Outcomes and predictors of improved survival after gamma knife radiosurgery for metastatic brain tumors originated from breast carcinoma

Clinical and pathological factors and outcome of central nervous system metastasis in breast cancer

Background

In Switzerland, approximately 6000 new breast cancer cases and 1300 deaths are reported annually. Brain metastasis from breast cancer (BMBC) has a major effect on prognosis. This study aimed to identify prognostic factors for overall survival (OS) in a cohort of Swiss patients with BMBC. This study evaluated the prognosis on older BMBC, which has not been completely addressed in the literature.

Methods

We performed a retrospective chart review analysis with the primary endpoint of OS after a diagnosis of BMBC. The study population was divided into 2 groups based on an OS cut-off value of 12 months after diagnosis. Univariate and multivariate analyses of several risk factors, including age, were performed. To evaluate differences in OS according to age, we performed a secondary analysis to examine the prognostic value of clinical symptoms, metastatic pattern, and lymph node involvement in an older (≥65 years) vs. younger (<65 years) cohort.

Results

From 1989 to 2019, 55 patients were identified as having BMBC, among whom 47 patients were confirmed to be dead. The median patient age was 58 years (range 25-83 years). Comorbidities were present in 45 (81.8%) patients. The median survival in the OS <12 and OS ≥12 months groups was 4.3 and 30.7 months, respectively (p<0.001). Multivariate analysis revealed no significant differences in terms of comorbidities, medication use, M-stage, and symptomatology between the 2 groups. Additionally, there was no significant difference in OS in the 2 subgroups of patients aged <65 and ≥65 years.

Discussion

We concluded that age should not be a decisive factor in therapy planning for advanced breast cancer patients with BMBC.

Cerebral metastases

Brain metastases are common and deadly. Over the last 25 years GKNS has been established as an invaluable treatment. It may be used as a primary treatment or after either surgery or WBRT. Patients are assessed using one of a number of available scales. GKNS may be repeated for new metastases and for unresponsive tumors. Prescription doses are usually between 18 and 20Gy. The use of advanced MR techniques to highlight sensitive structures like the hippocampi have extended the efficacy of the treatment. More recently GKNS has been used with different target therapies with improved results. More recently frameless treatments have become more popular in this group of very sick patients. GKNS controls tumors in between 80% and over 95% of cases and may even be used for brainstem tumors.

The Use of Treatment Response Assessment Maps in Discriminating Between Radiation Effect and Persistent Tumoral Lesion in Metastatic Brain Tumors Treated with Gamma Knife Radiosurgery

BACKGROUND

Traditional imaging modalities are not useful in the follow-up of irradiated metastatic brain tumors, because radiation can change imaging characteristics. We aimed to assess the ability of treatment response assessment maps (TRAMs) calculated from delayed-contrast magnetic resonance imaging (MRI) in differentiation between radiation effect and persistent tumoral tissue.

METHODS

TRAMs were calculated by subtracting three-dimensional T1 MRIs acquired 5 minutes after contrast injection from the images acquired 60-105 minutes later. Red areas were regarded as radiation effect and blue areas as persistent tumoral lesion. Thirty-seven patients with 130 metastatic brain tumors who were treated with Gamma Knife radiosurgery and who underwent TRAMs perfusion-weighted MRI were enrolled in this retrospective study.

RESULTS

The median age was 58 years and the most common primary diagnosis was lung cancer (n = 21). The median follow-up period of patients was 12 months. The overall local control rate was 100% at 1 year and 98.9% at 2 years. The median progression-free survival was 12 months. The mean overall survival was 27.3 months. The radiologic and clinical follow-up showed a clinicoradiologic diagnosis of a persistent tumoral lesion in 3 tumors (2.3%) and radiation effect in 127 tumors (97.7%). There was a fair agreement between clinicoradiologic diagnosis and TRAMs analysis (κ = 0.380). The sensitivity and positive predictive value of TRAMs in diagnosing radiation effect were 96.06% and 99.2%, respectively. TRAMs showed comparable results to perfusion-weighted MRI, with a diagnostic odds ratio of 27.4 versus 20.7, respectively.

CONCLUSIONS

The presented results show the ability of TRAMs in differentiating radiation effect and persistent tumoral lesions.

Propensity for different vascular distributions and cerebral edema of intraparenchymal brain metastases from different primary cancers

PURPOSE

This study seeks to ascertain whether different primary tumor types have a propensity for brain metastases (BMs) in different cerebral vascular territories and cerebral edema.

METHODS

Consecutive adult patients who underwent surgical resection of a BM at a tertiary care institution between 2001 and 2011 were retrospectively reviewed. Only patients with the most common primary cancers (lung, breast, skin-melanoma, colon, and kidney) were included. Preoperative MRIs were reviewed to classify all tumors by cerebral vascular territory (anterior cerebral artery-ACA, lenticulostriate, middle cerebral artery-MCA, posterior cerebral artery-PCA, posterior fossa, and watershed), and T2-weighted FLAIR widths were measured. Chi square analyses were performed to determine differences in cerebral vascular distribution by primary tumor type, and one-way ANOVA analyses were performed to determine FLAIR signal differences.

RESULTS

669 tumors from 388 patients were classified from lung (n = 316 BMs), breast (n = 144), melanoma (n = 119), renal (n = 47), and colon (n = 43). BMs from breast cancer were less likely to be located in PCA territory (n = 18 [13%]; χ² = 6.10, p = 0.01). BMs from melanoma were less likely to be located in cerebellar territory (n = 11 [9%]; χ² = 14.1, p < 0.001), and more likely to be located in lateral (n = 5 [4%]; χ² = 4.56, p = 0.03) and medial lenticulostriate territories (n = 2 [2%]; χ² = 6.93, p = 0.009). BMs from breast and melanoma had shorter T2-FLAIR widths, with an average [IQR] of 47.2 [19.6-69.2] mm (p = 0.01) and 41.2 [14.4-62.7] mm (p = 0.002) respectively. Conversely, BMs from renal cancer had longer T2-FLAIR widths (64.2 [43.6-80.8] mm, p = 0.002).

CONCLUSIONS

These findings suggest that different primary tumor types could have propensities for different cerebral vascular territories and cerebral edema.

Brain metastasis in breast cancer: a comprehensive literature review

This comprehensive review provides information on epidemiology, size, grade, cerebral localization, clinical symptoms, treatments, and factors associated with longer survival in 14,599 patients with brain metastasis from breast cancer; the molecular features of breast cancers most likely to develop brain metastases and the potential use of these predictive molecular alterations for patient management and future therapeutic targets are also addressed. The review covers the data from 106 articles representing this subject in the era of modern neuroimaging (past 35 years). The incidence of brain metastasis from breast cancer (24 % in this review) is increasing due to advances in both imaging technologies leading to earlier detection of the brain metastases and introduction of novel therapies resulting in longer survival from the primary breast cancer. The mean age at the time of breast cancer and brain metastasis diagnoses was 50.3 and 48.8 years respectively. Axillary node metastasis was noted in 32.8 % of the patients who developed brain metastasis. The median time intervals between the diagnosis of breast cancer to identification of brain metastasis and from identification of brain metastasis to death were 34 and 15 months, respectively. The most common symptoms experienced in patients with brain metastasis consisted of headache (35 %), vomiting (26 %), nausea (23 %), hemiparesis (22 %), visual changes (13 %) and seizures (12 %). A majority of the patients had multiple metastases (54.2 %). Cerebellum and frontal lobes were the most common sites of metastasis (33 and 16 %, respectively). Of the primary tumors for which biomarkers were recorded, 37 % were estrogen receptor (ER)+, 41 % ER-, 36 % progesterone receptor (PR)+, 34 % PR-, 35 % human epithelial growth factor receptor 2 (HER2)+, 41 % HER2-, 27 % triple negative and 18 % triple positive (TP). Treatment in most patients consisted of a multimodality approach often with two or more of the following: whole brain radiation therapy (52 %), chemotherapy (51 %), stereotactic radiosurgery (20 %), surgical resection (14 %), trastuzumab (39 %) for HER2 positive tumors, and hormonal therapy (34 %) for ER and/or PR positive tumors. Factors that had an impact on prognosis included grade and size of the tumor, multiple metastases, presence of extra-cranial metastasis, triple negative or HER2+ biomarker status, and high Karnovsky score. Novel therapies such as application of agents to reduce tumor angiogenesis or alter permeability of the blood brain barrier are being explored with preliminary results suggesting a potential to improve survival after brain metastasis. Other potential therapies based on genetic alterations in the tumor and the microenvironment in the brain are being investigated; these are briefly discussed.