Radiation Oncology: A Question-Based Review

Palliative Care: Brain Metastases

Bronwyn R. Stall and Kevin Camphausen

image Background

What is the most common intracranial tumor?

Brain mets is the most common intracranial tumor.

What is the annual incidence of brain mets in the U.S.?

170,000 cases/yr of brain mets in the U.S.

What cancers are associated with hemorrhagic brain mets?

Hemorrhagic brain mets is associated with melanoma, renal cell carcinoma, and choriocarcinoma.

What does the term solitary brain met connote?

A solitary brain met is only 1 brain lesion.

What cancers are most likely to metastasize to the brain?

Cancers associated with brain mets: lung (40%–50%), breast (15%), melanoma (10%), and unknown −primary (5%–10%)

What is more common type of brain met: single or multiple?

Most pts have multiple brain mets rather than a single lesion.

How do pts with brain mets present?

Presentation of pts with brain mets: Sx of ↑ ICP (HA, n/v), weakness, change in sensation, mental status changes, and seizure

Where do most brain mets occur?

Most brain mets arise in the gray/white matter −junction due to narrowing of blood vessels. (Delattre J et al., Arch Neurol 1988)

Are most brain mets infra- or supratentorial?

The majority of brain mets are supratentorial.

What is the distribution of brain mets within the brain?

The distribution of brain mets correlates with relative weight and blood flow:

1.     Cerebral hemispheres: 80%

2.     Cerebellum: 15%

3.     Brain stem: 5%

(Delattre J et al., Arch Neurol 1988)

If a pt presents with brain mets without a prior Dx of cancer, what is the most likely source?

Pts presenting with brain mets without a prior Dx of cancer most often have a lung primary.

What is the overall median time from initial cancer Dx to development of brain mets?

The median overall time from initial cancer Dx to development of brain mets is 1 yr.

Do most pts with brain mets die from their CNS Dz?

No. ~30%–50% of pts with brain mets die from their CNS Dz.

image Workup/Staging

Describe the workup of a brain met.

Brain met workup: H&P focus on characterization of any neurologic Sx, evaluation for infectious causes (fever, CBC), careful neurologic exam, MRI brain +/− gadolinium, assessment for status of extracranial Dz, determination of Karnofsky performance status (KPS), and neurosurgery consult

What imaging test is 1st line in evaluating brain mets?

MRI is 1st line in the evaluation of brain mets.

What is the DDx for a new lesion in the brain?

Brain lesion DDx: mets, infection/abscess, hemorrhage, primary brain tumor, infarct, tumefactive demyelinating lesion, and RT necrosis

What imaging features are suggestive of brain mets?

Imaging features suggestive of brain mets include lesions at gray/white matter junction, multiple lesions, ring-−enhancing lesions, and significant vasogenic edema

What is triple-dose gadolinium, and why is it used?

Triple-dose gadolinium: 0.3 mmol/kg. It is used to increase the sensitivity of MRI.

image Treatment/Prognosis

The RTOG recursive partitioning analysis (RPA) divides brain mets pts into how many prognostic classes?

The RTOG RPA divides brain mets pts into 3 −prognostic classes.

What prognostic factors are included in the RPA for brain mets?

Prognostic factors included in the RPA for brain mets include KPS, control of the primary, age <65 yrs, and the presence of mets outside the CNS. (Gaspar L et al., IJROBP 1997)

What pts are included in class I according to the RTOG for brain mets?

Brain met RPA class I: KPS ≥70, age <65 yrs, −primary controlled, and no extracranial mets

What pts are included in class II according to the RTOG for brain mets?

Brain mets RPA class II: KPS ≥70 with 1 of the following—primary uncontrolled, age >65 yrs, or extracranial mets

What pts are included in class III according to the RTOG for brain mets?

Brain mets RPA class III: KPS <70

What is the MS time for RTOG RPA classes I, II, and III?

MS according to the RTOG brain met RPA:

1.     Class I: 7.2 mos

2.     Class II: 4.2 mos

3.     Class III: 2.3 mos

What is the Sperduto Index?

The Sperduto Index is a graded prognostic assessment based on age, KPS, # of brain mets, and the presence or absence of extracranial mets developed from an analysis of 1,960 pts in the RTOG database. Criteria is based on a point system:

1.     0 points: age >60 yrs, KPS <70, >3 brain mets, presence of extracranial mets

2.     0.5 points: age 50–59 yrs, KPS 70–80, 2 CNS mets

3.     1 point: age <50 yrs, KPS 90–100, 1 CNS met, no extracranial mets

The sum of points predicts MS in mos:

1.     0–1 point: 2.6 mos

2.     1.5–2.5 points: 3.8 mos

3.     3 points: 6.9 mos

4.     3.5–4 points: 11 mos

(Sperduto P et al., IJROBP 2007)

In pts with untreated brain mets, what is the MS?

MS of untreated brain mets is 1 mo.

What Tx may be used for brain mets?

Brain met Tx: steroids, surgery, fractionated RT (WBRT), and SRS

In pts with brain mets treated with steroids alone, what is the MS?

MS in pts with brain mets treated with steroids alone is 2 mos.

How are steroids for brain mets typically prescribed?

Steroid dose for newly diagnosed brain mets: 4 mg dexamethasone q6hrs; may give initial loading dose of 10 mg.

Why are steroids used for symptomatic brain mets?

In pts with symptomatic brain mets, steroids reduce leakage from tumor vessels, therefore decreasing edema and mass effect.

What pharmacologic Tx should always accompany steroid Tx?

When prescribing steroids, also provide GI prophylaxis with a proton-pump inhibitor or H2 blocker.

Should anticonvulsants be used prophylactically?

No. In accordance with guidelines from the American Academy of Neurology, pts with newly diagnosed brain tumors should not be started on prophylactic anticonvulsants. (Glantz M et al. Neurology 2000)

Is there any randomized data on the dose for WBRT?

Yes. The RTOG conducted several RCTs from 1970–1995 of WBRT alone, assessing different fractionation schemes. The 1st 2 trials (RTOG 6901 and 7361) included >1,800 pts randomized to 40 Gy/20, 40 Gy/15, 30 Gy/15, 30 Gy/10, or 20 Gy/5. No significant difference was found in response rates, length of response, or OS. The MS in the 1st study was 4.1 mos and 3.4 mos in the 2nd. (Borgelt B et al., IJROBP 1980)

2 ultra-rapid fractionation schemes were also tested on these studies and reported separately; 10 Gy/1 (RTOG 6901) and 12 Gy/2 (RTOG 7361) in 26 and 33 pts, respectively. These schedules were associated with worse toxicity and time to neurologic progression than the standard fractionation. (Borgelt B et al., IJROBP 1981)

2 studies showed no MS advantage to giving a higher total dose. RTOG 7606 randomized 255 pts to 30 Gy/10 vs. 50 Gy/20. MS was 4.1 and 3.9 mos, respectively (Kurtz J et al., IJROBP 1981). RTOG 9104 −randomized 429 pts to 30 Gy/10 vs. 54.4/1.6 Gy bid. MS was 4.5 mos in both arms (Murray K et al., IJROBP 1997).

What dose and fractionation schemes are considered standard for WBRT?

The most standard WBRT dose is 30 Gy/10. Pts with a good KPS and longer life expectancy may be treated to 37.5 Gy/15, 40 Gy/20, or 50 Gy/20.

What % of brain met pts have Sx improvement with WBRT?

WBRT improves Sx from brain mets in ~60% of cases.

What is the rate of CR to WBRT for brain mets?

~25% of pts have a CR to WBRT for brain mets.

Should Bx be recommended if a new Dx of brain mets is suspected?

Yes. Bx should be considered if a new Dx of brain mets is suspected, as 11% of pts enrolled on the 1st Patchell trial were found not to have metastatic Dz on Bx despite MRI or CT findings consistent with −metastatic Dz. (Patchell R et al., NEJM 1990)

What data supports surgery + RT rather than Bx + RT for brain mets?

The 1st Patchell study for brain mets randomized 48 pts with 1 brain met and KPS ≥70 to surgery + WBRT vs. Bx + WBRT. WBRT in both arms was 36 Gy in 3 Gy/fx. Pts treated with surgery had a longer MS (40 wks vs. 15 wks, p <0.01), longer functional independence (38 wks vs. 8 wks), and ↓LR (20% vs. 52%, p <0.02). (Patchell R et al., NEJM 1990)

Did the Netherlands trial of WBRT +/− surgery support or refute the Patchell study?

The Noordijk study supported the findings of the 1st Patchell study. It randomized 63 pts to WBRT alone or surgery + WBRT. WBRT was 40 Gy in 2 Gy bid fx. Pts treated with surgery had improved MS (10 mos vs. 6 mos, p = 0.04) and longer functional independence (7.5 mos vs. 3.5 mos, p = 0.06). (Noordijk E et at., IJROBP 1994)

Does adj WBRT after surgical resection of a brain met improve OS?

No. Postop WBRT following resection of a brain met does not improve survival. In the 2nd Patchell study for brain mets, 95 pts following surgical resection of a −single met were randomized to no further Tx or WBRT (50.4 Gy in 1.8 Gy/fx). WBRT decreased LR (10% vs. 46%), decreased the rate of any brain failure (18% vs. 70%), and decreased the rate of neurologic death (14% vs. 44%) but did significantly change MS (48 wks vs. 43 wks). (Patchell R et al., JAMA 1998)

Are there any current studies assessing the benefit of WBRT?

YesEORTC 22952 is an RCT of pts with 1–3 brain mets s/p surgery or SRS randomized to no further Tx vs. WBRT. This was closed to accrual in 2007 but has not yet been reported.

What was the 1st randomized study of WBRT +/− an SRS boost?

The 1st RCT of WBRT +/− an SRS boost was conducted at the University of Pittsburgh. 27 pts with KPS ≥70 and 2–4 mets ≤2.5 cm that were at least 5 mm from the chiasm were randomized to WBRT (30 Gy/12 fx) +/− a 16-Gy boost. The trial closed early b/c of significant difference in brain control. The SRS arm had a longer time to LF (36 mos vs. 6 mos, p = 0.0005) and longer time to any brain failure (34 mos vs. 5 mos, p = 0.002) but no difference in OS (11 mos vs. 7.5 mos, p = 0.11). (Kondziolka D et al., IJROBP 1999)

According to RTOG 9508, which pts had a survival advantage with the addition of an SRS boost to WBRT?

RTOG 9508 randomized 331 pts with 1–3 brain mets to WBRT + SRS boost vs. WBRT alone. WBRT on both arms was 37.5 Gy in 2.5 Gy/fx. The SRS boost dose was dependent on size in accordance with RTOG 9005. On univariate analysis, the addition of SRS improved the MS for pts with a single brain met (6.5 mos vs. 4.9 mos, p = 0.39). On subgroup multivariate analysis (MVA), RPA class I pts had improved survival with the SRS boost, as did pts with a lung cancer primary. (Andrews D et al., Lancet 2004)

What is the main determinant in selecting the Rx dose for SRS Tx of a brain met?

The SRS Rx dose for a brain met is determined by size in accordance with the results of RTOG 9005, a dose escalation study: 24 Gy if <2-cm diameter, 18 Gy if 2–3 cm, and 15 Gy if 3–4 cm. (Shaw H et al., IJROBP 1996)

What retrospective data supports the omission of upfront WBRT in pts treated with SRS for brain mets?

Sneed et al. compiled a database from 10 U.S. −institutions to assess the effect of omitting upfront WBRT in pts treated with SRS for brain mets. 983 pts were analyzed and excluded pts treated with surgery (159 pts) and pts with a >1-mo interval between WBRT and SRS (179 pts). Of 569 evaluable pts, 268 had SRS alone and 301 had upfront WBRT + SRS. When adjusted for RPA class, there was no difference in survival. (IJROBP 2002)

According to randomized data, what is the effect of delaying WBRT after SRS for pts with 1–4 brain mets?

JROSG99-1 showed that the omission of WBRT after SRS for 1–4 brain mets does not affect survival but increases the risk of intracranial relapse (46% with SRS + WBRT vs. 76.4% with SRS alone) and thus increases the need for salvage Tx. (Aoyama H et al., JAMA 2006)

What was the 1st ran−domized trial to assess the effect of delaying WBRT after SRS?

The 1st trial to assess the effect of delaying WBRT after SRS was JROSG99-1. 132 pts with 1–4 mets were randomized to SRS or WBRT + SRS. The SRS dose was based on size (lesions ≤2 cm to 22–25 Gy; lesions >2 cm to 18–20 Gy) and randomization (30% SRS dose reduction for pts on the WBRT arm). The WBRT dose was 30 Gy in 10 fx. (Aoyama H et al., JAMA 2006)

In the RTOG SRS dose escalation study, did pts rcv WBRT?

No. In RTOG 9005, an SRS dose escalation study, pts did not rcv WBRT.

Is there any data on SRS dosing with planned WBRT?

Yes. Retrospective data from the University of −Kentucky showed that optimal control of brain mets ≤2 cm was achieved with SRS of 20 Gy + WBRT. Pts treated with >20 Gy SRS + WBRT had higher rates of grade 3–4 neurotoxicity. (Shehata M et al., IJROBP 2004)

Is there any data comparing surgery + WBRT with SRS alone?

Yes. Retrospective data from Germany comparing RPA class I–II pts with 1–2 brain mets treated either with surgery + WBRT or SRS alone suggests that SRS is as effective. Of 206 pts treated from 1994–2006, 94 pts had SRS alone (18–25 Gy), and 112 pts had resection + WBRT (30 Gy/10 or 40 Gy/20). At 12 mos, there was no difference in OS (~50% in both groups), LC, or brain control. There was no difference according to the RPA group. (Rades D et al., Cancer 2007)

Can pts treated with WBRT for brain mets be reirradiated?

Yes. Reirradiation following WBRT may be considered in pts who initially responded well to WBRT and then develop worsening neurologic function at least 4 mos after initial WBRT. (Cooper J et al., Radiology 1990)

What dose should be used for reirradiation after WBRT for brain mets?

The optimal dose for reirradiation after WBRT is unknown. 20 Gy in 10 fx is often used.

How are the fields arranged for WBRT?

WBRT is delivered using opposed lat fields; a post gantry tilt of 3–5 degrees is used to avoid divergence into the eyes; and multileaf collimation or custom blocks are used to ensure adequate coverage of the cribriform plate, temporal lobe, and brain stem while protecting the eyes, nasal cavity, and oral cavity. The inf border is generally set at C1-2.

Should surgery be used for recurrent tumors?

Yes. Retrospective data from MDACC have suggested that reoperation for recurrent brain mets can prolong survival and improve QOL. MVA revealed several −negative prognostic factors: presence of systemic Dz, KPS <70, short time to recurrence (<4 mos), age ≥40 yrs, and breast and melanoma primaries. (Bindal R et al., J Neurosurg 1995)

What is the advantage of tumor bed radiosurgery after brain met resection?

Retrospective data from the University of Sherbrooke in Canada have suggested that SRS to the tumor bed following resection for brain mets achieves LC rates that are comparable to WBRT but does not impact the development of remote brain mets. 40 pts underwent resection → SRS at a median of 4 wks post resection. 73% achieved LC, and 54% developed new brain mets. (Mathieu D et al., Neurosurgery 2008)

image Toxicity

What are potential acute toxicities of WBRT?

Potential WBRT acute toxicities: alopecia, fatigue, HA, n/v, ototoxicity

What are potential long-term toxicities of WBRT?

Potential WBRT chronic toxicities: thinned hair, decline in short-term memory, altered executive function, leukoencephalopathy, brain atrophy, normal pressure hydrocephalus, RT necrosis

What is the relationship between WBRT-induced brain met shrinkage and neurocognitive function?

WBRT-induced brain met shrinkage correlates with improved neurocognitive function. This was demonstrated in an analysis of 208 pts with brain mets randomized to WBRT alone on a phase III trial of WBRT +/− motexafin gadolinium. Pts with a good response (>45% tumor volume reduction at 2 mos) to WBRT had a longer time to decline in neurocognitive function. (Li J et al., JCO 2007)

What daily fx size in WBRT is associated with RT necrosis?

WBRT administered in fx sizes >300 cGy/day are associated with RT necrosis. (DeAngelis L et al., Neurology 1989)

Name the potential acute toxicities of SRS for brain mets.

Potential acute toxicities of SRS for brain mets: HA, nausea, dizziness/vertigo, seizure

What is the risk of symptomatic RT necrosis after SRS for brain mets?

There is an ~5% risk of symptomatic RT necrosis secondary to SRS for brain mets. This is usually treated with steroids but may require surgery.