Radiosurgical Boost for Primary High-Grade Gliomas

Abstract & Commentary

Synopsis: Glioblastoma multiforme represents approximately one quarter of brain tumors in adults. Median survival is typically 10-12 months. Efforts at improving survival using radiotherapy dose escalation have been largely disappointing, including attempts at interstitial brachytherapy, hyperfractionation, and higher 3D boost doses. This study from Brazil reports that, in their small series, patients with the best pretreatment performance status exhibited significantly improved survival following a radiosurgery boost to residual disease after external beam radiation.

Source: Prisco FE, et al. J Neurooncol. 2002;57: 151-160.

Despite many attempts to escalate the radiation dose in glioblastoma multiforme (GBM), investigators continue to explore various means of overcoming this tumor’s propensity to recur locally. It is acknowledged that 80% of these lesions recur within 2 centimeters of the initial tumor volume, with the majority in the high-dose region. This suggests a role for dose escalation, though no dose-response above 60 Gy in 30 fractions was seen in several randomized trials reported for the past 20 years. Prisco and colleagues reasoned that some of the dose constraints associated with normal tissue tolerance, ie, neurotoxicity, might be overcome by using stereotactic radiosurgery (SRS) as a boost in irradiated patients. In SRS, high doses can be delivered to the tumor while at the same time delivering minimal doses to the normal surrounding tissue owing to an inherently steep dose gradient.

Between 1993 and 1998, 38 patients were treated for high-grade gliomas at the Hospital Israelita Albert Einstein in Sao Paulo, Brazil. Prisco et al performed a retrospective analysis to assess the efficacy of a SRS boost, which was used in fifteen patients. Thirty-two patients had GBM and 6 had anaplastic astrocytomas. Regarding the GBM patients only, the median age was 51 years (r, 21-78). Twenty-eight underwent resection and 4 had a biopsy only. Seventeen patients had a Karnofsky Performance Status (KPS) of > 80, and the rest were < 80. Median external beam radiotherapy dose was 60 Gy (r, 53-66). The SRS and control GBM groups were well balanced demographically, except for tumor volume. Mean initial tumor volume was 67cc for the control group and 37cc for the SRS group (P = .04).

The decision regarding whether to boost with SRS was made based on MRI findings following 60 Gy. Those patients with < 4 cm residual contrast-enhancing tumor > 5 mm from the optic chiasm were boosted. Median time to the boost was 5 weeks (r, 1-13). Median tumor volume treated with SRS was 15 cc (r, 3-70). Median SRS dose was 10 Gy (r, 8-12.5). Median maximum dose was 15 Gy (r, 12.5-24 Gy).

There was no significant difference in the external beam RT dose to either group. One patient in each group was lost to follow-up. Follow-up for living patients ranged from 10.5-67.9 months. Median actuarial overall survival for the entire group was 15 months. It was 11.6 months for the control group, and 21.4 months for the SRS group (P = .025). Among only patients with a KPS > 80, median actuarial survival was 11 months for the controls, and 53.9 months for the SRS patients. Comparison limited to patients in the control group with initial tumor volumes comparable to the SRS patients still resulted in a significant advantage for the SRS group (21.4 vs 14.1 months; P = .03). Multivariate analysis to assess the importance of age, KPS, extent of surgery, and SRS boost on survival indicated that only SRS was a significant independent prognostic factor.

Prisco et al cited small sample size and the inherent selection bias associated with retrospective analyses as obvious limitations of their study. Nevertheless, given the survival benefit seen in their patients with the highest KPS, they postulated that SRS may overcome radioresistance due to a more intense radiobiological effect of a single large-dose treatment.

Comment by Edward J. Kaplan, MD

Radiosurgery was first described by Leksell in 1951.1 He used a battery of orthovoltage beams to concentrate high doses of radiation at the intersection of the beams with relative sparing of surrounding normal tissue. Today, the same approach is accomplished with either x-rays from linear accelerator-based systems (eg, X-Knife), or from gamma rays generated by a cobalt-60 based system (Gamma Knife). One advantage of SRS is its minimally invasive nature, as opposed to interstitial brachytherapy which requires free access to brain parenchyma.

Data from previous Radiation Therapy Oncology Group (RTOG) trials are commonly used to gauge the effectiveness of therapies being evaluated in ongoing trials. For example, even the most favorable patients (ie, those < 50 years old with a KPS of > 90) had a median overall survival of 18 months in 1 publication from 1993 using pooled data from 3 prior RTOG series including 1500 patients.2 Other interesting results can be gleaned from some of the older papers, like a report on RTOG 83-02 that described an improvement in survival for SRS-eligible patients compared with ineligible patients, though no SRS was administered.3 Similar findings for brachytherapy-eligible vs. ineligible patients have been reported.4

Modern reports regarding SRS boosting of GBM patients were published from Harvard and the University of Maryland by Shrieve5 and Nwokedi,6 respectively. Shrieve used an X-Knife system to deliver a median minimum peripheral dose of 12 Gy (r, 6-24) to a median tumor volume of 9.4 cc (r, 0.86-72). Treatments were normalized to a median 85% isodose (r, 60-100). Only tumors < 4 cm on preoperative MRI, exclusive of edema, received a SRS boost. Median actuarial survival was 19.9 months. Nwokedi et al used a Gamma Knife to deliver a median dose of 17 Gy (r, 10-28) to a median tumor volume of 18.5 cc (r, 1.6- 62.2). Treatments were normalized to a median 50% isodose line (r, 45-65). Median actuarial overall survival was significantly improved on multivariate analysis for the 31 patients receiving the SRS boost compared to the 33 that did not (25 vs 13 months; P = .03).

Data from RTOG 93-05, a Phase III trial comparing the use of SRS followed by RT and BCNU vs. RT and BCNU alone, will be reported in an oral presentation at ASTRO. Results from this study that enrolled 200 patients may help us better understand whether there is a real role for SRS boosts in GBM patients. The doses used in this trial depended on the size of the lesion, from 24 Gy for tumors < 2 cm to 15 for lesions 31-40 mm. Doses were prescribed to the 50- 90% isodose lines. Another ongoing Phase II trial, RTOG BR 0023, is designed to assess the efficacy of accelerated RT and a weekly SRS boost during the last half of treatment followed by carmustine. It will be interesting to see what direction we are sent in when the results are out. I believe we may have to repeat some of this work using functional imaging to supplement MRI imaging so that we can better define our target before we can say that dose escalation is not the answer, or is the answer for only those patients with smaller lesions. 

Dr. Kaplan is Acting Chairman, Department of Radiation Oncology, Cleveland Clinic Florida, Ft. Lauderdale, FL; Medical Director, Boca Raton Radiation Therapy Regional Center, Deerfield Beach, FL.


1. Leksell L. Acta Chir Scand. 1951;102:316.

2. Curran WJ, et al. J Natl Cancer Inst. 1993;85:704-710.

3. Curran WJ, et al. J Clin Oncol. 1993;11:857-862.

4. Florell RC, et al. J Neurosurg. 1992;76:179-183.

5. Shrieve DS, et al. J Neurosurg. 1999;90:72-77.

6. Nwokedi E, et al. Neurosurgery. 2002;50:41-47.