Patients with brain metastases are traditionally treated with surgery, whole-brain radiotherapy (WBRT), or stereotactic radiosurgery (SRS).1, 2 In the past decade, the use of SRS in the treatment of brain metastases has grown considerably. Recent studies have revealed an overall survival advantage with the incorporation of SRS after WBRT.3
Although SRS is commonly used in the treatment of patients with brain metastases, the application of a stereotactic head frame is associated with patient discomfort and a reluctance to undergo the procedure.4 Along with the development of single-fraction SRS treatment, radiosurgery systems were developed to allow fractionated stereotactic radiotherapy (SRT) approaches for the treatment of intracranial tumors. Because of the radiobiologic advantage of fractionated treatment on sensitive structures in the brain,5 relocatable head frames were developed to allow for stereotactic delivery over many days and weeks without the need for repeated invasive head frame applications. Such systems were found to provide accuracy equivalent to that of single-fraction systems, while providing fractionated protection of eloquent structures.1
With the growth of the use of fractionated SRT in the brain, some investigators began to explore the use of SRT for the treatment of brain metastases.6-8 Although these reported experiences have been encouraging, to our knowledge, no clear treatment guidelines have been provided to date due to the paucity of data. Beginning in 2004, our group began to treat brain metastases with hypofractionated stereotactic radiotherapy (HSRT). The current study describes our experience.
- Top of page
- MATERIALS AND METHODS
- Conflict of Interest Disclosures
For the treatment of brain metastases, WBRT followed by a planned SRS boost has gained much support from several randomized trials. Kondziolka et al first reported that WBRT with SRS had a 92% local control rate at 1 year compared with 0% with WBRT alone in 27 patients with 2 to 4 brain metastases each.11 In the Radiation Therapy Oncology Group (RTOG) 9508 study, 333 patients with 1 to 3 brain metastases each were randomized to WBRT plus SRS versus WBRT alone. The results demonstrated better local control rates at 1 year in the group treated with WBRT and SRS compared with the patients treated with WBRT alone (82% vs 71%; P = .01).3
SRS is typically delivered in a single fraction using an invasive head frame. Although the frame provides accurate localization and minimizes motion, patients are often reluctant to undergo standard head frame placement because of pain and general discomfort. Because of the radiobiologic advantage of fractionated treatment on sensitive structures in the brain,5 radiosurgery systems were developed with relocatable head frames to allow for stereotactic delivery over multiple fractions without the need for repeated invasive head frame applications. Such fractionated approaches have been found to provide accuracy equivalent to that of single-fraction SRS, while providing protection for eloquent structures through fractionation.
The results of this retrospective study demonstrate that HSRT is a very effective and safe treatment approach that is comparable to SRS at controlling brain metastases. Although radiographic shrinkage occurred in 25 (48.1%) of the 52 treated lesions, durable, long-term local tumor control was still maintained in a large percentage of the remaining patients who did not demonstrate a radiographic response. In this respect, the median overall survival time was 10.8 months and the actuarial local tumor control rates at 6 months and 1 year were 93.9% and 68.2%, respectively. These results are similar to those of previous single-fraction SRS studies.7, 12 Compared with WBRT plus SRS as used in the RTOG 9508 trial (which reported local control rates of 82% and 71%, respectively, at 3 months and 1 year),3 the local tumor control rates produced by HSRT in the current series were comparable.
With regard to overall survival, concurrent chemotherapy (P < .001) and the number of treated lesions (P = .036) were found to be statistically significant factors that affected overall survival. Concurrent chemotherapy and having a large number of treated lesions were found to be negatively correlated with overall survival. This was considered to be because patients treated with concurrent chemotherapy have more extensive or bulkier extracranial disease and patients with more metastatic lesions have a more advanced stage of disease. Both cases were found to be worse outcomes.
Previous investigators have demonstrated that maximum tumor dimension (in mm) and volume (in cc) are prognostic factors for local tumor control and toxicity.13 We found similar results in the current study and there was a significant difference in the local tumor control rates among the treatment groups when segregated by maximum tumor dimension (≤20 mm vs >20 mm and ≤30 mm vs >30 mm; P = .033). Tumors measuring ≤20 mm were found to have better actuarial local control rates at 1 year than tumors measuring >20 mm (81.5% vs 37.5%). With regard to tumor volume, patients with a tumor volume ≤1.0 cc were found to have a better tumor control rate (P = .028). Actuarial local tumor control rates for tumors with volumes ≤1.0 cc and >1.0 cc at 6 months were 100% and 87.5%, respectively. Furthermore, local control rates for tumors with volumes of ≤1.0 cc and >1.0 cc at 1 year were 100% and 46.7%, respectively. This finding also may be closely supported by the discovery that even the treatment group with tumor volumes ≤2.0 cc had a better chance of achieving local tumor control than those with tumor volumes >2.0 cc (P = .068). Similarly, Varlotto et al reported that a larger tumor size significantly decreased the rate of local tumor control in patients treated with SRS (P = .0029).14 In their study, tumors with volumes ≤2 cc and tumors with volumes >2 cc had actuarial local control rates of 95.2% and 83.3%, respectively, at 1 year. For larger tumor volumes, the results of the current study demonstrated a trend toward even poorer tumor control. With a larger patient cohort, this likely would have been statistically significant. The current study demonstrated a difference in the local tumor control rate between tumors with volumes ≤3.0 cc (95%) and tumors with volumes >3.0 cc (75%) (P = .085). Aoyama et al reported that the actuarial local tumor control rates of tumors with volumes ≤3.0 cc and >3.0 cc at 1 year were 96% and 59%, respectively.15 We also found that a maximum tumor dimension of <30 mm was statistically significant for local tumor control rates (P = .009) but, considering the sample size of tumors with a maximum dimension of >30 mm in the current study, it was a less reliable finding.
Early in our experience, we started with a HSRT prescription of 25 Gy delivered in 5 fractions. This dose selection was chosen based on the inferior outcomes of administering 20 Gy in 4 to 5 fractions as shown by Ernst-Stecken et al and Shepherd et al.12, 16 Furthermore, previous investigators had shown the safety and efficacy of higher doses within the range of 30 Gy to 40 Gy.17 Vordermark et al demonstrated that a HSRT dose of 30 Gy resulted in longer overall survival than doses of <30 Gy while being able to be delivered safely and effectively.18 For the upper limit, Shepherd et al previously reported that a HSRT dose of 30 Gy to 35 Gy was tolerable but a dose of >40 Gy was a significant predictor of radiation damage.16 In the current study, the total HSRT dose was not found to be a statistically significant factor for local tumor control (P = .684), and there was no major difference in tumor control noted between the treatment groups who received HSRT doses of 20 Gy, 25 Gy, and ≥30 Gy. Crude local control rates were 100%, 90.3%, and 88.9%, respectively, in these 3 groups. However, we were unable to demonstrate that other dose classifications (20 Gy, 25 Gy, and ≥30 Gy) were statistically significant for local tumor control because of our limited sample size. Therefore, there is a great need for a comparison of the multiple levels of doses between 25 Gy and 35 Gy in a prospective study.
Investigators have previously examined the influence of WBRT in conjunction with SRS on tumor control.19 Lindvall et al compared WBRT with a HSRT boost with HSRT alone and reported that WBRT combined with HSRT demonstrated better local control than HSRT alone for brain metastases (100% vs 84%).8 However, we found that HSRT alone or WBRT with a HSRT boost had actuarial local tumor control rates of 100% and 92.9%, respectively, at 6 months. Thus, we were unable to demonstrate an effect on local tumor control from WBRT. In our experience, the influence of WBRT was difficult to ascertain potentially because of the disproportionate number of patients in our sample (86.5% [45 of 52 patients]) who had received prior WBRT with an HSRT boost.
Previous investigators have examined the possibly favorable radiobiologic effect of HSRT because it allows for the delivery of a higher dose with fewer adverse events than SRS.6-8 In the current study, 1 patient of 27 had a grade 3 adverse event (3.7%), a finding that was similar to previous studies.11 In the Eastern Cooperative Oncology Group 6397 SRS study, there was 1 grade 3 seizure and 2 other grade 3 adverse events (fatigue and neutropenia, respectively) reported among 31 patients.19 When we examined the influence of higher total doses on toxicity, no relation was found. None of the 4 tumor lesions that received 36 Gy in 6 fractions developed grade 3 toxicity or radiation necroses. Future studies including a larger number of patients and close follow-up will be needed to understand this issue better.
In the current study, we found HSRT to be comparable to single-fraction SRS in terms of local tumor control and toxicity, and thus we believe it provides an alternative treatment choice for patients with brain metastases. We also confirmed that tumor size is a strong prognostic factor for local tumor control. We found fraction dose and prior WBRT were not statistically significant factors with regard to local tumor control with our limited sample size. Therefore, we believe there is a need for a larger prospective study to establish dosing guidelines for HSRT and to pave the way for a randomized trial to compare single-fraction SRS with a hypofractionated approach.