menu ☰
menu ˟

Radiosurgery to the Postoperative Tumor Bed for Metastatic Carcinoma Versus Whole Brain Radiation After Surgery



Type: Article
Region: Republic of Ireland
Northern Ireland


The treatment paradigm from postoperative whole brain radiation therapy (WBRT) to post-operative stereotactic radiosurgery (SRS) to the tumor bed has shifted with little data to evaluate whether each treatment modality confers equivalent tumor control and survival outcomes.


Patients with surgical resection of single brain metastases from January 2010 to December 2014 were treated postoperatively with either WBRT or SRS. Retrospective patient data was compared for local control, distant brain recurrence, overall survival, and radiation complications.


Forty-six received WBRT, and 37 received tumor bed SRS. Twelve of 35 (34%) SRS patients experienced local recurrence compared to 17 of 31 (55%) WBRT patients (p = 0.09). The median survival was 440 days (14.7 months) for SRS and 202 days (6.7 months) for WBRT (p = 0.062, log-rank). SRS demonstrated improved survival benefit in the first six months (p = 0.0034; Wilcoxon). Radiation-related adverse changes after SRS (22%) were not statistically different from WBRT (8.7%) (p = 0.152). Age (p = 0.08), systemic cancer status (p = 0.30), Graded Prognostic Assessment (p = 0.28), number of brain metastases at diagnosis (p = 0.65), tumor volume at diagnosis (p = 0.13), new brain lesions (p = 0.74) and neurologic versus systemic cause of death (p = 0.11) did not differ between the groups.


Following surgical resection, tumor bed SRS can be used effectively in lieu of WBRT to treat brain metastases with comparable local control and distant control and without significantly more adverse events.


Traditionally, brain metastases have been treated with surgical resection followed by whole brain radiation therapy (WBRT) to decrease the rates of local recurrence, distant brain recurrence and neurologic cause of death [1-2]. Since it treats the entire brain, WBRT is thought to control tumor progression by treating identifiable metastases and otherwise unidentifiable micrometastases. However, WBRT has become increasingly implicated with global cognitive impairment that persists after cessation of treatment [3-4], thereby expanding concerns regarding quality of life despite reducing tumor burden. Furthermore, WBRT may not increase overall survival time [5-6]. Consequently, shifts in the treatment paradigm for brain metastases from postoperative WBRT to postoperative stereotactic radiosurgery (SRS) to the tumor bed are occurring [7].

SRS delivers high-dose radiations to a discrete volume within the brain. Because SRS spares healthy brain parenchyma, it confers a theoretically favorable alternative to WBRT and is being increasingly utilized in the management of brain metastases. SRS confers less cognitive decline in patients relative to WBRT [5, 8-10] and improved quality of life [11]. A number of studies have examined the ability of SRS to the post-resection tumor bed to control metastatic brain disease after tumor resection [12-27]. All of these studies, however, look only at populations of patients treated with SRS and do not compare outcomes between postoperative WBRT and SRS. Questions remain as to whether tumor bed SRS imparts long-term survival advantages equivalent to those of WBRT [9].

Based on these gaps in the understanding of the potential benefits of SRS to the post-resection tumor bed, we sought to compare the outcomes of patients who received surgical resection followed by either WBRT or SRS for brain metastases. We hypothesized that the outcomes related to tumor control of tumor-bed SRS would be clinically equivalent to those of postoperative WBRT without an increase in complications.

Materials & Methods

Patient data

All patients who received either surgical resection followed by WBRT or SRS for a presumed metastatic brain tumor from January 1, 2010 to December 31, 2014 at the University Hospital were retrospectively identified from patient treatment databases of the senior author [National Science Library (NSL)]. Patients were excluded if they were from the Veterans' Administration system or state prisoners (due to different Institutional Review Board processes), if they had no documentable follow-up after radiation treatment, or if they died within three months of treatment.

Waiver of consent was granted by the University of Missouri Health Sciences Institutional Review Board (IRB #1210881). Electronic medical records were reviewed for variables collected at the time of diagnosis, including age, sex, status of systemic disease outside of the brain (no evidence, stable, or progressive), number of brain metastases, site of primary malignancy (lung, breast, melanoma, colorectal, other), and tumor volume. Graded Prognostic Assessment (GPA) [28] was also determined. The data collected following treatment included chemotherapy received (yes or no), development of adverse radiation changes, management of adverse radiation changes (observation, steroids, or surgery), same site recurrence, distant recurrence within the central nervous system (CNS), and cause of death (neurologic or systemic).

Once the data was collected and collated, SRS patients were compared to WBRT patients for the outcome measures of local control, new brain metastases, overall survival, and adverse radiation changes.

Radiation treatment

Patients were chosen for SRS or WBRT at the discretion of the treating radiation oncologist and neurosurgeon in a non-randomized fashion. 

All SRS patients were treated with SRS using the Varian Trilogy TX linear accelerator system (Varian, Palo Alto, CA) at the Ellis Fischel Cancer Center, part of the University of Missouri Health System. Each patient underwent neuronavigational magnetic resonance imaging (MRI). In addition, the patient was immobilized with an Aquaplast Mask (Civco, Coralville, IA) which was attached to the couch. When appropriate, the patient was administered intravenous (IV) contrast dye, and axial images were obtained throughout the head for the planning computed tomography (CT) scan. Initially a bite block was constructed as a part of the immobilization apparatus, but this was discontinued in 2013 in favor of using Vision RT (London, UK). The planning CT scan was fused to the planning magnetic resonance image (MRI). The gross tumor volume (GTV) was contoured to the leading edge of the enhancement on gadolinium-enhanced T1-weighted axial MRI images at the site of the previously defined tumor for which the surgery was performed. The treatment plan was then optimized by dosimetry and physics using Varian Eclipse (Varian, Palo Alto, CA). Each plan used seven or eight static fields with a mini-multileaf collimator with 2.5 mm leaves. The prescription tumor volume (PTV) included a 2 mm expansion of the GTV as is usually done with frameless SRS [12-14, 16, 18, 25-27]. The dose to the margin was 1500 centigray (cGy) prescribed to the 80% isodose line for single fraction SRS. If the spherical diameter equivalent of the GTV exceeded 3 cm, then fractionated radiosurgery was performed in which the dose to the margin was 2100 cGy prescribed in three fractions to the 80% isodose line. Dose constraints to critical structures for single fraction SRS included the following: eyes - 800 cGy, optic nerves - 800 cGy, optic chiasm - 800 cGy, and brainstem - 1000 cGy.

The treatment plans were reviewed by the radiation oncologist and the neurosurgeon. Prior to each treatment, quality assurance was performed by physics on the multi-leaves, isocenter and cone beam. Immediately prior to each treatment, a cone beam CT was performed and compared to the reference CT and adjustments if any were made. A six degrees of freedom couch (Civco, Coralville, IA) was added when the Vision RT system was installed, which allowed for rotational adjustments of the couch at the time of treatment. Using the Vision RT system, the patientâ?Ts position during treatment delivery was monitored.

For whole brain radiation therapy, treatments were performed at Ellis Fischel Cancer Center or a nearby local community hospital. All patients were immobilized using the Aquaplast Mask. CT images were obtained for simulation. Typically, treatments were delivered using lateral or anterior oblique fields using photons. The doses were 3000 cGy delivered in 10 or 12 fractions.



Rights: Public
Suggested citation:

Cureus. (2017) Radiosurgery to the Postoperative Tumor Bed for Metastatic Carcinoma Versus Whole Brain Radiation After Surgery [Online]. Available from: [Accessed: 18th March 2019].


View your saved citations and reading lists


Northern Ireland Cancer Network (NICaN)
Click here to view all the resources gathered from this organisation's website.