Assessing the Association of Targeted Therapy and Intracranial Metastatic Disease

Key Points

Question  What is the association of targeted therapy with survival for patients with intracranial metastatic disease (IMD)?

Findings  In this cohort study of 26 676 patients with IMD, prolonged survival was observed in patients with IMD and ERBB2 (formerly HER2)–positive breast cancer, EGFR-positive lung and bronchus cancer, or BRAF-positive melanoma who received targeted therapy compared with those who did not. In patients with metastatic ERBB2-positive breast cancer or EGFR-positive lung and bronchus cancer, but not BRAF-positive melanoma, shorter survival was observed in patients with IMD vs those without.

Meaning  Targeted therapies seem to be associated with improved survival in patients with IMD and ERBB2-positive breast cancer, EGFR-positive lung and bronchus cancer, or BRAF-positive melanoma.

Importance  Targeted therapies have been hypothesized to prolong survival in the treatment of patients with intracranial metastatic disease (IMD) but, paradoxically, increase IMD incidence by improving systemic disease control and prolonging survival from the primary tumor. The real-world benefits of targeted therapy in treating patients with IMD are unclear, as clinical trials have excluded patients with IMD and lacked end points that report intracranial outcomes.

Objective  To assess the association of targeted therapy and IMD with patient survival.

Design, Setting, and Participants  This retrospective cohort study included all patients in Ontario, Canada, who received a diagnosis of IMD from April 2005 to January 2018 with primary diagnoses of breast cancer, lung or bronchus cancer, or melanoma and control patients who were matched by primary disease without IMD. The data were analyzed between March and October 2020.

Exposures  EGFR-, ERBB2 (HER2-), or BRAF-targeted therapy or IMD status.

Main Outcomes and Measures  Kaplan-Meier and multivariable Cox regression analyses were performed to compare overall survival (OS) between patient subcohorts divided by primary disease and stratified by targeted therapy receipt or IMD status.

Results  In this cohort of 26 676 patients with IMD and breast cancer, lung and bronchus cancer, or melanoma, 57% of patients were women, and the median age at IMD diagnosis was 64 years (interquartile range, 56-72 years). Post-IMD targeted therapy was associated with prolonged OS in patients with ERBB2-positive breast cancer (hazard ratio [HR], 0.41; 95% CI, 0.33-0.50), EGFR-positive lung cancer (HR, 0.28; 95% CI, 0.23-0.34), and BRAF-positive melanoma (HR, 0.20; 95% CI, 0.14-0.29) compared with those who did not receive post-IMD targeted therapy. The presence of IMD was associated with shorter OS in patients with metastatic ERBB2-positive breast cancer (HR, 1.80; 95% CI, 1.56-2.08) and metastatic EGFR-positive lung cancer (HR, 1.22; 95% CI, 1.08-1.39) but not metastatic BRAF-positive melanoma (HR, 1.11; 95% CI, 0.77-1.61) compared with those without IMD.

Conclusions and Relevance  The findings of this cohort study suggest an association between real-world use of targeted therapies and prolonged OS in patients with IMD in the setting of ERBB2-positive breast cancer, EGFR-positive lung cancer, and BRAF-positive melanoma. Including patients with IMD in clinical trials and using end points that interrogate IMD will be critical to determine the role of targeted therapies in treating patients with IMD.

Intracranial metastatic disease (IMD) is a common and feared complication of primary cancer.¹ Most brain metastases arise from lung cancer (20%-56% of IMD cases), breast cancer (5%-20%), and melanoma (7%-16%).^1,2 The risk of IMD is known to be disease subtype–specific. For example, patients with EBBR2 (formerly HER2)–positive breast cancers have a higher risk of IMD than patients with hormone receptor–positive/ERBB2-negative disease.³

The rising incidence of IMD over the past 2 decades has been attributed to increased detection with magnetic resonance imaging and, paradoxically, improved systemic disease control.⁴ Targeted therapies have contributed substantially to the latter and have become the standard of care for some cancer subtypes.^5-7 While the brain has historically been considered a sanctuary site for cancer progression given the poor penetration of systemic therapies across the blood-brain barrier, antibody therapies have shown intracranial access.^8-10 However, there is little evidence to interrogate the clinical association of targeted therapies with IMD, as clinical trials have typically excluded patients with IMD and lacked end points that report intracranial outcomes. We hypothesized that targeted therapy may be associated with prolonged survival in patients with IMD. To address this, we investigated the association of ERBB2-, EGFR-, and BRAF-targeted therapy with IMD survival in a real-world setting for patients with metastatic ERBB2-positive breast cancer, EGFR-positive lung and bronchus cancer, and BRAF-positive melanoma, respectively.

This retrospective cohort study included all patients with primary cancer and IMD in Ontario, Canada, who received a diagnosis between April 1, 2005, and January 24, 2018, with follow-up through January 24, 2019, and patients with the same primary cancers without IMD during the same period. The use of data in this project was authorized under section 45 of Ontario’s Personal Health Information Protection Act, which does not require review by a research ethics board

We compared (1) post-IMD overall survival (OS) by receipt of post-IMD targeted therapy and (2) the association of IMD status with OS in patients with stage 4 disease at diagnosis. Additional outcomes are presented in the eMethods and eResults in the Supplement.

We described and adjusted for variables, including sex, age, cancer-directed surgery, systemic therapy, radiotherapy, extracranial metastatic sites, pathological and clinical stage at diagnosis, and IMD synchronicity. Molecular disease subtype was ascertained by receipt of targeted therapy. For example, patients with breast cancer who received ERBB2-targeted therapy were assumed to have ERBB2-positive disease.

Kaplan-Meier analysis and Cox proportional hazards models were used to estimate outcomes. Adjusted estimates were produced for comparisons in which the stratifying variable was significant at univariable analysis. A P value of .05 was used to determine significance for all tests. Analyses were performed using SAS, version 7.15 (SAS Institute) and R, versions 3.1.2 and 4.0.0 (R Core Team).

The most common primary cancers among patients with IMD in Ontario were lung cancer (13 238 [57.5%]), breast cancer (2880 [12.5%]), and melanoma (1213 [5.26%]). Records for 22 676 patients with IMD were retrieved.

Baseline characteristics in patients with IMD and ERBB2-positive breast cancer stratified by receipt of post-IMD ERBB2-targeted therapy are described in eTable 1 in the Supplement. Patients who received post-IMD targeted therapy were significantly more likely to receive subsequent therapies and to have synchronous IMD. Similar baseline characteristics for patients with EGFR-positive lung and bronchus cancer or BRAF-positive melanoma are shown in eTables 2 and 3 in the Supplement. Bivariate tables for additional end points are available in eTables 4, 5, 6, 7, 8, and 9 in the Supplement.

Patients with EBBR2-positive breast cancer, EGFR-positive lung and bronchus cancer, and BRAF-positive melanoma were followed for a median of 7.26 (interquartile range [IQR], 2.17-19.7) months, 4.90 (IQR, 0.95-18.3) months, and 4.27 (IQR, 0.87-10.2) months after IMD diagnosis, respectively (eTable 10 in the Supplement). Cox regression among patients with IMD and EBBR2-positive breast cancer showed the following variables to be prognostic for shorter OS in multivariable analysis: older age at IMD diagnosis, pre-IMD nontargeted systemic therapy, and 2 or more extracranial metastatic sites; conversely, the following variables were associated with prolonged OS: post-IMD nontargeted systemic therapy, brain-targeted radiotherapy, and post-IMD EBBR2-targeted therapy (eTable 11 in the Supplement). Similar associations were shown for patients with IMD and EGFR-positive lung and bronchus cancer (eTable 12 in the Supplement) or BRAF-positive melanoma (eTable 13 in the Supplement).

Receipt of post-IMD ERBB2-targeted therapy was associated with prolonged OS (adjusted hazard ratio [HR], 0.41; 95% CI, 0.33-0.5; Figure 1A). A subgroup analysis showed that post-IMD OS was prolonged in patients who received a post-IMD ERBB2-targeted tyrosine kinase inhibitor (adjusted HR, 0.46; 95% CI, 0.33-0.64) or antibody-based ERBB2-targeted therapy (adjusted HR, 0.40; 95% CI, 0.32-0.49), compared with no post-IMD ERBB2-targeted therapy. Among patients with lung and bronchus cancer, post-IMD EGFR-targeted therapy was associated with prolonged OS (adjusted HR, 0.28; 95% CI, 0.23-0.34; eTable 12 in the Supplement; Figure 1B). Among patients with melanoma, post-IMD BRAF-targeted therapy was also associated with prolonged OS (adjusted HR, 0.20; 95% CI, 0.14-0.29; eTable 13 in the Supplement; Figure 1C).

In patients with stage 4 ERBB2-positive breast cancer, the presence of IMD conferred a worse OS from primary diagnosis (adjusted HR, 1.8; 95% CI, 1.56-2.08; Figure 2A; eTable 14 in the Supplement). Shorter OS from primary diagnosis was also observed in patients with stage 4 EGFR-positive lung and bronchus cancer with vs without IMD (adjusted HR, 1.22; 95% CI, 1.08-1.39; Figure 2B; eTable 15 in the Supplement). The presence of IMD was not associated with OS in patients with stage 4 BRAF-positive melanoma (HR, 1.11; 95% CI, 0.77-1.61; eTable 16 in the Supplement). Results for additional analyses are available in the eFigures 1-7 and eTables 17-32 in the Supplement.

In this study’s real-world cohort, prolonged post-IMD OS was observed for patients with ERBB2-positive breast cancer, EGFR-positive lung and bronchus cancer, and BRAF-positive melanoma who respectively received post-IMD ERBB2-, EGFR-, and BRAF-targeted therapy compared with those who did not. This finding is considerable given the extent of disease in these patients and current unfavorable estimates of the benefits of genome-driven therapy.¹¹ Diagnosis of IMD was associated with shorter OS in patients with stage 4 ERBB2-positive breast cancer and EGFR-positive lung and bronchus cancer, suggesting that IMD may be associated with mortality in these patients. Among patients with stage 4 BRAF-positive melanoma, IMD diagnosis had no association with OS, which may indicate that IMD is not a key driver of mortality for this patient population. This finding conflicts with evidence supporting American Joint Committee on Cancer staging (eighth edition), which introduced a new category (M1day) to reflect the negative prognostic effect of IMD in patients with melanoma.¹²

This study has several limitations. First, molecular subtype was ascertained by receipt of targeted therapy. This proxy definition for disease subtype may have resulted in the exclusion of patients with synchronous IMD. Second, several confounders were present (for example, post-IMD nontargeted systemic therapy was reported more frequently among patients who were receiving post-IMD targeted therapy), although adjusted analyses were performed to address these differences. Third, key data were missing that would have facilitated cohort comparisons, including tumor response, causes of death, Karnofsky performance status, and intracranial tumor burden.¹³

This study finds a positive association between post-IMD targeted therapy and prolonged OS in patients with ERBB2-positive breast cancer, EGFR-positive lung and bronchus cancer, and BRAF-positive melanoma. Diagnosis of IMD was associated with shorter OS in patients with metastatic ERBB2-positive breast cancer and metastatic EGFR-positive lung and bronchus cancer, which supports efforts to assess intracranial screening for IMD in these patients. Select patients with melanoma at high risk for IMD may still benefit from screening. Future studies are needed to clarify the association of targeted therapy with IMD incidence and the role for targeted therapy in IMD management.

Accepted for Publication: March 30, 2021.

Published Online: June 3, 2021. doi:10.1001/jamaoncol.2021.1600

Corresponding Author: Sunit Das, MD, PhD, St Michael’s Hospital, 30 Bond St, Toronto, ON M5B 1W8, Canada (sunit.das@utoronto.ca).

Author Contributions: Mr Erickson had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Erickson, Das.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Erickson, Das.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Erickson, Habbous.

Obtained funding: Das.

Administrative, technical, or material support: Lofters, Das.

Supervision: Habbous, Lofters, Das.

Conflict of Interest Disclosures: Dr Lofters serves as the Provincial Primary Care Lead of Cancer Screening for Ontario Health (Cancer Care Ontario) outside the submitted work. Dr Jerzak reported personal fees from Amgen, AstraZeneca, Apo Biologix, Eli Lilly, Eisai, Exact Sciences, Knight Therapeutics, Merck, Myriad Genetics Inc, Pfizer, Roche, Novartis, and Purdue Pharma as well as grants from AstraZeneca and Eli Lilly outside the submitted work. Dr Das reported being a member of the advisory board of the Subcortical Surgery Group; nonfinancial support from Xpan Medical and Theralase; personal fees from the Congress of Neurological Surgeons, American Association of Neurological Surgeons, Integra, and Society for NeuroOncology; grants from Alkermes, Medicenna, and AbbVie; and serving as the Provincial Lead for CNS Tumours for Ontario Health (Cancer Care Ontario) outside the submitted work. No other disclosures were reported.

Funding/Support: Dr Das is supported by an Early Researcher Award from the Province of Ontario. Mr Erickson is supported by the Graduate Diploma in Health Research program at the University of Toronto. Dr Lofters is supported by a Canadian Institutes of Health Research New Investigator Award, as chair in Implementation Science at the Peter Gilgan Centre for Women’s Cancers at Women’s College Hospital in partnership with the Canadian Cancer Society, and as a clinician scientist by the University of Toronto Department of Family and Community Medicine. This study was supported by ICES, which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care. Portions of this article are based on data and information provided by Cancer Care Ontario (CCO) and the Canadian Institute for Health Information (CIHI).

Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The opinions, results, view, and conclusions reported in this paper are those of the authors and do not necessarily reflect those of CCO or CIHI. No endorsement by ICES (previously the Institute for Clinical Evaluative Sciences), the Ontario Ministry of Health and Long-Term Care, CCO, or CIHI is intended or should be inferred.

Additional Contributions: We thank Alex Marchand-Austin, MSc, for cutting the patient cohorts from the data holdings at ICES, where he was a salaried employee at ICES, the facility from which data were obtained. His contributions fell within the remit of that employment. He was not directly compensated by any member of our team. We also thank Lillian Su, MD, Stanford University, for her guidance and mentorship. She was not compensated for her contributions.