafers) are a treatment option. Post surgical MRI should be obtained within 72 hours of the procedure with and without contrast.

If major tumor removal is deemed too risky, a stereotactic biopsy should be performed to establish the diagnosis.

Adjuvant therapy after surgical intervention depends on the tumor pathology and performance status of the patient:

For anaplastic astrocytoma the options for a patient with good KPS (≥ 70) can choose EBRT, chemotherapy or chemoradiation in a clinical trial context. Chemotherapy outside of clinical trials is controversial. Patients with poor performance status can be managed by radiation, chemotherapy or best supportive care.

For Glioblastoma Multiforme the options for a patient with poor performance status include:

  1. radiation
  2. chemotherapy
  3. best supportive care
  4. chemo-radiation only if no carmustine wafers are used.

The addtion of TMZ in younger patients (age < 70) is recommended, and is a consensus recommendation in the elderly. If carmustine wafers were implanted adding TMZ to younger patients is also a consensus recommendation without major disagreement but no high level evidence supports this.

Because radiation therapy can cause additional blood-brain barrier disturbances, increased corticosteroid doses may be required, depending on the mass effect and extent of disease. Post-treatment scans may appear worse during the first 3 months after completion of RT even though there is no progression of disease. The first post treatment MR should be obtained at 2 - 4 weeks for this reason. MRI scans should be obtained every 2 - 4 months post treatment for 2 - 4 years.

Radiation Therapy

General guidelines for radiation therapy based on studies provide for an initial volume of radiation to 46 Gy to the GTV1 = MRI T1+T2/FLAIR. The CTV1 = GTV1 + 1.25 cm. PTV adds an additional 0.5 to the CTV (=1.75 cm total). The reduced volume is treated to an additional 14 Gy with a reduced volume to the GTV2=T1/Gad tumor bed. CTV2=GTV2+0.75 cm and a 0.5 cm PTV (= total 1.25 cm). Post-operative imaging is the target volume.

BTSG 69-01 in 1978 demonstrated doubled survival with RT compared with best supportive care. This study randomized patients to observation, BCNU, whole brain radiation therapy or WBRT + BCNU. There was no difference between WBRT and WBRT + BCNU, but radiation therapy was better than no radiation therapy.

The Scandinavian Glioblastoma Studgy Group (SGSG 1981) used 45 Gy + bleomycin compared to 45 Gy + observation. Median survival in the radiation arms doubled to 10 months from 5 months.

French Data from Keime-Buibert NEJM 2007) supported radiation in the elderly GBM population. Keime-Guibert looked at patients ≥ 70 years old with KPS > 70 treated to 50.4 Gy compared with observation. Median survival improved to 29 weeks from 16.9 weeks. There was no difference between cognition and quality of life.

Roa and Bauman (JCO 2004, IJROBP 1994) studies support the use of hypofractionation in feasibility studies. They suggest that 40 Gy in 15 fractions (2.67 Gy/fraction) or 30 Gy in 10 fractions (3 Gy/fraction) in GBM patients with poor performance status (> 60 and KPS < 50) would be feasible. Both studies reported no difference in outcome compared with standard fractionation.

BTCG 80-01 examined limited field radiation (Shapiro 1989 J. Neurosurg) and found no survival difference in limited compared with whole brain radiation therapy. This was a prospective study with two arms:

  1. WBRT to 60 Gy
  2. WBRT to 43 Gy → conedown to 60 Gy.

This study also demonstrated no difference between BCNU as a single agent compared to a multi-agent protocol. Because of this study, a number of follow on studies were performed that support currently recommended regimens.

Hochberg in 1980 compared CT correlation with post-mortem tissue and found that the CT abnormality + 2 cm margin encompassed the tumor by 83%. Recurrence by imaging also occured within 2 cm of the margin in primary disease in 90% of the cases.

Kelly in 1987 (J. Neurosurg) correlated MRI and CT imaging with stereotactic biopsy in untreated gliomas. the study found that isolated tumor cells extended at least as far as the T2 abnormalities suggesting tht T1 enhancement is equivalent to GTV and T2 is equivalent to subclinical disease.

Data supporting current RT dose of 60 Gy presently used in high grade gliomas includes a combined analysis of 3 BTSG trials, reported in 1979. Four dose regimens were compared: < 45 Gy, 50 Gy, 55 Gy and 60 Gy. The median survival increased with increasing dose from 4 months → 7 months → 9 months → 10 months. Additional data from the MRC-UK reported in 1991 in a randomized controlled trial of 474 patients compared 45 Gy to 60 Gy without chemotherapy and found median survival increased from 9 months to 12 months with higher dose.

Dose escalation beyond 60 Gy has not demonstrated a benefit. The RTOG 7401 did a study of > 600 patients increasing dose from 60 Gy to 70 Gy and found no benefit for 70 gy over 60 Gy. Chan (2002 JCO) escalated dose to 90 Gy and found no survival benefit. Of those who failed at 90 Gy, 91% failed in field.

RTOG 8302 examined hyperfractionation in > 700 patients in a randomized Phase I trial to 64.8 Gy compared with 81 Gy BID. There was no benefit. RTOG 9006 repeated this study and also found no benefit.

RTOG 9305 examined stereotactic boost for HGG and found no benefit nor increased toxicity.

MRC Glioma Meta-analysis Trialists Group (2002 Lancet) examined chemotherapy (pre-TMZ) and found a benefit to adding chemotherapy. This study demonstrated a small improved median PFS of 7.5 months v. 6 months with the addition of chemotherapy to radiation therapy with a reduced risk of death of 15% and a 6% increase of OS-1. There was no RT dose-response in doses > 60 Gy or < 60 Gy.

The Fine meta-analysis in 1993 also showed improved median survival with chemotherapy at 12 months v. 9.4 months.

Radiation Therapy and Temozolomide

There is significant evidence based on the EORTC/NCIC studies reported b Stupp in 2005 (NEJM) and a 5 year update in 2009 (Lancet) that temozolomide increased OS-5 to 10% compared with 2% without temozolomide. Mirimanoff in 2006 (JCO) updated and revised the GBM RPA Classifications due to the benefit of TMZ but found no improvement in RPA Class V (age > 50; KDP≥ 70 or KPS < 70 and normal mental status. The only significant benefit to radiation plus temozolomide was in Classes III-IV: age younger than 50 or able to work.

MGMT methylation status is important in identifying response to TMZ + radiation. Methylated MGMT gives greater response as reported by Hegi, Mirimanoff. The OS-4 in unmethylated MGMT is 0% in RT alone compared with 11% in RT+TMZ. OS-4 in methylated MGMT improves to 5% with RT alone compared with 22% with RT + TMZ . All findings are statistically significant.

Gliasite Brachytherapy

The Gliasite device is implanted and remote afterloaded. The recommended dose is 60 Gy at 5 - 10 mm at 50 cGy/hour using I-125.

Chemotherapy

Gliodel wafers are implantable wafers which are FDA approced for recurrent disease with re-resection. Improved survival is reported by Brem in 1995 with an improvement in median survival of 8 months v. 6 months. It is also approved in the newly diagnosed adjuvant setting. Median survival was increased to 13.9 months compared to 11.6 months.

Anaplastic Astrocytoma/ WHO Grade III Disease

The German NOAH-04 Study by Wick (2009 JCO) compared adjuvant RT with adjuvant chemotherapy and found similar outcomes. NOAH-04 showed similar outcomes with adjuvant RT compared with chemotherapy consisting of procarbazine/lomustine/vincristine (PCV) or TMZ in WHO Grade III anaplastic astrocytomas. This study showed the same PFS/OS for all arms, either RT alone or 2 chemotherapy agents alone.

Good predictors included:

The toxicities included Grade 3 - 4 hematologic toxicity significantly higher in PCV than for TMZ.

Additional studies of oligodendrogliomas compared sequential PCV → RT compared with RT alone. RTOG 9402/INT-0149 (Cairncross 2006 JCO) showed no OS benefit. There was improved PFS with chemotherapy but at a significant toxicity cost. 1p19q del conferred better outcomes.

EORTC 26882 investigated sequential RT → BCNU compared with RT alone in anaplastic astrocytoma. This study demonstrated no overall survival or progression free survival difference.

EORTC 26951 tested the role of adjuvant PCV (procarbazine/lomustine/vincristine) after RT in oligodendrogliomas. This study showed the same overall survival but prolonged progression free survival. 1p19q deletions did better. There was no difference in long term quality of life after PCV.

RTOG 9813 is an ongoing study investigating RT with TMZ v. nitrosurea in anaplastic astrocytoma.No results were available as of 2011.

RTOG 9131 is comparing TMZ + RT in anaplastic oligodendrogliomas. This study is examining neoadjuvant TMZ for 6 months followed by RT+ concurrent TMZ. An interim report suggests that the combination therapy is well tolerated and response correlates with 1p19q deletion.

RTOG 0525 is testing dose intensification of TMZ after TMZ+radiation therapy. This study is a randomized study examining patients treated with RT + TMZ → 1 month break → randomizing to TMZ on days 1 - 21 compared to standard treatment of days 1 - 5 for up to 12 cycles, depending on response.

Toxicity

Radionecrosis is of substantial concern. Radionecrosis is characterized by similar findings on original diagnosis, with:

These findings occur at greater than 6 months post radiation therapy. MRI spectroscopy may aid in differentiating recurrence against radionecrosis.

Temozolomide has primarily hematologic toxicity -- thrombocytopenia. Stupp reported in the EORTC studies Grade 3 -4 toxicities.

Imaging follow up of HGG includes MRI 1 month post radiation with weekly CBC while on temozolomide. After the initial MRI, repeat MRIs every other month are necessary for ongoing follow up. 80% - 90% of all recurrences are local. Up to 50% of patients may show pseudoprogression after RT + TMZ. MGMT methylation (favorable) increases the incidence of pseudoprogression after TMZ + RT.