Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Peterson, A. M.
Right arrow Articles by Kondziolka, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Peterson, A. M.
Right arrow Articles by Kondziolka, D.
(Radiology. 1999;211:807-814.)
© RSNA, 1999

Neuroradiology

MR Imaging Response of Brain Metastases after Gamma Knife Stereotactic Radiosurgery1

Arleen M. Peterson, MD, Carolyn Cidis Meltzer, MD, E. Jane Evanson, FRCR, John C. Flickinger, MD and Douglas Kondziolka, MD

1 From the Departments of Radiology, Division of Neuroradiology (A.M.P., C.C.M., E.J.E.), Psychiatry (C.C.M.), Radiation Oncology (J.C.F., D.K.), and Neurological Surgery (J.C.F., D.K.), University of Pittsburgh Medical Center, Rm PUH D-132, 200 Lothrop St, Pittsburgh, PA 15213-2582. Received March 20, 1998; revision requested June 17; revision received August 7; accepted November 5. Address reprint requests to A.M.P.


   Abstract
TOP
 Abstract
Introduction
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
PURPOSE: To characterize the magnetic resonance (MR) imaging response of brain metastases after gamma knife stereotactic radiosurgery and determine whether imaging features and tumor response rates correlate with local tumor control and survival.

MATERIALS AND METHODS: Serial MR examinations were performed in 48 patients (25 men, 23 women; mean age, 58 years) with 78 lesions. Pretreatment and follow-up enhancing lesion volumes and imaging features were assessed. Rates of response to stereotactic radiosurgery were calculated. Prognostic imaging features affecting local control and survival were analyzed.

RESULTS: Local tumor control was achieved in 66 (90%) of 73 metastases at 20 weeks after stereotactic radiosurgery; 61% maintained local control at 2 years. A homogeneous baseline enhancement pattern and initial good response rate (>50% lesion volume reduction) predicted local control. Five metastases demonstrated a transient volume increase after treatment. The median survival time after stereotactic radiosurgery was 53 weeks and correlated with systemic disease burden and primary tumor type.

CONCLUSION: Baseline homogeneous tumor enhancement and initial good response correlate with local control. Initial lesion growth does not preclude local control and may represent radiation-related change. Recognition of these serial MR imaging findings may guide image interpretation and influence treatment in patients with stereotactic radiosurgery–treated metastases.

Index terms: Brain, MR, 13.121411, 13.12143 • Brain neoplasms, secondary, 13.381, 13.3810, 13.3816, 13.38181, 13.382, 13.3820, 13.3826, 13.38281 • Stereotaxis, 13.1267


   Introduction
TOP
 Abstract
 Introduction
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Stereotactic radiosurgery is a noninvasive procedure for the treatment of brain metastases and an alternative to surgical resection (1). Lars Leksell, MD, is credited with the concept, terminology, and initial development of the technology (2). With computed tomography (CT) or magnetic resonance (MR) guided stereotactic localization of an intracranial lesion, in stereotactic radiosurgery, a Gamma Knife (Elekta Instruments, Atlanta, Ga) is used to deliver a single high dose of cobalt 60 gamma radiation to a radiographically discrete target (3). Radiosurgery can also be performed with a modified linear accelerator or cyclotron. With stereotactic radiosurgery, there is a steep dose gradient at the target periphery, which markedly reduces the dose of radiation to the surrounding normal brain tissue (4). Unlike complete surgical resection, which eliminates the tumor at the time of the operation, stereotactic radiosurgery causes the tumor to undergo necrosis or growth arrest over time (5).

The goals of stereotactic radiosurgery are local tumor control, improved quality of life, and prolonged survival (1,6,7). Local tumor control is defined as the absence of a substantial increase in tumor volume (<25%) at follow-up MR imaging (8). Metastatic lesions are particularly well suited for treatment with stereotactic radiosurgery because they are usually small (<3 cm), well circumscribed, spherical, and have radiographically distinct enhancing margins (7). Compared with surgical resection, less invasive stereotactic radiosurgery has the added advantage of enabling one to treat multiple metastases in one session.

To our knowledge, the potential prognostic MR imaging features as indicators of local tumor control in patients with stereotactic radiosurgery–treated brain metastases have not been previously defined. Although high local control rates that range from 82% to 96% have been achieved with stereotactic radiosurgery (6,8–14), detailed descriptions of the serial MR imaging changes that stereotactic radiosurgery–treated metastases undergo have not been reported. To evaluate the MR imaging response of brain metastases to stereotactic radiosurgery, we retrospectively reviewed the pretreatment (ie, baseline) and serial posttreatment follow-up MR images in patients with stereotactic radiosurgery–treated brain metastases. The objectives of this study were threefold: (a) to characterize the change in MR imaging appearance over time and response of stereotactic radiosurgery–treated metastatic brain lesions, (b) to determine whether baseline imaging features, including size and enhancement characteristics, and differences in initial tumor response rates correlate with long-term maintenance of local tumor control, and (c) to determine whether imaging response correlates with patient survival.


   MATERIALS AND METHODS
TOP
 Abstract
 Introduction
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 References
 
The medical records of 227 consecutive patients with intracranial metastases treated with stereotactic radiosurgery at our institution during a recent 5-year interval were retrospectively reviewed. Patients with extraaxial metastases, patients who underwent baseline and follow-up examinations with CT rather than with MR imaging, and patients who underwent imaging follow-up at outside institutions were excluded from the study. In our study, a subgroup of 48 patients (with 78 metastases) whose pretreatment MR examination and at least one follow-up MR examination were performed at our institution were examined. The study cohort consisted of 25 men (52%) and 23 women (48%) with a median age of 58 years (age range, 30–83 years). A summary of the primary cancer types is shown in Table 1.


View this table:
[in this window]
[in a new window]
  		TABLE 1. Primary Cancer Types
 
The median stereotactic radiosurgery dose delivered to the tumor margin was 16 Gy (range, 14–25 Gy). This was prescribed to the 40%–70% isodose line, which corresponded to the lesion periphery as defined by using contrast material–enhanced MR imaging. The median central dose delivered to the tumor was 16 Gy divided by 0.4–0.7. The dose prescription was based on the tumor volume, prior history of whole-brain radiation therapy, and tumor location. All except five patients received whole-brain radiation therapy either prior to or in conjunction with stereotactic radiosurgery. The mean dose of whole-brain radiation therapy was 3,000 cGy (range, 2,000–5,000 cGY), which was given in 10 fractions.

Stereotactic radiosurgery was performed in 31 (65%) patients with solitary metastasis, 11 (23%) patients with two metastases, four (8%) patients with three metastases, and two patients (4%) with four metastases. Four (8%) patients who were initially treated with stereotactic radiosurgery for solitary metastases underwent subsequent stereotactic radiosurgery for five metachronous metastases. The median interval between the diagnosis of the brain metastasis and the stereotactic radiosurgery was 60 days.

MR imaging was performed with a 1.5-T imaging unit (Signa; GE Medical Systems, Milwaukee, Wis) by using a standard head coil. The baseline MR imaging protocol performed immediately prior to stereotactic radiosurgery consisted of T1-weighted (400–800/12–26 [repetition time msec/echo time msec]) sagittal and axial spin-echo images obtained before contrast material administration and T1-weighted (400–800/12–26) axial and coronal spin-echo images obtained after the administration of either 0.1 mmol/kg gadoteridol (Prohance; Bracco, Princeton, NJ) or gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ) with a stereotactic head frame in place. The MR imaging protocol for follow-up examinations included axial fast spin-echo T2-weighted (2,500–4,000/85–102 [effective]) and intermediate-weighted (2,000/17–20 [effective]) sequences.

Three neuroradiologists (A.M.P., C.C.M., E.J.E.) served as case readers. For each baseline and follow-up MR imaging examination, two readers independently measured the maximum enhancing metastatic lesion diameter in three orthogonal planes (anterior-posterior [d1], transverse [d2], and superior-inferior [d3]), assessed the degree of mass effect (none, mild [effacement of local sulci], moderate [effacement of adjacent ventricle and sulci], or severe [midline and/or uncal herniation), and evaluated the lesion enhancement characteristics (none [observed on posttreatment images only], homogeneous, or heterogeneous [also subclassified as rim-enhancing and patchy]). Interobserver discrepancies were resolved by consensus.

Lesion volumes were calculated according to the following mathematical formula for the volume of an ellipsoid: volume = 4/3 • {pi}(d1/2 • d2/2 • d3/2). The imaging features were recorded at baseline and at all follow-up examinations performed during the following five intervals after stereotactic radiosurgery, when available: 0–10 weeks, 10–20 weeks, 20–40 weeks, 40–80 weeks, and more than 80 weeks. For those patients who underwent more than one MR examination in a given follow-up interval (seven patients [nine metastatic lesions] underwent two MR examinations, and one patient [four metastatic lesions] underwent four MR examinations in one of their follow-up intervals), the mean lesion volume from examinations performed in that interval was recorded. The lesion location was also noted.

The numbers of patients and metastatic lesions imaged during each of the five follow-up intervals are shown in Figure 1. Forty-eight metastases were imaged within 10 weeks, and 43 metastases (25 imaged for the first time and 18 imaged for the second time) were imaged between 11 and 20 weeks after stereotactic radiosurgery. Therefore, 73 metastases were initially imaged within 20 weeks after stereotactic radiosurgery.



View larger version (18K):
[in this window]
[in a new window]
  		Figure 1a. {block} = first follow-up, {square} = second follow-up, = third follow-up, = fourth follow-up. (a) Time course of serial MR imaging in all patients who underwent imaging during each of the five follow-up intervals after stereotactic radiosurgery. (b) Time course of serial MR imaging of all lesions in which imaging was performed during each of the five follow-up intervals after stereotactic radiosurgery.

 


View larger version (21K):
[in this window]
[in a new window]
  		Figure 1b. {block} = first follow-up, {square} = second follow-up, = third follow-up, = fourth follow-up. (a) Time course of serial MR imaging in all patients who underwent imaging during each of the five follow-up intervals after stereotactic radiosurgery. (b) Time course of serial MR imaging of all lesions in which imaging was performed during each of the five follow-up intervals after stereotactic radiosurgery.

 
The MR imaging response of each stereotactic radiosurgery–treated metastasis was grouped into one of the following categories according to its change in volume relative to baseline: (a) good response, indicating a greater than 50% volume reduction; (b) partial response, indicating a 25%–50% volume reduction; (c) nonsubstantial response, indicating a less than 25% volume reduction or up to 25% volume increase; and (d) lesion growth, indicating a 25% or greater volume increase. Tumors with local control—that is, a volume increase of less than 25%—were those tumors with good, partial, or nonsubstantial responses.

Calculations of initial tumor control rates were performed in metastastic lesions imaged within 20 weeks after stereotactic radiosurgery (n = 73). Long-term local control rates were calculated from additional follow-up imaging performed in 50 of these lesions.

Actuarial survival curves were calculated by using the method of Kaplan and Meier (15). Univariate comparisons of survival and local control between different groups were performed by using the log-rank test (16). Multivariate analyses were performed by using the proportional hazards model (17).


   RESULTS
TOP
 Abstract
 Introduction
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 References
 
The range of follow-up intervals for all patients was 4–125 weeks (median follow-up, 30 weeks) after stereotactic radiosurgery. The metastases' locations are listed in Table 2.


View this table:
[in this window]
[in a new window]
  		TABLE 2. Metastasis Locations
 
The average baseline lesion volume (ie, sum of all lesion volumes divided by number of lesions) was 2,775 mm3 (range, 4–12,244 mm3). The enhancement pattern was highly correlated with the baseline lesion volume (P < .001). The average baseline volume of homogeneously enhancing metastases was 926 mm3, whereas the rim-enhancing and patchy lesions tended to be larger, with average volumes of 4,102 mm3 and 4,616 mm3, respectively (Table 3).


View this table:
[in this window]
[in a new window]
  		TABLE 3. Baseline Lesion Enhancement Patterns
 
Local tumor control was observed in 66 (90%) of the 73 metastases imaged within 20 weeks after stereotactic radiosurgery. Of the 73 lesions, 41 (56%) exhibited a good response, 15 (20%) exhibited a partial response, and 10 (14%) exhibited a nonsubstantial response. Five metastases were not imaged until after 20 weeks following stereotactic radiosurgery. All five of these lesions exhibited a good response to radiosurgery.

Local tumor control rates were also determined in the subgroup of 50 metastatic lesions in which both initial follow-up MR imaging (within 20 weeks after stereotactic radiosurgery) and long-term follow-up MR imaging (more than 20 weeks after stereotactic radiosurgery) were performed (Table 4). Local tumor control was observed in 46 (92%) of these 50 metastases at initial follow-up MR imaging. Twenty-eight (56%) of the 50 metastases either disappeared completely or exhibited a considerable volume reduction (ie, good response) within 20 weeks after stereotactic radiosurgery. Eleven (22%) of these treated tumors exhibited a partial response, seven (14%) exhibited a nonsubstantial response, and four (8%) increased in size after treatment (Table 4, Fig 2).


View this table:
[in this window]
[in a new window]
  		TABLE 4. Initial and Long-term Responses of 50 Metastases to Stereotactic Radiosurgery
 


View larger version (16K):
[in this window]
[in a new window]
  		Figure 2. Initial MR imaging response to stereotactic radiosurgery. Local control was achieved in 92% of lesions at initial imaging (ie, within 20 weeks after stereotactic radiosurgery). {block} = good response, {square} = partial response, = nonsubstantial response, = lesion growth.

 
Increased enhancing tumor volume within the stereotactic radiosurgery field (ie, lesion growth) was noted in 18 lesions (36%) (four imaged within 20 weeks after stereotactic radiosurgery and 14 imaged more than 20 weeks after stereotactic radiosurgery), for an approximate long-term local control rate of 64%. The 2-year tumor control rate was 61% ± 8.3 (Fig 3).



View larger version (13K):
[in this window]
[in a new window]
  		Figure 3. Actuarial local control rates of 78 metastases in 48 patients after gamma knife stereotactic radiosurgery.

 
Long-term local control rates were calculated in the same group of 50 lesions in which subsequent MR imaging was performed beyond 20 weeks after stereotactic radiosurgery. Most of the lesions that demonstrated a good response at the initial (<=20 weeks) posttreatment MR examination (n = 28) maintained local control at subsequent MR examinations (n = 24) (P < .002) (Figs 4, 5). Furthermore, lesions with an initial good response not only maintained local control, but they also maintained a good response (n = 19) at later MR examinations—that is, they remained stable in size or further decreased in size. Six (54%) of the 11 metastatic lesions with an initial partial response and four (57%) of the seven lesions with an initial nonsubstantial response showed subsequent lesion growth (Table 4, Fig 6). Of the 16 patients (28 metastases) who underwent only one post–stereotactic radiosurgery follow-up MR imaging examination, three had subsequent lesion growth.



View larger version (14K):
[in this window]
[in a new window]
  		Figure 4. Correlation of initial response with actuarial tumor control. Good response (GR) (ie, >50% volume decrease relative to baseline) at the initial MR examination was strongly correlated with maintenance of long-term local control. IR = insignificant (ie, nonsubstantial) response, PR = partial response.

 


View larger version (128K):
[in this window]
[in a new window]
  		Figure 5a. Maintenance of local control of two treated metastases after an initial good response to radiosurgery. Axial contrast-enhanced T1-weighted (400/15) MR images of lung cancer metastatic to (a, c, and e) the left frontal lobe and (b, d, and f) the right parietal convexity. Note that the superior extent of the left frontal lobe metastasis (arrow) in a, c, and e is seen in b, d, and f (right arrow). (a, b) Baseline MR images obtained at consecutive levels demonstrate homogeneously enhancing masses (arrows). (c, d) Initial follow-up images at 5 weeks after stereotactic radiosurgery demonstrate an interval volume decrease (54% in the left lobe, 64% in the right lobe) (arrows). (e, f) Long-term follow-up images at 45 weeks show stable volume reduction in the left frontal lesion (right arrow in e and f) and further (82%) volume reduction in the right parietal lesion (left arrow in f).

 


View larger version (143K):
[in this window]
[in a new window]
  		Figure 5b. Maintenance of local control of two treated metastases after an initial good response to radiosurgery. Axial contrast-enhanced T1-weighted (400/15) MR images of lung cancer metastatic to (a, c, and e) the left frontal lobe and (b, d, and f) the right parietal convexity. Note that the superior extent of the left frontal lobe metastasis (arrow) in a, c, and e is seen in b, d, and f (right arrow). (a, b) Baseline MR images obtained at consecutive levels demonstrate homogeneously enhancing masses (arrows). (c, d) Initial follow-up images at 5 weeks after stereotactic radiosurgery demonstrate an interval volume decrease (54% in the left lobe, 64% in the right lobe) (arrows). (e, f) Long-term follow-up images at 45 weeks show stable volume reduction in the left frontal lesion (right arrow in e and f) and further (82%) volume reduction in the right parietal lesion (left arrow in f).

 


View larger version (145K):
[in this window]
[in a new window]
  		Figure 5c. Maintenance of local control of two treated metastases after an initial good response to radiosurgery. Axial contrast-enhanced T1-weighted (400/15) MR images of lung cancer metastatic to (a, c, and e) the left frontal lobe and (b, d, and f) the right parietal convexity. Note that the superior extent of the left frontal lobe metastasis (arrow) in a, c, and e is seen in b, d, and f (right arrow). (a, b) Baseline MR images obtained at consecutive levels demonstrate homogeneously enhancing masses (arrows). (c, d) Initial follow-up images at 5 weeks after stereotactic radiosurgery demonstrate an interval volume decrease (54% in the left lobe, 64% in the right lobe) (arrows). (e, f) Long-term follow-up images at 45 weeks show stable volume reduction in the left frontal lesion (right arrow in e and f) and further (82%) volume reduction in the right parietal lesion (left arrow in f).

 


View larger version (155K):
[in this window]
[in a new window]
  		Figure 5d. Maintenance of local control of two treated metastases after an initial good response to radiosurgery. Axial contrast-enhanced T1-weighted (400/15) MR images of lung cancer metastatic to (a, c, and e) the left frontal lobe and (b, d, and f) the right parietal convexity. Note that the superior extent of the left frontal lobe metastasis (arrow) in a, c, and e is seen in b, d, and f (right arrow). (a, b) Baseline MR images obtained at consecutive levels demonstrate homogeneously enhancing masses (arrows). (c, d) Initial follow-up images at 5 weeks after stereotactic radiosurgery demonstrate an interval volume decrease (54% in the left lobe, 64% in the right lobe) (arrows). (e, f) Long-term follow-up images at 45 weeks show stable volume reduction in the left frontal lesion (right arrow in e and f) and further (82%) volume reduction in the right parietal lesion (left arrow in f).

 


View larger version (139K):
[in this window]
[in a new window]
  		Figure 5e. Maintenance of local control of two treated metastases after an initial good response to radiosurgery. Axial contrast-enhanced T1-weighted (400/15) MR images of lung cancer metastatic to (a, c, and e) the left frontal lobe and (b, d, and f) the right parietal convexity. Note that the superior extent of the left frontal lobe metastasis (arrow) in a, c, and e is seen in b, d, and f (right arrow). (a, b) Baseline MR images obtained at consecutive levels demonstrate homogeneously enhancing masses (arrows). (c, d) Initial follow-up images at 5 weeks after stereotactic radiosurgery demonstrate an interval volume decrease (54% in the left lobe, 64% in the right lobe) (arrows). (e, f) Long-term follow-up images at 45 weeks show stable volume reduction in the left frontal lesion (right arrow in e and f) and further (82%) volume reduction in the right parietal lesion (left arrow in f).

 


View larger version (147K):
[in this window]
[in a new window]
  		Figure 5f. Maintenance of local control of two treated metastases after an initial good response to radiosurgery. Axial contrast-enhanced T1-weighted (400/15) MR images of lung cancer metastatic to (a, c, and e) the left frontal lobe and (b, d, and f) the right parietal convexity. Note that the superior extent of the left frontal lobe metastasis (arrow) in a, c, and e is seen in b, d, and f (right arrow). (a, b) Baseline MR images obtained at consecutive levels demonstrate homogeneously enhancing masses (arrows). (c, d) Initial follow-up images at 5 weeks after stereotactic radiosurgery demonstrate an interval volume decrease (54% in the left lobe, 64% in the right lobe) (arrows). (e, f) Long-term follow-up images at 45 weeks show stable volume reduction in the left frontal lesion (right arrow in e and f) and further (82%) volume reduction in the right parietal lesion (left arrow in f).

 


View larger version (173K):
[in this window]
[in a new window]
  		Figure 6a. Coronal contrast-enhanced T1-weighted (400/11) MR images of lung carcinoma metastatic to the left parietal convexity show local tumor recurrence after an initial partial response to radiosurgery. (a) Baseline image demonstrates a patchy enhancing lesion (arrow). The cursor indicates the center of the stereotactic head frame. (b) Initial follow-up image at 20 weeks after treatment demonstrates an interval 37% volume decrease (arrow). (c) Follow-up image at 44 weeks after treatment shows a 163% lesion volume increase (arrow).

 


View larger version (172K):
[in this window]
[in a new window]
  		Figure 6b. Coronal contrast-enhanced T1-weighted (400/11) MR images of lung carcinoma metastatic to the left parietal convexity show local tumor recurrence after an initial partial response to radiosurgery. (a) Baseline image demonstrates a patchy enhancing lesion (arrow). The cursor indicates the center of the stereotactic head frame. (b) Initial follow-up image at 20 weeks after treatment demonstrates an interval 37% volume decrease (arrow). (c) Follow-up image at 44 weeks after treatment shows a 163% lesion volume increase (arrow).

 


View larger version (168K):
[in this window]
[in a new window]
  		Figure 6c. Coronal contrast-enhanced T1-weighted (400/11) MR images of lung carcinoma metastatic to the left parietal convexity show local tumor recurrence after an initial partial response to radiosurgery. (a) Baseline image demonstrates a patchy enhancing lesion (arrow). The cursor indicates the center of the stereotactic head frame. (b) Initial follow-up image at 20 weeks after treatment demonstrates an interval 37% volume decrease (arrow). (c) Follow-up image at 44 weeks after treatment shows a 163% lesion volume increase (arrow).

 
By using univariate analysis, local control was found to be significantly more often achieved in lesions that demonstrated baseline homogeneous enhancement prior to stereotactic radiosurgery than in those with heterogeneous (ie, rim-enhancing and patchy) enhancement (P < .03) (Fig 7). In the multivariate model, the difference in local control rates between homogeneously and heterogeneously enhancing lesions was not statistically significant (P = .18). Multivariate analysis of baseline lesion volume and primary tumor type demonstrated no correlation with local tumor control (Table 5).



View larger version (17K):
[in this window]
[in a new window]
  		Figure 7. Effective enhancement pattern correlated with actuarial local tumor control. The baseline homogeneous enhancement pattern correlated with local tumor control.

 

View this table:
[in this window]
[in a new window]
  		TABLE 5. Multivariate Analysis of Potential Prognostic Factors Affecting Local Tumor Control
 
The mass effect associated with stereotactic radiosurgery–treated lesions progressively decreased with each follow-up interval. Thirty-two percent (n = 25) of 78 lesions had no mass effect at baseline, whereas this fraction grew to 62% (31 of 50 lesions) at 20 or more weeks after stereotactic radiosurgery. The percentages of metastases that demonstrated mild and moderate mass effect decreased slightly with each follow-up interval (47% [37 of 78 lesions] and 18% [14 of 78 lesions], respectively, at baseline to 30% [15 of 50 lesions] and 8% [four of 50 lesions], respectively, at 20 or more weeks after stereotactic radiosurgery). Severe mass effect was present at baseline in two lesions and resolved in both cases during the first follow-up interval after stereotactic radiosurgery.

The enhancement patterns of the metastases changed after stereotactic radiosurgery. The percentage of homogeneous lesions decreased after stereotactic radiosurgery, whereas the percentage of heterogeneous (ie, rim-enhancing and patchy) lesions increased over time (Fig 8).



View larger version (14K):
[in this window]
[in a new window]
  		Figure 8. Lesion enhancement pattern versus time following stereotactic radiosurgery. F/U = follow-up, ------ = heterogeneous enhancement pattern, —•— = homogeneous enhancement pattern.

 
Five of the 50 metastases (in five patients) in which both initial and long-term follow-up MR imaging examinations were performed in our series demonstrated a transient increase in volume before shrinking (Fig 9). Four of these five lesions reached sizes greater than 25% above the baseline volume. None of these lesions was erroneously recorded as having lesion growth. All five lesions subsequently decreased in size; three decreased either to or below the baseline volume. An associated transient increase in T2-weighted signal intensity changes (compatible with vasogenic edema) and mass effect accompanied the transient volume increases. There was no correlation between transient volume increase and lesion histology (ie, primary cancer type), location, or size. The primary cancer types from which the metastases with transient volume increases originated included lung carcinoma (n = 1), renal cell carcinoma (n = 1), melanoma (n = 2), and thyroid carcinoma (n = 1). The interval between stereotactic radiosurgery and resolution of transient lesion growth ranged from 22 to 59 weeks, with the peak in lesion growth occurring between 4 and 24 weeks.



View larger version (170K):
[in this window]
[in a new window]
  		Figure 9a. Radiation effect. Coronal enhanced T1-weighted (400/15) MR images of renal cell carcinoma metastatic to the frontal lobe. (a) Baseline image demonstrates a homogeneously enhancing lesion (arrowhead). (b) Initial follow-up image at 8 weeks after treatment demonstrates an interval 156% volume increase (arrowhead). Note the interval increased mass effect with displacement and partial effacement of the right frontal horn (arrow). (c) Follow-up image at 40 weeks after treatment shows decrease in enhancing lesion volume (arrowhead) and mass effect on the right frontal horn (arrow). (d) Follow-up image at 60 weeks after treatment shows further decrease in enhancing lesion volume (arrowhead). The mass effect has resolved. The volume loss is now evident, as manifested by the mild dilatation of the right frontal horn (arrow).

 


View larger version (173K):
[in this window]
[in a new window]
  		Figure 9b. Radiation effect. Coronal enhanced T1-weighted (400/15) MR images of renal cell carcinoma metastatic to the frontal lobe. (a) Baseline image demonstrates a homogeneously enhancing lesion (arrowhead). (b) Initial follow-up image at 8 weeks after treatment demonstrates an interval 156% volume increase (arrowhead). Note the interval increased mass effect with displacement and partial effacement of the right frontal horn (arrow). (c) Follow-up image at 40 weeks after treatment shows decrease in enhancing lesion volume (arrowhead) and mass effect on the right frontal horn (arrow). (d) Follow-up image at 60 weeks after treatment shows further decrease in enhancing lesion volume (arrowhead). The mass effect has resolved. The volume loss is now evident, as manifested by the mild dilatation of the right frontal horn (arrow).

 


View larger version (160K):
[in this window]
[in a new window]
  		Figure 9c. Radiation effect. Coronal enhanced T1-weighted (400/15) MR images of renal cell carcinoma metastatic to the frontal lobe. (a) Baseline image demonstrates a homogeneously enhancing lesion (arrowhead). (b) Initial follow-up image at 8 weeks after treatment demonstrates an interval 156% volume increase (arrowhead). Note the interval increased mass effect with displacement and partial effacement of the right frontal horn (arrow). (c) Follow-up image at 40 weeks after treatment shows decrease in enhancing lesion volume (arrowhead) and mass effect on the right frontal horn (arrow). (d) Follow-up image at 60 weeks after treatment shows further decrease in enhancing lesion volume (arrowhead). The mass effect has resolved. The volume loss is now evident, as manifested by the mild dilatation of the right frontal horn (arrow).

 


View larger version (172K):
[in this window]
[in a new window]
  		Figure 9d. Radiation effect. Coronal enhanced T1-weighted (400/15) MR images of renal cell carcinoma metastatic to the frontal lobe. (a) Baseline image demonstrates a homogeneously enhancing lesion (arrowhead). (b) Initial follow-up image at 8 weeks after treatment demonstrates an interval 156% volume increase (arrowhead). Note the interval increased mass effect with displacement and partial effacement of the right frontal horn (arrow). (c) Follow-up image at 40 weeks after treatment shows decrease in enhancing lesion volume (arrowhead) and mass effect on the right frontal horn (arrow). (d) Follow-up image at 60 weeks after treatment shows further decrease in enhancing lesion volume (arrowhead). The mass effect has resolved. The volume loss is now evident, as manifested by the mild dilatation of the right frontal horn (arrow).

 
The overall median survival time (± SD) after stereotactic radiosurgery was 53 weeks (± 7.7) (95% CI, 38 weeks, 67 weeks). The actuarial survival curve is shown in Figure 10. Although the initial response of stereotactic radiosurgery–treated metastases was predictive of maintenance of local tumor control, a marked influence on survival was not observed. Primary cancer type and presence of systemic disease were the main factors that affected survival (Table 6). The median survival for patients with breast carcinoma (n = 5) was not reached; three (60%) of these patients were alive at the conclusion of the study. The median survival time for patients with melanoma and renal cell carcinoma (n = 16) was 15.7 weeks. The median survival time for patients with other carcinomas (n = 27) was 23.6 weeks. The baseline (ie, pretreatment) tumor volume, baseline enhancement pattern, patient age, and multiplicity of brain metastases had no effect on survival.



View larger version (14K):
[in this window]
[in a new window]
  		Figure 10. Actuarial survival calculated from the time of stereotactic radiosurgery.

 

View this table:
[in this window]
[in a new window]
  		TABLE 6. Multivariate Analysis of Potential Prognostic Factors Affecting Patient Survival
 

   DISCUSSION
TOP
 Abstract
 Introduction
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
References
 
The local tumor control rate in over 250 brain metastases treated with radiosurgery, which ranged from 82% to 96%, has been the focus of several retrospective series in patients with solitary metastases (9,10,12–14) and the subject of prior review (7,11). In two large multiinstitutional trials (6,8), with 116 and 122 patients, 2-year actuarial tumor control rates of 67% and 77%, respectively, were reported for solitary metastases treated by using stereotactic radiosurgery. However, these reports did not detail the relationship between baseline MR imaging features and initial imaging responses and the persistence of local tumor control. Identification of specific imaging features that are predictive of tumor control may be useful for patient treatment.

Although the overall local control rate of 90% was initially achieved within 20 weeks of stereotactic radiosurgery, further sequential analysis of lesion volumes revealed a continuation of response (either a continued volume decrease or stabilization of volume size) for several months after stereotactic radiosurgery, as would be expected owing to the slow removal of tumor cells. Volumetric analysis also provided insight into the relationship between initial rate of tumor response and maintenance of tumor control. Good response at the initial postradiosurgical MR examination was strongly correlated with maintenance of local control (ie, lesions with good response, partial response, or nonsubstantial response) on subsequent postradiosurgical MR studies.

Smaller lesions tended to enhance homogeneously, probably reflecting a lack of intratumoral necrosis prior to treatment. In our series, the baseline homogeneously enhancing appearance of lesions was associated with significantly better local control compared with the baseline heterogeneously enhancing appearance (P < .03). However, initial (pretreatment) size of lesions alone did not correlate with local control. The larger, more heterogeneously enhancing lesions may have outgrown their blood supply and become centrally necrotic and thus more hypoxic than their homogeneous counterparts. We hypothesize that the homogeneously enhancing lesions responded better to treatment primarily because uniform contrast enhancement reflects uniform oxygen distribution throughout the mass, and this, in turn, promotes a uniform response to radiation treatment (18–20).

The extent of mass effect produced by the metastatic lesions decreased over time after stereotactic radiosurgery. This finding was not unexpected, given the high local control rates that were achieved with stereotactic radiosurgery. Because the lesion-associated T2-weighted signal intensity abnormalities probably reflected a combination of peritumoral edema and changes due to whole-brain radiation therapy, measurements of T2-weighted changes were not obtained.

The percentage of lesions that were initially homogeneous decreased over time, but the percentage of heterogeneous lesions increased over time, because the initially homogeneous lesions tended to become heterogeneous after stereotactic radiosurgery. This finding was probably due to radiation-induced necrosis in the initially homogeneous tumors (21,22).

The mechanism for the relatively uncommon stereotactic radiosurgery–induced transient increase in tumor volume (ie, radiation effect) is unknown. Transient imaging changes following both whole-brain radiation therapy and stereotactic radiosurgery have been previously described in patients treated for intracranial neoplasms and arteriovenous malformations (23–25). Graeb et al (26) described three cases in which whole-brain radiation therapy was followed by CT changes that initially were suggestive of tumor progression (increased enhancement and edema formation) but regressed 3–6 months later. Bakardjiev et al (27) identified transient treatment-related MR imaging changes that mimicked tumor progression in 43% of pediatric patients who were treated with stereotactic radiosurgery for low-grade astrocytoma. Perhaps a single large dose of radiation results in disruption of the blood-brain barrier, with resultant increased capillary permeability, vasogenic edema, and abnormal contrast enhancement, which are mechanisms that have been attributed to late radiation injury following conventional whole-brain radiation therapy (21, 28,29). Similarly, defects in the blood-brain barrier that result from a single large stereotactic fraction of radiation have been reported in experimental studies in cats (30) and baboons (31) treated with stereotactic radiosurgery.

Although the exact mechanism for transient increases in enhancing volume following radiosurgery is unknown, the importance of this finding is that such changes can be distinguished from tumor progression by using follow-up MR imaging. On the basis of our data, lesions that exhibited transient growth achieved the maximum volume at 4–24 weeks after stereotactic radiosurgery. Subsequent volume reduction was first observed between 22 and 59 weeks after treatment. This suggests that prolonged MR imaging follow-up may be necessary to differentiate between true tumor growth and radiation effect.

In our series, the overall median survival time after stereotactic radiosurgery was 53 weeks, similar to that in published reports (6–8,13,32), in which the median survival time after stereotactic radiosurgery ranged from 44 to 56 weeks. In previous series (6,8,32), as in the current study, the primary factor that affected survival was the presence of active extracranial systemic disease. The primary tumor type influenced survival; patients with breast carcinoma had the longest survival times after stereotactic radiosurgery, and those with melanoma and renal cell carcinoma had the shortest (6,8). Baseline lesion volume, initial response rate, and baseline enhancement pattern did not correlate with survival.

Although the primary tumor type influenced survival, it did not appear to affect the local control rate. The results of previous reports (13,14,32–35) have indicated that stereotactic radiosurgery is effective in treating brain metastases, regardless of their histology, including those that are resistant to conventional whole-brain radiation therapy, such as those that originate from melanoma and renal cell carcinoma.

Stereotactic radiosurgery has become a standard procedure for the treatment of patients with brain metastases. Relative to whole-brain radiation therapy, stereotactic radiosurgery results in longer survival and higher local control rates. Compared with resection, radiosurgery is associated with lower morbidity and decreased cost (36). Knowledge of specific imaging features that are predictive of tumor control may be useful for patient treatment. In this study, a homogeneous baseline enhancement pattern correlated with local control. Good response at the initial posttreatment MR examination was strongly correlated with local control (ie, good response, partial response, or nonsubstantial response) at subsequent posttreatment MR imaging. On the basis of our observation that a small percentage of lesions may undergo a transient volume increase, we hypothesize that initial lesion growth does not necessarily preclude local control. Conversely, most metastases that exhibited initial lesion growth continued to grow and did not demonstrate a radiation effect. In summary, although an initial good response after radiosurgery is predictive of long-term local control, lesion growth does not always indicate tumor progression. Recognition of serial MR imaging findings after stereotactic radiosurgery is important and will allow the practicing radiologist to accurately interpret the response in patients with stereotactic radiosurgery–treated brain metastases.


   Acknowledgments
 
We thank Karol Rosengarth for secretarial assistance.


   Footnotes
 
Author contributions: Guarantor of integrity of entire study, A.M.P.; study concepts, A.M.P., C.C.M., D.K.; study design, A.M.P., C.C.M.; definition of intellectual content, A.M.P.; literature research, A.M.P.; clinical and experimental studies, A.M.P.; data acquisition, A.M.P., C.C.M., E.J.E.; data analysis, A.M.P., C.C.M., J.C.F.; statistical analysis, J.C.F.; manuscript preparation, A.M.P.; manuscript editing, A.M.P., C.C.M., D.K.; manuscript review, A.M.P., C.C.M.


   References
TOP
 Abstract
 Introduction
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Flickinger J, Kondziolka D. Radiosurgery instead of resection for solitary brain metastases: the gold standard redefined. Int J Radiat Oncol Biol Phys 1996; 35:185-186.[Medline]
  2. Leksell L. The stereotaxic method and radiosurgery of the brain. Acta Chir Scand 1951; 102:316-319.[Medline]
  3. Leksell D. Stereotactic radiosurgery. Neurol Res 1987; 9:60-68.[Medline]
  4. Loeffler JS, Alexander EI, Wen PY, Fine HA, Kooy HM, Black PM. Radiosurgery for brain metastases: five-year experience at the Brigham and Women's Hospital. In: Lunsford L, eds. Stereotactic update. New York, NY: Elsevier, 1992; 383-392.
  5. Morantz RA, Wara WM. Gamma knife radiosurgery in the treatment of brain tumors. Cancer Control 1995; :300-308.
  6. Auchter R, Lamond J, Alexander E, et al. A multiinstitutional outcome and prognostic factor analysis of radiosurgery for resectable single brain metastases. Int J Radiat Oncol Biol Phys 1996; 35:27-35.[Medline]
  7. Loeffler JS, Larson DA, Shrieve DC, Flickinger JC. Radiosurgery for the treatment of intracranial lesions. Important Adv Oncol 1995; :141-156.
  8. Flickinger JC, Kondziolka D, Lunsford LD, et al. A multi-institutional experience with stereotactic radiosurgery for solitary brain metastases. Int J Radiat Oncol Biol Phys 1994; 28:797-802.[Medline]
  9. Engenhart R, Romahn J, Gademann G, et al. Indications for radiosurgery in treatment of brain metastases. In: Lunsford L, eds. Stereotactic radiosurgery update. New York, NY: Elsevier, 1992; 393-398.
  10. Kihlstrom L, Karlsson B, Lindquist C. Gamma knife surgery for cerebral metastases: implications for survival based on 16 years experience. Stereotact Funct Neurosurg 1993; 61:45-50.
  11. Mehta M, Rozental J, Levin A, et al. Defining the role of radiosurgery in the management of brain metastases. Int J Radiat Oncol Biol Phys 1992; 24:619-625.[Medline]
  12. Loeffler J, Alexander EI. Radiosurgery in the definitive management of CNS malignancies. In: Meyer J, Vaeth J, eds. Organ conservation in curative cancer treatment. Frontiers in Radiation Therapy Oncology. Vol 27. New York, NY: Karger, 1993; 227-244.
  13. Coffey RJ, Flickinger JC, Bissonette DJ, Lunsford LD. Radiosurgery for solitary brain metastases using the cobalt-60 gamma unit: methods and results in 24 patients. Int J Radiat Oncol Biol Phys 1991; 20:1287-1295.[Medline]
  14. Adler J, Cox R, Kaplan I, Martin D. Stereotactic radiosurgical treatment of brain metastases. J Neurosurg 1992; 76:444-449.[Medline]
  15. Kaplan ES, Meier P. Nonparametric estimation from incomplete observation. J Am Stat Assoc 1958; 53:457-480.
  16. Peto R, Pike MC, Armitage P, et al. Design and analysis of randomized clinical trials requiring prolonged observation of each patient: analysis and examples. Br J Cancer 1977; 35:1-29.[Medline]
  17. Cox DR. Regression models and life tables. J Royal Stat Soc 1982; 34:187-220.
  18. Hall E. The oxygen effect and reoxygenation. In: Hall E, eds. Radiobiology for the radiologist. 4th ed. Philadelphia, Pa: Lippincott, 1994; 134-152.
  19. Valk E, Mathis C, Prados M, Gilbert J, Budinger T. Hypoxia in human gliomas: demonstration by PET with flourine-18-fluoromisonidazole. J Nucl Med 1992; 33:2133-2137.[Abstract/Free Full Text]
  20. Koh WJ, Rasey JS, Evans ML, et al. Imaging of hypoxia in human tumors with [F-18] fluoromisonidazole. Int J Radiat Oncol Biol Phys 1992; 22:199-212.[Medline]
  21. Valk P, Dillon W. Radiation injury of the brain. Am J Neuroradiol 1991; 12:45-62.[Abstract]
  22. Sheline G, Wara W, Smith V. Therapeutic irradiation and brain injury. Int J Radiat Oncol Biol Phys 1980; 6:1215-1228.[Medline]
  23. Morikawa M, Numaguchi Y, Rigamonti D, et al. Radiosurgery for cerebral arteriovenous malformations: assessment of early phase magnetic resonance imaging and significance of gadolinium-DTPA enhancement. Int J Radiat Oncol Biol Phys 1996; 34:663-675.[Medline]
  24. Guo W, Lindquist C, Karlsson B, Kihlstrom L, Steiner L. Gamma knife surgery of cerebral arteriovenous malformations. Int J Radiat Oncol Biol Phys 1993; 25:315-323.[Medline]
  25. Loeffler JS, Alexander EI. Stereotactic radiosurgery. In: Alexander E, Loeffler JS, Lunsford D, eds. Radiosurgery for the treatment of intracranial metastases. New York, NY: McGraw Hill, 1993; 197-206.
  26. Graeb D, Steinbok P, Robertson W. Transient early computed tomographic changes mimicking tumor progression after brain tumor irradiation. Radiology 1982; 144:813-817.[Abstract/Free Full Text]
  27. Bakardjiev AI, Barnes PD, Goumnerova LC, et al. Magnetic resonance imaging changes after stereotactic radiation therapy for childhood low grade astrocytoma. Cancer 1996; 78:864-873.[Medline]
  28. Marks J, Baglan R, Prassad S, Blank W. Cerebral radionecrosis: incidence and risk in relation to dose, time, fractionation and volume. Int J Radiat Oncol Biol Phys 1981; 7:243-252.[Medline]
  29. Curnes J, Laster D, Ball M, Moody D, Witcofski R. MRI of radiation injury to the brain. Am J Neuroradiol 1986; 147:119-124.
  30. Buatti JM, Friedman WA, Theele DP, Bova FJ, Mendenhall WM. The lazaroid U74389g protects normal brain from stereotactic radiosurgery–induced radiation injury. Int J Radiat Oncol Biol Phys 1996; 34:591-597.[Medline]
  31. Lunsford LD, Altschuler EM, Flickinger JC, Wu A, Martinez AJ. In vivo biological effects of stereotactic radiosurgery: a primate model. Neurosurgery 1990; 27:373-382.[Medline]
  32. Alexander EI, Moriarty T, Davis R, et al. Stereotactic radiosurgery for the definitive noninvasive treatment of brain metastases. J Natl Cancer Inst 1995; 87:34-40.[Abstract/Free Full Text]
  33. Caron JL, Souhami L, Podgorsak EB. Dynamic stereotactic radiosurgery in the palliative treatment of cerebral metastatic tumors. J Neurooncol 1992; 12:173-179.[Medline]
  34. Engenhart R, Kimmig B, Hover K, et al. Long-term follow-up for brain metastases treated by percutaneous stereotactic single high-dose irradiation. Cancer 1993; 71:1353-1361.[Medline]
  35. Kihlstrom L, Karlsson B, Lindquist C, Norén G, Rahn T. Gamma knife surgery for brain metastases. Acta Neurochir (Wien) 1991; 52:87-89.
  36. Rutigliano M, Lunsford L, Kondziolka D, Strauss M, Khanna V, Green M. The cost effectiveness of stereotactic radiosurgery versus surgical resection in the treatment of solitary metastatic brain tumors. Neurosurgery 1995; 37:453-455.



This article has been cited by other articles:


Home page
 RadioGraphicsHome page
M. B. Fukui, T. M. Blodgett, C. H. Snyderman, J. J. Johnson, E. N. Myers, D. W. Townsend, and C. C. Meltzer
Combined PET-CT in the Head and Neck: Part 2. Diagnostic Uses and Pitfalls of Oncologic Imaging
RadioGraphics, July 1, 2005; 25(4): 913 - 930.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
M. Essig, M. Waschkies, F. Wenz, J. Debus, H. R. Hentrich, and M. V. Knopp
Assessment of Brain Metastases with Dynamic Susceptibility-weighted Contrast-enhanced MR Imaging: Initial Results
Radiology, July 1, 2003; 228(1): 193 - 199.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Roentgenol.Home page
G. A. Tung, G. Noren, J. M. Rogg, and I. M.D. Jackson
MR Imaging of Pituitary Adenomas After Gamma Knife Stereotactic Radiosurgery
Am. J. Roentgenol., October 1, 2001; 177(4): 919 - 924.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Peterson, A. M.
Right arrow Articles by Kondziolka, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Peterson, A. M.
Right arrow Articles by Kondziolka, D.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE