As the most common intracranial tumor, brain metastases (BMs) occur in up to 40% of all oncological patients.1,2 Local treatments for BM include stereotactic radiosurgery (SRS), microsurgical resection, and whole-brain radiation therapy (WBRT).1,3 In contrast to microsurgical resection, Gamma Knife radiosurgery (GKRS) is a noninvasive neurosurgical method, which also allows treatment in multimorbid patients with contraindications to general anesthesia during surgery.4–6 Furthermore, SRS is the only local treatment method for multiple disseminated and thereby nonresectable BMs.1 In general, microsurgical resection is considered the treatment of choice for BMs exceeding 2.5 or 3 cm in diameter.6,7 However, the dose-staged technique may enable radiosurgical treatment of larger metastases. This novel method allows the application of a high cumulative dose for the treatment of complex BMs.8–10
The aim of this study was to evaluate clinical outcomes in patients with large or high-risk BMs who were treated with different two-fraction dose-staged GKRS regimens.
Methods
Study Population and Patient Characteristics
To investigate data from a larger cohort, patient data from two study centers were retrospectively gathered. The combined data were from 142 patients who had been treated with two-fraction dose-staged GKRS between June 2015 and January 2020. The study population included patients treated at study centers in two different countries (center I: 113/142, 80%; center II: 29/142, 20%). The inclusion criteria were age > 18 years and having undergone dose-staged GKRS treatment for at least 1 high-risk BM. This study complied with the ethical principles of the Declaration of Helsinki and was approved by the institutional review board. Patient consent was not obtained due to the retrospective study design. Detailed clinical characteristics of the study patients are displayed in Table 1.
Baseline characteristics of the study population
| GKRS1 All Pts (n = 142) | Dose Escalation (n = 41, 29%) | Dose Maintainance (n = 72, 51%) | Dose De-escalation (n = 29, 20%) | p Value (btwn groups) | |
|---|---|---|---|---|---|
| Age, yrs | 63 (34–89) | 59 (42–86) | 63 (34–88) | 69 (35–89) | 0.009 |
| Female/male ratio | 68:74 | 18:23 | 38:34 | 12:17 | 0.486 |
| KPS score, % | 80 (40–100) | 80 (60–90) | 80 (40–100) | 80 (50–90) | 0.975 |
| KPS score groups | 0.869 | ||||
| ≥80% | 92 (65%) | 26 (63%) | 46 (64%) | 20 (69%) | |
| >80% | 50 (35%) | 15 (37%) | 26 (36%) | 9 (31%) | |
| ECM at BMdiagnosis | 0.532 | ||||
| Yes | 96 (68%) | 26 (63%) | 48 (67%) | 22 (75%) | |
| No | 46 (32%) | 15 (37%) | 24 (33%) | 7 (25%) | |
| Primary tumor | 0.086 | ||||
| Lung cancer* | 67 (47%) | 19 (47%) | 34 (47%) | 14 (48%) | |
| Breast cancer | 21 (15%) | 9 (22%) | 11 (15%) | 1 (3%) | |
| Melanoma | 18 (13%) | 1 (2%) | 12 (17%) | 5 (18%) | |
| GU cancer | 13 (9%) | 9 (22%) | 3 (4%) | 1 (3%) | |
| GU cancer | 10 (7%) | 1 (2%) | 4 (6%) | 5 (18%) | |
| Multiple/other | 13 (9%) | 2 (5%) | 8 (11%) | 3 (10%) | |
| Pre-GKRS1 CNS treatment† | 0.194 | ||||
| None | 127 (90%) | 33 (81%) | 68 (95%) | 26 (90%) | |
| Resection | 9 (6%) | 5 (12%) | 2 (3%) | 2 (7%) | |
| GKRS | 1 (1%) | — | — | 1 (3%) | |
| Resection & GKRS | 2 (1%) | 1 (2%) | 1 (1%) | — | |
| WBRT &/or fRT | 3 (2%) | 2 (5%) | 1 (1%) | — | |
| Dose-staged BM localization | 0.103 | ||||
| Multiple | 16 (12%) | 7 (17%) | 9 (13%) | — | |
| Frontal | 26 (18%) | 5 (12%) | 18 (25%) | 3 (10%) | |
| Parietal | 12 (9%) | 4 (10%) | 6 (8%) | 2 (7%) | |
| Temporal | 13 (9%) | 5 (12%) | 5 (7%) | 3 (10%) | |
| Occipital | 13 (9%) | 4 (10%) | 6 (8%) | 3 (10%) | |
| Central | 26 (18%) | 3 (7%) | 16 (22%) | 7 (25%) | |
| Cerebellar | 23 (16%) | 6 (15%) | 9 (13%) | 8 (28%) | |
| Basal ganglia/brainstem/other | 13 (9%) | 7 (17%) | 3 (4%) | 3 (10%) |
ECM = extracranial metastasis; fRT = fractionated radiotherapy; pts = patients.
Values are presented as median (range) or number (%) of patients unless otherwise indicated. Boldface type indicates statistical significance.
Included only non–small cell lung cancer patients. Small cell lung cancer patients were excluded from the study.
Mainly performed for distant BMs.
Radiosurgical Technique
Patient treatments were planned with the GammaPlan system and performed with the Leksell Gamma Knife Perfexion (Elekta AB). The planning sequences were performed on a 1.5-T or 3-T MRI and always included gadolinium contrast-enhanced T1-weighted MRI sequences in axial and coronal planes. Multiplanar T2-weighted MRI sequences were additionally performed as appropriate. The target was defined as a contrast-enhanced tumor mass on T1 sequences. The whole tumor mass was covered without an additional margin. All metastases visualized on the planning MRI were treated with GKRS.11 As previously described, the dose-staged GKRS treatment method is applied in two reduced fractions to high-risk BMs.9 In our study, 166 BMs from 142 patients were treated by use of two-fraction dose-staged GKRS treatment. Our study population was divided into three treatment groups: dose escalation, defined as an increase in marginal dose after the first GKRS treatment (GKRS1) to the second GKRS treatment (GKRS2); dose maintenance, defined as no differences in marginal dose between GKRS1 and GKRS2; and dose de-escalation, defined as a decrease in marginal dose from GKRS1 to GKRS2. In both study centers, all three dose strategies were applied. All detailed radiosurgical parameters and localization of the dose-staged BMs in all three different dose-strategy groups for each treatment group are presented in Table 2, which shows an overview of the radiosurgical parameters. In 16/142 (11%) patients, more than 1 BM was treated with the dose-staged treatment strategy. The median time between the staged GKRS1 and GKRS2 treatments was 32 days (range 21–74 days).
Radiosurgical parameters of the three different dose-strategy groups
| Margin Dose Changes Btwn GKRS1 & GKRS2 | ||||
|---|---|---|---|---|
| Dose Escalation (n = 53/166; 32%)* | Dose Maintainance (n = 84/166; 50%)* | Dose De-escalation (n = 29/166; 17%)* | p Value | |
| Tumor volume, cm3 | ||||
| GKRS1 | 9.8 (0.9–19.1) | 6.7 (0.5–17.2) | 6.2 (0.9–13.5) | <0.001 |
| GKRS2 | 5.9 (0.4–17.3) | 3.7 (0.4–20.7) | 3.8 (0.7–11.4) | <0.001 |
| Last FU | 3.1 (0.1–15.4) | 1.7 (0.1–25.0) | 1.1 (0.1–7.8) | 0.101 |
| Max BM diameter, cm | ||||
| GKRS1 | 3.0 (1.3–4.3) | 2.5 (0.9–3.7) | 2.4 (0.8–3.0) | <0.001 |
| GKRS2 | 2.5 (1.0–3.8) | 2.0 (0.6–3.5) | 1.9 (0.7–2.7) | <0.001 |
| Last FU | 2.1 (0.1–4.6) | 1.5 (0.1–4.2) | 1.4 (0.3–4.0) | 0.010 |
| Isodose line, % | ||||
| GKRS1 | 50 (40–60) | 50 (40–60) | 50 (45–50) | 0.658 |
| GKRS2 | 50 (40–55) | 50 (48–60) | 50 (40–50) | 0.037 |
| Prescription dose, Gy | ||||
| GKRS1 | 12 (10–14) | 14 (10–16) | 14 (12–16) | <0.001 |
| GKRS2 | 15 (12–18) | 14 (10–16) | 13 (10–15) | <0.001 |
| Central dose, Gy | ||||
| GKRS1 | 24.3 (19.9–32.5) | 28.0 (20.0–35.6) | 28.0 (24.0–33.3) | <0.001 |
| GKRS2 | 30.0 (24.0–36.4) | 28.0 (22.2–49.3) | 26.0 (20.0–31.1) | <0.001 |
| Localization of dose-staged BM | 0.099 | |||
| Frontal | 8 (15%) | 22 (26%)) | 4 (14%) | |
| Parietal | 10 (19%) | 8 (10%) | 2 (7%) | |
| Temporal | 9 (17%) | 7 (8%) | 3 (10%) | |
| Occipital | 6 (11%) | 7 (8%) | 3 (10%) | |
| Central | 4 (8%) | 20 (24%) | 6 (21%) | |
| Cerebellar | 9 (17%) | 16 (19%) | 8 (28%) | |
| Basal ganglia/brainstem/other | 7 (13%) | 4 (5%) | 3 (10%) | |
| Primary tumor | 0.001 | |||
| Lung cancer | 21 (40%) | 38 (45%) | 15 (53%) | |
| Breast cancer | 13 (24%) | 11 (13%) | 1 (3%) | |
| Melanoma | 1 (2%) | 15 (18%) | 5 (17%) | |
| GU cancer | 15 (28%) | 3 (4%) | 1 (3%) | |
| GI cancer | 1 (2%) | 5 (6%) | 5 (17%) | |
| Multiple/other | 2 (4%) | 12 (14%) | 2 (7%) | |
| Time btwn GKRS1 & GKRS2, days | 34 (21–62) | 32 (23–74) | 32 (23–42) | 0.286 |
FU = follow-up.
Values are presented as median (range) or number (%) of BMs unless otherwise indicated. Boldface type indicates statistical significance.
Patients with multiple dose-staged BMs in whom the GKRS2 treatment could not be performed for 2 BMs (2/168, 1%) due to local progression. Since the other BMs were treated with the dose-staged method, only these 2 BMs were excluded from this subanalysis.
Outcome Evaluation and Statistical Analysis
Actual observed survival was compared with predicted survival determined according to three different prognostic outcome scales: general and disease-specific graded prognostic assessment (GPA), recursive partitioning analysis (RPA), and Score Index for Radiosurgery (SIR).12–14
According to our clinical standard procedure, all radiosurgically treated patients were clinically and radiologically followed up for a 3-month interval. However, as occurs in daily clinical practice, not all patients adhered to their follow-up appointments.
Based on standard MRI sequences (FLAIR/T2-weighted, T1-weighted pregadolinium- and postgadolinium-based contrast agent) and the Response Assessment for Neuro-Oncology (RANO) criteria, progression was defined as an increase of at least 20% in the longest tumor diameter.15 Radiation reaction (RR) was defined as progressive surrounding edema and radiation necrosis (RN) as a progressive ring-enhancing lesion with surrounding edema.16 Intralesional hemorrhage was identified either as a novel intralesional increase of CT density compatible with hemorrhage or as a progressive or novel additional (T2-weighted hypointense or T1-weighted hyperintense) signal alteration.17,18
For the evaluation of maximum BM diameter at the last follow-up, BMs without any residual contrast enhancement or with only glial tissue changes and a minimum tumor diameter of 0.1 cm were used.
Categorical data were presented as counts and percentages, and continuous parameters as median and range. The chi-square test was used to analyze the counts. The Wilcoxon signed-rank test was used to compare the predicted survival calculated according to prognostic scores with the observed survival and to evaluate differences in BM volume. For comparison between different treatment groups, Kruskal-Wallis or Mann-Whitney U-tests were applied. Survival after the GKRS1 treatment was estimated with the Kaplan-Meier method and compared with the log-rank test. Tumor control rates at 3, 6, and 12 months were calculated with life tables. For all tests, p < 0.05 was considered statistically significant. Statistical analyses were carried out with IBM SPSS Statistics for Windows version 24 (IBM Corp.) and GraphPad Prism version 8.1.2 (GraphPad Software).
Results
Overall Clinical Outcome After Dose-Staged Radiosurgical Treatment
Due to missing follow-up data, 2/142 (1%) patients were excluded from the clinical outcome analyses. Overall, the majority of patients (91/140, 65%) had died at the time of last follow-up. The estimated median overall survival after BM diagnosis was 14.0 months (95% CI 8.2–19.8). Compared to the predicted survival based on prognostic scores, the observed median survival time was significantly longer in our study population (Table 3). The estimated median overall survival after GKRS1 was 11.0 months (95% CI 7.0–15.0). In our study population, the longest median survival after GKRS1 was observed in lung and breast cancer patients (Table 3). Furthermore, the estimated median survival after the radiosurgical treatment did not show any significant differences between the different dose-strategy treatment groups (p = 0.449).
Predicted and observed overall survival by primary tumor
| Survival, mos | p Value | ||
|---|---|---|---|
| Predicted by Prognostic Scores | Observed | ||
| Overall (n = 140) | |||
| GPA general | 3.8 (2.6–11.0) | 14.0 (8.2–19.8) | <0.001 |
| GPA specific | 5.5 (2.6–25.3) | <0.001 | |
| RPA | 4.5 (2.3–7.7) | <0.001 | |
| SIR | 6.0 (2.1–8.8) | <0.001 | |
| Lung cancer (n = 66) | |||
| GPA general | 3.8 (2.6–11.0) | 15.2 (7.3–23.1) | <0.001 |
| GPA specific | 5.5 (2.6–14.8) | <0.001 | |
| RPA | 4.5 (2.3–7.7) | <0.001 | |
| SIR | 6.0 (2.1–8.8) | <0.001 | |
| Melanoma (n = 18) | |||
| GPA general | 3.8 (2.6–3.8) | 8.1 (0.0–35.9) | 0.002 |
| GPA specific | 4.7 (3.0–13.2) | 0.039 | |
| RPA | 4.5 (2.3–4.5) | 0.003 | |
| SIR | 6.0 (2.1–6.0) | 0.003 | |
| GU cancer (n = 13) | |||
| GPA general | 3.8 (2.6–6.9) | 9.0 (5.1–12.9) | 0.004 |
| GPA specific | 7.3 (3.3–11.3) | NS | |
| RPA | 4.5 (2.3–7.7) | 0.003 | |
| SIR | 6.0 (2.1–8.8) | 0.028 | |
| GI cancer (n = 10) | |||
| GPA general | 3.8 (2.6–3.8) | 7.0 (0.0–17.2) | NS |
| GPA specific | 4.4 (3.1–6.9) | NS | |
| RPA | 4.5 (2.3–7.7) | NS | |
| SIR | 6.0 (2.1–6.0)) | NS | |
NS = not significant. Values are presented as median (range) unless otherwise indicated. Boldface type indicates statistical significance. Predicted survival durations according to the different prongostic outcome scales compared with the actual observed survival duration in 140/142 (99%) patients. Two patients (1%) were excluded due to missing follow-up data. Survival after BM diagnosis was calculated for the entire study population as well as for the different primary tumor entities.
Radiological Outcome and Complications After Dose-Staged Radiosurgical Treatment
In 142 patients, 166 BMs were treated by two-fraction dose-staged GKRS. Of patients with multiple staged BMs (16/142, 11%), the GKRS2 treatment could not be performed for 2 BMs (2/168, 1%) in 2 patients (2/142, 1%). In both cases (BM from melanoma in 1 case and from lung cancer in 1 case), emergency resection was performed before the planned GKRS2 treatment, due to intralesional hemorrhage and/or RR resulting in neurological symptoms.
Documented radiological follow-up was available for 118/166 (71%) dose-staged BMs (Fig. 1). In the vast majority of dose-staged BMs (107/118, 91%), the BM size remained stable or decreased between the GKRS1 treatment and the last follow-up (Table 4).
Flowchart depicting the study inclusion algorithm. FU = follow-up.
Tumor control rates at last follow-up after dose-staged treatment
| BM Status at Last FU | Total (n = 118) | Dose Escalation (n = 32/118; 27%) | Dose Maintainance (n = 67/118; 57%) | Dose De-escalation (n = 19/118; 16%) |
|---|---|---|---|---|
| Stable/ decreased | 105 (89%) | 28 (88%) | 60 (90%) | 17 (90%) |
| Increased | 13 (11%) | 4 (12%) | 7 (10%) | 2 (10%) |
BM status at last follow-up evaluated for 118/166 (71%) dose-staged BMs (after exclusion of patients lost to follow-up and without documented radiological follow-up data). BM volume decreased in the majority (105/118, 89%) of dose-staged BMs. These findings were observed in all three treatment groups.
For all 166 dose-staged BMs, the median volume decreased significantly between the GKRS1 and GKRS2 treatments (7.4 vs 4.4 cm3; p < 0.001), and further until the time of the last follow-up MRI (1.8 cm3; p < 0.001; Fig. 2A). This significant decrease in BM volume was observed among all three treatment groups (Fig. 2B–D). Similar results were observed when evaluating maximum BM diameter instead of BM volume (Table 2). The BM diameter decreased significantly between the GKRS1 and GKRS2 treatments (p < 0.001 for all three treatment groups) and to the time of the last follow-up MRI (p < 0.001 for dose escalation and dose maintenance; p = 0.017 for dose de-escalation). Calculated tumor control rates after GKRS1 at 3, 6, and 12 months were 97%, 97%, and 92% for the dose escalation group; 97%, 95%, and 90% for the dose maintenance group; and 100%, 100%, and 91% for the dose de-escalation group, respectively.
Outcome after dose-staged treatment. A: The median volume of all dose-staged BMs at GKRS1, GKRS2, and last FU. B–D: The median volume of all dose-staged BMs at GKRS1, GKRS2, and last FU according to the three treatment groups (dose escalation [B], dose maintenance [C], dose de-escalation [D]).
Overall, for the vast majority (98/118, 83%) of dose-staged BMs, the treated patients did not show any significant postradiosurgical complications, including RR, RN, and intralesional hemorrhage. The remaining 17% (20/118) of dose-staged BMs were in patients diagnosed with either symptomatic RR (8/118, 7%), symptomatic RN (9/118, 7%), or intralesional hemorrhage (3/118, 3%). An additional subanalysis did not reveal any significant differences for postradiosurgical complications or local progression among the three treatment groups (RR, p = 0.412; RN, p = 0.854; local progression, p = 0.952; and intralesional hemorrhage, p = 0.456).
Outcome After Dose-Staged Radiosurgical Treatment in Relation to Primary Tumor
The influence of the types of primary tumors on patient outcomes was evaluated. Primary tumor groups comprised lung cancer, breast cancer, melanoma, and pooled gastrointestinal (GI)/genitourinary (GU) cancer. Patients with multiple or other primary tumors (13/142, 9%) were excluded from these subanalyses. Detailed outcome data are displayed in Fig. 3.
Outcome after dose-staged treatment according to primary tumor and treatment groups. A–C: Lung cancer. D–F: Breast cancer. G–I: Melanoma. J–L: GI and GU cancer.
In lung cancer patients (Fig. 3A–C), the median BM volume decreased significantly between the GKRS1 and GKRS2 treatments in all three treatment groups. In addition, the median BM volume also decreased significantly from GKRS1 to the last follow-up in all three treatment groups. However, the significant volume decrease from GKRS2 to last follow-up was only observed in the dose maintenance group (p < 0.001; Fig. 3B). Among lung cancer patients, calculated overall local tumor control rates at 3, 6, and 12 months after GKRS1 were 98%, 98%, and 92%, respectively.
In breast cancer patients (Fig. 3D–F), the median BM volume decreased significantly from GKRS1 to GKRS2 as well as to the last follow-up in both the dose escalation and dose maintenance strategy groups. Interestingly, a further significant decrease in volume after GKRS2 to last follow-up was only observed in the dose escalation group (p = 0.009; Fig. 3D). Among breast cancer patients, calculated overall local tumor control rates at 3, 6, and 12 months after GKRS1 were 100%, 94%, and 85%, respectively.
In melanoma patients (Fig. 3G–I), significant volume decreases between GKRS1 and GKRS2 (p = 0.047) as well as to the last follow-up (p = 0.016) were observed in the dose maintenance group. From GKRS2 to last follow-up, a tendency toward volume decrease was seen (p = 0.056; Fig. 3H). In the dose de-escalation group, no significant volume decrease was observed from GKRS1 to GKRS2 or from GKRS2 to last follow-up (Fig. 3I). Among melanoma patients, calculated overall local tumor control rates at 3, 6, and 12 months after GKRS1 were 89%, 89%, and 89%, respectively.
In GI and GU cancer patients (Fig. 3J–L), only the dose escalation group showed significant decreases between GKRS1 and GKRS2 (p < 0.001) as well as from GKRS2 to the last follow-up (p = 0.043; Fig. 3J). Of note, at the time of GKRS1, patients with GI or GU cancer had a significantly worse median Karnofsky Performance Status (KPS) score compared to patients with lung cancer (p = 0.003), breast cancer (p = 0.043), or melanoma (p = 0.004).
Discussion
Rationale Behind Different Dose Strategies
Microsurgical resection is still considered the treatment of choice for large, symptomatic BMs.19 However, surgical procedures are often deemed high risk in oncological patients due to the general clinical condition of the patient.1,9,20 In recent years, SRS has emerged as an effective local treatment method with low complication rates and almost no neurotoxicity, even in patients with multiple BMs.1,6 Still, the application of a single-fraction radiosurgical treatment for large, high-risk BMs is challenging due to a decreased responsiveness to radiation and an increased risk of complications.19 The consequent application of a decreased radiation dose often results in poor local tumor control.21
To minimize neurotoxicity and to maintain the advantages of SRS, Higuchi et al. introduced a new treatment concept with three-staged fractionated SRS in 2009.21,22 In a study by Kim et al., an optimal dose-fraction scheme for BM > 3 cm was investigated to balance local tumor control and radiation-induced toxicity.23 Their study revealed that 27 Gy in 3 fractions might be the most beneficial regimen, compared to 24 or 30 Gy in 3 fractions. However, the decision-making regarding the dose-fraction scheme was mainly performed in the randomized study setting, without evaluation of other potentially influencing factors, such as primary tumor.23
Based on this treatment approach, an alternative radiosurgical treatment method with two-fraction dose-staged GKRS was developed.10 As we and others have previously published, we usually perform dose-staged GKRS treatment for large or high-risk BMs in eloquent areas with an interval of 4 weeks.9 However, data on the optimal dose-fraction strategy for large, high-risk BMs are still missing.22,24 Previous studies have reported on different dose strategies for the two-fraction dose-staged treatment, with prescription doses ranging between 10 and 16 Gy.9,22,24 Moreover, Serizawa et al. performed a retrospective multi-institutional study to compare three-staged and two-staged GKRS.24 However, no significant differences in terms of overall survival, tumor progression, or radiosurgical complications were observed. Therefore, Serizawa et al. recommended the two-staged GKRS treatment regimen for the treatment of large BMs, considering the burdens on patients, the costs, and the total treatment interval.24
The definitions of large, high-risk BMs reported in previous studies range widely. In summary, BMs with a maximum diameter of 2 to 3 cm and upward were considered to be large and therefore eligible for two-fraction dose-staged treatment.9,22,24 In the present study, we defined large BMs as ≥ 5 cm3, but have treated BMs up to 21 cm3. However, we also applied dose-staged GKRS to BMs that were localized in highly eloquent brain regions and multiple smaller BMs in close proximity to each other.9 Depending on the clinical and radiological parameters at GKRS1, the radiological findings at GKRS2, and the expertise of the radiosurgeon, marginal doses were prescribed as appropriate. Thus, we assigned patients to the following dose-strategy treatment groups: dose escalation, dose maintenance, and dose de-escalation.
Overall, the dose escalation group showed the largest median BM volume at GKRS1. Thus, lower initial prescription doses had to be used to cover the whole tumor mass and to reduce the risk of neurotoxicity. Among these patients, 30% of BMs originated from GI or GU cancers, which are generally described to be radioresistant.25,27 At GKRS2, the median volume of these dose-staged BMs decreased significantly, allowing the application of higher prescription doses for further treatment. The dose maintenance group results reflect the hitherto existing standard for two-fraction dose-staged treatment, whereas the dose de-escalation strategy was often applied for lung cancer BMs or in case of a progressing edema after GKRS1.9 Thus, the decision of maintaining or changing the marginal dose between GKRS1 and GKRS2 was made individually, based on the type of primary tumor, the BM treatment volume at GKRS1, and the radiological response of the BM displayed on planning MR images at GKRS2. Consequently, our data provide the basis for an individual treatment approach that may be chosen according to the tumor histology and treatment volume but may also be tailored to the findings on the planning MR images for the second fraction at GKRS2.
Overall Clinical and Radiological Outcome
Overall, excellent local tumor control rates were achieved among all dose-strategy groups, similar to previously reported data for single-fraction as well as two-fraction radiosurgical BM treatments.9,25,26 Of note, tumor volume of dose-staged BMs decreased significantly from GKRS1 to GKRS2 treatment, but also to the last available follow-up, regardless of the different dose-strategy groups.
The significant decrease in volume that had already occurred after GKRS1 is evidence for the key advantage of the two-fraction dose-staged GKRS method, given that the risk of RN has been reported to increase with tumor volume.27 Thus, at the time of the GKRS2 fraction not only the treatment volume but also the volume of normal brain exposed to radiation is drastically reduced.27
To further evaluate the outcome of our patients, the survival after BM diagnosis was compared to the predicted survival after well-known prognostic scores.12–14 Of note, our patients showed a significantly longer observed survival period than those predicted after analysis of all four scores.
Outcome According to Primary Tumor
It has been well established that the type of primary tumor influences the outcome of radiosurgically treated BM, yet studies investigating the differences in outcome according to the primary tumor among dose-staged BMs are scarce. However, patients with dose-staged BMs from GI cancer have been reported to have a lower volume reduction rate than BMs from other primary tumors.25 To adjust the prescription dose at GKRS1 only according to the primary tumor seems to defeat the objective of the dose-staged treatment option for particularly large BMs. Thus, we suggest that different dose strategies should be applied for dose-staged management.
For lung cancer patients, a significant decrease in tumor volume from GKRS1 to GKRS2 and from GKRS1 to last follow-up was observed among all three dose-strategy groups. However, a further significant decrease in tumor volume after GKRS2 until last follow-up was only achieved in the dose maintenance group. Thereby, dose maintenance might represent the most effective method in patients with lung cancer BMs. These findings might be influenced by the smaller number of patients in the dose de-escalation and dose escalation lung cancer patient groups. However, there is a trend toward a further decrease of tumor volume after GKRS2 in the dose escalation group as well.
In patients with BMs from breast cancer, only the dose escalation group showed a significant further decrease after GKRS2 until the last follow-up. However, there was a trend toward the same results in the dose maintenance group. Significantly greater volume reductions in large BMs were generally achieved in patients with breast cancer than in patients with lung or GI cancer.25 Based on these high response rates and significant tumor volume reduction after GKRS1, the dose escalation strategy may represent an optimal treatment method for large breast cancer BMs.
BMs from melanoma are more radioresistant than those from lung or breast cancer.11,26,28 Among melanoma patients in our study, the vast majority were treated by use of a dose maintenance strategy, resulting in a significant decrease in tumor volume from GKRS1 to GKRS2 and to the last follow-up. Other dose strategies were applied in too few cases to allow a conclusion. Still, the two-fraction dose-staged treatment achieved excellent overall rates of tumor control that are in line with previously published single-fraction SRS data.11,26,28
Because of the small number of patients in our study, data for BMs originating from GU and GI cancer were pooled. Significant decreases in tumor volume from GKRS1 to GKRS2 and from GKRS1 to last follow-up were achieved only in the dose escalation group. However, our analysis suffers from a rather high rate of patients having been lost to follow-up or died due to peripheral disease progression before follow-up, especially in the GI cancer group. BMs from both GU and GI primary tumors are considered radioresistant. Unfavorable survival and tumor control rates after fractionated WBRT for these cancers have been described in the literature.27,29 In contrast, SRS for these BMs, especially with higher prescription doses, has achieved excellent results.27,29
Postradiosurgical Complications
After radiosurgical treatment, RN, RR, and intralesional hemorrhage are possible adverse radiation effects (AREs).9,30 Occurrences of these postradiosurgical complications are associated with several factors, including prior SRS treatment to the same lesion as well as maximum BM diameter and volume.25,30
As previously reported in a large study cohort with 2200 single-fraction treated BMs, the increased risk of worst-case AREs was clearly associated with increasing BM volume.30
As mentioned above, two-fraction dose-staged BM management allows reduction of radiation exposure of the normal brain tissue, leading to a reduction of neurotoxicity.27 However, the evaluations of postradiosurgical complications for dose-staged BMs are still limited to smaller study cohorts. In a study by Ito et al., the cumulative incidence for RN was 4.2%. Furthermore, no significant differences in risk of AREs could be seen between different primary tumors.25 Among all dose-strategy treatment groups, the majority of our patients did not show any significant postradiosurgical complications. Moreover, the postradiosurgical complication rates in our study were in line with those for previously reported single-fraction SRS and also dose-staged SRS complication rates.25,30
Study Limitations
One of our study limitations is the retrospective study design. Furthermore, the subanalyses for the origins of each primary tumor or postradiosurgical complications were limited when separating the study population into three different dose-strategy groups.
Conclusions
In patients with large or high-risk BMs, all three dose-staged GKRS treatment strategies represent effective local treatment methods with excellent local tumor control rates. Complication rates are acceptable and in line with complication rates previously reported for single-fraction methods as well as for dose-staged radiosurgery. Depending on the primary tumor origin and initial treatment volume, different dose-strategy options are suggested, although further prospective studies may be necessary. Still, our data provide the basis for an individual treatment approach that may be chosen according to the tumor histology and treatment volume but may also be tailored to the findings at GKRS2.
Disclosures
The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.
Author Contributions
Conception and design: Frischer, Cho, Kranawetter, Untersteiner, Ilyalov. Acquisition of data: Frischer, Cho, Medvedeva, Kranawetter, Untersteiner, Hirschmann, Lepilina, Baulin, Ertl, Gatterbauer, Ilyalov. Analysis and interpretation of data: Frischer, Cho, Kranawetter, Untersteiner, Buschmann, Ertl, Marik, Dorfer, Rössler, Ilyalov. Drafting the article: Frischer, Cho, Kranawetter, Untersteiner. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Frischer. Statistical analysis: Frischer, Cho, Untersteiner. Study supervision: Frischer, Rössler, Ilyalov.
References
- 3↑
Gerosa M, Nicolato A, Foroni R, Tomazzoli L, Bricolo A. Regional treatment of metastasis: role of radiosurgery in brain metastases—gamma knife radiosurgery. Ann Oncol. 2004;15(suppl 4):iv113–iv117.
- 4
Minniti G, Clarke E, Lanzetta G, et al. Stereotactic radiosurgery for brain metastases: analysis of outcome and risk of brain radionecrosis. Radiat Oncol. 2011;6:48.
- 5
Yamamoto M, Serizawa T, Shuto T, et al. Stereotactic radiosurgery for patients with multiple brain metastases (JLGK0901): a multi-institutional prospective observational study. Lancet Oncol. 2014;15(4):387–395.
- 6↑
Kaal EC, Niël CG, Vecht CJ. Therapeutic management of brain metastasis. Lancet Neurol. 2005;4(5):289–298.
- 7↑
Shen CJ, Lim M, Kleinberg LR. Controversies in the therapy of brain metastases: shifting paradigms in an era of effective systemic therapy and longer-term survivorship. Curr Treat Options Oncol. 2016;17(9):46.
- 8
Niranjan A, Monaco E, Flickinger J, Lunsford LD. Guidelines for multiple brain metastases radiosurgery. Prog Neurol Surg. 2019;34:100–109.
- 9↑
Frischer JM, Fraller A, Mallouhi A, et al. Evaluation of dose-staged Gamma Knife radiosurgical treatment method for high-risk brain metastases. World Neurosurg. 2016;94:352–359.
- 10↑
Yomo S, Hayashi M, Nicholson C. A prospective pilot study of two-session Gamma Knife surgery for large metastatic brain tumors. J Neurooncol. 2012;109(1):159–165.
- 11↑
Gatterbauer B, Hirschmann D, Eberherr N, et al. Toxicity and efficacy of Gamma Knife radiosurgery for brain metastases in melanoma patients treated with immunotherapy or targeted therapy—a retrospective cohort study. Cancer Med. 2020;9(11):4026–4036.
- 12
Sperduto PW, Berkey B, Gaspar LE, Mehta M, Curran W. A new prognostic index and comparison to three other indices for patients with brain metastases: an analysis of 1,960 patients in the RTOG database. Int J Radiat Oncol Biol Phys. 2008;70(2):510–514.
- 13
Gaspar L, Scott C, Rotman M, et al. Recursive partitioning analysis (RPA) of prognostic factors in three Radiation Therapy Oncology Group (RTOG) brain metastases trials. Int J Radiat Oncol Biol Phys. 1997;37(4):745–751.
- 14
Weltman ESJ, Oliveira VC, Brandt RA, et al. Score index for stereotactic radiosurgery of brain metastases. J Radiosurg. 1998;1(2):89–97.
- 15↑
Lin NU, Lee EQ, Aoyama H, et al. Response assessment criteria for brain metastases: proposal from the RANO group. Lancet Oncol. 2015;16(6):e270–e278.
- 16↑
Stockham AL, Tievsky AL, Koyfman SA, et al. Conventional MRI does not reliably distinguish radiation necrosis from tumor recurrence after stereotactic radiosurgery. J Neurooncol. 2012;109(1):149–158.
- 19↑
Kirkpatrick JP, Soltys SG, Lo SS, Beal K, Shrieve DC, Brown PD. The radiosurgery fractionation quandary: single fraction or hypofractionation?. Neuro Oncol. 2017;19(suppl 2):ii38–ii49.
- 20↑
Lovo EE, Torres LB, Campos FJ, et al. Two-session radiosurgery as initial treatment for newly diagnosed large, symptomatic brain metastases from breast and lung histology. Cureus. 2019;11(8):e5472.
- 21↑
Higuchi Y, Serizawa T, Nagano O, et al. Three-staged stereotactic radiotherapy without whole brain irradiation for large metastatic brain tumors. Int J Radiat Oncol Biol Phys. 2009;74(5):1543–1548.
- 22↑
Ginalis EE, Cui T, Weiner J, Nie K, Danish S. Two-staged stereotactic radiosurgery for the treatment of large brain metastases: single institution experience and review of literature. J Radiosurg SBRT. 2020;7(2):105–114.
- 23↑
Kim KH, Kong DS, Cho KR, et al. Outcome evaluation of patients treated with fractionated Gamma Knife radiosurgery for large (> 3 cm) brain metastases: a dose-escalation study. J Neurosurg. 2020;133(3):675–684.
- 24↑
Serizawa T, Higuchi Y, Yamamoto M, et al. Comparison of treatment results between 3- and 2-stage Gamma Knife radiosurgery for large brain metastases: a retrospective multi-institutional study. J Neurosurg. 2018;131(1):227–237.
- 25↑
Ito D, Aoyagi K, Nagano O, Serizawa T, Iwadate Y, Higuchi Y. Comparison of two-stage Gamma Knife radiosurgery outcomes for large brain metastases among primary cancers. J Neurooncol. 2020;147(1):237–246.
- 26↑
Lippitz B, Lindquist C, Paddick I, Peterson D, O’Neill K, Beaney R. Stereotactic radiosurgery in the treatment of brain metastases: the current evidence. Cancer Treat Rev. 2014;40(1):48–59.
- 27↑
Angelov L, Mohammadi AM, Bennett EE, et al. Impact of 2-staged stereotactic radiosurgery for treatment of brain metastases ≥ 2 cm. J Neurosurg. 2018;129(2):366–382.
- 28↑
Powell JW, Chung CT, Shah HR, et al. Gamma Knife surgery in the management of radioresistant brain metastases in high-risk patients with melanoma, renal cell carcinoma, and sarcoma. J Neurosurg. 2008;109(suppl):122–128.
- 29↑
Hasegawa T, Kondziolka D, Flickinger JC, Lunsford LD. Stereotactic radiosurgery for brain metastases from gastrointestinal tract cancer. Surg Neurol. 2003;60(6):506–515.
- 30↑
Sneed PK, Mendez J, Vemer-van den Hoek JG, et al. Adverse radiation effect after stereotactic radiosurgery for brain metastases: incidence, time course, and risk factors. J Neurosurg. 2015;123(2):373–386.
