Spinal cord radiation tolerance

Abstract

Each case of radiation myelopathy represents a unique occurrence where an individual’s tolerance was exceeded. Therefore, each case offers the challenge to determine why this patient’s tolerance was exceeded at a dose that other patients tolerate. In radiation oncology, we believe that higher doses invariably result in higher incidences of late complications and that this dose-response is largely a result of the increasing probability of the killing of cells. We also hold that cell death from radiation is a stochastic event and consequently complications of radiation are stochastic events in as much as they depend on cell death. However, dose is not the only factor that is causally related to radiation myelopathy, and it is not just the unlucky patient who experiences radiation myelopathy at doses tolerated by other patients. The factors that may reduce a patient’s radiation tolerance may be identified through case control analysis, proportional hazards analysis, or certain multivariate regression methods. The proper methods of analysis depend on the exact nature set. In all cases, however, there must be some patients identified who share characteristics with the myelopathy cases, but who did not develop the injury. The article in this issue by Wong et al. (8) is a report of a survey of the experience of the Princess Margaret Hospital/Ontario Cancer Institute which quotes data on radiation myelopathy cases alone. The information in this report would be of greater clinical use if some indication had been provided of the total number of patients from which the myelopathy cases were drawn. Nonetheless, there are some useful measures of the incidence of myelopathy provided in this study. There were 18 cases of radiation myelopathy after single courses of thoracic treatment. In 9 of these 18, a split course regimen ofgreater than 40 Gy in 10 fractions was given-a well-known myelopathy-causing technique used mainly in the 1970s and now obsolete (1, 2, 4, 7). Historically, the dose regimen of 40 Gy in 10 fractions resulted in a 5-l 5% observed incidence of radiation myelopathy. Three of the remaining cases of thoracic myelopathy were following multiple fractions per day; 5 others received 15 to 20 fractions with doses per fraction generally about 3 Gy. The only conventional fractionation schedule among the thoracic cases was 15 X 2 Gy followed by 10 X 2.5 Gy. These investigators found no case of radiation myelopathyfrom the Princess Margaret Hospital/Ontario Cancer Institute in a patient who received 50 Gy (either to the spinal cord or midline) in 25 ,fractions. The only case of cervical myelopathy after 25 fractions or more occurred in a patient who received a dose of 60 Gy. Another case received 47.7 Gy in 20 fractions, and all other cervical myelopathy cases involved high doses per fraction or multiple daily fractions. The evidence is now that 50 Gy in 25 fractions yields an incidence of radiation myelopathy less than 0.5% and probably less than 0.2% (5,6, 3). This estimate is probably valid for most cases where the dose is calculated to the midline as well. However, the dose to the spinal cord should be kept under 50 Gy (lower if possible) as long as the prescribed dose can be delivered reliably to the clinical target volume. This suggestion is offered so that the rare case, often involving patients with conditions that reduce the spinal cord tolerance, can be avoided. It is well understood that heterogeneity in a population decreases the steepness of dose-response functions. It is also clear that the doses typically delivered to the spinal cord during a course of radiation treatment are in the flat portion of the dose-response function. Therefore, any case of radiation myelopathy resulting from conventionally fractionated treatments of about 50 Gy or less will likely be idiosyncratic and possibly as much a result of an individually low tolerance as of the dose to the spinal cord. For this reason, it would be helpful to pursue and report any factors that lead to reduced tolerance in such cases. A partial list of these factors includes hypertension, hypotension, acquired or congenital spinal defects, vascular diseases or vascular damage from diseases, and of course certain chemotherapeutic agents, Being aware of any of