In brain metastases, the choice between whole brain radiation therapy (WBRT) and stereotactic radiosurgery (SRS) changes the target, the preservation strategy, and the kind of intracranial control the team is trying to achieve. This dedicated article rebuilds the chapter in practical clinical language, covering selection between WBRT and SRS, field design, postoperative cavity contouring, and dose schedules. For the broader framework, see our Target Volume Delineation and Field Setup – Complete Clinical Guide.
For a stereotactic setup example outside the cranial setting, see our article on Head and Neck SBRT: A Practical Clinical Guide.
WBRT versus SRS in brain metastases
The chapter does not try to settle every indication question between WBRT and SRS. Instead, it places the decision where it belongs: number and volume of brain metastases, performance status, and the use of prognostic tools, including the molecular graded prognostic assessment. That framing matters because the same diagnosis can present as diffuse intracranial disease, where broad coverage matters most, or as a limited lesion set, where focal precision drives the plan.
The tradeoff described in the text is direct. In general, SRS offers better preservation of neurocognitive function and quality of life, whereas WBRT improves distant and overall intracranial control rates. In practice, that contrast is helpful because it gives the team a clean way to explain the difference: one technique prioritizes functional preservation, the other strengthens coverage of microscopic disease and failures away from the index target.
Diffuse and numerous metastases push the case toward WBRT. Intact metastases and postoperative cavities, when small and adequately separated from critical structures, fit the SRS logic described later in the chapter.
WBRT: planning and target delineation principles
The WBRT section starts with the nonnegotiable basics. Planning uses a noncontrast CT scan from the vertex to the upper cervical spine, with axial slice thickness of 2.5 mm or less. The patient is supine, the head is kept neutral, immobilization is provided with a thermoplastic mask, and the requested field of view is 600 mm. It is a simple simulation recipe, but it is stable enough to reproduce the field edges that protect the lenses, soft tissue, and parotids while keeping the entire cranial contents covered.
When the goal includes cognitive preservation, the chapter points to two strategies: memantine and/or hippocampal avoidance WBRT. There is no room for approximation here. Hippocampal avoidance depends on fusing the planning CT with a gadolinium-enhanced MRI using a three-dimensional spoiled gradient sequence and axial slices of 1.25 to 1.5 mm to define the avoidance region. Under the RTOG 0933 contouring rules shown in the figure, only the SGZ portion of the hippocampi is contoured, then expanded volumetrically by 5 mm. The PTV is the whole brain tissue minus that 5 mm expansion, with no added setup margin.
In conventional arrangements, the typical technique is 3D-CRT with opposed lateral 6 MV photon beams and an MLC block. The chapter figures are useful because they show what usually goes wrong when the geometry is relaxed: undercoverage of the cribriform plate, poor temporal lobe coverage, or direct lens divergence. The figure below condenses the standard arrangement with the beam pair rotated slightly in RAO/LAO.
Suggested WBRT Fields
Table 28.1 lays out the clinical scenarios, the selected technique, and the geometric details that actually change the field design. It also shows when planning moves away from conventional lateral opposed beams and into IMRT or VMAT with hippocampal avoidance.
| Aspect | Conventional WBRT | Leptomeningeal disease | Lymphoma/leukemia | Scalp sparing | HA-WBRT |
|---|---|---|---|---|---|
| Clinical scenario | Diffuse brain metastases, numerous lesions, or lesions that are simply “too many to count”; PCI for small-cell lung cancer. | Leptomeningeal disease. | CNS prophylaxis for ALL; high-risk CNS leukemia. | Cosmetic outcome is a priority and the technique may produce a “reverse Mohawk.” | Diffuse brain metastases, numerous lesions, or lesions that are “too many to count”; PCI for small-cell lung cancer; exclusion if the lesion lies within 5 mm of the hippocampus. |
| Technique | 3D-CRT with lateral opposed beams rotated slightly off axis in RAO/LAO to avoid divergence into the lenses. | 3D-CRT with lateral opposed beams rotated slightly off axis in RAO/LAO to avoid divergence into the lenses. | 3D-CRT with lateral opposed beams rotated slightly off axis in RAO/LAO to avoid divergence into the lenses. | 3D-CRT with lateral opposed beams rotated slightly off axis in RAO/LAO to avoid divergence into the lenses. | IMRT/VMAT. |
| Target volumes and margins | Entire cranial contents. | Entire cranial contents plus optic nerves, retro-orbital region, and lamina cribosa. | Entire cranial contents plus optic nerves, retro-orbital region, retina, and, when ocular involvement is present, the whole globe. | Entire cranial contents. | CTV: whole brain parenchyma to the foramen magnum. PTV: CTV minus the hippocampi with a 5 mm expansion, with no setup margin. |
| Field edges and coverage | Superior: 2 cm flash. Posterior: 2 cm flash with or without a posterior neck MLC block to protect soft tissue. Inferior: bottom of C1. Anterior: MLC block from 2 cm flash to the anterior aspect of C1, blocking parotids and lenses. Cover the temporal lobes and cribriform plate. | Cover the temporal lobes and cribriform plate with an additional 8-10 mm margin for penumbra and daily setup. | Cover the temporal lobes and cribriform plate with an additional 8-10 mm margin for penumbra and daily setup; cover the posterior one-third of the globes if there is no ocular involvement on slit-lamp examination, or the entire bilateral globes if ocular involvement is present. | Set the MLC edge at the outer table of the calvarium. | No conventional geometric field edge is listed; the chapter describes inverse planning with hippocampal avoidance. |
| Normal structure constraints | — | — | — | — | Brain metastases: hippocampi D100% ≤ 9 Gy and Dmax ≤ 16 Gy; optic nerves and chiasm Dmax ≤ 30 Gy. PCI for small-cell lung cancer: hippocampi D100% ≤ 7.5 Gy and Dmax ≤ 13.5 Gy; optic nerves and chiasm Dmax ≤ 25 Gy. |
Source: Target Volume Delineation and Field Setup, 2nd Edition (Table 28.1)
The conventional technique variations prevent a common error: treating every whole-brain case as the same geometry. In leptomeningeal disease, the field sits farther from the cribriform plate. In CNS lymphoma or leukemia, orbital coverage changes. In the scalp-sparing approach, the MLC edge moves to the outer table of the calvarium.
WBRT Dose and Fractionation
Table 28.2 captures a simple but important point: the prescription changes with the clinical setting. The chapter keeps the common schedules for brain metastases and leptomeningeal disease, but it also includes reirradiation, PCI, and CNS prophylaxis.
| Clinical scenario | Dose and fractionation |
|---|---|
| WBRT, leptomeningeal disease | 30 Gy in 10 fractions (most common); 37.5 Gy in 15 fractions (RTOG); 30 Gy in 12 fractions; 20 Gy in 5 fractions for poor prognosis. |
| WBRT reirradiation | 20-25 Gy in 10 fractions, with a minimum interval of 4-6 months. |
| PCI for small-cell lung cancer | 25 Gy in 10 fractions (most common). |
| CNS prophylaxis for ALL | 12 Gy in 8 fractions. |
| High-risk CNS leukemia | ≥18 Gy in 9-10 fractions, with dose based on the intensity of systemic therapy. |
Source: Target Volume Delineation and Field Setup, 2nd Edition (Table 28.2)
Setup verification in WBRT is also stated clearly. The chapter uses weekly orthogonal films with MV imaging. Daily kV is usually reserved for IMRT-based WBRT. That fits the increase in geometric complexity when the goal shifts from treating the entire cranial contents to treating the whole brain while actively sparing the hippocampi inside explicit numeric constraints.
The HA-WBRT figure makes the central message clear: coverage remains diffuse, but the plan is now shaped by memory-related structures and by explicit optic pathway constraints.
SRS: planning and target delineation principles
The SRS section goes straight to the clinical use cases. The chapter includes single-fraction SRS and fractionated SRS in 2 to 5 fractions for both intact brain metastases and postoperative cavities. Fractionation depends on target size or target volume and on the distance to critical structures. The equipment can be a frame-based or frameless cobalt-based Leksell Gamma Knife or a LINAC-based platform.
For target delineation and treatment planning, the preferred imaging is a volumetric contrast-enhanced T1 MRI with 1 to 2 mm slices. If the patient cannot tolerate MRI or has an incompatible implanted device, contrast-enhanced CT is the fallback. In LINAC-based SRS, a thin-slice CT is acquired and co-registered, and daily imaging becomes mandatory. That detail is practical, not ornamental. With margins this tight, daily imaging is part of the treatment definition.
For intact metastases, the rule is spare and clean: GTV equals the contrast-enhancing lesion on T1 MRI, and CTV is GTV plus 0 mm. The chapter keeps this simple because the advantage of SRS lives in selectivity. The postoperative cavity discussion is where the contouring strategy becomes more nuanced.
Method 1, as described by Soltys et al., applies a uniform 2 mm expansion around the cavity seen on postcontrast MRI. Method 2, described by Soliman et al., includes the entire enhancing region, the surgical cavity, and the surgical tract. If the original tumor touched dura, a 5 to 10 mm margin is added along the bone flap beyond the initial region of contact. If there was no dural contact, the bone-flap margin drops to 1 to 5 mm. If there was venous sinus contact, another 1 to 5 mm margin is added along the sinus.
The chapter figure dedicated to the postoperative cavity translates these differences well. The case is a 24 mm left temporal lobe cavity after gross total resection of a 33 mm rectal metastasis with preoperative dural contact but no venous sinus contact. Single-fraction SRS was chosen because the cavity was small, under 3 cm, and sufficiently far from delicate brain structures. Because treatment used Leksell Gamma Knife, PTV was equal to CTV with 0 mm additional expansion. The figure also labels the right optic nerve, left optic nerve, and brainstem.
SRS Target Delineation
Table 28.3 presents two clearly different postoperative cavity philosophies. One uses a uniform 2 mm expansion. The other includes the enhancing region, the surgical cavity, the surgical tract, and extensions along the bone flap or venous sinus when the preoperative tumor had contact there.
| Situation | GTV | CTV |
|---|---|---|
| Unresected brain metastases | Contrast-enhancing lesion on T1-weighted MRI. | GTV + 0 mm. |
| Postoperative cavity after gross total resection (method 1) [2] | n/a | 2 mm expansion around the resection cavity borders visualized on postcontrast MRI. |
| Postoperative cavity after gross total resection (method 2) [3] | n/a | Entire contrast-enhancing region, surgical cavity, and surgical tract seen on postoperative MRI; 5-10 mm margin along the bone flap beyond the initial region of preoperative tumor contact when the tumor contacted dura; 1-5 mm margin along the bone flap when the tumor did not contact dura; 1-5 mm margin along the venous sinus when the tumor contacted a venous sinus. |
Source: Target Volume Delineation and Field Setup, 2nd Edition (Table 28.3)
The second example describes four new breast cancer metastases, 6 to 20 mm in size and 0.07 to 1.92 cm3 in volume, after prior WBRT of 30 Gy in 10 fractions. Because all were under 3 cm and adequately separated from delicate structures, treatment again used single-fraction Gamma Knife, with GTV defined by post-gadolinium T1 enhancement and 0 mm expansion to both CTV and PTV.
SRS Dose and Organ-at-Risk Constraints
Table 28.4 makes the practical logic easy to see. Smaller lesions and small cavities stay in the single-fraction range; once volume grows or the relationship to critical structures changes, 3- and 5-fraction schedules become the safer framework.
| Scheme | 1 Fraction | 3 Fractions | 5 Fractions |
|---|---|---|---|
| PTV dose (postoperative cavity) | 20 Gy (<4.2 cm3) 18 Gy (≥4.2 to <8.0 cm3) 17 Gy (≥8.0 to <14.4 cm3) 15 Gy (≥14.4 to <20 cm3) 14 Gy (≥20 to <30 cm3) 12 Gy (≥30 cm3) |
27 Gy (<30 cm3) | 30 Gy (≥30 cm3 to <5 cm) |
| PTV dose (unresected metastases) | 24 Gy (<1 cm) 22 Gy (≥1.0 to <2.0 cm) 18 Gy (≥2.0 to <3.0 cm) 15 Gy (≥3.0 to <4.0 cm) |
27 Gy | 30 Gy |
| Brainstem constraint | V12 Gy < 1 cm3 V18 Gy < 0.5 cm3 |
23.1 Gy maximum V23 Gy < 0.5 cm3 |
28 Gy maximum |
| Optic apparatus constraint | 9 Gy maximum | 17.4 Gy maximum V13.8 Gy < 0.2 cm3 |
23 Gy maximum V20 Gy < 0.2 cm3 |
Source: Target Volume Delineation and Field Setup, 2nd Edition (Table 28.4)
The dose table makes one practical difference easy to see. For small postoperative cavities, the chapter allows 20 Gy in 1 fraction when the volume is under 4.2 cm3, then steps down the prescription as the volume increases through progressively larger groups up to 30 cm3 or more. When the cavity is larger, but still under 5 cm, the schedule moves to 30 Gy in 5 fractions. In unresected metastases, the logic changes from volume to diameter: 24 Gy for lesions under 1 cm, 22 Gy for lesions from 1.0 to under 2.0 cm, 18 Gy from 2.0 to under 3.0 cm, and 15 Gy from 3.0 to under 4.0 cm.
The brainstem and optic apparatus constraints also show why the distance to critical structures matters so much. In 1 fraction, the optic apparatus is limited to 9 Gy maximum, while the brainstem is held to V12 Gy under 1 cm3 and V18 Gy under 0.5 cm3. In 3 fractions, the brainstem limit becomes 23.1 Gy maximum with V23 Gy under 0.5 cm3, and the optic apparatus is held to 17.4 Gy maximum with V13.8 Gy under 0.2 cm3. In 5 fractions, the maxima rise to 28 Gy for the brainstem and 23 Gy for the optic apparatus, with optic V20 Gy under 0.2 cm3.
If there is one clear thread across both parts of the chapter, it is this: in brain metastases, technique and target volume are inseparable. WBRT treats the whole cranial contents and then adapts for leptomeningeal spread, lymphoma, leukemia, scalp sparing, or hippocampal avoidance. SRS depends on thin-slice MRI, minimal margins, and strict selection by size, volume, and proximity to critical structures. Reading the two strategies side by side helps avoid the most common planning error in this disease: treating very different clinical problems as if they should receive the same geometric solution.
