Abstract
Background
Breast cancer (BC) is the second most common cancer that metastasizes to the brain. Particularly up to half of patients with human epidermal growth factor receptor 2 (HER2)-positive (HER2+) metastatic breast cancer (mBC) may develop brain metastases over the course of the disease. Nevertheless, little is known about the prevalence and the outcome of brain and leptomeningeal metastases (BLMM) in HER2-low BC. We compared the cumulative incidence of BLMM and associated outcomes among patients with HER2-low, HER2-negative (HER2−) and HER2+ mBC.
Patients and methods
This cohort study was conducted from the Epidemiological Strategy and Medical Economics (ESME) mBC database and included patients treated for mBC between 2012 and 2020 across 18 French comprehensive cancer centers and with known HER2 and hormone receptor (HR) status. The cumulative incidence of BLMM after metastatic diagnosis was estimated using a competing risk methodology with death defined as a competing event.
Results
19 585 patients were included with 6118 (31.2%), 9943 (50.8%) and 3524 (18.0%) being HER2-low, HER2− and HER2+ mBC, respectively. After a median follow-up of 48.6 months [95% confidence interval (CI) 47.7-49.3 months], BLMM were reported in 4727 patients: 1192 (25.2%) were diagnosed with BLMM at first metastatic diagnosis and 3535 (74.8%) after metastatic diagnosis. Multivariable analysis adjusted for age, histological grade, metastases-free interval and HR status showed that the risk of BLMM at metastatic diagnosis was similar in patients with HER2− compared to HER2-low mBC [odds ratio (OR) (95% CI) 1.00 (0.86-1.17)] and higher in those with HER2+ compared to HER2-low [OR (95% CI) 2.23 (1.87-2.66)]. Similar results were found after metastatic diagnosis; the risk of BLMM was similar in HER2− compared to HER2-low [subdistribution hazard ratio (sHR) (95% CI) 1.07 (0.98-1.16)] and higher in the HER2+ group [sHR (95% CI) 1.56 (1.41-1.73)].
Conclusions
The prevalence and evolution of BLMM in HER2-low mBC are similar to those in patients with HER2− tumors. In contrast to patients with HER2+ mBC, the prognosis of BLMM remains dismal in this population.
Key words: breast cancer, metastasis, brain metastases, HER2-positive, HER2-low
Highlights
-
•
Risk of BLMM was similar in HER2− patients compared to HER2-low mBC and higher in those with HER2+ compared to HER2-low.
-
•
The risk of BLMM was similar in HER2− compared to HER2-low and higher in the HER2+ group.
-
•
Median overall survival was 6.6 (5.8-7.4), 5.5 (5.0-5.9) and 24.4 (21.8-26.9) m in HER-low, HER2− and HER2+ mBC patients.
-
•
Prognosis of BLMM remains dismal in HER2-low mBC patients.
-
•
Development of new treatment is eagerly awaited for patients with HER2-low or HER2− mBC.
Introduction
Breast cancer (BC) is the second most common cancer that metastasizes to the brain, and particularly the human epidermal growth factor receptor 2-positive (HER2+) subtype is characterized by a high rate of central nervous system involvement.1, 2, 3 It has been reported that up to half of patients with HER2+ metastatic breast cancer (mBC) may develop brain metastases (BM) over the course of the disease.1,4 Nevertheless, little is known about the prevalence and the outcome of brain and leptomeningeal metastases (BLMM) in HER2-low BC. Indeed, very few studies have specifically studied this population. A recent Turkish retrospective analysis showed a higher rate of BM in HER2-low patients (8.5%) compared to HER2-negative (HER2−) (5.1%), which is still lower than the HER2+ population (10.1%).5 In addition, HER2-low disease was associated with a worse prognosis as compared to HER2− disease. In the phase II DAISY trial that evaluated the efficacy of trastuzumab deruxtecan (T-DXd) in three cohorts of patients, with HER2+, HER2-low and HER2− BC, the proportion of patients with BM at baseline in the HER2-low cohort was higher than the HER2− and lower than the HER2+ cohort.6,7
A growing number of new drugs, mainly antibody–drug conjugates (ADCs), vaccines and bispecific antibodies, are currently under investigation in patients with HER2-low BC,8 with a wide spectrum of action and different potential to cross the altered blood–brain barrier, better referred to as the blood–tumor barrier upon development of BM.9 Recently, in Destiny-Breast04,10 T-DXd showed significantly increased survival outcomes in patients with HER2-low mBC previously treated with one or two lines of chemotherapy. Furthermore, the phase II, multicohort DEBBRAH trial, presented at SABCS 2023, has shown a significant activity of T-DXd also in patients with BLMM and HER2-low BC (cohort 5), with a median progression-free survival (PFS) of 8.9 months and a median overall survival (OS) of 13.3 months.11
Therefore, to develop further effective drugs and treatment strategies in HER2-low BC, it is first essential to determine the actual prevalence of brain and leptomeningeal involvement in this specific subgroup. Leveraging the Epidemiological Strategy and Medical Economics (ESME) mBC data platform, a large multicenter French real-life database using a retrospective data collection process, we aimed to assess the cumulative incidence of BLMM over the clinical evolution of mBC in patients with HER2-low tumors as compared to those with HER2− and HER2+ mBC, and to determine the associated survival outcomes.
Methods
Study overview
This cohort study was conducted on the ESME mBC platform. ESME is population-based registry collecting individual data on all consecutive patients treated for an mBC between January 2008 and December 2020 across 18 French comprehensive cancer centers. Data were collected until 5 February 2022. Main eligibility criteria were: patients aged 18 years or older, presenting a diagnosis of mBC, who initiated their treatment between January 2012 and December 2020. Patient demographic data, tumor characteristics, outcomes and treatment strategies were collected. This study was conducted according to the guidelines of the Declaration of Helsinki, and the reporting followed the European Society for Medical Oncology Guidance for Reporting Oncology real-World evidence (GROW).12
The ESME research program is managed by Unicancer in accordance with current best practice guidelines and rules. It is supervised by an independent steering committee. The ESME mBC database was authorized by the French data protection authority (initial authorization no. DE-2013-117 and subsequent amendment in accordance with GDPR). An independent scientific committee specifically approved the present work. The ESME mBC cohort is registered on www.clinicaltrials.gov under the number NCT03275311. All data were obtained retrospectively. All patients have approved the use of their data.
Study population
HER2 status was defined at the time of mBC as the first HER2 status available within 6 months from metastatic diagnosis if patients had a biopsy of one of the metastatic sites, otherwise HER2 status was defined on the primary tumor as the first HER2 status available from initial diagnosis. If two or more HER2 status were available at the same date, the positive HER2 status was considered. Based on the expression of HER2, patients were categorized into three groups: the HER2-positive group (HER2+), which showed HER2 overexpression in an immunohistochemistry (IHC) test (score 3+) or gene amplification in an in situ hybridization (ISH) assay; the HER2-low group, characterized by HER2 IHC scores of 1+ or 2+ along with negative ISH results; and the HER2-negative group (HER2−), as indicated by an IHC score of 0.
Study endpoints
The incidence of BLMM was the primary endpoint and was analyzed over two periods: at metastatic diagnosis and during the follow-up (after metastatic diagnosis). At metastatic diagnosis, the incidence of BLMM was defined as the presence of BLMM at first metastatic diagnosis or within the following month (binary endpoint). After metastatic diagnosis, the incidence of BLMM was analyzed using a competing risk methodology as a time-to-event endpoint defined as the time from metastatic diagnosis to first BLMM (TTB) considering death as a competing event. For this analysis, patients with BLMM at metastatic diagnosis (or within the following month) and patients alive with <1 month of follow-up were excluded from the competing risk analysis. Additional analyses were carried out by treatment line in the metastatic setting. The time to BLMM during the first, second and third line were defined as the time from treatment initiation to first BLMM, considering progression outside BLMM and death as competing events, in patients who received at least one, two or three treatment line(s), respectively. In patients who developed a BLMM, OS after BLMM was defined as the time from first BLMM to death and PFS after BLMM was defined as the time from first treatment initiation after BLMM to progression or death. Patients without event were censored at their last follow-up. The metastasis-free interval (MFI) was defined as the time between the initial diagnosis of BC and the occurrence of the first metastases (cerebral or extra-cerebral). First-line treatment was defined as the primary antitumor regimen administered for metastatic disease. In the event of a treatment modification, the subsequent therapy was termed second-line, and so forth. Each treatment line persisted until the onset of disease progression (either clinically or radiologically), unacceptable toxicity, death or patients’ decision of treatment discontinuation. The ESME database was not designed to specifically study data from BM alone versus leptomeningeal metastases alone versus both. The only comparison possible was brain versus CNS/meningeal versus other.
Statistical analysis
Data were summarized by median and range (min-max) for quantitative variables and by frequencies and percentages for qualitative variables on the overall population and by HER2 group. The presence of BLMM at first metastatic diagnosis was presented by frequency and percentage in each group (HER2-low, HER2-0 and HER2+). Univariable analysis was carried out using the Kruskal–Wallis test for quantitative variables and the chi-squared or Fisher’s exact test for qualitative variables. Logistic regression model was used to carry out multivariable analysis. The odds ratios (ORs) were estimated with their 95% confidence interval (95% CI) taking as reference the HER2-low group. Survival rates of OS and PFS after BLMM were estimated using the Kaplan–Meier method. Univariable and multivariable analyses were carried out using the log-rank test and the Cox proportional hazards model, respectively. The hazard ratios (HRs) with their 95% CI were estimated using the Cox proportional hazards model taking as reference the HER2-low group. Cumulative incidences of each first event type (BLMM, death before BLMM and extracranial progression before BLMM by treatment line) were estimated using competing risks methodology. Univariable and multivariable analyses were carried out using the Fine and Gray model. The subdistribution HRs (sHR) were estimated with their 95% CI taking as reference the HER2-low group. Subgroups analyses were carried out according to the HR status (HR+ and HR−).
All statistical tests were two-sided and differences were considered significant at the 5% level. Statistical analyses were carried out using Stata statistical software version 16 (College Station, TX, StataCorp LLC).
Results
Characteristics of the population and treatments
Among the 30 459 patients from the ESME database, 19 585 with known HER2 and HR status and treated for mBC between 2012 and 2020 were included in this study (Supplementary Figure S1, available at https://doi.org/10.1016/j.esmoop.2024.103447), of whom 6118 (31.2%), 9943 (50.8%) and 3524 (18.0%) had HER2-low, HER2− and HER2+ mBC, respectively. Median age at metastatic diagnosis was 61 years (range 19-103 years) for the overall population, 62 years (22-103 years), 61 years (22-97 years) and 57 years (19-97 years) for HER2-low, HER2− and HER2+ subgroups, respectively, and median MFI was 31.6 months (range −2.8-545.9 months) for HER2-low, 32.7 months (−3.0-596.1 months) for HER2− and 3.3 months (−3.0-534.4 months) for HER2+. HR-positive disease accounts for 85.5%, 75.6% and 61.5% of HER2-low, HER2− and HER2+ subgroups, respectively. In the HER2-low subgroup, 27.1% of the patients had a grade III tumor at initial diagnosis compared to 32.6% and 45.9% in the HER2− and HER2+ subgroups, respectively. In our cohort, HER2 status on metastatic sites and primary tumor were different in 26.5% of cases: 17.1% with increased expression on metastatic sites (13.3% changed from HER2− to HER2-low, 2.1% from HER2− to HER2+ and 1.7% from HER2-low to HER2+) and 9.4% with decreased expression (1.7% changed from HER2+ to HER2-low, 1.5% from HER2+ to HER2− and 6.2% from HER2-low to HER2−).
In the overall population, 1192 patients (6.1%) had a BLMM at first metastatic diagnosis: 278 (4.5%), 540 (5.4%) and 374 (10.6%) in the HER2-low, HER2− and HER2+ groups, respectively. A multivariable logistic regression adjusted for age, histological grade, MFI and HR status showed that the risk of BLMM at first metastatic diagnosis was similar in patients with HER2− compared to those with HER2-low mBC [OR (95% CI) 1.00 (0.86-1.17)] and higher in HER2+ compared to the HER2-low group [OR (95% CI) 2.23 (1.87-2.66)] (Supplementary Table S1, available at https://doi.org/10.1016/j.esmoop.2024.103447). After a median follow-up of 48.6 months (95% CI 47.7-49.3 months), BLMM were reported in 4727 (24.1%) patients: the 1192 mentioned already (25.2%, BLMM at first metastatic diagnosis) and 3535 (74.8%) that developed BLMM later on, during the further disease evolution (after metastatic diagnosis). Clinical details of the 4727 patients with BLMM are shown on Table 1. When we restricted the analysis to the population with BLMM, the incidence of BLMM at diagnosis was equivalent across the three subgroups. Indeed, BLMM were present at first metastatic diagnosis in 22.7%, 24.0% and 29.7% in HER2-low, HER2− and HER2+ subgroups, respectively. Median age at first BLMM diagnosis was 60 years (range 27-88 years), 58 years (25-95 years) and 55 years (20-91 years) for patients with HER2-low, HER2− and HER2+ mBC, respectively. For patients who developed BLMM after metastatic diagnosis (n = 3535), the interval from mBC to BLMM was 20.9 months (1.0-105.2 months), 16.8 months (1.0-104.0 months) and 18.0 months (1.0-93.5 months) in the HER2-low, HER2− and HER2+ subgroups, respectively.
Table 1.
Clinical characteristics of patients with BLMM and systemic treatments administrated after diagnosis of BLMM (n = 4727)
| Total |
HER2− |
HER2-low |
HER2+ |
|
|---|---|---|---|---|
| (N = 4727) | (n = 2247) | (n = 1222) | (n = 1258) | |
| Age at dBLMM | ||||
| Median (range) | 58 (20-95) | 58 (25-95) | 60 (27-88) | 55 (20-91) |
| Histological grade of initial diagnosis, n (%) | ||||
| Grade I | 307 (6.9%) | 178 (8.4%) | 85 (7.3%) | 44 (3.7%) |
| Grade II | 2299 (51.4%) | 1058 (49.6%) | 668 (57.6%) | 573 (48.4%) |
| Grade III | 1869 (41.7%) | 896 (42.0%) | 406 (35.0%) | 567 (47.9%) |
| Missing | 252 | 115 | 63 | 74 |
| Number of metastatic sites at metastatic diagnosis | ||||
| Median (range) | 2 (1.0-11.0) | 2 (1.0-10.0) | 2 (1.0-11.0) | 2 (1.0-9.0) |
| Metastasis-free interval (MFI) | ||||
| Median (range) | 24.7 (−2.8 to 591.7) | 27.2 (−2.8 to 591.7) | 30.3 (−1.9 to 482.0) | 14.7 (−2.7 to 400.5) |
| Hormone receptor status at dBLMM, n (%) | ||||
| Negative | 1746 (36.9%) | 868 (38.6%) | 297 (24.3%) | 581 (46.2%) |
| Positive | 2981 (63.1%) | 1379 (61.4%) | 925 (75.7%) | 677 (53.8%) |
| Timing of BLMM, n (%) | ||||
| At mBC diagnosis | 1192 (25.2%) | 540 (24.0%) | 278 (22.7%) | 374 (29.7%) |
| After mBC diagnosis | 3535 (74.8%) | 1707 (76.0%) | 944 (77.3%) | 884 (70.3%) |
| Interval from mBC to BLMM in months for patients with BLMM after mBC diagnosis (n = 3535) | ||||
| Median (range) | 18.0 (1.0-105.2) | 16.8 (1.0-104.0) | 20.9 (1.0-105.2) | 18.0 (1.0-93.5) |
| Number of line before dBLMM | ||||
| Median (range) | 1.0 (0.0-16.0) | 1.0 (0.0-11.0) | 2.0 (0.0-16.0) | 1.0 (0.0-12.0) |
| At least one treatment line administrated following dBLMM, n (%) | 3577 (75.7%) | 1723 (76.7%) | 908 (74.3%) | 946 (75.2%) |
| Number of lines for patients treated after dBLMM (n = 3577), n (%) | ||||
| Median (range) | 2 (1-12) | 2 (1-11) | 2 (1-12) | 2 (1-11) |
| 1rst line | 1329 (37.1%) | 615 (35.6%) | 309 (34.0%) | 405 (42.8%) |
| 2nd line | 926 (25.9%) | 399 (23.2%) | 202 (22.2%) | 325 (34.3%) |
| 3rd line | 508 (14.2%) | 265 (15.4%) | 124 (13.7%) | 119 (12.6%) |
| 4th line or more | 814 (22.8%) | 444 (25.8%) | 273 (30.1%) | 97 (10.3%) |
| Type of treatment administrated for patients treated after dBLMM (n = 3577), n (%) | ||||
| Endocrine therapy | 1026 (28.7%) | 492 (28.6%) | 307 (33.8%) | 227 (24.0%) |
| Chemotherapy | 2747 (76.8%) | 1447 (84.0%) | 743 (81.8%) | 557 (58.9%) |
| Targeted therapy | 1064 (29.7%) | 269 (15.6%) | 148 (16.3%) | 647 (68.4%) |
| Antibody–drug conjugate | 234 (6.5%) | 7 (0.4%) | 6 (0.7%) | 221 (23.4%) |
| Immunotherapy | 32 (0.9%) | 22 (1.3%) | 9 (1.0%) | 1 (0.1%) |
BLMM, brain and leptomeningeal metastases; dBLMM, first BLMM diagnosis; HER2, human epidermal growth factor receptor 2; mBC, metastatic breast cancer.
Risk of developing BLMM according to HER2 status
A competing risk methodology was used to evaluate the incidence of BLMM during the disease evolution (after metastatic diagnosis). Thus, patients with BLMM at metastatic diagnosis and patients with <1 month of follow-up were excluded from this analysis. A total of 18 365 patients were included in the competing risks population, of whom 5828 (31.7%), 9390 (51.1%) and 3147 (17.2%) had HER2-low, HER2− and HER2+ mBC, respectively (Supplementary Figure S1, available at https://doi.org/10.1016/j.esmoop.2024.103447). The risk of developing BLMM after the first metastatic diagnosis was slightly higher in HER2− compared to HER2-low [sHR (95% CI) 1.15 (1.06-1.24)], as well as the risk of death before BLMM [sHR (95% CI) 1.09 (1.04-1.14)], while patients with HER2+ mBC compared to those with HER2-low had a higher risk of developing BLMM [sHR (95% CI) 1.95 (1.78-2.14)], but a lower risk of death before BLMM [sHR (95% CI) 0.46 (0.42-0.50)] (Figure 1). In multivariable analysis including HER2 and HR status, age, histological grade, MFI and number and type of metastatic sites at metastatic diagnosis, the risk of BLMM was similar in the HER2− versus the HER2-low subgroup [sHR (95% CI) 1.07 (0.98-1.16)] and higher in the HER2+ versus the HER2-low subgroup [sHR (95% CI) 1.56 (1.41-1.73)] (Supplementary Table S2, available at https://doi.org/10.1016/j.esmoop.2024.103447). In the subgroup of HER2-low patients, the cumulative incidence of BLMM at 5 years was 17.6% (95% CI 16.3% to 18.8%) in HR+ and 27.6% (24.2% to 31.0%) in HR− (Supplementary Figure S2, available at https://doi.org/10.1016/j.esmoop.2024.103447).
Figure 1.
Cumulative incidence of BLMM and death before BLMM by HER2 status estimated using the competing risks method in the competing risks population (n = 18 365). (A) Cumulative incidence of BLMM (event of interest). (B) Cumulative incidence of death before BLMM (competing event).
BLMM, brain and leptomeningeal metastases; CI, confidence interval; HER2, human epidermal growth factor receptor 2.
An additional analysis was carried out by treatment line. The risk of BLMM was also higher and the risk of progression outside BLMM or death before BLMM was lower in HER2+ compared to HER2-low (Table 2). HER2− status was associated with a higher risk of BLMM compared to HER2-low only during the first and second lines of treatment.
Table 2.
Univariable and multivariable analyses of competing risks TTB by treatment line in the competing risks population (n = 18 365)
| BLMM |
Progression outside BLMM and death before BLMM |
|||||
|---|---|---|---|---|---|---|
| Cumulative incidence at 12 months | sHR (95% CI) | Adjusted sHR (95% CI) | Cumulative incidence at 12 months | sHR (95% CI) | Adjusted sHR (95% CI) | |
| During first line | ||||||
| HER2-low (n = 5787) | 2.7 (2.3-3.1) | 1.00 | 1.00 | 48.7 (47.4-50.0) | 1.00 | 1.00 |
| HER2− (n = 9299) | 3.5 (3.1-3.9) | 1.24 (1.05-1.46) | 1.12 (0.94-1.32) | 53.2 (52.1-54.2) | 1.06 (1.02-1.10) | 1.02 (0.98-1.06) |
| HER2+ (n = 3109) | 6.3 (5.4-7.1) | 3.87 (3.29-4.55) | 2.70 (2.26-3.23) | 35.7 (34.1-37.4) | 0.54 (0.51-0.57) | 0.54 (0.51-0.57) |
| During second line | ||||||
| HER2-low (n = 3622) | 3.7 (3.1-4.4) | 1.00 | 1.00 | 73.6 (72.1-75.0) | 1.00 | 1.00 |
| HER2− (n = 5779) | 4.8 (4.2-5.4) | 1.32 (1.08-1.60) | 1.25 (1.02-1.53) | 73.3 (72.1-74.4) | 1.01 (0.97-1.05) | 1.00 (0.95-1.04) |
| HER2+ (n = 1217) | 8.5 (7.0-10.2) | 2.91 (2.31-3.66) | 2.41 (1.88-3.10) | 58.6 (55.7-61.4) | 0.64 (0.60-0.69) | 0.60 (0.55-0.65) |
| During third line | ||||||
| HER2-low (n = 2421) | 5.0 (4.1-5.9) | 1.00 | 1.00 | 80.3 (78.6-81.9) | 1.00 | 1.00 |
| HER2− (n = 3775) | 5.1 (4.4-5.8) | 0.99 (0.79-1.24) | 0.94 (0.75-1.19) | 80.1 (78.8-81.4) | 1.04 (0.98-1.09) | 1.02 (0.97-1.08) |
| HER2+ (n = 647) | 6.2 (4.5-8.3) | 1.40 (1.00-1.96) | 1.19 (0.82-1.73) | 74.9 (71.2-78.2) | 0.84 (0.76-0.92) | 0.76 (0.69-0.85) |
Patients with BLMM at treatment initiation and patients with <1 month of follow-up after treatment initiation were excluded from each analysis: 170 for first line, 7747 for second line and 11 522 for third line.
Adjusted sHR were calculated from the multivariable Fine and Gray model including HER2 status, age, histological grade, number and type of metastatic sites, MFI and hormone receptor status.
BLMM, brain and leptomeningeal metastases; CI, confidence interval; HER2, human epidermal growth factor receptor 2; sHR, subdistribution hazard ratios; TTB, time to BLMM defined as the time between treatment initiation and BLMM (event of interest) or progression outside BLMM and death before BLMM (competing event) or last contact (censored data).
Outcome of patients after BLMM according to HER2 status
The median OS (mOS) after BLMM diagnosis was 6.6 months (95% CI 5.8-7.4 months), 5.5 months (95% CI 5.0-5.9 months) and 24.4 months (95% CI 21.8-26.9 months) in the HER2-low, HER2− and HER2+ subgroups, respectively (Figure 2). Multivariable analyses (including HER2 status, HR status, histological grade at initial diagnosis, number of metastatic sites at metastatic diagnosis, age at BM diagnosis, MFI and interval between metastatic diagnosis and BM) showed that compared to patients with HER2-low mBC, the risk of death after BLMM was lower in those with HER2+ mBC [HR (95% CI) 0.39 (0.35-0.43)] and higher in those with HER2− mBC [HR (95% CI) 1.08 (1.00-1.17)] (Table 3). In the HER2-low group, mOS was 5.3 months (4.2-6.8 months) and 7.0 months (5.9-7.9 months) in HR− and HR+ patients, respectively (Supplementary Figure S3, available at https://doi.org/10.1016/j.esmoop.2024.103447). The mPFS after diagnosis of BLMM was 3.3 months (95% CI 3.0-3.6 months), 3.1 months (95% CI 3.0-3.3 months) and 6.2 months (95% CI 5.7-6.9 months) in the HER2-low, HER2− and HER2+ subgroups, respectively (Figure 2), with differences also between patients with HR+/HER2-low and HR-/HER2-low mBC: 2.7 months (95% CI 2.3-3.1 months) and 3.6 months (95% CI 3.2-3.9 months), respectively (Supplementary Figure S3, available at https://doi.org/10.1016/j.esmoop.2024.103447). Multivariable analyses showed that compared to patients with HER2-low mBC, the risk of progression after BLMM was lower in those with HER2+ mBC [HR (95% CI) 0.56 (0.51-0.62)] and similar in those with HER2− mBC [HR (95% CI) 1.01 (0.93-1.11)] (Table 3).
Figure 2.
Progression-free survival and overall survival after dBLMM by HER2 status. (A) Overall survival and (B) progression-free survival after BLMM by HER2 status.
CI, confidence interval; dBLMM, diagnosis of brain and leptomeningeal metastases; HER2, human epidermal growth factor receptor 2; HR, hazard ratio; OS, overall survival; PFS, progression-free survival.
Table 3.
Multivariable analysis of overall survival and progression-free survival after BLMM in patients with BLMM (n = 4727)
| OS (n = 4474) |
PFS (n = 3385) |
|||
|---|---|---|---|---|
| HR (95% CI) | P value | HR (95% CI) | P value | |
| HER2 status | ||||
| HER2-low | 1.00 | 1.00 | ||
| HER2− | 1.08 (1.00-1.17) | 0.0485 | 1.01 (0.93-1.11) | 0.7443 |
| HER2+ | 0.39 (0.35-0.43) | <0.0001 | 0.56 (0.51-0.62) | <0.0001 |
| Age at BLMM diagnosis, years | ||||
| <50 | 1.00 | 1.00 | ||
| 50-70 | 1.15 (1.06-1.24) | 0.0006 | 0.95 (0.88-1.03) | 0.2393 |
| >70 | 1.64 (1.47-1.82) | <0.0001 | 0.99 (0.88-1.12) | 0.9206 |
| Histological grade | ||||
| Grade I/II | 1.00 | 1.00 | ||
| Grade III | 1.06 (0.98-1.14) | 0.1300 | 1.10 (1.02-1.18) | 0.0146 |
| Number of metastatic sites | ||||
| <3 sites | 1.00 | 1.00 | ||
| ≥ 3 sites | 1.16 (1.07-1.24) | 0.0001 | 1.21 (1.12-1.31) | <0.0001 |
| Metastasis-free interval (MFI), months | ||||
| <6 | 1.00 | 1.00 | ||
| 6-24 | 1.41 (1.27-1.57) | <0.0001 | 1.44 (1.29-1.61) | <0.0001 |
| >24 | 0.97 (0.90-1.05) | 0.5061 | 1.00 (0.92-1.09) | 0.9078 |
| Hormone receptor status | ||||
| Negative | 1.00 | 1.00 | ||
| Positive | 0.73 (0.67-0.78) | <0.0001 | 0.82 (0.76-0.89) | <0.0001 |
| Interval from mBC to BLMM, months | ||||
| 0-1 | 1.00 | 1.00 | ||
| 1-12 | 2.12 (1.92-2.33) | <0.0001 | 1.77 (1.60-1.94) | <0.0001 |
| >12 | 2.30 (2.10-2.52) | <0.0001 | 1.79 (1.64-1.96) | <0.0001 |
Of the 4727 patients who developed a BLMM, 253 patients were excluded from the multivariable analysis for missing data on histological grade and MFI. For the PFS analysis, 1089 patients without treatment line after dBLMM were also excluded.
Bold = P value <0.05, significative value.
BLMM, brain and leptomeningeal metastases; CI, confidence interval; dBLMM, first BLMM diagnosis; HR, hazard ratio; HER2, human epidermal growth factor receptor 2; mBC, metastatic breast cancer; OS, overall survival; PFS, progression-free survival.
Discussion
In this large population-based study, the cumulative incidence of BM was 24.1% and compares with a prior analysis conducted on the ESME database.3 Interestingly, we observed no significant difference in the incidence of BLMM at metastatic diagnosis between patients with HER2-low and HER2− mBC with rates of 4.5% and 5.4%, respectively, while patients with HER2+ mBC exhibited a much higher rate of 10.6%. Multivariable analysis emphasized that HER2 overexpression or the absence of HR remains a significant factor contributing to the development of BLMM, consistently with prior studies showing that the increased odds of BLMM are associated with tumors that are estrogen receptor-negative, HER2+ or basal phenotype.13 In our analysis the risk was comparable between the HER2-low and HER2− subgroups [sHR (95% CI) 1.07 (0.98-1.16)], but notably higher in the HER2+ subgroup compared to the HER2-low subgroup [sHR (95% CI) 1.56 (1.41-1.73)]. Considering that, in line with published data, about two-thirds of patients with HER2-low BC had HR-positive tumors, these results validate that HER2-low status does not contribute to the development of BM, while emphasizing that the absence of estrogen receptor expression remains a key risk factor. Our findings align with a recent cohort study investigating the occurrence of BM in over 18 000 BC patients. The study similarly revealed no rise in the frequency of brain metastasis in cases with HER2-low compared to HER2− cases. Furthermore, among HER2-negative and HER2-low patients, HR status emerged as a major determinant of BLMM occurrence.
In terms of survival outcomes, our analysis did not reveal substantial differences between patients with HER2− and HER2-low mBC and diagnosed with BLMM. Indeed, following a diagnosis of BLMM, the mOS for HER2-low patients was 6.6 months (95% CI 5.8-7.4 months) compared to 5.5 months (95% CI 5.0-5.9 months) for HER2− patients. The prognosis for HER2-low patients may see improvement in the future. The recent DEBBRAH study, presented at SABCS 2023, suggests that new compounds, such as T-DXd, could have a significant impact on the survival outcomes of patients with HER2-low BM.11 These data support the rationale of drug development in this context, particularly with the exploration of new anti-HER2 ADCs and immune-stimulating antibody conjugates.
Consistent with previously published data, the survival of patients with HER2+ BC was significantly better than that of the other two groups. This is most likely due to more effective treatment option for these patients.14,15 Differences in the immune microenvironment have also been highlighted in a recent study comparing HER2+ and HER2− bBM.16 This could partially account for variations in treatment efficacy between subtypes and, consequently, differences in terms of survival.
Our study has certain limitations. Firstly, it exclusively includes a French population, which may not be representative of the global population. Additionally, its retrospective nature could potentially limit the robustness of the data. Secondly, as reported in a prior ESME analysis, the assessment of HER2 status did not undergo central review, and the majority of cases with ERBB2 status were determined based on the primary tumor. This might, in part, account for the relatively lower percentage of patients with HER2-low BC compared to published data.17 The fact that the ESME database has not been designed for the specific study of data from patients with BM alone versus leptomeningeal metastases alone versus both deprives us of valuable information. Furthermore, the inclusion of patients over several years means that some patients may not have received currently recommended treatments, such as CDK4/6 inhibitors for luminal cancers or T-DXd for HER2+ and HER2-low patients, which could impact survival data. Finally, in the ESME cohort, granular data on intracranial disease (size and localization of metastatic lesions, RANO criteria for evaluation of intracranial response) are not captured.
Conclusion
The present study underscores that patients with HER2-low and HER2− mBC have similar patterns of intracranial disease progression and similar survival outcomes after the diagnosis of BM. OS after intracranial progression remains substantially shorter for these patients as compared to those with HER2+ mBC, highlighting the urgent need to unravel the potential biological distinctions within BLMM among HER2−, HER2-low and HER2+ mBC cases to refine future treatment development.
Acknowledgements
We deeply thank all patients who participated in the ESME mBC database for granting access to their data and allowing this work. We thank the participating sites and each ESME local coordinator for managing the project at the local level (alphabetical order): Institut de Cancérologie de l’Ouest, Angers/Nantes; Institut Bergonié, Bordeaux; Centre François Baclesse, Caen; Centre Jean Perrin, Clermont Ferrand; Centre Georges-François Leclerc, Dijon; Centre Oscar Lambret, Lille; Centre Léon Bérard, Lyon; Institut Paoli-Calmettes, Marseille; Institut de Cancérologie de Montpellier; Institut de Cancérologie de Lorraine, Nancy; Centre Antoine Lacassagne, Nice; Institut Jean Godinot, Reims; Centre Eugène Marquis, Rennes; Centre Henri Becquerel, Rouen; Centre Paul Strauss ICANS, Strasbourg; Institut Claudius Regaud IUCT-O, Toulouse; Institut Curie, Paris/ Saint-Cloud, Gustave Roussy, Villejuif. We also thank the ESME central coordinating staff, the MBC Scientific Group and the ESME Strategic Committee for their ongoing support.
Funding
This work was supported by Unicancer (no grant number). Unicancer manages independently ESME databases (i.e. data collection, analysis and publication) and is the sole data controller for data processing. The ESME mBC database receives financial support from industrial partners.
Disclosure
WJ declares grants: AstraZeneca, Daiichi Sankyo; payment or honoraria for lectures, presentations, speakers bureaus, manuscript writing or educational events: AstraZeneca, Eisai, BMS, Lilly France, Daiichi Sankyo, MSD, Novartis, Pfizer, Roche, Seagen; support for attending meetings and/or travel: AstraZeneca, Novartis, Chugai Pharma, Pfizer, Eisai, Pierre Fabre, Glaxo Smithkline, Roche, Lilly France, Sanofi Aventis; participation on a data safety monitoring board or advisory board : AstraZeneca, Eisai, BMS, Lilly France, Daiichi Sankyo, MSD, Novartis, Pfizer, Roche, Seagen, Gilead. MA declares payment or honoraria for lectures, presentations, speaker’s bureaus, manuscript writing or educational events: Gilead, Daiichi Sankyo, Pfizer, Lilly, AstraZeneca, Novartis; support for attending meetings and/or travel: Gilead, Pfizer, Novartis; participation on a data safety monitoring board or advisory board: AstraZeneca, Novartis, Pfizer. FLD declares consulting fees: Novartis, Gilead; payment or honoraria for lectures, presentations, speakers bureaus, manuscript writing or educational events: Amgen, Daiichi, Lilly, Novartis, Seagen, AZ; support for attending meetings and/or travel: Novartis, Pfizer, Seagen; participation on a data safety monitoring board or advisory board: Daiichi, Lilly, Seagen, Pfizer, Novartis, Roche, Exact Sciences; receipt of equipment, materials, drugs, medical writing, gifts or other services : Lilly. TF declares payment or honoraria for lectures, presentations, speaker’s bureaus, manuscript writing or educational events: Roche, Jansen, Lilly. BP received fees as advisor/consultant from Pierre Fabre (self), Daiichi Sankyo (self), Merck Sharp & Dohme (institution), Seattle Genetics (institution), Eli Lilly (institution) and Novartis (institution); funding to institution for research support from Daiichi Sankyo and AstraZeneca; and travel expenses from AstraZeneca, Pfizer and Gilead. JSF declares consulting fees: Pfizer, Lilly, Novartis, AstraZeneca, Clovis Oncology, GSK, Gilead, Daiichi Sankyo, Seagen, Exact Science, MSD; payment or honoraria for lectures, presentations, speakers bureaus, manuscript writing or educational events: Lilly, Novartis, AstraZeneca, Gilead, Daiichi Sankyo, Seagen, MSD; support for attending meetings and/or travel: Pfizer, Lilly, Novartis, AstraZeneca, Clovis Oncology, GSK, Gilead, Daiichi Sankyo, Seagen, MSD. All other authors have declared no conflicts of interest.
Data sharing
The datasets analyzed for the current study were extracted from the ESME mBC data platform. In accordance with ethical and legal requirements related to the ESME data warehouse, individual data from the ESME databases cannot be made available. For any specific demand, please contact the corresponding author. Each demand will be examined on a case-by-case basis by the scientific committee.
Supplementary data
References
- 1.Brufsky A.M., Mayer M., Rugo H.S., et al. Central nervous system metastases in patients with HER2-positive metastatic breast cancer: incidence, treatment, and survival in patients from registHER. Clin Cancer Res. 2011;17(14):4834–4843. doi: 10.1158/1078-0432.CCR-10-2962. [DOI] [PubMed] [Google Scholar]
- 2.Hurvitz S.A., O’Shaughnessy J., Mason G., et al. Central nervous system metastasis in patients with HER2-positive metastatic breast cancer: patient characteristics, treatment, and survival from SystHERs. Clin Cancer Res. 2019;25(8):2433–2441. doi: 10.1158/1078-0432.CCR-18-2366. [DOI] [PubMed] [Google Scholar]
- 3.Darlix A., Louvel G., Fraisse J., et al. Impact of breast cancer molecular subtypes on the incidence, kinetics and prognosis of central nervous system metastases in a large multicentre real-life cohort. Br J Cancer. 2019;121(12):991–1000. doi: 10.1038/s41416-019-0619-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Leyland-Jones B. Human epidermal growth factor receptor 2–positive breast cancer and central nervous system metastases. J Clin Oncol. 2009;27:5278–5286. doi: 10.1200/JCO.2008.19.8481. [DOI] [PubMed] [Google Scholar]
- 5.Guven D.C., Kaya M.B., Fedai B., et al. HER2-low breast cancer could be associated with an increased risk of brain metastasis. Int J Clin Oncol. 2022;27:332–339. doi: 10.1007/s10147-021-02049-w. [DOI] [PubMed] [Google Scholar]
- 6.Diéras V., Deluche E., Lusque A., et al. Trastuzumab deruxtecan (T-DXd) for advanced breast cancer patients (ABC), regardless of HER2 status: a phase II study with biomarkers analysis (DAISY) Cancer Res. 2022;82:PD8. 02. [Google Scholar]
- 7.Epaillard N., Lusque A., Pistilli B., et al. 260P Antitumor activity of trastuzumab deruxtecan (T-DXd) in patients with metastatic breast cancer (mBC) and brain metastases (BMs) from DAISY trial. Ann Oncol. 2022;33:S656. [Google Scholar]
- 8.Tarantino P., Hamilton E., Tolaney S.M., et al. HER2-low breast cancer: pathological and clinical landscape. J Clin Oncol. 2020;38(17):1951–1962. doi: 10.1200/JCO.19.02488. [DOI] [PubMed] [Google Scholar]
- 9.Steeg P.S. The blood–tumour barrier in cancer biology and therapy. Nat Rev Clin Oncol. 2021;18(11):696–714. doi: 10.1038/s41571-021-00529-6. [DOI] [PubMed] [Google Scholar]
- 10.Modi S., Jacot W., Yamashita T., et al. Trastuzumab deruxtecan in previously treated HER2-low advanced breast cancer. N Engl J Med. 2022;387(1):9–20. doi: 10.1056/NEJMoa2203690. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Pérez-García J.M., Vaz Batista M., Cortez P., et al. Trastuzumab deruxtecan in patients with central nervous system involvement from HER2-positive breast cancer: The DEBBRAH trial. Neuro Oncol. 2023;25(1):157–166. doi: 10.1093/neuonc/noac144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Castelo-Branco L., Pellat A., Martins-Branco D., et al. ESMO guidance for reporting oncology real-World evidence (GROW) Ann Oncol. 2023;34(12):1097–1112. doi: 10.1016/j.annonc.2023.10.001. [DOI] [PubMed] [Google Scholar]
- 13.Pestalozzi B.C., Zahrieh D., Price K.N., et al. Identifying breast cancer patients at risk for Central Nervous System (CNS) metastases in trials of the International Breast Cancer Study Group (IBCSG) Ann Oncol. 2006;17(6):935–944. doi: 10.1093/annonc/mdl064. [DOI] [PubMed] [Google Scholar]
- 14.Lin N.U., Borges V., Anders C., et al. Intracranial efficacy and survival with tucatinib plus trastuzumab and capecitabine for previously treated HER2-positive breast cancer with brain metastases in the HER2CLIMB trial. J Clin Oncol. 2020;38(23):2610–2619. doi: 10.1200/JCO.20.00775. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Bailleux C., Eberst L., Bachelot T. Treatment strategies for breast cancer brain metastases. Br J Cancer. 2021;124(1):142–155. doi: 10.1038/s41416-020-01175-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Griguolo G., Tosi A., Dieci M.V., et al. A comprehensive profiling of the immune microenvironment of breast cancer brain metastases. Neuro-Oncol. 2022;24(12):2146–2158. doi: 10.1093/neuonc/noac136. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.de Calbiac O., Lusque A., Mailliez A., et al. Comparison of management and outcomes in ERBB2-low vs ERBB2-zero metastatic breast cancer in France. JAMA Netw Open. 2022;5(9) doi: 10.1001/jamanetworkopen.2022.31170. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.