Faculty Bibliography

The authors review their experience with the operative management of 19 consecutive cases of malignant astrocytoma of the spinal cord. There was a male to female ratio of 1.1:1, and the median age of the population was 14 years (range 1 to 32 years). The median duration of symptoms prior to definitive diagnosis was 7 weeks. Radical excision was carried out in all cases, with 18 patients (95%) receiving radiotherapy and 10 patients (53%) receiving chemotherapy as well. To date, 15 (79%) of the 19 patients in this series have died, with a median survival period of 6 months following surgery. No patient improved after operation. Hydrocephalus was present in 11 patients (58%), seven of whom underwent ventricular shunting procedures. Dissemination of disease was found in 11 patients (58%). Extraneural metastases did not occur in the absence of a ventricular shunt. The authors conclude that malignant astrocytomas of the spinal cord are heralded by a short history followed by rapid neurological deterioration and usually death. The rationale for operation is discussed, and an aggressive approach utilizing adjuvant therapy directed at the entire neuraxis is suggested

Although high-dose intravenous (IV) methotrexate (MTX) with leucovorin rescue (HDMTX) is effective for certain recurrent primary brain tumors, concern for inducing leukoencephalopathy has restrained its use as adjuvant therapy following therapeutic brain irradiation (RT). We have conducted a phase I to II clinical trial using four biweekly courses of HDMTX (8 g/m2) in a neoadjuvant setting in ten patients with newly diagnosed high-risk pediatric primary brain tumors. Four patients experienced an objective response after two to four courses of HDMTX alone (medulloblastoma, one; pineoblastoma, one; malignant cerebral astrocytoma, two). All ten patients subsequently received a course of therapeutic RT, and in seven cases, adjuvant chemotherapy with other agents. One patient acquired an acute transient encephalopathy before RT that completely resolved, and another developed a seizure disorder following RT associated with white matter abnormalities on a magnetic resonance imaging (MRI) scan. Five patients have survived a minimum of 33+ months, and four remain in continuous remission. The acute and delayed neurotoxicity of neoadjuvant HDMTX is acceptable, and we favor further use of this neoadjuvant approach in the context of a phase III trial

Thirty-three patients were treated with intravenous (IV) cisplatin (CPDD) of whom 32 were considered evaluable. There were 14 medulloblastomas, five primitive neuroectodermal tumors (PNET), nine gliomas, three ependymomas, and one germ cell tumor. The overall response rate was 13 of 32 (41%). Eleven responses (five complete [CR], five partial [PR], one mixed [MR]) were noted in the patients with medulloblastoma. The response rate within this group was 79%. Toxicity was tolerable, although it precluded further therapy in five patients.

From 1977 to 1986, 50 children aged 15 months to 18 years were treated for supratentorial malignant gliomas at the Memorial Sloan-Kettering Cancer Center and the New York University Medical Center. Thirteen patients had glioblastoma multiforme, 29 had anaplastic astrocytomas, and 8 had malignant gliomas. In 10 patients the tumor evolved from a low-grade lesion. Seven patients, including 2 patients with neurofibromatosis, developed multiple primary malignant neoplasms. The median time to tumor progression after surgery was 31 weeks, with local recurrence representing the mode of treatment failure in nearly all patients. Notable clinical features included symptomatic leptomeningeal metastasis (13 patients) and intratumoral hemorrhage (9 patients). The estimated median survival time for all 50 patients was 98 weeks, with a 3-year survival rate of 32%. A trend toward longer survival was seen in patients 12 years of age or younger at diagnosis. There was no apparent correlation between survival and tumor histology or tumor location. Recommendations for management are presented

A neoadjuvant (preradiotherapy) chemotherapy regimen consisting of either cyclophosphamide alone (60 to 80 mg/kg) or a modified multidrug regimen (vinblastine, bleomycin, cyclophosphamide, and cisplatin) was administered to 15 newly diagnosed patients with histologically confirmed, fully staged, primary germ-cell tumors (GCT's) of the central nervous system (CNS). There were 11 patients with germinomas and four with non-germinoma malignant GCT's. There were six females and nine males, whose median age was 13 years (range 4 months to 24 years). Seven germinoma patients (64%) had disseminated disease. For the germinoma patients, the subsequent radiotherapy dose was modified based on the response to the neoadjuvant chemotherapy, and craniospinal radiotherapy was given only to those with disseminated CNS disease at diagnosis. Ten of the 11 germinoma patients had complete disappearance of all evaluable disease after two courses of chemotherapy (cyclophosphamide in eight and multidrug in three) and one had a partial response. The planned dose of radiotherapy to the primary tumor was reduced from 5500 to 3000 rads, and the craniospinal dose was lowered from 3600 to 2000 rads. Ten patients remain in continuous disease-free remission 20+ to 89+ months after diagnosis (median follow-up period 47 months). All four patients with non-germinoma GCT's received the multidrug regimen, and two fo three patients with evaluable disease had a partial response. High-dose regional and craniospinal radiotherapy was administered thereafter, but only two patients remain in their first remission. Previously untreated germinoma is a highly chemosensitive disease and the neoadjuvant treatment strategy permits the identification of active chemotherapy regimens in newly diagnosed patients. Patients who have complete responses to neoadjuvant chemotherapy tolerate a significant radiotherapy dose reduction without compromising long-term survival, thereby allowing a reduction of some of the late effects of therapeutic radiation. Germinomas tend to disseminate early in the course of the disease and a pre-therapy staging evaluation permits individualized radiotherapy treatment planning

Carboplatin, a cisplatin analogue, was administered as an intravenous (IV) one-hour infusion in a 4-consecutive weekly dose schedule to 44 patients with recurrent childhood brain tumors. Twenty-four patients were registered on our phase I, and 20 on our phase II studies. The maximum tolerable dose derived from our phase I study was 210 mg/m2/wk in patients with solid tumors, and the recommended dose for subsequent pediatric phase II studies was 175 mg/m2/wk. This dose was administered to 14 patients in the phase I and all 20 patients in the phase II study. Nine of 36 (25%) evaluable patients in the combined studies experienced objective responses for a median duration of 10+ months. Seven of nine responders had received prior cisplatin. Disease-specific response rates were as follows: medulloblastoma, six of 14 (43%) with three complete (CR) and three partial responses (PR); pineoblastoma, one of one (PR); germinoma, one of two (CR); and brainstem glioma, one of eight (13%) (PR). Carboplatin had mild emetic effects but no significant auditory or renal toxicity. Thrombocytopenia (less than 49,000) was encountered in nine of 28 (32%) evaluable trials at a dose of 175 mg/m2/wk. Because of its low potential for auditory, renal, and emetic toxicity, ease of administration, and high disease-specific activity, carboplatin deserves further study in multiagent phase II and III trials, especially in chemotherapy-sensitive diseases such as medulloblastoma