Intramedullary spinal cord astrocytomas in children

Intramedullary spinal cord astrocytomas are uncommon tumors, accounting for less than 3% of all CNS neoplasms. In view of the rarity of these lesions, few single institution cohorts have incorporated sufficient numbers of patients with an adequate length of follow-up to permit more than an anecdotal assessment of the impact of various treatment parameters on outcome. This has led to particular controversy in the management of low-grade astrocytomas. Some authors strongly advocate a consistent policy of aggressive resection followed by observation for these tumors, with a high percentage of patients exhibiting short-term disease control and acceptable levels of morbidity 1,2. This has been the management approach followed at our institution as well. However, given the potentially indolent growth of these tumors, and the fact that progression can occur years after an initial resection, the absence of long-term follow-up precludes definitive conclusions regarding disease control after surgery alone. Conversely, other groups have favored more conservative surgery followed by irradiation 3, a strategy designed to diminish the risk of acute morbidity from surgery, but perhaps at a cost of increased late morbidity from radiation-induced sequelae. As a third management alternative, the report of Townsend et al. and others 4 favors the use of chemotherapy after conservative surgery, and the results in terms of disease control also seem acceptable. Certainly, the use of a well-tolerated chemotherapy regimen that has known activity against low-grade gliomas, as was applied in this report, represents a logical attempt to minimize surgical morbidity and also avoid the potentially neurotoxic effects of irradiation, and this seems to represent an excellent strategy in those instances in which an extensive resection is felt to be too risky. The latter issue is a “judgment call” and is influenced by the growth characteristics of the tumor, the experience of the surgeon, the reliability of intraoperative monitoring (as a way to avoid rather than merely detect irreversible neurological injury), and the risk tolerance of the patient and family. Even in experienced hands, severe neurological morbidity is an inherent risk of attempted tumor resection 5 and this serious risk needs to be weighed against the possibility of improved tumor control with more extensive tumor removal. In view of the small number of patients in the current series who were treated with conservative surgery and chemotherapy as an initial management approach and the relatively short interval of follow-up, it is difficult to be certain that the use of chemotherapy has actually influenced the natural history of the residual disease, which in some instances will remain stable without any adjuvant therapy. Nonetheless, the current report provides justification for a more extensive analysis of this management approach, and calls attention to the need for a controlled multi-institutional study evaluating various treatment strategies for children with low-grade spinal astrocytomas, who have historically not been enrolled on cooperative group low-grade glioma studies. Since there are fairly entrenched views at different institutions regarding the “optimal” management of these lesions, which would probably preclude treatment randomization, such an analysis would probably need to be a natural history study, to allow the collection of reliable baseline data. In the absence of such a study, the optimal management for low-grade spinal astrocytomas will continue to be conjectural and controversial. In contrast to the diversity in opinions regarding the management of low-grade astrocytomas, there is general agreement that high-grade spinal astrocytomas require multidisciplinary intervention. These tumors are clearly not amenable to gross total resection and often exhibit evidence of leptomeningeal dissemination at, or shortly following, diagnosis. Because these lesions are even less common than low-grade tumors, adjuvant therapies have generally been examined in parallel with studies of intracranial high-grade gliomas and the outcome results have been equally poor. One treatment protocol that specifically had a separate stratum for spinal cord high-grade gliomas, the Children's Cancer Group CCG-945 study, used “8-in-1” chemotherapy and irradiation, and concluded that this regimen was suboptimal for achieving disease control 6. At present, a variety of novel treatment approaches are being examined for high-grade gliomas, including concurrent chemotherapy and irradiation, and growth factor receptor inhibitors, alone and in conjunction with irradiation. Because these protocols are not specifically targeting spinal cord malignant gliomas, it is likely that only a handful of such patients would be accrued on any given study. Nonetheless, it is hoped that insights from the management of high-grade gliomas in general would be applicable to the management of spinal cord lesions in order to improve the otherwise discouraging prognosis of these tumors.