Abstract
Objectives
Patients with a variety of different primary cancers can develop intracranial metastases. Patients who develop intracranial metastases are often grouped into the same study population, and therefore an understanding of outcomes for patients with different primary cancers remain unclear.
Methods
Adults who underwent intracranial metastatic tumor surgery from 1997–2011 at a single institution were retrospectively reviewed. Primary pathologies were compared using Fisher’s exact and Student’s t-test, and Cox regression analysis was used to identify factors associated with survival.
Results
About 708 patients underwent surgery during the reviewed period, where 269 (38%) had non-small cell lung cancer (NSCLC), 106 (15%) breast cancer (BC), 72 (10%) gastrointestinal (GI) cancers, 88 (12%) renal cell cancer (RCC), and 88 (12%) melanoma. The most notable differences were that NSCLC patients were older, BC younger, BC had more primary tumor control, and NSCLC less extracranial spread. BC had longer survival, RCC had longer local progression free survival (PFS), and NSCLC had longer distal PFS. The factors independently associated with survival for NSCLC (female, recursive partitioning analysis (RPA) class, primary tumor control, solitary metastasis, tumor size, adenocarcinoma, radiation, discharge to home), BC (age, no skull base involvement, radiation), GI cancer (age, RPA class, Karnofsky performance scale (KPS), lack of preoperative motor deficit, non-esophageal tumors, non-hemorrhagic tumors, avoidance of new deficits), melanoma (preoperative seizures, solitary metastasis, smaller tumor size, discharge to home, chemotherapy), and RCC (KPS, chemotherapy) were distinctly different.
Discussion
These differences between patients with different primary cancers support the fact that patients with intracranial disease are not all the same and should be studied by their primary pathology.
Keywords: Breast, Gastrointestinal, Lung, Melanoma, Metastatic brain tumor, Recurrence, Renal, Surgery, Survival
Introduction
Metastatic brain cancer is the most common type of intracranial tumor in adults.1,2 It is estimated that there will be 100 000 to 180 000 new cases diagnosed in the United States each year.1,2 The most common primary sources include lung, breast, melanoma, renal cell (RCC), and gastrointestinal (GI) cancers.1,2 These primary cancers have varying degrees of sensitivity to radiation, dissimilar systemic disease burden, different chemotherapeutic regimens, and, most importantly, disparate prognoses.1,2 Despite these dissimilarities, these histologically different tumors are grouped into the same study population when they metastasize to the brain and require brain surgery.3–9 This grouping may be potentially erroneous, and hinder progress in understanding and developing optimal treatment for patients with different primary cancers.
Studies on metastatic brain cancer separated by primary pathology are few and limited.10–13 This limitation has made it difficult to understand survival prognoses and recurrence profiles for patients with different primary cancers. A better understanding may translate into more optimal treatment regimens, once these cancers metastasize to the brain, guided by primary pathology and not metastatic site. As a result, the goals of this study were to: (1) understand differences in pre, peri, and postoperative characteristics between patients with different primary cancers that metastasize to the brain, and (2) ascertain factors independently associated with survival for each of these primary pathologies.
Methods
Patient selection
Institutional Review Board approval was first obtained prior to initiating this study (36875). All adult patients (age > 18 years) who underwent intracranial surgery (resection or biopsy) for an intracranial metastasis (single or multiple) at a single, tertiary care institution between 1997 and 2011 were tabulated and reviewed. The pathology was determined by a senior neuro-pathologist in all cases.
Recorded variables
Seven hundred and eight patients underwent surgery for a single or multiple intracranial metastases during the reviewed period. The clinical, operative, and hospital course records of these patients were retrospectively reviewed. The information collected from clinical notes included patient demographics, comorbidities, presenting symptoms, neurological function, radiographic imaging findings, and use of adjuvant therapy. Patients were assigned a preoperative Karnofsky performance scale (KPS) by a reviewer blinded to patient outcomes at the clinical visit prior to surgery during a chart review.14 The presence of a motor deficit was defined as any decrease in strength, and a language deficit as any combination of receptive and/or expressive aphasia. A cognitive deficit was any complaint of decreased mental status or ability, and a vision deficit was defined as any decrease in visual acuity or field perception. Patients were also assigned a recursive partitioning analysis (RPA) classification group as previously defined.15
All MRI images were obtained and reviewed. The characteristics that were recorded included the lesion’s size (largest diameter based on gadolinium enhancement), specific lobe involvement, and number of intracranial metastases. Extent of resection was classified as gross-total resection (GTR), near total resection (NTR), and sub-total resection (STR) if there was no enhancement, a rim of enhancement, and residual nodular enhancement in the resection cavity on postoperative MRI, respectively.16 In addition to brain MRIs, patients generally underwent computed tomography scans of the chest, abdomen, pelvis, and spine with and without contrast or positron emission tomography scans before surgery to identify metastatic body sites.
Outcome definitions
The primary outcomes that were evaluated for each primary cancer were overall survival, local recurrence, and distal recurrence. Survival data was obtained from the social security index database. Time to death was defined as the time from cranial surgery to death, where patients whose deaths were unconfirmed were classified as lost to follow-up at the time of their last clinic visit. Local recurrence was defined as the recurrence or progression of tumor in the previous surgical cavity. Distal recurrence was defined as presence of new tumor in the brain not adjacent to the previous surgical cavity. Time to local or distal recurrence was defined as the time from surgery to MRI diagnosis of recurrence. Patients whose recurrence was unconfirmed were classified as lost to follow-up at the time of their last neuro-imaging.
General treatment strategy
In general, surgical resection was pursued when patients presented with intracranial metastases causing symptoms or at risk of causing symptoms either from location or swelling. Patients typically had at least 3 months of expected survival. Biopsy was pursued when only diagnosis was desired and/or for patients who presented with poor survival prognoses. The general goal of surgical resection was to achieve GTR of the tumor without causing new iatrogenic deficits. Surgery was pursued for multiple metastases when the metastases were easily accessible and/or causing symptoms.
Postoperative MRI with gadolinium was typically performed at 3-month intervals following surgery, or when new or exacerbating symptoms occurred. The use of adjuvant therapy, including whole brain radiation therapy (WBRT), stereotactic radiosurgery (SRS), and chemotherapy, were determined by a multi-disciplinary team (neurosurgery, neuro-oncology, medical oncology, and radiation oncology) as well as the patients and their families.
Statistical analysis
Patients were grouped by their histology, where each group consisted of > 50 patients. This number was assigned a priori in order to have sufficient numbers to perform multivariate analyses. Groups with less than fifty patients were excluded from the analyses.
Summary data were presented as mean±standard deviation for parametric data and as median [inter-quartile range (IQR)] for non-parametric data. The student’s t-test and Fisher’s exact test were used to make inter-group comparison for continuous and categorical data, respectively. Stepwise multivariate proportional hazards regression analyses were performed for each type of primary cancer to identify factors independently associated with survival, local recurrence, and distal recurrence. In these analyses, univariate analysis was first performed to identify potential associations between radiographic, preoperative, operative, perioperative, and pathologic variables with survival. All variables associated with survival in univariate analysis (P < 0.10) were then included in a step-wise multivariate proportional hazards regression model. For continuous variables, the variables were dichotomized to identify the greatest statistical association. Overall survival, local progression free survival (PFS), and distal PFS as a function of time were plotted using the Kaplan–Meier method. Log-rank analyses were used to compare Kaplan–Meier plots between each primary cancer (GraphPad Prism 5, La Jolla, CA, USA). Values with P < 0.05 in these analyses were considered statistically significant. JMP 9 (SAS, Cary, NC, USA) was used unless otherwise specified.
Results
Patient population and histological groups
A total of 708 patients underwent surgery for a single or multiple intracranial metastases during the reviewed period (Table 1). Of these 708 patients, 269 (38%) had non-small cell lung cancer (NSCLC), 106 (15%) had breast cancer (BC), 72 (10%) had GI cancers, 88 (12%) had RCC, and 122 (17%) had other primary cancers. Among NSCLC patients, 214 (80%) had documented pathological data, where 163 (76%) had adenocarcinoma and 51 (24%) had squamous cell carcinoma. Among GI cancer patients, 35 (49%) had colon cancer, 11 (15%) had pancreatic/bile duct cancer, 8 (11%) had esophageal cancers, 8 (11%) had stomach cancer, 4 (6%) had liver cancer, and 6 (8%) had salivary cancers. Among the 122 patients with other primary cancers, 34 (28%) had small cell lung cancer, 17 (14%) had primary bone cancers, 12 (10%) had bladder cancer, 10 (8%) had ovarian/testicular cancer, 9 (7%) had hematopoietic cancers, 10 (8%) had thyroid cancer, 9 (7%) had endometrial cancer, 8 (7%) had prostate cancer, 10 (8%) had unknown primary cancer, and 3 (2%) had vaginal/cervical cancer. These patients with other cancers were excluded from further analyses by histology.
Table 1.
Preoperative characteristics of patients undergoing surgery of intracranial metastases separated by primary cancer (non-small cell lung–NSCLC, breast, gastrointestinal–GI, melanoma, renal cell–RCC). The bolded values are statistically different than one or more of the other histological groups (P < 0.05)
| Study Population (n = 708)
| ||||||
|---|---|---|---|---|---|---|
| Characteristics | NSCLC
|
Breast
|
GI
|
Melanoma
|
RCC
|
All
|
|
n = 269
|
n = 106
|
n = 72
|
n = 88
|
n = 51
|
n = 708
|
|
| Number (%) | Number (%) | Number (%) | Number (%) | Number (%) | Number (%) | |
| Demographics | ||||||
| Age* | 60.5±10 | 53.2±12 | 57.6±14 | 58.3±12 | 60.9±12 | 58.4±12 |
| Male | 149 (55%) | 1 (0.9%) | 38 (53%) | 55 (63%) | 35 (69%) | 336 (47%) |
| KPS** | 80 (70–80) | 80 (70–80) | 80 (70–90) | 80 (70–90) | 80 (70–80) | 80 (70–80) |
| Headaches | 103 (38%) | 55 (52%) | 28 (39%) | 34 (39%) | 16 (31%) | 293 (41%) |
| Seizures | 50 (19%) | 18 (17%) | 7 (10%) | 11 (13%) | 8 (16%) | 110 (16%) |
| Motor deficit | 100 (37%) | 39 (37%) | 33 (46%) | 16 (18%) | 20 (39%) | 254 (36%) |
| Language deficit | 45 (17%) | 11 (10%) | 12 (17%) | 17 (19%) | 8 (16%) | 111 (16%) |
| Cognitive deficit | 57 (21%) | 13 (12%) | 17 (24%) | 23 (26%) | 10 (20%) | 146 (21%) |
| Vision deficit | 36 (13%) | 25 (24%) | 14 (19%) | 17 (19%) | 9 (18%) | 120 (17%) |
| Tumor characteristics | ||||||
| Control of primary tumor | 196 (73%) | 85 (80%) | 50 (69%) | 58 (66%) | 41 (80%) | 502 (71%) |
| Extracranial spread | 77 (29%) | 66 (62%) | 45 (63%) | 46 (52%) | 35 (69%) | 324 (46%) |
| No. of body met sites** | 1 (1–2) | 1 (1–3) | 2 (1–3) | 2 (1–3) | 2 (1–3) | 1 (1–3) |
| RPA Class | ||||||
| RPA Class 1 | 70 (26%) | 22 (21%) | 13 (18%) | 17 (19%) | 6 (12%) | 152 (21%) |
| RPA Class 2 | 143 (53%) | 67 (63%) | 43 (60%) | 56 (64%) | 38 (75%) | 421 (59%) |
| RPA Class 3 | 56 (21%) | 17 (16%) | 16 (22%) | 15 (17%) | 7 (14%) | 135 (19%) |
| Solitary metastasis | 182 (68%) | 56 (53%) | 45 (63%) | 52 (59%) | 37 (73%) | 460 (65%) |
| Radiographics | ||||||
| Tumor size* | 3.0±1.4 | 3.3±1.3 | 3.4±1.2 | 2.9±1.3 | 2.7±1.1 | 3.2±1.5 |
| Location | ||||||
| Frontal lobe | 102 (38%) | 37 (35%) | 16 (22%) | 34 (39%) | 21 (41%) | 256 (36%) |
| Temporal lobe | 30 (11%) | 15 (14%) | 12 (17%) | 18 (20%) | 5 (10%) | 102 (14%) |
| Parietal lobe | 62 (23%) | 25 (24%) | 14 (19%) | 20 (23%) | 8 (16%) | 154 (22%) |
| Occipital lobe | 40 (15%) | 10 (9%) | 13 (18%) | 18 (20%) | 11 (22%) | 101 (14%) |
| Cerebellum | 55 (20%) | 26 (25%) | 20 (28%) | 8 (9%) | 10 (20%) | 140 (20%) |
| Brainstem | 2 (0.7%) | 3 (3%) | 2 (3%) | 2 (2%) | 0 (0%) | 12 (2%) |
| Skull base | 11 (4%) | 4 (4%) | 2 (3%) | 0 (0%) | 1 (2%) | 25 (4%) |
| Bone involvement | 12 (4%) | 5 (5%) | 5 (7%) | 2 (2%) | 2 (4%) | 41 (6%) |
| No. of brain mets** | 1 (1–2) | 1 (1–2) | 1 (1–2) | 1 (1–3) | 1 (1–1) | 1 (1–2) |
| Hemorrhagic | 37 (14%) | 10 (9%) | 13 (18%) | 38 (43%) | 19 (37%) | 138 (19%) |
mean±standard deviation,
median (interquartile range); KPS = Karnofsky performance score; RPA = recursive partitioning analysis
Preoperative characteristics and differences between histological groups
The preoperative characteristics are summarized in Table 1. The average±standard deviation age of the entire cohort was 58.4±12 years. The preoperative RPA classes of the patients were RPA 1 in 152 (21%), RPA 2 in 421 (59%), and RPA 3 in 135 (19%). Radiographically, 460 (65%) presented with solitary metastasis, and the median [IQR] number of intracranial metastases was 1 [1,2].
Among the histological groups, NSCLC patients were older than patients with BC (P = 0.0001) and GI cancer (P = 0.04). Patients with BC were younger than patients with GI (P = 0.02), melanoma (P = 0.005), and RCC (P = 0.0003). Male patients were less commonly seen with BC than other cancers (P = 0.0001). With regards to symptoms, headaches were more common in BC than NSCLC (P = 0.02) and RCC (P = 0.02). Patients with melanoma less commonly had motor deficits than patients with NSCLC (P = 0.001), BC (P = 0.006), and GI cancers (P = 0.0003). Likewise, cognitive deficits were more common in melanoma than BC (P = 0.02), and vision deficits in BC than NSCLC (P = 0.02).
Primary tumor control was more common in breast cancer patients than melanoma (P = 0.03). NSCLC patients less commonly had extracranial spread than patients with BC (P = 0.0001), GI (P = 0.0001), melanoma (P = 0.0001), and RCC (P = 0.0001). NSCLC patients had on average less metastatic sites than patients with BC (P = 0.0001), GI (P = 0.0001), melanoma (P = 0.0001), and RCC (P = 0.0001). With regards to RPA class, NSCLC were more commonly RPA Class 1 than RCC (P = 0.03). Solitary metastasis was less common in BC than patients with NSCLC (P = 0.009) and RCC (P = 0.02).
Radiographically, the average size of a melanoma metastasis was smaller than BC (P = 0.04) and GI cancer (P = 0.02). RCC metastases were also smaller than breast (P = 0.008) and GI cancer (P = 0.004). Patients with melanoma less commonly had cerebellar metastases than NSCLC (P = 0.01), breast (P = 0.007), and GI cancers (P = 0.003). BC had less intracranial metastases than patients with NSCLC (P = 0.01) and RCC (P = 0.05). Melanoma patients presented with more intracranial metastases than NSCLC (P = 0.03) and GI cancers (P = 0.04). Hemorrhagic tumors were more common in melanoma patients than NSCLC (P = 0.0001), BC (P = 0.0001), and GI cancers (P = 0.0007). RCC were more commonly hemorrhagic than NSCLC (P = 0.0002), breast (P = 0.0001), and GI cancers (P = 0.01).
Perioperative characteristics and differences between histological groups
The perioperative outcomes are summarized in Table 2. GTR, NTR, and STR were achieved in 502 (71%), 126 (18%), and 59 (8%) patients. Biopsy was pursued in 21 (3%) patients. About 80 (11%) patients underwent surgery for multiple metastatic tumors. Following surgery, 65 (9%) had a new motor deficit, 20 (3%) had a language deficit, and 16 (2%) had a new vision deficit. At last follow-up, 13 (2%) incurred a wound infection, 10 (1%) an intracranial hemorrhage requiring re-operation, 29 (4%) deep vein thrombosis (DVT)/pulmonary embolism (PE), and 9 (1%) pneumonia.
Table 2.
Peri and postoperative characteristics of patients undergoing surgery of intracranial metastases separated by histology (non-small cell lung–NSCLC, breast, gastrointestinal–GI, melanoma, renal cell–RCC). The bolded values are statistically different than one or more of the other histological groups (P < 0.05)
| Study Population (n = 708)
| ||||||
|---|---|---|---|---|---|---|
| Characteristics | NSCLC
|
Breast
|
GI
|
Melanoma
|
RCC
|
All
|
|
n = 269
|
n = 106
|
n = 72
|
n = 88
|
n = 51
|
n = 708
|
|
| Number (%) | Number (%) | Number (%) | Number (%) | Number (%) | Number (%) | |
| Surgery | ||||||
| Gross-total resection | 198 (74%) | 73 (69%) | 53 (74%) | 62 (70%) | 40 (78%) | 502 (71%) |
| Near-total resection | 42 (16%) | 20 (19%) | 14 (19%) | 14 (16%) | 9 (18%) | 126 (18%) |
| Sub-total resection | 23 (9%) | 8 (8%) | 4 (6%) | 10 (11%) | 2 (4%) | 59 (8%) |
| Biopsy | 6 (2%) | 5 (5%) | 1 (1%) | 2 (2%) | 0 (0%) | 21 (3%) |
| En bloc resection | 18 (7%) | 10 (9%) | 7 (10%) | 6 (7%) | 8 (16%) | 60 (8%) |
| Cortical mapping | 9 (3%) | 5 (5%) | 1 (1%) | 0 (0%) | 3 (6%) | 21 (3%) |
| Multiple mets resected | 29 (11%) | 18 (17%) | 7 (10%) | 12 (14%) | 4 (8%) | 80 (11%) |
| New symptoms | ||||||
| Motor deficit | 19 (7%) | 10 (9%) | 8 (11%) | 9 (10%) | 7 (14%) | 65 (9%) |
| Language deficit | 4 (1%) | 3 (3%) | 3 (4%) | 5 (6%) | 2 (4%) | 20 (3%) |
| Vision deficit | 5 (2%) | 7 (7%) | 2 (3%) | 2 (2%) | 0 (0%) | 16 (2%) |
| Adjuvant therapy | ||||||
| Chemotherapy | 90 (33%) | 54 (51%) | 23 (32%) | 35 (40%) | 16 (31%) | 249 (35%) |
| Radiation therapy | 174 (65%) | 76 (72%) | 45 (63%) | 64 (73%) | 24 (47%) | 449 (63%) |
| Whole brain XRT | 131 (49%) | 60 (57%) | 33 (46%) | 45 (51%) | 16 (31%) | 340 (48%) |
| Stereotactic XRT | 91 (34%) | 37 (35%) | 19 (26%) | 32 (36%) | 17 (33%) | 222 (31%) |
| Clinical trials | 18 (7%) | 8 (8%) | 8 (11%) | 11 (13%) | 7 (14%) | 60 (8%) |
| Complications | ||||||
| Wound infection | 5 (2%) | 2 (2%) | 3 (4%) | 1 (1%) | 1 (2%) | 13 (2%) |
| Intracranial hemorrhage | 5 (2%) | 1 (0.9%) | 1 (1%) | 1 (1%) | 1 (2%) | 10 (1%) |
| Leptomeningeal disease | 17 (6%) | 17 (16%) | 2 (3%) | 8 (9%) | 0 (0%) | 51 (7%) |
| DVT/PE | 10 (4%) | 2 (2%) | 6 (8%) | 3 (3%) | 3 (6%) | 29 (4%) |
| Pneumonia | 2 (0.7%) | 1 (0.9%) | 0 (0%) | 1 (1%) | 0 (0%) | 9 (1%) |
| Survival | ||||||
| Deaths | 189 (70%) | 68 (64%) | 57 (79%) | 70 (80%) | 39 (76%) | 514 (73%) |
| Median survival | 37.1 | 13.1 | 6.3 | 6.9 | 11.5 | 9.3 |
| Local recurrence | 45 (17%) | 22 (21%) | 10 (14%) | 11 (13%) | 2 (4%) | 107 (15%) |
| 12-month local PFS rate | 74.1% | 72.5% | 70.9% | 83.5% | 100% | 84.3% |
| Distal recurrence | 73 (27%) | 41 (39%) | 23 (32%) | 39 (44%) | 20 (39%) | 226 (32%) |
| 12-month distal PFS rate | 63.0% | 48.1% | 39.4% | 36.5% | 63.1% | 54.6% |
mean±standard deviation,
median (interquartile range); DVT = deep vein thrombosis; PE = pulmonary embolism; XRT = radiation therapy.
Postoperative characteristics and differences between histological groups
The postoperative outcomes are summarized in Table 2. Five hundred and fourteen (73%) patients died at last follow-up, where the median survival was 9.3 months. The 6, 12, and 18-month survival rates were 60.7%, 41.5%, and 31.0%, respectively (Fig. 1). About 107 (15%) had local recurrence at the site of the surgical cavity, where the 6, 12, and 18-month local PFS was 93.3%, 84.3%, and 77.7%, respectively (Fig. 2). A total of 226 (32%) had distal intracranial recurrence following surgery, where the 6, 12, and 18-month distal PFS was 71.7%, 54.6%, and 47.0%, respectively (Fig. 3).
Figure 1.
Overall survival for patients with different primary cancers. (A) Kaplan–Meier survival curves for patients separated by histology (non-small cell lung–NSCLC, breast, gastrointestinal–GI, melanoma, renal cell–RCC). The median survival for all patients was 9.3 months, while the median survival was 37.1 months for patients with NSCLC, 13.1 months for breast cancer, 6.3 months for GI cancers, 6.9 months for melanoma, and 11.5 months for RCC. Patients with NSCLC had longer survival times as compared to patients with breast (P < 0.0001), GI (P < 0.0001), melanoma (P < 0.0001), and RCC (P < 0.0001). Likewise, patients with breast cancer had longer median survival than patients with GI (P = 0.005) and RCC (P = 0.04). (B) Diagram of factors associated with prolonged survival for patients with different primary pathologies.
Figure 2.
Local progression free survival (PFS) for patients with different primary cancers. Kaplan–Meier local PFS curves for patients separated by histology (non-small cell lung–NSCLC, breast, gastrointestinal–GI, melanoma, renal cell–RCC). The 24-month local PFS rates for all patients was 74.3%, while the rate was 67.4% for patients with NSCLC, 56.2%, for breast cancer 70.9% for GI cancer, 75.9% for melanoma, and 91.7% for RCC. Patients with RCC had significantly longer local PFS rates than patients with NCSLC (P = 0.02), breast cancer (P = 0.004), and GI cancer (P = 0.02).
Figure 3.
Distal progression free survival (PFS) for patients with different primary cancers. Kaplan–Meier distal PFS curves for patients separated by histology (non-small cell lung–NSCLC, breast, gastrointestinal–GI, melanoma, renal cell–RCC). The median distal PFS for all patients was 15.7 months, while the median survival was 25.7 months for patients with NSCLC, 11.7 months for breast cancer, 9.7 months for GI cancers, 8.5 months for melanoma, and 15.2 months for RCC. Patients with NSCLC had significantly longer distal PFS rates than patients with breast (P = 0.03) and melanoma (P = 0.002).
Leptomeningeal disease was more common following BC resection than NSCLC (P = 0.005), GI (P = 0.005), and RCC (P = 0.002). BC patients more commonly received chemotherapy than patients with NSCLC (P = 0.002), GI (P = 0.01), and RCC (P = 0.03). RCC patients less commonly received brain radiation therapy than NSCLC (P = 0.03), breast (P = 0.004), and melanoma (P = 0.003). RCC patients less commonly underwent WBRT than NSCLC (P = 0.03), BC (P = 0.004), and melanoma (P = 0.03).
The median survival (Fig. 1A) was longest for patients with NSCLC as compared to patients with BC (P < 0.0001), GI (P < 0.0001), melanoma (P < 0.0001), and RCC (P < 0.0001). Likewise, patients with BC had longer median survival than patients with GI (P = 0.005) and RCC (P = 0.04). The local PFS rate (Fig. 2) was longer in RCC patients than patients with NSCLC (P = 0.02), BC (P = 0.004), and GI cancer (P = 0.02). Distal recurrence occurred less frequently for NSCLC than BC (P = 0.02) and melanoma (P = 0.002). The distal PFS rates (Fig. 3) were longer for patients with NSCLC than BC (P = 0.03) and melanoma (P = 0.002).
Factors independently associated with overall survival by histological groups
The results are summarized in more detail in Table 3 and Fig. 1. For patients with NSCLC, the factors that remained independently associated with prolonged survival were: female gender, RPA Class 1, control of primary cancer, solitary brain metastasis, tumor size > 2 cm, adenocarcinoma pathology, discharge to home, and postoperative brain radiation. For patients with BC, the factors that remained independently associated with prolonged survival were: age < 65 years, no skull base involvement, and postoperative brain radiation. For patients with GI cancer, the factors that remained independently associated with prolonged survival were: age < 75 years, RPA Class < 3, KPS > 50, lack of motor deficit, non-esophagus primary, non-hemorrhagic tumors, and avoidance of language deficit. For patients with melanoma, the factors that remained independently associated with prolonged survival were: seizures, solitary metastasis, tumor size < 2 cm, discharge to home, and postoperative chemotherapy. For patients with RCC, the factors that remained independently associated with prolonged survival were: KPS > 60 and postoperative chemotherapy. When including all patients, the factors that remained independently associated with prolonged survival for patients with any type of primary cancer were: female gender, RPA Class 1, lack of motor deficit, lack of cognitive deficit, solitary brain metastasis, tumor size < 2 cm, discharge to home, postoperative chemotherapy, and postoperative brain radiation.
Table 3.
Multivariate associations with prolonged survival for patients undergoing surgery of intracranial metastases separated by histology (non-small cell lung–NSCLC, breast, gastrointestinal–GI, melanoma, renal cell–RCC)
| Multivariate associations with prolonged survival
| ||
|---|---|---|
| Variable | Relative risk (95%CI) | P-value |
| NSCLC | ||
| Female | 0.610 (0.420–0.880) | 0.008 |
| RPA Class 1 | 0.549 (0.322–0.899) | 0.02 |
| Control of primary cancer | 0.572 (0.387–0.851) | 0.006 |
| Solitary brain metastasis | 0.520 (0.356–0.767) | 0.001 |
| Tumor size < 2 cm | 0.680 (0.452–1.004) | 0.05 |
| Adenocarcinoma | 0.598 (0.386–0.934) | 0.02 |
| Discharge to home | 0.642 (0.421–0.992) | 0.05 |
| Post-op brain radiation | 0.388 (0.266–0.570) | < 0.0001 |
| Breast | ||
| Age < 65 years | 0.478 (0.258–0.954) | 0.04 |
| Lack of COPD | 0.208 (0.072–0.880) | 0.04 |
| No skull base involvement | 0.179 (0.053–0.820) | 0.03 |
| Post-op brain radiation | 0.388 (0.221–0.714) | 0.003 |
| GI | ||
| Age < 75 years | 0.285 (0.112–0.838) | 0.02 |
| RPA Class < 3 | 0.080 (0.020–0.411) | 0.005 |
| KPS > 50 | 0.208 (0.072–0.754) | 0.02 |
| Lack of seizures | 0.173 (0.068–0.514) | 0.003 |
| Lack of motor deficit | 0.407 (0.218–0.747) | 0.004 |
| Non-esophagus primary | 0.186 (0.062–0.631) | 0.009 |
| Non-hemorrhagic tumors | 0.392 (0.195–0.832) | 0.02 |
| Avoidance of language deficit | 0.053 (0.014–0.263) | 0.002 |
| Melanoma | ||
| Lack of coronary artery disease | 0.017 (0.001–0.399) | 0.02 |
| Seizures | 0.377 (0.124–0.936) | 0.03 |
| Solitary metastasis | 0.301 (0.175–0.515) | < 0.0001 |
| Tumor size < 2 cm | 0.562 (0.312–0.964) | 0.03 |
| Discharge to home | 0.562 (0.312–0.964) | 0.002 |
| Post-op chemotherapy | 0.383 (0.220–0.659) | 0.0005 |
| RCC | ||
| KPS > 60 | 0.274 (0.110–0.782) | 0.02 |
| Avoidance of post-op ICH | 0.274 (0.110–0.782) | 0.02 |
| Postoperative chemotherapy | 0.278 (0.103–0.637) | 0.008 |
| All metastases | ||
| Female | 0.619 (0.513–0.748) | < 0.0001 |
| RPA Class 1 | 0.683 (0.530–0.868) | < 0.0001 |
| Lack of motor deficit | 0.773 (0.637–0.941) | 0.01 |
| Lack of cognitive deficit | 0.654 (0.525–0.820) | 0.0003 |
| Solitary brain metastasis | 0.601 (0.494–0.733) | < 0.0001 |
| Tumor size < 2 cm | 0.711 (0.566–0.885) | 0.002 |
| Discharge to home | 0.623 (0.502–0.777) | < 0.0001 |
| Postoperative chemotherapy | 0.726 (0.593–0.887) | 0.002 |
| Postoperative brain radiation | 0.546 (0.448–0.666) | < 0.0001 |
Discussion
Prior studies on metastatic brain tumors separated by histology
Patients who develop intracranial metastases are generally considered to have dismal prognoses regardless of the extent of their systemic disease.1,2 Studies on patients with intracranial metastases often combine patients with different primary cancers into the same study population.3–9,17 This grouping of all patients intuitively appears to be flawed. This is because patients with different types of primary cancers have varying degrees of disease burden, undergo different treatment regimens, receive dissimilar systemic therapies, and have different median survival times.1,2 Moreover, these studies will be biased by the primary cancers that comprise the majority of the patient population, which is usually lung cancer in most studies.3–9,15,17 These key differences make it important to separate patients with different primary cancers when attempting to identify factors associated with survival, as well as selecting treatment regimens.
Studies evaluating outcomes for patients with metastatic brain disease separated by primary cancers are few and limited.10–13,18–23 Studies evaluating outcomes following surgical resection of these different primary cancers are even more sparse (Table 4).10–13,18–21,24 These prior studies and others are limited because they included patients who did not undergo surgery,10,23–26 only included solitary metastases,18,23,27 contained a significant number of patients who did not undergo adjuvant therapies,12,13,19–21,27 and did not perform multivariate analyses.11,18,24,26,27 As a result, the factors independently associated with survival for patients with different types of primary cancers remain unclear. This lack of clarity makes it difficult to prognosticate outcomes for patients with different types of primary cancers.
Table 4.
Summary of previous surgical studies on intracranial metastases where different primary cancers were separately evaluated
| Studies | Year | Non-surgical patients | No. surgical patients | NSCLC | Breast | RCC | Melanoma | GI |
|---|---|---|---|---|---|---|---|---|
| Present study | 2012 | No | 708 | X | X | X | X | X |
| Sundrasen et al. | 1985 | No | 50 | X | ||||
| Macchiarini et al. | 1991 | No | 37 | X | ||||
| Wronski et al.* | 1995 | No | 231 | X | ||||
| Arbit et al.* | 1995 | No | 109 | X | ||||
| Abrahams et al. | 2001 | Yes | 50 | X | ||||
| Wronski et al.* | 1997 | No | 70 | X | ||||
| Niwinska et al. | 2011 | Yes | 57 | X | ||||
| Cahill et al. | 2011 | Yes† | 264 | |||||
| Wronski et al.* | 1996 | No | 50 | X | ||||
| Salvati et al. | 1992 | No | 29 | X | ||||
| Shuch et al. | 2008 | Yes | 48 | X | ||||
| Wronski et al.* | 2000 | No | 91 | X | ||||
| Zacest et al | 2002 | No | 147 | X | ||||
| Raizer et al.* | 2008 | Yes | 126 | X | ||||
| Farnell et al. | 1996 | Yes | 11 | X | ||||
| Wronski et al.* | 1999 | No | 73 | X | ||||
| Jiang et al. | 2011 | Yes | 7 | X | ||||
| Noura et al. | 2012 | Yes | 17 | X | ||||
| Jiang et al. | 2012 | Yes | 6 | X |
From the same institution.
Included patients who underwent spine surgery. GI = gastrointestinal; NSCLC = non-small cell lung cancer; RCC = renal cell cancer.
Differences between metastatic brain tumor patients with different primary cancers
There were significant differences in the pre, peri, and postoperative characteristics between the pathological groups in this study. Among the preoperative differences, the most notable were seen in the components used to assign RPA classification (age, primary tumor control, extracranial spread, KPS).28 These key distinctions, as well as others found in this study, may make the RPA and other classification systems based on all patients with brain metastases less applicable or useful, especially for primary cancers that are less common.
In addition to these preoperative differences, there were also postoperative distinctions between different primary cancers. Patients with NSCLC and BC had the longest median survival. Rades et al. evaluated 195 patients who underwent surgery of a solitary metastasis, and found that a diagnosis of BC was independently associated with improved survival.28 Tendulkar et al. studied 271 patients who underwent solitary metastatic tumor resection, and found that lung cancer was independently associated with improved survival.29 These studies and others, however, are limited because they group all patients with metastatic tumors together and only include solitary lesions.28,29 Large-scale studies attempting to evaluate differences in outcomes for patients with different primary cancers therefore remain unclear.
Factors independently associated with survival
Non-small cell lung cancer
It is estimated that 30–80% of patients with NSCLC will develop brain metastases.3–9,17 Among brain metastases, NSCLC represents the most common source of primary cancers.3–9,17 Once brain metastases develop, median survival is less than 6 months without treatment. Surgical resection and WBRT are associated with median survival times of 8–19 months.20 Previous studies have found that the factors associated with shorter survival were: no surgery of their primary tumor, male gender, infra-tentorial location, extracranial spread, and age > 60 years.20 This study also found gender and radiation to be associated with survival, but also adds RPA Class, control of primary tumor, solitary brain metastasis, tumor size, adenocarcinoma, and discharge to home. This study also shows that among NSCLCs, adenocarcinomas have better prognoses than squamous cell carcinomas. This may be important because these histologies are typically grouped into the same NSCLC category.4,15,30,31 In addition, solitary and smaller tumors were each independently associated with improved survival, which may signify less intracranial disease burden and thus improved survival.32,33 Moreover, this study also emphasizes the importance of preserving functional status, where patients who were able to be discharged to home had improved survival.
Breast cancer
BC is the second most common cause of intracranial metastases.34 It is estimated that 10–30% of patients with BC will develop brain metastases.34 Median survival times vary from 5 to 16 months; however, the majority of these studies involved WBRT.23,32,35 Niwinska and colleagues evaluated 100 patients with solitary BC metastases, and found that RPA Class 1, neurosurgical intervention, and chemotherapy were associated with improved survival.23 Wronski et al. evaluated 70 patients and found that WBRT and absence of leptomeningeal disease were associated with improved survival.21 The present study found that younger age, no skull base involvement, and postoperative brain radiation were associated with prolonged survival. Younger patients typically have a better ability to tolerate neurological insults associated with surgery, radiation, and chemotherapy.4 Skull base lesions are relatively rare among patients with metastatic disease, and are typically treated with radiation therapy because of their presumed poor prognosis.36 Lastly, as seen in previous primarily non-surgical studies, the use of postoperative radiation was associated with improved survival.35
Gastrointestinal cancer
Gastrointestinal cancers include a variety of different primary cancers including colorectal, hepatocellular, and gastric cancers.37 These cancers metastasize less frequently to the brain, where approximately 2–4% of patients with these primary cancers will develop brain metastases.37 The median survival for patients who develop brain metastases ranges from 4 to 8 months.12,37 Jung et al. evaluated 126 patients, of which 20 underwent surgical resection.38 In this mixed treatment population, RPA class and chemotherapy were associated with prolonged survival.38 Wronski et al. studied 73 patients operated on for metastatic colorectal carcinoma between 1974 and 1993, and found that only a cerebellar location was associated with poorer survival.12 Among patients with GI primary cancers, younger age, lower RPA class, improved KPS, lack of preoperative motor deficit, non-esophageal tumors, non-hemorrhagic tumors, and lack of postoperative language deficits were independently associated with improved survival in this study. The most significant age, RPA class, and KPS associated with survival were 75, Class 3, and 50, respectively. These cutoffs are relatively low because it includes older patients and less functional patients than prior RPA grading systems.4,15,33 Surgical intervention can therefore lead to improved outcomes for even GI patients with presumed poor prognoses.
Melanoma
Melanoma is considered the third most common type of cancer.39 It is estimated that less than 10% of patients with melanoma will develop brain metastases.39 However, brain involvement increases to 45–70% of patients with advanced melanoma.39 Moreover, melanoma, unlike other cancers, has a greater propensity for micrometastases where patients can harbor numerous metastases without neurological symptoms.39 The median survival ranges from 2 to 8 months in previous studies.19,40 Raizer et al. evaluated 355 patients with melanoma, of which only 126 patients underwent surgery.40 They found that age, number of intracranial metastases, extracranial spread, and presence of neurological symptoms were associated with poorer survival.40 Wronski et al. from the same institution evaluated 91 patients who underwent intracranial melanoma resection, and reported similar findings.19 This study found that preoperative seizures, solitary metastasis, smaller size tumors, discharge to home, and use of postoperative chemotherapy were independently associated with improved survival.
Renal cell cancer
RCC accounts for approximately 1–3% of all cancers.41 Despite this relative rarity, it accounts for a significant number of intracranial metastases.41 It is estimated that as high as 30% of patients with RCC will develop brain metastases.41 The median survival times for patients who develop RCC brain metastases ranges from 4 to 23 months.13,27,42 Salvati et al. evaluated 29 patients with solitary RCC who were operated on from 1975 to 1988, and found that radiation did not significantly improve survival.27 Wronski et al. evaluated 50 patients operated on between 1974 and 1993, and found that a supratentorial location, lack of preoperative neurological deficit, and left-sided lesions were associated with prolonged survival.13 This study found that preoperative KPS and use of postoperative chemotherapy were independently associated with prolonged survival.
Strengths and limitations
We believe this study provides several useful findings for patients who undergo surgical resection of intracranial metastases. First, studies comparing characteristics between patients with different primary cancers have yet to be done. This study shows there are important distinctions between patients with different primary cancers. These dissimilarities may preclude them from being included into the same study population. Second, studies attempting to obtain factors independently associated with survival for patients with different types of primary cancers are few and limited. The majority of these studies include patients who did not undergo surgical resection and/or had solitary metastases. These features are not necessarily applicable to most patients with metastatic disease. Lastly, this study may provide useful information that may help guide treatment strategies aimed at optimizing outcomes for patients with metastatic disease. This study shows that each primary cancer has different set of factors independently associated with survival. These factors can be taken into account when selecting optimal treatment regimens aimed at prolonging survival.
This study, however, has some limitations. One limitation is that these findings only apply to patients who underwent surgical resection of their intracranial metastasis. Second, this study was limited to patients with the most common types of primary cancer (lung, breast, GI, melanoma, and RCC). Patients with less common types of intracranial metastases were excluded from further analyses because of small patient numbers. Additionally, this study did not analyze the prognostic implication of molecular markers and genotypes including hormonal receptors, epidermal growth factor receptors, and HER-2, among others. These molecular markers may be better associated with survival for some types of primary cancer, but were not analyzed in this study. Furthermore, a significant number of patients in this study did not undergo GTR, WBRT, and/or SRS. The findings of this study may be altered in the context of receiving the most aggressive treatment regimens. Lastly, this study is inherently limited by its retrospective design. It is therefore not appropriate to infer direct causal relationships. However, we tried to create a uniform patient population by utilizing strict inclusion criteria in order to provide more relevant information for patients who underwent intracranial metastatic tumor surgery. We also performed multivariate analyses to control for potential confounding variables. Given these statistical controls and relatively precise outcome measures, we believe our findings offer useful insights into the management of patients with different types of primary cancers. However, prospective studies are ultimately needed to provide better data to guide clinical decision-making.
Conclusions
Patients with a variety of different primary cancers can develop intracranial metastases. When patients with different primary cancers develop metastatic disease to body sites, they are not grouped into the same category. It is therefore peculiar why patients who develop intracranial metastases are all lumped together. This study shows that patients with different types of primary cancers have disparate patient-related characteristics and factors independently associated with survival. Patients with different primary cancers who develop intracranial metastases therefore represent distinct clinical groups with disparate characteristics and outcomes, and should be studied according to their primary cancer and not their metastatic site.
Footnotes
Disclosures
None
References
- 1.Eichler AF, Loeffler JS. Multidisciplinary management of brain metastases. Oncologist. 2007;12:884–98. doi: 10.1634/theoncologist.12-7-884. [DOI] [PubMed] [Google Scholar]
- 2.Ewend MG, Morris DE, Carey LA, Ladha AM, Brem S. Guidelines for the initial management of metastatic brain tumors: role of surgery, radiosurgery, and radiation therapy. J Natl Compr Canc Netw. 2008;6:505–13. doi: 10.6004/jnccn.2008.0038. quiz 514. [DOI] [PubMed] [Google Scholar]
- 3.Agarwal JP, Wadasadawala T, Munshi A, Chadda P, Apsani R, Upasani M, et al. Validation of recursive partitioning analysis classification in patients with brain metastases from non-small cell lung cancer treated with short-course accelerated radiotherapy. Clin Oncol (R Coll Radiol) 2010;22:837–43. doi: 10.1016/j.clon.2010.05.022. [DOI] [PubMed] [Google Scholar]
- 4.Barnholtz-Sloan JS, Yu C, Sloan AE, Vengoechea J, Wang M, Dignam JJ, et al. A nomogram for individualized estimation of survival among patients with brain metastasis. Neuro Oncol. 2012;14:910–8. doi: 10.1093/neuonc/nos087. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Ewend MG, Brem S, Gilbert M, Goodkin R, Penar PL, Varia M, et al. Treatment of single brain metastasis with resection, intracavity carmustine polymer wafers, and radiation therapy is safe and provides excellent local control. Clin Cancer Res. 2007;13:3637–41. doi: 10.1158/1078-0432.CCR-06-2095. [DOI] [PubMed] [Google Scholar]
- 6.Grossman R, Mukherjee D, Chang DC, Purtell M, Lim M, Brem H, et al. Predictors of inpatient death and complications among postoperative elderly patients with metastatic brain tumors. Ann Surg Oncol. 2011;18:521–8. doi: 10.1245/s10434-010-1299-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Patchell RA, Tibbs PA, Regine WF, Dempsey RJ, Mohiuddin M, Kryscio RJ, et al. Postoperative radiotherapy in the treatment of single metastases to the brain: a randomized trial. JAMA. 1998;280:1485–9. doi: 10.1001/jama.280.17.1485. [DOI] [PubMed] [Google Scholar]
- 8.Patchell RA, Tibbs PA, Walsh JW, Dempsey RJ, Maruyama Y, Kryscio RJ, et al. A randomized trial of surgery in the treatment of single metastases to the brain. N Engl J Med. 1990;322:494–500. doi: 10.1056/NEJM199002223220802. [DOI] [PubMed] [Google Scholar]
- 9.Rades D, Hornung D, Veninga T, Schild SE, Gliemroth J. Single brain metastasis: radiosurgery alone compared with radiosurgery plus up-front whole-brain radiotherapy. Cancer. 2012;118:2980–5. doi: 10.1002/cncr.26612. [DOI] [PubMed] [Google Scholar]
- 10.Jiang XB, Yang QY, Sai K, Zhang XH, Chen ZP, Mou YG. Brain metastases from colorectal carcinoma: a description of 60 cases in a single Chinese cancer center. Tumour Biol. 2011;32:1249–56. doi: 10.1007/s13277-011-0229-7. [DOI] [PubMed] [Google Scholar]
- 11.Sundaresan N, Galicich JH. Surgical treatment of single brain metastases from non-small-cell lung cancer. Cancer Invest. 1985;3:107–13. doi: 10.3109/07357908509017493. [DOI] [PubMed] [Google Scholar]
- 12.Wronski M, Arbit E. Resection of brain metastases from colorectal carcinoma in 73 patients. Cancer. 1999;85:1677–85. doi: 10.1002/(sici)1097-0142(19990415)85:8<1677::aid-cncr6>3.0.co;2-c. [DOI] [PubMed] [Google Scholar]
- 13.Wronski M, Arbit E, Russo P, Galicich JH. Surgical resection of brain metastases from renal cell carcinoma in 50 patients. Urology. 1996;47:187–93. doi: 10.1016/S0090-4295(99)80413-0. [DOI] [PubMed] [Google Scholar]
- 14.Dutta D, Vanere P, Gupta T, Munshi A, Jalali R. Factors influencing activities of daily living using FIM-FAM scoring system before starting adjuvant treatment in patients with brain tumors: results from a prospective study. J Neurooncol. 2009;94:103–10. doi: 10.1007/s11060-009-9810-y. [DOI] [PubMed] [Google Scholar]
- 15.Gaspar L, Scott C, Rotman M, Asbell S, Phillips T, Wasserman T, et al. Recursive partitioning analysis (RPA) of prognostic factors in three Radiation Therapy Oncology Group (RTOG) brain metastases trials. Int J Radiat Oncol Biol Phys. 1997;37:745–51. doi: 10.1016/s0360-3016(96)00619-0. [DOI] [PubMed] [Google Scholar]
- 16.McGirt MJ, Chaichana KL, Gathinji M, Attenello FJ, Than K, Olivi A, et al. Independent association of extent of resection with survival in patients with malignant brain astrocytoma. J Neurosurg. 2009;110:156–62. doi: 10.3171/2008.4.17536. [DOI] [PubMed] [Google Scholar]
- 17.Ewend MG, Williams JA, Tabassi K, Tyler BM, Babel KM, Anderson RC, et al. Local delivery of chemotherapy and concurrent external beam radiotherapy prolongs survival in metastatic brain tumor models. Cancer Res. 1996;56:5217–23. [PubMed] [Google Scholar]
- 18.Macchiarini P, Buonaguidi R, Hardin M, Mussi A, Angeletti CA. Results and prognostic factors of surgery in the management of non-small cell lung cancer with solitary brain metastasis. Cancer. 1991;68:300–4. doi: 10.1002/1097-0142(19910715)68:2<300::aid-cncr2820680215>3.0.co;2-s. [DOI] [PubMed] [Google Scholar]
- 19.Wronski M, Arbit E. Surgical treatment of brain metastases from melanoma: a retrospective study of 91 patients. J Neurosurg. 2000;93:9–18. doi: 10.3171/jns.2000.93.1.0009. [DOI] [PubMed] [Google Scholar]
- 20.Wronski M, Arbit E, Burt M, Galicich JH. Survival after surgical treatment of brain metastases from lung cancer: a follow-up study of 231 patients treated between 1976 and 1991. J Neurosurg. 1995;83:605–16. doi: 10.3171/jns.1995.83.4.0605. [DOI] [PubMed] [Google Scholar]
- 21.Wronski M, Arbit E, McCormick B. Surgical treatment of 70 patients with brain metastases from breast carcinoma. Cancer. 1997;80:1746–54. doi: 10.1002/(sici)1097-0142(19971101)80:9<1746::aid-cncr8>3.0.co;2-c. [DOI] [PubMed] [Google Scholar]
- 22.Niwinska A, Murawska M. New breast cancer recursive partitioning analysis prognostic index in patients with newly diagnosed brain metastases. Int J Radiat Oncol Biol Phys. 2012;82:2065–71. doi: 10.1016/j.ijrobp.2010.10.077. [DOI] [PubMed] [Google Scholar]
- 23.Niwińska A, Pogoda K, Murawska M, Niwiński P. Factors influencing survival in patients with breast cancer and single or solitary brain metastasis. Eur J Surg Oncol. 2011;37:635–42. doi: 10.1016/j.ejso.2011.05.002. [DOI] [PubMed] [Google Scholar]
- 24.Noura S, Ohue M, Shingai T, Fujiwara A, Imada S, Sueda T, et al. Brain metastasis from colorectal cancer: prognostic factors and survival. J Surg Oncol. 2012;106:144–8. doi: 10.1002/jso.23055. [DOI] [PubMed] [Google Scholar]
- 25.Farnell GF, Buckner JC, Cascino TL, O’Connell MJ, Schomberg PJ, Suman V. Brain metastases from colorectal carcinoma. The long term survivors. Cancer. 1996;78:711–6. doi: 10.1002/(SICI)1097-0142(19960815)78:4<711::AID-CNCR3>3.0.CO;2-H. [DOI] [PubMed] [Google Scholar]
- 26.Jiang XB, Ke C, Zhang GH, Sai K, Chen ZP, Mou YG. Brain metastases from hepatocellular carcinoma: clinical features and prognostic factors. BMC Cancer. 2012;12:49. doi: 10.1186/1471-2407-12-49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Salvati M, Scarpinati M, Orlando ER, Celli P, Gagliardi FM. Single brain metastases from kidney tumors. Clinico-pathologic considerations on a series of 29 cases. Tumori. 1992;78:392–4. doi: 10.1177/030089169207800610. [DOI] [PubMed] [Google Scholar]
- 28.Rades D, Kueter JD, Gliemroth J, Veninga T, Pluemer A, Schild SE. Resection plus whole-brain irradiation versus resection plus whole-brain irradiation plus boost for the treatment of single brain metastasis. Strahlenther Onkol. 2012;188:143–7. doi: 10.1007/s00066-011-0024-9. [DOI] [PubMed] [Google Scholar]
- 29.Tendulkar RD, Liu SW, Barnett GH, Vogelbaum MA, Toms SA, Jin T, et al. RPA classification has prognostic significance for surgically resected single brain metastasis. Int J Radiat Oncol Biol Phys. 2006;66:810–7. doi: 10.1016/j.ijrobp.2006.06.003. [DOI] [PubMed] [Google Scholar]
- 30.Nieder C, Andratschke N, Grosu AL, Molls M. Recursive partitioning analysis (RPA) class does not predict survival in patients with four or more brain metastases. Strahlenther Onkol. 2003;179:16–20. doi: 10.1007/s00066-003-1028-x. [DOI] [PubMed] [Google Scholar]
- 31.Regine WF, Rogozinska A, Kryscio RJ, Tibbs PA, Young AB, Patchell RA. Recursive partitioning analysis classifications I and II: applicability evaluated in a randomized trial for resected single brain metastases. Am J Clin Oncol. 2004;27:505–9. doi: 10.1097/01.coc.0000135379.36325.de. [DOI] [PubMed] [Google Scholar]
- 32.DeAngelis LM, Mandell LR, Thaler HT, Kimmel DW, Galicich JH, Fuks Z, et al. The role of postoperative radiotherapy after resection of single brain metastases. Neurosurgery. 1989;24:798–805. doi: 10.1227/00006123-198906000-00002. [DOI] [PubMed] [Google Scholar]
- 33.Hartford AC, Paravati AJ, Spire WJ, Li Z, Jarvis LA, Fadul CE, et al. Postoperative stereotactic radiosurgery without whole-brain radiation therapy for brain metastases: potential role of preoperative tumor size. Int J Radiat Oncol Biol Phys. 2013;85:650–5. doi: 10.1016/j.ijrobp.2012.05.027. [DOI] [PubMed] [Google Scholar]
- 34.Lee YT. Breast carcinoma: pattern of metastasis at autopsy. J Surg Oncol. 1983;23:175–80. doi: 10.1002/jso.2930230311. [DOI] [PubMed] [Google Scholar]
- 35.Liu MT, Hsieh CY, Wang AY, Chang TH, Pi CP, Huang CC, et al. Prognostic factors affecting the outcome of brain metastases from breast cancer. Support Care Cancer. 2006;14:936–42. doi: 10.1007/s00520-006-0029-6. [DOI] [PubMed] [Google Scholar]
- 36.Pan J, Liu AL, Wang ZC. Gamma knife radiosurgery for skull base malignancies. Clin Neurol Neurosurg. 2013;115:44–8. doi: 10.1016/j.clineuro.2012.04.013. [DOI] [PubMed] [Google Scholar]
- 37.Cascino TL, Leavengood JM, Kemeny N, Posner JB. Brain metastases from colon cancer. J Neurooncol. 1983;1:203–209. doi: 10.1007/BF00165604. [DOI] [PubMed] [Google Scholar]
- 38.Jung M, Ahn JB, Chang JH, Suh CO, Hong S, Roh JK, et al. Brain metastases from colorectal carcinoma: prognostic factors and outcome. J Neurooncol. 2011;101:49–55. doi: 10.1007/s11060-010-0214-9. [DOI] [PubMed] [Google Scholar]
- 39.Schouten LJ, Rutten J, Huveneers HA, Twijnstra A. Incidence of brain metastases in a cohort of patients with carcinoma of the breast, colon, kidney, and lung and melanoma. Cancer. 2002;94:2698–705. doi: 10.1002/cncr.10541. [DOI] [PubMed] [Google Scholar]
- 40.Raizer JJ, Hwu WJ, Panageas KS, Wilton A, Baldwin DE, Bailey E, et al. Brain and leptomeningeal metastases from cutaneous melanoma: survival outcomes based on clinical features. Neuro Oncol. 2008;10:199–207. doi: 10.1215/15228517-2007-058. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Motzer RJ, Bander NH, Nanus DM. Renal-cell carcinoma. N Engl J Med. 1996;335:865–75. doi: 10.1056/NEJM199609193351207. [DOI] [PubMed] [Google Scholar]
- 42.Shuch B, La Rochelle JC, Klatte T, Riggs SB, Liu W, Kabbinavar FF, et al. Brain metastasis from renal cell carcinoma: presentation, recurrence, and survival. Cancer. 2008;113:1641–8. doi: 10.1002/cncr.23769. [DOI] [PubMed] [Google Scholar]