Abstract
Background/Aim: Most patients with resected brain metastases receive post-operative radiotherapy. This study investigated outcomes of fractionated stereotactic radiotherapy (FSRT) alone or whole-brain irradiation plus simultaneous integrated boost (WBI+SIB) in the post-operative setting. Patients and Methods: Forty-four patients receiving FSRT alone (n=32) or WBI+SIB (n=12) after resection of 1-3 brain metastases from 2014-2022 were analyzed. Twelve factors were evaluated for local control (LC), distant brain control (DBC), and overall survival (OS). Results: On univariate and multivariate analyses, single brain metastasis was associated with improved LC and DBC. Longer interval between tumor diagnosis and radiotherapy, single brain metastasis, and Karnofsky performance score >80 were associated with improved OS. WBI+SIB showed a trend towards better DBC. Conclusion: Several independent predictors of outcomes after FSRT or WBI+SIB following resection of brain metastases were identified. Given similar survival in the post-operative setting between FSRT and WBI+SIB, potential toxicity remains a significant factor in treatment recommendations.
- Brain metastases
- resection
- post-operative radiotherapy
- fractionated stereotactic radiotherapy
- whole-brain irradiation
- simultaneous integrated boost
Resection of brain metastases is performed in many patients with a single, symptomatic, or few lesions (1, 2). In a randomized trial from 1998 including 95 patients, whole-brain irradiation (WBI) after resection of a single brain metastasis led to significantly improved local control (LC) and freedom from distant brain metastases (distant brain control, DBC) (3). Since then, many patients with resectable brain metastases received postoperative WBI. However, WBI is associated with a considerable risk of neuro-cognitive decline. Therefore, stereotactic radiotherapy (SRT), which can be administered as a single-dose stereotactic radiosurgery (SRS) or as fractionated stereotactic radiotherapy (FSRT), is increasingly used for the treatment of brain metastases (4-12). Initially, SRT was investigated for intact (unresected) brain metastases. In a retrospective study of 186 patients with 1-3 brain metastases, SRS alone resulted in better LC than WBI alone and was not inferior with respect to DBC and overall survival (OS) (6). Aiming to further improve outcomes of patients with a limited number of brain metastases, several studies investigated the combination of SRS and WBI and compared this approach to SRS alone (4, 5, 7-10). In these studies, the addition of WBI resulted in better intracerebral control (and sometimes also better LC) but did not improve OS. Moreover, in two randomized trials, WBI led to significantly higher rates of neuro-cognitive deficits (4, 5). Therefore, radiation oncologists have become more reserved regarding the addition of WBI to SRS or FSRT. This holds true also for radiotherapy following resection of brain metastases (13). In single-arm studies that investigated post-operative SRS or FSRT alone, the 1-year LC rates ranged between 61% and 97.5% and the 1-year DBC rates between 36% and 69% (14-22). In a randomized trial, post-operative SRS significantly improved 1-year LC (72% vs. 43% in the observation arm, p=0.015) (23). Very few studies compared SRS to WBI after resection of brain metastases (24-26). Similar to the treatment of unresected brain metastases, WBI improved intracerebral control but not OS. Moreover, in these studies WBI had no significant impact on LC of the resected lesions (24-26). In a randomized trial, WBI led to a significant increase in neuro-cognitive decline (25, 26).
Two additional aspects should be considered for radiotherapy after resection of brain metastases. One aspect is that the risk of radiation necrosis after treatment of larger lesions can be reduced by using FSRT instead of SRS (27, 28). The second aspect is that WBI can be combined with a simultaneous integrated boost (SIB) to the metastatic sites (29-31). In several studies, this approach was shown to be feasible and considered at least as effective as SRT and WBI plus SRT regarding LC and intracerebral control (30-37). If reasonable, WBI+SIB can be combined with hippocampus-sparing techniques to further reduce the risk of neuro-cognitive decline (38). This study investigates the outcomes of post-operative radiotherapy with FSRT or WBI+SIB after resection of 1-3 brain metastases.
Patients and Methods
Forty-four patients treated with FSRT alone (n=32) or WBI+SIB (n=12) after resection of 1-3 brain metastases between 2014 and 2022 were included in this retrospective study, which was approved by the Ethics Committee of the University of Lubeck (22-059). FSRT was administered using a conventional linear accelerator (Varian Medical Systems, Palo Alto, CA, USA). Different dose-fractionation regimens of FSRT were used, including conventionally fractionated regimens with 5 daily fractions of 2 Gy per week up to 36-50 Gy (n=7), hypo-fractionated regimens with 5 daily fractions of 2.5 Gy per week up to 35-45 Gy (n=6) or 5 daily fractions of 3 Gy per week up to 30-48 Gy (n=14), and ultra-hypo-fractionated regimens with 3-8 fractions of 4-7.5 Gy up to 22.5-40 Gy (n=5). In one patient, FSRT was terminated after 7 of 18 fractions of 2 Gy; 31 patients (97%) received the complete planned treatment. In these 31 patients, the biologically effective doses (BED) ranged between 32.0 and 60.0 Gy12 (median=48.8 Gy12). For WBI+SIB, three different regimens were used, where the SIB was always 0.5 Gy per fraction. The regimens were 14 × (2.5 Gy WBI + 0.5 Gy SIB) in eight patients, 15 × (2.5 Gy WBI + 0.5 Gy SIB) in one patient, and 18 × (2.0 Gy WBI + 0.5 Gy SIB) in three patients. The BEDs at the metastatic sites were 52.5 Gy12, 56.3 Gy12 and 54.4 Gy12, respectively (median=52.5 Gy12). In two of these patients, WBI+SIB was combined with a hippocampus-sparing technique.
Twelve factors (Table I) were evaluated for local control (LC, 43 patients) of the resected lesions, distant brain control (DBC, 43 patients), and overall survival (OS, 44 patients). These factors included the interval between first diagnosis of the malignant disease and radiotherapy following resection of brain metastases (≤11 vs. ≥12 months, median=11.5 months), number of resected brain metastases (1 vs. 2-3), extent of resection (subtotal vs. total), type of radiotherapy (FSRT alone vs. WBI+SIB), systemic treatment within 6 months before radiotherapy (no vs. yes), age at radiotherapy (≤61 vs. ≥62 years, median=61 years), sex (female vs. male), Karnofsky performance score (KPS ≤80 vs. >80), type of primary tumor (breast cancer vs. non-small cell lung cancer vs. upper gastrointestinal cancers vs. colorectal cancer vs. other types), infratentorial metastases (no vs. yes), cumulative size of metastatic lesions (≤40 vs. >40 mm, median=40 mm), and extracranial metastases (no vs. yes).
Patient characteristics.
LC, DBC, and OS were calculated from the last day of FSRT or WBI+SIB. For the univariate analyses, we used the Kaplan-Meier method and the log-rank test (BlueSky Statistics 10 GA from BlueSky Statistics LLC, Chicago, IL, USA). p-Values <0.05 were considered significant, and p-values ≤0.12 indicated a trend. The factors that were found significantly associated (p<0.05) with LC, DBC or OS were included in multivariate analyses (Cox proportional hazards model). Again, p-values <0.05 were considered significant.
Results
LC-rates at 6 months, 1 year and 2 years were 90%, 80% and 70%, respectively. On univariate analyses, improved LC was significantly associated with single brain metastasis (p=0.016, Table II) that maintained significance in the multivariate analysis [hazard ratio (HR)=6.19, 95% confidence interval (CI)=1.31-31.61, p=0.028).
Local control (LC) at 6 months, 1 year, and 2 years after radiotherapy.
DBC-rates at 6 months, 1 year and 2 years were 74%, 64% and 64%, respectively. On univariate analyses, improved DBC was significantly associated with single brain metastasis (p<0.001, Table III). Moreover, compared to FSRT, WBI+SIB showed a tend towards improved DBC (p=0.119). In the multivariate analysis of DBC, single brain metastasis remained significant (HR=13.74, 95%CI=3.41-55.29, p<0.001).
Distant brain control (DBC) at 6 months, 1 year, and 2 years after radiotherapy.
In the entire cohort, median OS was 18 months. On univariate analyses, improved OS was significantly associated with interval between tumor diagnosis and radiotherapy ≥12 months (p=0.022), single brain metastasis (p=0.005), and KPS >80 (p=0.018) (Table IV). Trends towards better OS were found for total resection of brain metastases (p=0.120), systemic treatment prior to radiotherapy (p=0.058), and absence of extracranial metastases (p=0.115). In the Cox proportional hazards model, longer interval between tumor diagnosis and radiotherapy (HR=0.32, 95%CI=0.14-0.76, p=0.010), single brain metastasis (HR=5.69, 95%CI=1.99-16.26, p=0.001), and KPS >80 (HR=0.39, 95%CI=0.16-0.95, p=0.039) proved to be independent predictors of OS. When considering these independent predictors, 6-month OS rates for patients with 3, 2, 1, and 0 unfavorable factors (interval tumor diagnosis to radiotherapy ≤11 months, 2-3 brain metastases, KPS ≤80) were 0%, 76%, 88%, and 100%, respectively. One-year OS rates were 0%, 36%, 62%, and 89%, respectively, and 2-year OS rates were 0%, 14%, 56%, and 89%, respectively.
Overall survival (OS) at 6 months, 1 year, and 2 years after radiotherapy.
Discussion
More than two decades ago, a randomized trial demonstrated that the outcomes of patients after resection of a single brain metastasis can be improved with post-operative WBI (2). However, WBI can lead to significant neuro-cognitive decline (4, 5). SRS or FSRT alone are standards of care that offer a less toxic post-operative alternative to conventional WBI. In two randomized trials in patients with 1-3 unresected brain metastases, the probability of experiencing neuro-cognitive decline at 3 or 4 months following radiotherapy was significantly lower after SRS alone than after SRS plus WBI (4, 5). In another randomized trial of patients with one resected brain metastasis and a maximum size of the resection cavity of 50 mm, decline in neuro-cognitive function was significantly less frequent and the interval to cognitive decline was longer after SRS alone administered to the resection cavity when compared to WBI (25). Moreover, in all three trials, WBI was associated with significantly better intracerebral control but did not improve OS (4, 5, 25). Therefore, many radiation oncologists prefer post-operative SRS or FSRT alone. FSRT is often favored over SRS for large post-operative targets to reduce the risk of radiation necrosis (27, 28). This is a question of an open Alliance trial (https://clinicaltrials.gov/ct2/show/NCT04114981). In addition to replacing WBI by SRS or FSRT, other options to reduce the risk of neuro-cognitive decline include the use of hippocampus-sparing WBI-techniques (38). In a phase II trial, the rates of neuro-cognitive decline at 4 months (vs. baseline) were 7% after hippocampus-sparing WBI and 30% in a historical control group receiving WBI without hippocampus-sparing (p<0.001) (38). The risk of neuro-cognitive decline after hippocampus-sparing WBI can be further improved with administration of memantine (39). Another more recent approach is lower-dose WBI plus a SIB, if possible combined with a hippocampus-sparing technique (29-37).
Modern radiation techniques such as volumetric-modulated arc therapy can generate steep dose gradients between metastases and normal tissues, which allows high doses to metastases and better sparing of other structures, e.g., the hippocampi (40). In our institution, both FSRT alone and WBI+SIB have been used during the last decade for post-operative treatment of brain metastases, depending on several factors including the current literature and the treating physicians’ preference. Both treatments proved to be effective with respect to longer-term LC. DBC was non-significantly better after WBI+SIB (81% vs. 56% after 2 years). However, this advantage did not translate into improved OS. These findings are consistent with randomized trials that demonstrated improved intracerebral control with similar survival after WBI alone (without SIB) when compared to SRS alone for unresected or resected brain metastases (4, 5, 25). Our results suggest that the addition of a SIB to WBI does not change this situation. Therefore, SRS alone appears preferable after resection of 1-3 brain metastases.
Our study identified several independent predictors of treatment outcomes that may contribute to personalization of the treatment. Interval between tumor diagnosis and radiotherapy ≥12 months, number of resected brain metastases, and KPS were independent predictors of OS. Since patients with all three unfavorable factors had a 6-month survival rate of 0%, it is questionable whether these patients benefit from aggressive treatments including resection and radiotherapy. Short-course palliative radiotherapy alone, corticosteroids, or best supportive care appear appropriate for this population (41). However, patients with all three favorable factors achieved a 2-year survival rate of 89%. These patients do benefit from post-operative irradiation. When considering these recommendations, the limitations of our study including the small sample size and the retrospective design must be considered. A prospective randomized trial comparing post-operative SRS or FSRT alone to WBI+SIB for LC, DBC, OS, and late toxicity in terms of neuro-cognitive decline and radiation necrosis is needed in order to define the optimal radiation treatment after resection of brain metastases.
In summary, several independent predictors of outcomes after FSRT or WBI+SIB following resection of brain metastases were identified. These prognostic factors may contribute to the personalization of the treatment for these patients. WBI+SIB showed a trend for better DBC when compared to FSRT alone but did not improve LC and OS. Thus, FSRT alone appears appropriate for many patients assigned to post-operative radiotherapy after resection of 1-3 brain metastases. A randomized trial comparing post-operative SRS or FSRT alone and WBI+SIB is warranted.
Footnotes
Authors’ Contributions
D.R. and J.G. designed the study. L.J. collected the data that were analyzed by N.Y.Y. and D.R. The article was drafted by N.Y.Y. and D.R. and finally reviewed and approved by all Authors.
Conflicts of Interest
The Authors state that there are no conflicts of interest related to this study.
- Received April 4, 2023.
- Revision received April 14, 2023.
- Accepted April 18, 2023.
- Copyright © 2023 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).