Malignant cells in cerebrospinal fluid (CSF): the meaning of a positive CSF cytology.

We reviewed the correlation between malignant cells in cerebrospinal fluid (CSF) (positive cytology) and pathologic findings at autopsy. The purpose was to discover: (1) the incidence of negative CSF cytology in patients with CNS malignancy, (2) the incidence of false-positive cytology, and (3) the relationship between a true-positive cytology and the distribution of malignant tumor at autopsy. Of 117 patients with CNS tumor and premortem cytologic examination of the CSF, 31 (26 percent) were positive and 86 (74 percent) were negative. Only 1 of 66 patients with tumor that did not reach the leptomeninges had a positive cytology. Of 51 patients with leptomeningeal tumor at autopsy, cytology was positive in 30 (59 percent) and negative in 21 (41 percent). Five potentially "false-positive" cytologies were encountered: three patients were treated, and tumor may have been eradicated; in two patients with lymphoma, inflammatory cells associated with infection were apparently mistaken for malignant cells. These data indicate that a positive CSF cytology is a reliable indicator of CNS malignancy and almost always reflects leptomeningeal tumor.

Sterilization of cerebral metastases by radiation therapy.

It is generally acknowledged that radiation therapy (RT) provides effective palliative treatment for the majority of patients with brain metastases. It is less well known that RT can be "curative." In 10 years, we examined five patients with brain metastases treated with whole-brain radiotherapy and with no residual tumor in the brain at autopsy. These five patients represented about 3 percent (5 of 187) of patients treated for brain metastases and later examined at autopsy. Some metastatic brain tumors are eradicated by RT, and all nonterminal patients should be offered such treatment.

Human brain tumor transplantation into nude mice.

Seven human brain tumors were transplanted into the brains (6/7 takes) and subcutaneous tissues (7/7 takes) of athymic nude mice. Compared to experimental animal brain tumors, these tumors, taken directly from patients in the operating room and transplanted, grew more slowly in the mice; their growth rates following explant generally paralleled those in the patients. A rough correlation was seen between the degree of the tumor's malignancy and both successful take and rate of growth following explant. The tumors' growth rates increased during serial transplantation after explant. Two tumors developed into long-term serial lines; both came from gliosarcomas. Preliminary chemotherapy experiments with these two lines demonstrated different chemosensitivities. One line was very sensitive to the nitrosoureas and resistant to procarbazine; the other line was more sensitive to procarbazine than to the nitrosoureas. This model permits study of the biologic behavior of human brain tumors growing intracerebrally and subcutaneously in nude mice.

Metastic tumor of the brain: development of an experimental model.

The injection of a suspension of Walker 256 carcinoma cells into the carotid artery of rats produced a model of hematogenously spread cerebral metastases. Most animals died from massive extracerebral tumors of the head and jaw; brain tumors were present in only one-quarter. External carotid artery ligation prior to tumor inoculation did not increase the incidence of fatal brain tumor. When cyclophosphamide, 15 mg/kg, was injected as a single dose on the fourteenth day after tumor inoculation, most of the extracerebral tumor disappeared. Fifty percent of the animals so treated were cured of tumor, but 90% of the remainder died of large intracerebral tumors. Severe cytopathic changes resulting from cyclophosphamide were present in extracerebral or choroid plexus tumors but were mild or nonexistent in intracerebral tumors. These selective effects of cyclophosphamide suggest that water-soluble agents are less effective against tumor in the brain than against extracerebral tumors despite the fact that metastatic tumor breaks down the blood-brain barrier.

Central nervous system infections in patients with cancer. Changing patterns.

Central nervous system infections in patients with cancer at Memorial Sloan-Kettering Cancer Center from 1971 to 1974 were surveyed and compared with a previous survey from 1955 to 1970. The two periods were similar in that: 1) There was a high incidence of CNS infection in patients with lymphoma, leukemia, and head and spine tumors, 2) specific organisms tended to infect patients with certain primary neoplasms, and 3) fungal infections were common and difficult to diagnose. The two periods differed in that: 1) the overall incidence of CNS infection was lower in 1971 to 1974, 2) there was a decreased incidence of cryptococcal meningitis, and 3) there was an increase in Listeria monocytogenes meningitis. Early recognition of CNS infection and aggressive therapy appears to increase survival.

Cerebral necrosis following radiotherapy of extracranial neoplasms.

We have examined 6 patients with delayed cerebral necrosis following irradiation of extracranial neoplasms. Four of the 6 patients received 1,760 rets (or less) tumor dose. The initial symptoms attributable to radiation necrosis appeared 4 to 31 months after irradiation and were those of a focal supratentorial mass. Cerebral angiography delineated an avascular frontal or temporal lesion in all 6 patients; in 1 case a magnification study revealed narrowing, irregularity, and occlusion of small cortical vessels. Four of our 6 patients underwent craniotomy with partial or complete surgical extirpation of necrotic brain tissue. Two operated patients are alive and without disabling neurological symptoms 30 and 25 months, respectively, after the operation. The characteristic neuropathological features of delayed radiation necrosis of brain suggest that vascular injury rather than neuronal or glial damage is of primary pathogenetic significance.

Meningeal carcinomatosis: development of an experimental model.

An experimental model of meningeal carcinomatosis has been produced by intracisternal inoculation of Walker 256 carcinoma cell suspension into Wistar rats. The tumor grows rapidly and is fatal in about 15 days if 10(6) cells are injected. The histopathological pattern observed is similar to that seen in diffuse leptomeningeal involvement of systemic cancer in human beings. The model will be useful for investigating the pathophysiology of the neurological disability produced by meningeal carcinomatosis and the efficacy of chemotherapeutic agents.

Computer tomography: the "resolving" intracerebral hemorrhage.

Two patients with a total of 4 intracerebral hemorrhages were examined by serial computer tomography (CT) over a period of up to 6 months, revealing a progressive decrease in lesion volume and density which eventually became less than that of the surrounding brain tissue. These changes were interpreted as complete clot resorption with cystic residue. However, postmortem examination revealed residual hematomas with no evidence of recent bleeding. Three of these correspond in size to those noted on the initial CT scans. Intraparenchymal hemorrhages have stages of organization and resolution which may simulate normal brain density without significant volumetric alteration.

Neurologic manifestations of intravascular coagulation in patients with cancer. A clinicopathologic analysis of 12 cases.

Among 1,459 autopsied patients with cancer, 12 had multifocal infarcts of the brain that appeared to be caused by intravascular coagulation. Most of these patients were women with leukemia or lymphoma, and all had a clinical course in which neurologic signs and symptoms were prominent. All had evidence of generalized brain disease (delirium and stupor or coma), and several also had focal brain disease (focal seizures, hemiparesis). All patients had laboratory evidence of coagulation abnormalities, although these were often not severe when neurologic symptoms began. Pathologically, there were multifocal hemorrhagic or ischemic infarcts in the distribution of several cerebral vessels, without a systemic source for cerebral emboli. Fibrin thrombi were identified in cerebral vessels and in vessels of several other organs. The clinical findings fit the pathologic picture, and in most instances the correct diagnosis might have been made earlier had it been considered.