Experimental models of brainstem tumors: development of a neonatal rat model.

OBJECTIVE: Brainstem tumor models are required to advance the treatment for diffuse pontine gliomas in children. The feasibility of creating an experimental rodent model by inoculating newborn pups with tumor cells was examined. The study was performed to create an animal model for diffuse brainstem tumors. METHODS: Eighty-two Fischer rat pups aged 12-24 h were anesthetized by hypothermia. The brainstem was injected with saline to identify anatomical coordinates for subsequent tumor cell challenges. The newborn pups were then inoculated with F98 (n=30) or 9L (n=30) glioma cells. Animals were returned to their mother for nursing. Tumor growth was assessed by survival and histopathology. RESULTS: Twenty-one percent of the saline-treated animals (17 out of 82) and 5% of the tumor cell-challenged pups (3 out of 60) were eliminated by their mothers. Inoculations with 9L and F98 cells produced brainstem tumors in 83% (24 out of 29) and 93% of animals (26 out of 28) respectively that were evaluated. CONCLUSIONS: Our results demonstrate that neonatal rat models for brainstem tumors can be prepared using known injection coordinates and orthotopic cell lines. Decreasing rates of maternal removal during the course of the work suggests that the method involves a learning curve.

New approach to tumor therapy for inoperable areas of the brain: chronic intraparenchymal drug delivery.

Because the brainstem has little functional redundancy, diffuse lesions have been regarded as inoperable. To determine whether local drug therapy can prolong survival in a rodent model of a tumor in such eloquent tissue, lethal doses of F98 and 9L tumor cells were injected into the brainstems of Fischer 344 rats. Five days after inoculations, 0.5 mg/ml solutions of carboplatin were infused at 1 microl/h for 7 days. Compared to control groups that survived 13-17 days with F98 tumors and 22-23 days with 9L tumors, animals locally infused with 0.1 mg of carboplatin survived 27-30 days (Prob > Chi Sq = 0.0003), and 32 days (Prob > Chi Sq = 0.01), respectively. Measurements of tissue platinum levels at autopsy suggested that infusions distributed pharmacologically relevant levels of carboplatin through a volume of tissue at least 0.5 cm in diameter. The results suggest that chronic low-flow infusions provide a promising approach to therapy for CNS lesions in tissues considered to be inoperable.