The authors hypothesized that chemotherapy infusions directly into the fourth ventricle might potentially play a role in treating malignant fourth ventricular tumors. The study tested the safety and pharmacokinetics of short- and long-term infusions of methotrexate into the fourth ventricle in a new nonhuman primate model.
Six rhesus monkeys underwent posterior fossa craniectomy and catheter insertion into the fourth ventricle. In Group I (3 animals), catheters were externalized, and lumbar drain catheters were placed simultaneously to assess CSF distribution after short-term methotrexate infusions. In 2 animals, methotrexate (0.5 mg) was infused into the fourth ventricle daily for 5 days. Serial CSF and serum methotrexate levels were measured. The third animal had a postoperative neurological deficit, and the experiment was aborted prior to methotrexate administration. In Group II (3 animals), catheters were connected to a subcutaneously placed port for subsequent long-term methotrexate infusions. In 2 animals, 4 cycles of intraventricular methotrexate, each consisting of 4 daily infusions (0.5 mg), were administered over 8 weeks. The third animal received 3 cycles, and then the experiment was terminated due to self-inflicted wound breakdown. All animals underwent detailed neurological evaluations, MRI, and postmortem histological analysis.
No neurological deficits were noted after intraventricular methotrexate infusions. Magnetic resonance images demonstrated catheter placement within the fourth ventricle and no signal changes in the brainstem or cerebellum. Histologically, two Group I animals, one of which did not receive methotrexate, had several small focal areas of brainstem injury. Two Group II animals had a small (≤ 1-mm) focus of axonal degeneration in the midbrain. Intraventricular and meningeal inflammation was noted in 4 animals after methotrexate infusions (one from Group I and all three from Group II). In all Group II animals, inflammation extended minimally into brainstem parenchyma. Serum methotrexate levels were undetectable or negligible in both groups, ranging from 0.00 to 0.06 μmol/L. In Group I, the mean peak methotrexate level in fourth ventricle CSF exceeded that in the lumbar CSF by greater than 10-fold. Statistically significant differences between fourth ventricle and lumbar AUC (area under the concentration-time curve) were detected at peaks (p = 0.04) but not at troughs (p = 0.50) or at all time collection points (p = 0.12). In Group II, peak fourth ventricle CSF methotrexate levels ranged from 84.62 to 167.89 μmol/L (mean 115.53 ± 15.95 μmol/L [SD]). Trough levels ranged from 0.06 to 0.55 μmol/L (mean 0.22 ± 0.13 μmol/L).
Methotrexate can be infused into the fourth ventricle in nonhuman primates without clinical or radiographic evidence of injury. Observed inflammatory and other histological changes had no clinical correlate. This approach may have pharmacokinetic advantages over current treatment paradigms. Further experiments are warranted to determine if fourth ventricular chemotherapy infusions may benefit patients with malignant fourth ventricular tumors.