|Home | About | Journals | Submit | Contact Us | Français|
Brain metastases from gynecologic primary cancers are rare events, but they can be a cause of morbidity and mortality when they occur.
This is a single institution retrospective study on patients with brain metastases from gynecologic primary cancer who received Gamma Knife stereotactic radiosurgery (SRS). Between 2000 and 2013, a total of 33 patients with brain metastases from gynecologic primary including cervical (n=2), endometrial (n=6) and ovarian cancers (n=25) were treated with SRS at our institution. Electronic medical records were reviewed to determine survival, patterns of failure and cause of death.
Overall survival at 1, 2 and 5 years for the entire population was 47.1%, 21.7%, and 14.5%, respectively. There was no difference in survival between the primary cancers (log-rank p = 0.33). 36.4% patients died of neurologic death. Local failure at 1 and 2 years for the entire population was 10.4% and 14.3%, respectively. There was no difference in local failure between the primary cancers. Distant brain failure at 1, 2 and 5 years for the entire population was 20.6%, 27.7%, and 31.3%, respectively. On multivariate Cox Proportional Hazards analysis, age was the only predictor of overall survival (HR = 1.03, p = 0.01). Ovarian cancer patients had decreased risk of distant brain failure (HR = 0.17, p=0.005), whereas cervical cancer patients had an increased risk of distant brain failure (HR = 35.7, p = 0.001).
SRS represents a feasible treatment option for patients with brain metastases from gynecologic cancer. Younger age is a positive prognostic factor. Ovarian cancer patients have lower risk of distant brain failure.
Stereotactic radiosurgery (SRS) is a treatment modality for brain metastases that has improved quality of life outcomes over traditional whole brain radiotherapy (WBRT) in large randomized trials1,2. Patient selection is important with regards to triaging patients to either SRS or WBRT with important factors that include number of brain metastases, performance status and status of disease in the rest of the body. Recently, histology-specific management of brain metastases has become more commonplace because of how patterns of failure3 and survival4 differ based on the different natural histories of various brain metastasis histologies.
Because of their rarity, it has been unclear how brain metastases from gynecologic primary cancers will be integrated into such a histology-specific approach. Brain metastases from gynecologic primary cancers occur in only 1-3% of patients with a gynecologic malignancy5 and represent only 1% of the patients who have brain metastases6. With gynecologic cancers, it is uncertain as to whether brain metastases represent a predisposition of cancer towards brain or whether it represents a late manifestation of a high burden of systemic cancer.
The current study represents a single institution retrospective review of patients with brain metastases from gynecologic primaries treated with SRS. Survival outcomes, patterns of failure and predictive factors were assessed in order to characterize features of brain metastases that may be specific to certain cancer types.
This study was approved by the Wake Forest School of Medicine Institutional Review Board. Patients were identified from a radiation oncology departmental database of patients who have received stereotactic radiosurgery for brain metastases. Patients were included in the study if they had history of primary cancers from gynecologic malignancies including ovarian cancer, cervical cancer, or endometrial cancer. Patients were included in the study if they have received prior WBRT or prior surgical resection of a brain metastasis. Patient characteristics were subsequently acquired from the electronic medical records. Between 2000 and 2013, a total of 33 consecutive patients with brain metastases from gynecologic primaries were treated with SRS. Overall survival was determined via the social security death index. Patient characteristics are summarized in Table 1.
SRS was performed using the Gamma Knife Model B (2000-2005), C (2005-2009), and Perfexion (2009-13). Prior to SRS, the patient underwent high-resolution MRI of the brain with contrast. Treatment planning was performed using the Leksell GammaPlan Treatment Planning System (Elekta AB, Stockholm, Sweden). A median dose of 18 Gy (range 10-24 Gy) was generally prescribed to the 50% isodose line. Dose prescriptions were generally based on guidelines published by Shaw et al7. A total of 10 patients were treated with cavity-directed SRS in which the resection bed of a resected brain metastasis was treated with adjuvant SRS as previously described by Jensen et al8.
Patients were followed with serial MRI of the brain 4-8 weeks after initial SRS procedure and then every 3 months subsequently. Local failures were defined either pathologically or as a 25% increase in size of the lesion on follow-up MRI with increased perfusion on perfusion-weighted imaging. Distant brain failures were determined to be new metastases that developed outside of the prior SRS volume. Neurologic death was defined as previously described by Mcytre et al and included patients who died with either progressive neurologic dysfunction at time of death, and those who had severe neurologic dysfunction who died of intercurrent disease9.
Time to event data were calculated from the time of stereotactic radiosurgery. Survival curves were generated using the Kaplan Meier method. The log rank test was performed to assess differences in survival curves. Cox proportional hazards regression was performed to assess for factors that predicted for clinical outcomes. Variables such as age, number of initial brain metastases, margin dose, and RPA class were all analyzed as continuous variables (not dichotomized). KPS was dichotomized into >=70 vs. <70. All statistics were performed using R version 3.2.1 software (R Foundation for Statistical Computing, Vienna, Austria).
Overall survival at 1, 2 and 5 years for the entire population was 47.1% (CI: 32.1-69.2%), 21.7% (CI: 10.8-43.9%), and 14.5% (CI: 5.9-35.7%), respectively. Median overall survival for all patients was 12 months (range 1-77 months). Kaplan Meier plots for overall survival are shown in Figure 1. Overall survival for patients with ovarian cancer was 51.9% (CI: 34.9-77.4%) and 23.6% (CI: 11.0-50.5%) at 1 and 2 years, respectively, while for non-ovarian gynecologic cancer OS was 31.2% (CI: 10.2-95.5%) and 15.6% (CI: 2.6-92.7%) at 1 and 2 years, respectively (log-rank p = 0.33). Median overall survival for cervical cancer was 5 months. For endometrial cancer, the median overall survival was 6 months (range 1-25 months). Ovarian cancer patients had a median overall survival of 16 months (range 1-77 months).
For the entire population, 36.4% patients died of neurologic death. 32.0% (8/25), 50% (1/2), and 50.0% (3/6), with ovarian, cervical, and endometrial cancer died of neurologic death (versus those still alive or deceased of non-neurologic or indeterminate causes).
Local failure at 1 and 2 years for the entire population was 10.4% and 14.3%, respectively. Cumulative incidence plots for local failure are shown in Figure 2. Local failure at 1 and 2 years for ovarian cancer was 9.5% and 14.5%, respectively; for non-ovarian gynecologic primaries 1 year local failure was 12.5% while 2 year local failure was not able to be estimated due to small patient number and limited follow-up.
Distant brain failure at 1, 2 and 5 years for the entire population was 20.6%, 27.7%, and 31.3%, respectively. Cumulative incidence plots for distant brain failure are shown in Figure 3. Distant failure at 1, 2 and 5 years for ovarian cancer was 9.9%, 14.8%, and 19.7%, respectively; for non-ovarian gynecologic primaries DBF at 1 year was 50.0% while 2 and 5 year DBF were not able to be estimated due to small patient number and limited follow-up.
11 of 33 patients underwent a second SRS procedure to treat distant brain failure. Two of twenty-four patients who did not receive WBRT prior to SRS ultimately received WBRT. One of these patients experienced multifocal distant brain failure, and the other experienced a local failure.
Cox proportional hazards regression was performed to determine factors that predicted for survival, and competing risks regression was performed for local and distant failure. Results are summarized in Table 2. Age (HR = 1.03, p = 0.01) and RPA class (2.11, p=0.04) were the only predictors of overall survival. Ovarian cancer patients (HR = 0.17, p=0.005) and patients with worse RPA class (HR = 0.2, p=0.005) had decreased risk of distant brain failure, whereas cervical cancer patients had an increased risk of distant brain failure (HR = 35.7, p = 0.001). The only factor associated with local failure was number of initial brain metastases (HR = 1.25, p=0.03).
Three of 33 patients experienced CTCAE grade 3 toxicity and required hospitalization. A single patient experienced intractable nausea 24 hours after SRS and required admission for intravenous steroids and anti-emetics. She recovered soon after and was treated with a steroid taper. The second patient experienced mental status changes also 24 hours after SRS and required admission for intravenous steroids. She recovered after a 4 day hospitalization and was able to taper off of steroids. The third patient experienced radiation necrosis that required surgical intervention 1 year post-SRS.
Because of the relative rarity of brain metastases from gynecologic primaries, standard treatments are undefined. At the current time, these patients are managed with a generalized paradigm for patients with brain metastases, while histology-specific management is essentially ignored because of a paucity of data. Prior studies have shown that some histologies of brain metastases are more prone to local failure after SRS10, and others are more prone to distant brain failure6. Other histologies are predisposed to dying of neurologic death11. In the current series, it appeared that patients with ovarian primaries were less likely to develop new brain metastases after SRS. This may be because extracranial ovarian cancer is in general more responsive to chemotherapy, leading to fewer events of re-seeding the brain with metastases.
There are several prior series of patients with brain metastases from gynecologic primaries12-15, but few series assessing the use of SRS for this population16. The role of SRS in this population is relevant because few patients with gynecologic primaries were likely included on the available prospective trials, but the use of SRS can have significant quality of life implications. From the current series, it appears that SRS is tolerable and that several patients experienced extended survival. Age was the only prognostic factor for overall survival, though this is an important finding because patients with better prognosis are often better managed with an SRS approach as opposed to with WBRT. An overview of previous studies is outlined in Table 3 below, outlining the effectiveness of SRS treatment and its implications on the future consideration and treatment of brain metastases from primary gynecologic cancers. Cognitive toxicity from WBRT tends to progress over time without a plateau17 and there are few effective treatments once this toxicity manifests18.
There are several limitations to the current analysis. As a retrospective analysis, its findings are limited to hypothesis generation. Because of the rarity of metastases from gynecologic primaries, we were limited in our statistical power to detect anything but large differences. Ultimately, it would be helpful for practitioners to have brain metastases from gynecologic primaries be included in such prognostic indices as the graded prognostic assessment19.
Authors’ disclosure of potential conflicts of interest
The authors have nothing to disclose.
Conception and design: Michael Chan
Data collection: Michael Chan, Hannah Johnston
Data analysis and interpretation: Emory McTyre
Manuscript writing: Cristina K. Cramer, M.D., Glenn J. Lesser, M.D., Jimmy Ruiz, M.D., J. Daniel Bourland, Ph.D., Kounosuke Watabe, Ph.D., Hui-Wen Lo, Ph.D., Shadi Qasem, M.D., Adrian W. Laxton, M.D., Stephen B. Tatter, M.D., Ph.D., Michael D. Chan, M.D.
Final approval of manuscript: Hannah Johnston