1.1 Background of Disease

Neoplasms of the CNS have certain unique characteristics that distinguish them from tumors elsewhere in the body, and they require a different approach in choosing and evaluating therapies. The brain tumor kills either by uncontrolled growth in confined space, leading to compression of vital areas not necessarily involved with tumor, or by direct invasion of critical neurological centers. Associated with the tumor is surrounding cerebral edema that can have the same effect as the expanding tumor mass. Critical evaluation of the effectiveness of therapy directed against the tumor can be measured in terms of survival or response. If time-to-tumor progression is the desired endpoint, then careful evaluation of each patient is required, as deterioration sometimes reflects secondary, nontumor-related causes, for example, fever, infection, hypoxia, steroid myopathy, etc.

Primary gliomas do not metastasize except in very rare instances. Hence, criteria for operability relate to the position, depth, and infiltration of the tumor. Surgery is not advised for small gliomas located in critical areas such as the brain stem or deep midline tumors; however, these areas can be approached with stereotactic methods for biopsy, and hence, a histologic diagnosis should be available in almost every case of primary intracranial tumor. Tumors of the frontal or temporal lobes are more easily removed with large volume resections. Unfortunately, even these latter groups remain incurable by surgery because of the infiltrative nature of the tumor.

It is estimated that there will be 12,100 deaths and 17,500 new cases of primary brain tumors in 1993. The mean age of patients with glioblastoma is 52 years, for astrocytoma and anaplastic astrocytoma 48 years, and for oligodendroglioma 42 years. There is a very slight increase in incidence in males over that of females. Etiologic factors in the development of human gliomas are uncertain. It is known that malignant gliomas are associated with neurofibromatosis, tuberous sclerosis, and rarely, with other genetically inherited diseases.¹ There is some suggestion of occupational exposures as a potential cause of gliomas, and although incriminating, as yet these have not been correlated definitely with specific causes of glioblastoma.^2,3 Other potential causes of glial neoplasms include viral agents that are known to induce brain tumors in animal models, and there is a very strong association with the Epstein-Barr virus and primary CNS lymphoma.⁴

1.2 Modality History

When the data on surgery, radiotherapy, and chemotherapy of malignant glioma is reviewed, it is quite clear that each modality has something to add. However, in general none of the modalities can produce a cure except in rare instances and new approaches are needed to improve overall survival. Surgery is indicated both to make a diagnosis and to remove tumor in as complete a method as possible. However, generally it is impossible to completely remove all microscopic disease, even with what appears to be a total removal on postoperative scans, due to the infiltrative nature of these lesions. Surgery alone will not control disease for longer than several months in patients with glioblastoma and probably not longer than four or five months in patients with anaplastic astrocytoma. The extent of surgery may have an impact on overall survival. In a recent study by the Brain Tumor Study Group, a direct relationship existed between the extent of postoperative tumor volume and survival.⁵ This effect was independent of age, performance status, and histology. Andreou et al correlated the amount of postoperative tumor burden with both the length and quality of survival⁶ Residual tumor volumes less than 4.5 cm were associated with 70% chance of greater than 700 days survival. Smaller tumor volumes increased this chance. Other such studies have confirmed this observation of longer, better-quality survival with more extensive resection.^7,8

Radiation therapy has proven to be an effective adjunct to surgery in improving survival for malignant gliomas. In fact, radiation therapy represents the single most effective treatment for glioblastoma after surgical resection. The initial studies conducted by the Brain Tumor Study Group in 1969 documented this survival advantage, with radiation therapy showing a statistically significant increase in survival when comparing radiation therapy to surgery alone or to chemotherapy alone following surgery.⁹ Many subsequent trials from the Brain Tumor Study Group and from studies at NCOG have documented the survival advantage, especially in subsets of patients with malignant glioma. In fact, the patients with nonglioblastoma appear to benefit the most with the addition of radiation therapy. It is clear from all of these studies that patients with anaplastic astrocytoma have a far better survival than those with glioblastoma, with median survivals in this group of patients closer to 20-24 months compared to 12 months for patients with glioblastoma.¹⁰

However, neither radiation therapy or surgery appears to influence the ultimate outcome of patients with malignant gliomas, including those with anaplastic astrocytomas. It would seem that any further advancement in the treatment of this disease would come most likely from chemotherapy. At present, the nitrosoureas and procarbazine are the principle drugs demonstrating significant activity against these malignancies. Procarbazine and CCNU, as single agents, have shown antitumor activity against malignant gliomas, achieving response rates of 40-50% and median durations of response of greater than six months in patients with glioblastoma and as long as 12 months in patients with anaplastic astrocytoma.^11,12,13 The combination of procarbazine, CCNU, and vincristine (PCV) was found to be active against recurrent malignant gliomas as well as other primary brain tumors, achieving combined response or disease stabilization rates of 60% with median durations of 9-10 months.^14,15 Following this initial clinical response and further experimental trials in the Brain Tumor Research Center Laboratories at UCSF, a revision of this protocol was adopted and was designated as PCV#3.¹⁶ A second clinical trial using this revised PCV #3 showed 84% of patients achieved some degree of objective response or disease stabilization, and the median time to tumor progression was 28 weeks, with 30% of patients still alive at one year. An evaluation of NCOG Study 6G61, a prospective randomized trial comparing the three drug regimen of procarbazine, CCNU, and vincristine with BCNU alone as an adjuvant to surgery and radiation therapy, showed a distinct survival advantage in the subset of patients with nonglioblastomas.¹⁷ These patients had a statistically significant increase in median survival compared to those patients who were treated with BCNU alone. The median survival of patients treated with PCV was 157 weeks versus only 82 weeks for patients treated with BCNU alone. There was no survival advantage seen with the three drug combination over that of BCNU alone in patients with glioblastoma multiforme. Given this prospective trial, we feel that PCV is the chemotherapy treatment of choice for patients with anaplastic astrocytomas.

1.2.1 BUdR(Bromodeoxyuridine)

Radiotherapy plays an important clinical role in the management of cancer. The interaction of radiation and drugs to selectively increase tumor-cell kill without increasing the normal tissue toxicity has been a major goal of radiation oncologists for decades.

The radiosensitization potential of halopyrimidines has been known for over 20 years.¹⁸ Halogenated pyrimidine analogs are incorporated only into actively dividing cells and substitute for thymidine in DNA. As a result of this substitution, the tumor cells are more sensitive to the lethal effect of radiation. Two agents, BUdR and IUdR, are currently available in clinical research trials. The sensitivity to irradiation of the cells that have incorporated BUdR into DNA increases about two to three times, as measured by a single exposure to irradiation in vitro in such experimental tumors as D98a3, D93S, HExp 1, Ascitic P-388, L-cell, and Hepatoma 129.^19,20,21 BUdR alone has virtually no cytocidal or antimetabolite effects on nonirradiated cells. Even though Kriss and his coworkers observed rapid debromination of BUdR after IV therapy,²² Krant treated 14 patients with IV BUdR and irradiation and thought that they observed some increase in radiation response.²³ In addition, Goffinet and Brown (1977) found that the hepatic dehalogenation of the drug was blocked by 5-diazouracil (DAZU) in order to ascertain relative importance in the degradation of IV administered analogues. His studies showed that following IV infusion, apparently enough BUdR bypasses the hepatic vessels to permit tumor radiosensitization despite dilution of the drug by the systemic circulation.²⁴

Primarily because of the work of Kriss,^24,25 most clinical trials were with intra-arterial BUdR. Sano^26,27 (27 patients) and Hoshino²⁸ (161 patients) reported a considerable therapeutic gain for brain tumors treated with the combination of intra-arterial BUdR radiation. They recognized that in the central nervous system, rapidly growing tumors could be the only replicating tissue since DNA synthesis is very low in endothelial or glial cells and virtually nonexistent in neural tissue. Their treatment consisted of daily infusion of 0.6 to 1 grams of BUdR per day for four to seven weeks with a total dose of 25 to 45 grams and a daily radiation dose of 150 to 200 rads up to a total of 6,000 rads. The results obtained with 107 cases were rather impressive. For grade 3 and grade 4 gliomas, the one-year survival rates after a combination therapy of BUdR and radiotherapy were 88% and 68% respectively, whereas survival for the latter after surgery alone was about 5%. At two years after treatment, the corresponding survival values were 72%, 44%, and 0%, and at three years, 54%, 18%, and 0%.^26,27,28 Unfortunately, intracarotid or intravertebral artery infusions can cause increased morbidity and potential mortality and have limited applicability to a large patient population. Recent studies at the NCI have focused on the pharmacokinetics of long-term IV infusions of BUdR and the potential applicability of this approach to the clinic.^29,30 BUdR infusions of 1.5 g/m² over 12 hours on consecutive days resulted in plasma levels of 1-5 x 10^-6M. When given five days per week of weeks 1, 2, 5, and 6 of a six-week radiation course, only minor drug-induced toxicities such as skin rash, increase in radiation dermatitis, and fingernail changes were observed. The toxicity of continuous infusion bromodeoxyuridine was evaluated in the most recent NCOG trial 6G-82-1 and was felt to be very tolerable during radiation therapy. With continuous infusion BUdR no grade 4 myelosuppression or grade 4 thrombocytopenia occurred. Only one grade 3 infection, no episodes of grade 4 infection, and no episodes of hemorrhage occurred in the study. Major toxicities that were encountered with the combined BUdR and radiation therapy were primarily related to skin toxicity due to the photosensitization of patients with BUdR and to some allergic events felt to be secondary to this drug. However, most of these events were grade 2 or 3 toxicities with only rare grade 4 episodes. Overall, the treatment was extremely well tolerated at the doses of 800 mg/m² per day used in this trial. In summary, we feel that the use of continuous infusion of BUdR is very well tolerated given in the manner of continuous intravenous delivery.

1.3 Rationale

Our strong desire to enhance palliation and to increase survival for patients with anaplastic astrocytoma has prompted us to design an aggressive multimodality program. In 1983, we began a trial previously described, that utilized surgery, external irradiation with BUdR, and postradiation chemotherapy (NCOG Protocol 6G-82-1). This treatment regimen has shown significantly promising results for patients with anaplastic astrocytoma. The most recent evaluation of this trial in patients with a diagnosis of nongliobastoma shows a median survival of 190 weeks with a median time to tumor progression of 108 weeks.³⁸ No significant advantage was seen in terms of median survival for patients with glioblastoma. As a result of this experience, we feel that patients with anaplastic astrocytoma should be treated uniquely and differently than those with glioblastoma. Because PCV chemotherapy (NCOG Protocol 6G61) is more effective than BCNU in this group of patients, we feel that the chemotherapy used in this study should include PCV. Because of the data previously described with the use of BUdR as a radiosensitizer, we feel that it should be evaluated in a prospective randomized trial. This study would be a prospective randomized phase III study evaluating patients with a diagnosis of highly anaplastic astrocytomas, treated with radiation therapy and PCV chemotherapy, with the randomization to receive BUdR or no sensitizer during the course of radiation therapy. The intent of the study is to specifically analyze whether BUdR increases to a significant degree median survival and time to tumor progression over that of conventional radiation therapy plus chemotherapy.

If WBC <= 2000 or platelets <= 75,000, discontinue BUdR therapy and notify study chair. If BUdR is stopped due to lowered blood counts, BUdR can be restarted only with WBC >= 3,000 and platelets >= 100,000 and it must be administered at 50% of the previously delivered dose.

If any of the toxicity parameters reach a WBC nadir of < 1500, ANC < 750, or platelets < 50,000, drugs can be restarted only after counts rise and stabilize at WBC >= 4000 and ANC >= 2000 and platelets >= 125, 000 (or higher). The doses for restarting therapy should be 80 mg/m² for CCNU and 30 mg/m² for procarbazine. For other hematologic toxicites, therapy can be resumed once the WBC reaches >= 3000 or ANC >= 1500 or platelets >= 100,000. Drug doses for patients with chronically (> 8 weeks) depressed marrow should be discussed with the study chair.

7.3.2 Dermatologic Toxicity/Oral Toxicity - BUdR/Procarbazine

BUdR should be discontinued completely for a generalized rash and not resumed unless it is certain that the rash is due to other drugs. Additionally, grade 3 or 4 mucositis would require discontinuing the drug for one week, then resuming it the next week at 50% of the previous dose if the mucositis has resolved. If the mucositis fails to resolve within one week, BUdR would be discontinued until resolution of the mucositis. If mucositis recurs following reinstitution of BUdR, the drug should be discontinued. For grade 1 or 2 mucositis, the dose of BUdR should be decreased by 25%. If mucositis fails to resolve or worsens, then the drug should be discontinued until resolution of mucositis occurs and restarted at 25% of the previous dose (or 50% of the original dose).

Rashes felt to be secondary to Procarbazine will be permanently discontinued. Dr. Prados should be contacted if there is any question.

7.3.3 Neurologic Toxicity - Vincristine

The dose of vincristine should be decreased to 0.7 mg/m² upon diminution of muscle strength to 60% of normal or persistent severe paresthesias (grade 3 neurotoxicity: peripheral nerve). For any other neurotoxicity (PN) at grade 3 or grade 4 (see table that follows), vincristine should be discontinued. Vincristine will not be resumed for the duration of the study, even if all symptoms disappear.

7.3.4 Hepatic or Non-Hematologic Toxicity (6/30/95)

BUdR should be stopped for grade >= 3 hepatic toxicity. Monitor with weekly blood tests until recovery takes place. BUdR may be resumed at 50% of the starting dose; if liver enzymes recover to grade <= 1.

If, after the 50% reduction, there is a second episode of grade >= 3 liver toxicity, BUdR will be permanently discontinued. Other grade >= 3 non-hematologic toxicities must be reported to Dr. Prados.

7.4 Toxicity Documentation and Ancillary Care

7.4.1 Toxicities and the time of their onsets in relation to drug administration will be recorded in the toxicity section of the flow sheet. Toxicity criteria are located in Appendix III.

7.4.2 All drugs (including doses) and transfusions given a patient during this protocol study will be recorded in the chemotherapy flow sheets.

7.4.3 The use of glucocorticoids will not exclude patients from this study. The attending physician should maintain the lowest dosage of steroids consistent with good medical management. No reduction in the dose of steroids at the start of radiotherapy should be made until completion of radiotherapy; escalation during the interval between drug treatments should be made only when absolutely necessary. Antacids are to be used with steroids.

7.4.4 The use of anticonvulsants is preferred for all patients with clinical evidence of neurotoxicity. Dilantin is the preferred anticonvulsant .

7.4.5 The use of analgesics or antiemetics will not exclude patients from study. Compazine or other antiemetics may be given to those patients who experience gastrointestinal toxicity.

7.4.6 The use of other cytotoxic agents is prohibited during therapy.

7.4.7 Patients must have all necessary general medical supportive care. In particular, care must be taken that patients do not become dehydrated due to vomiting or diarrhea and that their serum electrolytes remain within acceptable limits.

7.5 Adverse Reactions Reporting/RTOG Members (2/14/95)

7.5.1 Grades 4 and 5 unknown reactions. Phone report to RTOG, Study Chairman, and IDB within 24 hours. A written report to follow within 10 working days.

7.5.2 Grades 4 and 5 known reactions and Grades 2 and 3 unknown reactions. Written report to IDB within 10 working days. Copies to Study Chairman and RTOG.

7.5.3 Acute myeloid leukemia (AML). The report must include the time from original diagnosis to development of AML, characterization such as FAB subtype, cytogentics, etc. and protocol identification.

7.5.4 Phone report to Investigational Drug Branch (IDB) within 24 hours (301/230-2330, available 24 hours, recorded after hours).

7.5.5 Address for submitting ADR reports:

7.5.6 Further information for reporting Adverse Drug Reactions is detailed in Appendix IV.

7.6 Adverse Drug Reaction Reporting/SWOG Members (2/14/95)

7.6.1 All Southwest Oncology Group (SWOG) investigators are responsible for reporting of adverse drug reactions according to the NCI and Southwest Oncology Group Guidelines. SWOG investigators must:

Call the SWOG Operations Office 210/677-8808 within 24 hours of any suspected adverse event deemed either drug-related, or possibly drug-related.

Instructions will be given as to the necessary steps to take depending on whether the reaction was previously reported, the grade (severity) of the reaction, study phase, and whether the reaction was caused by investigational and/or commercial agent(s). The SWOG Operations Office will immediately notify the RTOG Headquarters Data Management Staff as listed in the RTOG reporting guidelines.

7.6.2 Within 10 days the investigator must send the completed (original) DCT form (for regimens using investigational agents) or the FDA 3500 form (for regimens using only commercial agents) to the NCI:

Investigational Drug Branch
P.O. Box 30012
Bethesda, Maryland 20824

7.6.3 In addition, within 10 days the investigator must send:
- a copy of the above report,
- all data records for the period covering prestudy through the adverse event, and
- documentation of IRB notification to the following address:

7.7 Adverse Drug Reaction Reporting/NCCTG Members (2/14/95)

7.7.1 Treatment that contains both investigational and commercial agents should be reported according to the investigational guidelines. However, if the reaction is clearly a known reaction of the commercially available agent involved, it should be reported according to the commercial agent guidelines. If further clarification is necessary, call the NCCTG medical oncology pharmacist (507/284-2701).

7.7.2 NCCTG Reporting for Commercial Drugs:

Fax, then report in writing to NCCTG Operations Office (no telephone calls necessary) within five working days:

1) Any ADR that is both serious and unexpected: life threatening (grade 4) or fatal (grade 5).
2) Any increased incidence of a known ADR that has been reported in the package insert or the literature.
3) Acute myeloid leukemia (AML). The report must include the time from original diagnosis to development of AML, characterization such as FAB subtype, cytogentics, etc. and protocol identification.
4) Any death on study, if clearly related to the commercial agent(s). The ADR report must be documented on the ADR form (Form FDA 3500) and the original mailed to:

5) The NCCTG Operations Office will immediately forward a copy of the ADR form to RTOG and IDB if deemed a reportable ADR.

7.7.3 NCCTG Reporting for Investigational Drugs:

Report by telephone to Investigational Drug Branch (IDB) directly, 301/230-2330, available 24 hours (recorder after hours), and to the Operations Office within 24 hours; also fax (507/284-1902) a copy of the DCT Adverse Reaction Form for Investigational Drugs to the Operations Office. If a weekend or holiday, the Operations Office must be notified the next working day:

1) All life-threatening events (grade 4) that may be due to drug administration.
2) All fatal events (grade 5) while on study.
3) First occurrence of any previously unknown clinical event (regardless of grade).
4) The original of the written report (DCT Adverse Reaction Form for Investigational Agents) to follow within five working days of the event to:

10.3 Following histologic review, Dr. Davis will send a completed Pathology Review Form to RTOG Headquarters. Slides will be forwarded to Dr. James S. Nelson at Louisiana State University Medical Center for a supplemental review and will be retained by Dr. Nelson until the institution pathologist requests their return. A copy of the Pathology Review Form will be sent to the clinical investigator and to the pathologist who made the initial diagnosis. If there is a disagreement as to diagnosis or classification, Dr. Davis will contact the initial pathologist to inform him/her of this difference of opinion. A diagnosis of glioblastoma multiforme will render a case ineligible.

10.4 NCCTG Members (2/14/95)

10.4.1 For NCCTG patients, pathology review is mandatory prior to study entry to confirm eligibility. It should be initiated as soon as the diagnosis has been made. Pathology review will be performed by Dr. Bernd Scheithauer, Mayo Clinic. All material for pre-entry review will be sent directly to:

To be eligible for this study, the patient's tumor must be pure grade III astrocytoma according to the WHO classification.

10.4.2 The following materials will be sent to the NCCTG Operations Office

H& E stained slides
polyglycene unstained slides
paraffin block
brain smears, if done
pathology reports
NCCTG Pathology Reporting Form - Brain Tumor
RTOG transmittal form

10.4.3 After the review, a call will be made to the institution notifying them whether or not the case is eligible and will be confirmed by fax. Material for eligible cases will be forwarded to RTOG Headquarters along with the RTOG transmittal form. A representative slide and the paraffin block will be retained by the NCCTG Operations Office. Material for ineligible cases will be returned to randomizing institution.

10.4.4 The institution entering the patient will be required to send two tubes of peripheral blood, 15 cc, one in EDTA and the other in Heparin, which can be drawn at the time of the next scheduled blood count assessment.
Note: A specimen mailing kit will be sent from the Mayo Clinic-Rochester to the NCCTG institution with specific instructions on how to properly ship the blood samples. Do not ship blood samples until the kit is received.

10.5 Fresh Tissue Repository (RTOG only) (6/30/95)

10.5.1 Fresh specimens are eligible for RTOG 93-08, the fresh/frozen tumor repository study.

10.6 Fixed Tumor Repository Study (Optional) (6/30/95)

10.6.1 Patients entered on this study are also eligible for the Fixed Tumor Repository Study.

10.6.2 To receive an additional RTOG case credit, the following must be provided to RTOG:

10.6.2.1 One additional paraffin block of tumor or 15 unstained slides (maximum thickness of 5 microns each). Block/slides must be clearly labeled with the pathology identification number that agrees with the Pathology Report.

10.6.2.2 Pathology report documenting that submitted block or slides contain tumor.

10.6.2.3 A Pathology Submission Form must be included and must clearly identify the enclosed materials as being for the Fixed Tumor Repository.

10.6.3 To encourage compliance, your Pathology Department could be reimbursed for obtaining blocks or cutting slides.

10.6.4 Patient consent form should give the Pathology Department authority and responsibility to comply with this request (pathology blocks belong to the patient from whom tissue has been removed).

10.6.5 Materials will be sent to:

10.6.6 SWOG and NCCTG members wishing to participate should submit the materials specified in Section 10.6.2 to their group offices (see Section 12.2.1) for forwarding to RTOG.