Local control and overall survival for adjuvant stereotactic radiosurgery in patients with residual or recurrent disease

Preoperative Versus Postoperative Radiosurgery of Brain Metastases: A Meta-Analysis

OBJECTIVE

While postoperative resection cavity radiosurgery (post-SRS) is an accepted treatment paradigm for brain metastasis (BM) patients who undergo surgical resection, there is emerging interest in preoperative radiosurgery (pre-SRS) followed by surgical resection as an alternative treatment paradigm. Here, we performed a meta-analysis of the available literature on this matter.

METHODS

Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines, a search of all studies evaluating pre-SRS and post-SRS was completed. Local recurrence (LR), overall survival (OS), radiation necrosis (RN), and leptomeningeal disease (LMD) were evaluated from the available data. Moderator analysis and pooled effect sizes were performed using a proportional meta-analysis with R using the metafor package. Statistics are presented as mean [95% confidence interval].

RESULTS

We identified 6 pre-SRS and 33 post-SRS studies with comparable tumor volume (4.5-17.6 cm3). There were significant differences in the pooled estimates of LR and LMD, favoring pre-SRS over post-SRS. Pooled aggregate for LR was 11.0% [4.9-13.7] and 17.5% [15.1-19.9] for pre- and post-SRS studies (P = 0.014). Similarly, pooled estimates of LMD favored pre-SRS, 4.4% [2.6-6.2], relative to post-SRS, 12.3% [8.9-15.7] (P = 0.019). In contrast, no significant differences were found in terms of RN and OS. Pooled estimates for RN were 6.4% [3.1-9.6] and 8.9% [6.3-11.6] for pre- and post-SRS studies (P = 0.393), respectively. Pooled estimates for OS were 60.2% [55.8-64.6] and 60.5% [56.9-64.0] for pre- and post-SRS studies (P = 0.974).

CONCLUSIONS

This meta-analysis supports further exploration of pre-SRS as a strategy for the treatment of BM.

Outcome comparison of patients who develop leptomeningeal disease or distant brain recurrence after brain metastases resection cavity radiosurgery

Background

To compare the outcomes between patients with leptomeningeal disease (LMD) and distant brain recurrence (DBR) after stereotactic radiosurgery (SRS) brain metastases (BM) resection cavity.

Methods

Twenty-nine patients having single-fraction SRS after BM resection who developed either LMD (n = 11) or DBR (n = 18) as their initial and only site of intracranial progression were retrospectively reviewed.

Results

Patients developing LMD more commonly had a metachronous presentation (91% vs 50%, P = .04) and recursive partitioning class 1 status (45% vs 6%, P = .02). There was no difference in the median time from SRS to the development of LMD or DBR (5.0 vs 3.8 months, P = .68). The majority of patients with LMD (10/11, 91%) developed the nodular variant (nLMD). Treatment for LMD was repeat SRS (n = 4), whole-brain radiation therapy (WBRT; n = 5), resection + WBRT (n = 1), and no treatment (n = 1). Treatment for DBR was repeat SRS (n = 9), WBRT (n = 3), resection + resection cavity SRS (n = 1), and no treatment (n = 5). Median overall survival (OS) from time of resection cavity SRS was 15.7 months in the LMD group and 12.7 months in the DBR group (P = .60), respectively. Median OS in salvage SRS and salvage WBRT were 25.4 and 5.0 months in the nLMD group (P = .004) while 18.7 and 16.2 months in the DBR group (P = .30), respectively.

Conclusions

Following BM resection cavity SRS, nLMD recurrence is much more frequent than classical LMD. Salvage SRS may be considered for selected patients with nLMD, reserving salvage WBRT for patients with extensive intracranial disease without compromising survival. Further study with larger numbers of patients is needed.

Timing of Adjuvant Fractionated Stereotactic Radiosurgery Affects Local Control of Resected Brain Metastases

PURPOSE

For resected brain metastases (BMs), stereotactic radiosurgery (SRS) is often offered to minimize local recurrence (LR). Although the aim is to deliver SRS within a few weeks of surgery, a variety of socioeconomic, medical, and procedural issues can cause delays. We evaluated the relationship between timing of postoperative SRS and LR.

METHODS AND MATERIALS

We retrospectively identified a consecutive series of patients with BM managed with resection and SRS or fractionated SRS at our institution from 2012 to 2018. We assessed the correlation of time to SRS and other demographic, disease, and treatment variables with LR, local recurrence-free survival, distant recurrence, distant recurrence-free survival, and overall survival.

RESULTS

A total of 133 patients met inclusion criteria. The median age was 64.5 years. Approximately half of patients had a single BM, and median BM size was 2.9 cm. Gross total resection was achieved in 111 patients (83.5%), and more than 90% of patients received fractionated SRS. The median time to SRS was 37.0 days, and the LR rate was 16.4%. Time to SRS was predictive of LR. The median time from surgery to SRS was 34.0 days for patients without LR versus 61.0 days for those with LR (P < .01). The LR rate was 2.3% with SRS administered ≤4 weeks postoperatively, compared with 23.6% if SRS was administered >4 weeks postoperatively (P < .01). Local recurrence-free survival was also improved for patients who underwent SRS at ≤4 weeks (P = .02). Delayed SRS was also predictive of distant recurrence (P = .02) but not overall survival.

CONCLUSIONS

In this retrospective study, the strongest predictor of LR after postoperative SRS for BM was time to SRS, and a cutoff of 4 weeks was a reliable predictor of recurrence. These findings merit investigation in a prospective, randomized trial.