Current experience with 12 patients studied prospectively suggests a new approach in the diagnosis and treatment of primary central nervous system (CNS) lymphoma, integrating the techniques of needle brain biopsy, immunohistochemical staining for monoclonal antibody and chemotherapeutic drug delivery in association with blood-brain barrier modification. Computed tomography (CT)-guided needle biopsy of deep parenchymal lesions contributed to the diagnosis in six patients. Immunohistochemical staining methods detected monoclonal immunoglobulins in those patients so tested. Following diagnosis, the patients have been treated with multi-agent chemotherapy in conjunction with osmotic blood-brain barrier modification (five without antecedent cranial irradiation) with an initial complete response rate by CT scan in nine patients, a median follow-up of 19 months from diagnosis, and a 1-year survival of 75%. This experience emphasizes the value of CT-guided stereotaxic or CT-guided needle biopsy, which limits the need for therapy without a diagnosis or the need for a major craniotomy in what are commonly deep, paraventricular lesions. Immunoperoxidase cytochemical stains can detect monoclonal immunoglobulin characteristic of CNS B-cell malignant lymphomas and provide an important diagnostic aid when only modest quantities of tissue or cells are obtained. Finally, chemotherapy administered in conjunction with osmotic blood-brain barrier modification results in a clinical response rate and survival that are at least as effective as radiotherapy as a primary therapeutic modality.

Medulloblastomas represent 20% of malignant brain tumors of childhood. Although, they show multiple, non-random genomic alterations, no common, early genetic event involving all histologic types of medulloblastomas have been described. Nineteen medulloblastomas were analyzed using chromosomal comparative genomic hybridization (cCGH). Nine tumors with the most frequent number of genetic changes were further analyzed using bacterial artificial chromosome array CGH (aCGH). With aCGH, the frequency of gains and losses were higher than with cCGH. Chromosome 2p gains spanning 2p11-2p25 including N-myc locus, 2p24.1 were detected in 5/9 (55%) tumors while 14q12 gains were detected in 6/9 (67%) tumors. The 14q12 locus overlapped with the FOXGI gene locus. Quantitative real time PCR showed a 2-7-fold copy gain for FOXG1 in all the nine tumors. Protein expression was demonstrated by immunohistochemistry in all histologic types. The expression of FOXG1 and p21cip1 showed an inverse relationship. FOXG1 copy gain (>2 to 21 folds) was seen in 93% (55/59) of a validating set of tumors and showed a positive correlation with protein expression (Spearman's rank order correlation coefficient = 0.276; P = 0.038) representing the first report of FOXG1 dysregulation in medulloblastoma. Modulation of FOXG1 expression in DAOY cell line using siRNA showed a modest decrease in proliferation with a 2-fold upregulation of p21cip1. Current reports indicate that FOXG1 represses TGF-beta induced expression of p21cip1 and cytostasis, and forms a transcriptional repressor complex with Notch signaling induced hes1. Our findings are consistent with a role for FOXG1 in the inhibition of TGF-beta induced cytostasis in medulloblastoma.

In the past, chemotherapeutic treatment of patients with high grade malignant gliomas following surgery and radiation has not added significantly to the 12-14 month median survival rate. Over four years, 37 patients with high grade malignant gliomas underwent 246 treatment procedures with a combination of methotrexate, cyclophosphamide, and procarbazine given in association with hyperosmolar mannitol-induced transient breakdown of the blood-brain barrier. These patients have demonstrated a median survivorship of 22 months after considering age, Karnofsky Performance Score, and necrosis by the Cox Proportional Hazards model. The study group had a mean age of 43 years, and mean Karnofsky Performance Score of 67%. Sixty-five percent of the procedures had well-documented barrier disruption. Sixteen percent remained in complete remission while 24 patients (65%) had partial or temporary remission. Progression-free intervals after blood-brain barrier disruption/chemotherapy ranged from 1-47 (mean 15) months. Neurotoxicity has been minimal with one peri-procedural mortality and five patients suffering an increase in neurologic deficit after a procedure. The results of this study are consistent with and further extend the reported literature on this method of brain tumor therapy as described in other centers. Chemotherapy in conjunction with osmotic disruption of the blood-brain barrier may provide the pharmacokinetic advantage sufficient to significantly improve survival in patients with high grade malignant glioma.