Drs. Chang, Adler, and Hancock provide a thorough overview of the delivery devices, as well as the clinical utilization, of radiosurgery in malignant and nonmalignant central nervous system (CNS) lesions. As is well known in radiosurgical practice, the ideal targets for these single, very precise, high doses of radiation would be small, space-occupying lesions somewhat separate from radiosensitive structures. However, questions remain as to the optimal dosing and fractionation schedule when balancing risk-benefit ratios for the treatment of certain diseases mentioned in the article by Chang et al. These questions have arisen because we now know that multiple fractionation regimens allow for greater cumulative dosage, as well as improved sparing of normal tissues.
One of the major difficulties when evaluating new medical interventions, however, is summarized in the simple question, will it be an improvement over the current standard of care? For patients with pituitary adenoma following subtotal resection or with continued hormonal abnormalities, external-beam radiotherapy has been used effectively for many decades. Tsang et al, when evaluating their experience at the Princess Margaret Hospital during the 1970s and 80s, found a greater than 90% 10-year actuarial local control rate in patients given postoperative radiation therapy. In the multivariate analysis, patients at risk for poor local control included older patients and those with larger tumors (the latter group being those who would be ineligible for radiosurgery). As mentioned in the article by Tsang et al, only 20% of cases of hypopituitarism are attributable to radiation. Other centers have reported similar findings.
Fractionated radiation has also proven to be highly effective in patients with hormonally active tumors. Recently, Estrada et al evaluated fractionated radiation to 50 Gy following unsuccessful transsphenoidal hypophysectomy in patients with Cushings disease. Of these patients, 83% achieved a remission during a median follow-up of 42 months (with the vast majority of remissions occurring during the first 2 years). None of the patients had a relapse of Cushings disease after remission was achieved.
As for morbidity related to external-beam radiotherapy in patients with pituitary adenomas, long-term follow-up shows that up to 50% may need hormonal replacement at some time.[4,5] However, many of these hormonal deficiencies also occur following surgical intervention. Long-term visual complications are relatively rare following radiation therapy when surgical damage is taken into account, and appear to occur in less than 1% of patients treated with radiation therapy.[4,5]
With respect to second malignancies caused by moderate-dose radiation in the pituitary adenoma population, Bliss et al found only one malignant brain tumor among patients treated over a 30-year period in Edinburgh. This rate of incidence falls well within the expected number of malignant tumors in the normal population. The likelihood that radiosurgery will prove more effective or safer than external-beam radiation therapy seems doubtful.
Acoustic neuromas are benign lesions that, in the short term, are well controlled with radiosurgery. The difficulty with radiosurgical intervention is the morbidity related to higher doses. However, the higher doses should afford greater long-term control. In a radiosurgical series with longer follow-up, reported facial and/or trigeminal nerve morbidities ranged from 19% to 33%.[7-9] One approach to decreasing the frequency and severity of neuropathies was to decrease the dose. Unfortunately, while the number of complications may decrease with lower doses of single-fraction radiosurgery, it will probably be at the expense of overall control rates.
Use of standard-fractionation or hypofractionated stereotactic radiotherapy should, again, allow a return to the radiobiological principles learned over decades for sparing of normal tissues (by using smaller daily fractions) and also allow an increase in the overall dose. Standard fractionation to total doses of 54 Gy has achieved 100% local control (albeit with short follow-up, ie, median survival of just over 2 years). In addition, no patient developed new cranial nerve deficits, and only one patient developed a worsening of a preexisting fifth cranial nerve neuropathy. Hypofractionation also has achieved, at least with short-term follow-up, excellent control of acoustic neuromas, with less than a 5% incidence of hearing loss or transient facial weakness.
Questions remain as to the ultimate role of "radiosurgery." Presently, a wide array of lesions are being treated with large, single doses of radiation in an attempt to localize effects. However, can we be sure that the localized effect on brain tumors will not, with long-term follow-up, impact regional tissues possibly as a result of effects on the microvasculature?
Radiosurgery has been shown to be a boon in the treatment of a number of lesions previously unapproachable by standard methods (eg, arteriovenous malformations). Current randomized studies are attempting to confirm the excellent results achieved in certain brain tumors (rivaling the best surgical series). However, it remains to be shown whether single large doses of radiation are the optimal therapy for many other lesions occurring in patients with many decades of life ahead of them, when compared with standard therapy and newer technologies. Radiation hardware for external-beam use can now deliver radiation with a precision and accuracy equal to radiosurgery units, while achieving fractionation, conformation, and homogeneous doses to the target volume, thereby allowing the physician to mesh radiobiological principles with modern technology.