Abstract P3-13-04: Clinical features and genetic alterations in patients (pts) with HER2-positive breast cancer (BC) and central nervous system (CNS) metastases (mets) in the real world

Abstract Introduction: Pts with HER2-positive metastatic BC (mBC) often develop CNS mets, which is associated with poor prognosis and negative impact on quality of life. Resistance to anti-HER2 treatment can lead to disease progression; furthermore, many therapies do not cross the blood-brain barrier, leading to an unmet need for effective CNS mets treatment. In this real world study, we explored the molecular profile and clinical features of pts with CNS and non-CNS mets to better understand tumor biology in these populations. Methods: This retrospective, cross-sectional, observational study evaluated pts with CNS mets or non-CNS mets at the time of HER2-positive mBC diagnosis using the Flatiron Health-Foundation Medicine Clinico-Genomic Database (FH-FMI CGDB). The database comprised de-identified data from ~280 US cancer clinics (~800 sites of care), and included pts with FMI comprehensive genomic profiling assays covering >300 cancer-related genes. Descriptive analyses were conducted for baseline demographic, clinical, and biomarker characteristics, as well as genetic alterations including single nucleotide variants (SNVs), copy number alterations (CNAs), and rearrangements. T-tests were used to compare continuous variables, chi-squared tests for categorical variables, and non-parametric tests for non-normal distributions. Results: Of 8204 pts with BC in the CGDB, 6111 females 18+ yrs were diagnosed with mBC on or after 1-1-2011, of whom 919 had HER2-positive mBC, and 314 had an FMI assay up to 120 days after mBC diagnosis. This study population included 45 pts with CNS mets and 269 pts with non-CNS mets (Table). FMI data were derived from the primary tumor (29% from breast) or a metastatic lesion (19%, 12%, and 4% from liver, lymph nodes, and brain, respectively). Pts with CNS vs non-CNS mets were significantly younger at initial and at mBC diagnosis (p=0.04). At mBC diagnosis, a higher percentage of pts with CNS vs non-CNS mets had ≥3 sites of mets (p=0.01) and mets sites in adrenal glands (p=0.01), while there was a trend toward lower rates of lung and liver mets (not significant [NS]). Pts with CNS mets were more likely to be hormone receptor-negative vs pts with non-CNS mets (NS). Missing data for HER2 results by IHC or FISH, tumor grade, and histology limited the interpretation of any observed differences. Pts with CNS vs non-CNS mets were more likely to have SNVs in TP53 (82% vs 64%, p=0.03) and CNAs in ERBB2 (60% vs 43%, p<0.05). The most common mutations in pts with CNS mets were TP53 (82%) and PIK3CA (29%), followed by ERBB2, BRCA1, ESR1, CDH1, MLL2, and ATRX (<9% each). The most common CNAs in pts with CNS mets were ERBB2 (60%) and MYC (20%), followed by CCND1, FGF19, FGF3, FGF4, and CCNE1 (13% each). Rearrangements were rare and no clear differences were observed between pts with CNS and non-CNS mets. Conclusions: Despite its small sample size, our analysis is one of the largest real world genomic studies in HER2-positive pts with CNS mets to date. Pts with CNS mets at mBC diagnosis showed potential distinguishing clinical and genetic features, including younger age, a higher proportion with adrenal gland mets and ≥3 sites of mets, and a higher prevalence of SNVs in TP53 and CNAs in ERBB2. Cautious interpretation is needed due to the small sample, unique pt population, and heterogeneity in tissue location. Further analyses can help elucidate the biology of CNS mets in these pts with a high unmet need. Table. Demographic, clinical, and genetic characteristics of the study populationVariableCNS mets (n=45)Non-CNS mets (n=269)Total (N=314)p-valueMedian age at mBC diagnosis, years (interquartile range [IQR])55 (47-62)60 (51-68)59 (49-68)0.04Median age at early BC diagnosis, years (IQR)51 (43-60)55 (48-64)55 (46-64)0.04<45 years14 (31)54 (20)68 (22)45-64 years24 (53)148 (55)172 (55)≥65 years7 (16)67 (25)74 (24)Race, n (%)White23 (51)174 (65)197 (63)0.09Black/Asian4 (9)31 (12)35 (11)Other10 (22)43 (16)53 (17)Missing8 (18)21 (8)29 (9)Group stage at initial diagnosis, n (%)aI<4 (<9)25 (9)26 (8)0.12II22 (49)88 (33)110 (35)III11 (24)76 (28)87 (28)IV6 (13)59 (22)65 (21)Missing5 (11)21 (8)26 (8)Number of metastatic sites, n (%)111 (24)110 (41)121 (39)0.01213 (29)90 (34)103 (33)≥321 (47)69 (26)90 (29)SNVs (selected a priori), n (%)bTP5337 (82)173 (64)210 (67)0.03PIK3CA13 (29)97 (36)110 (35)0.45ERBB2<4 (<9)21 (8)24 (8)1ESR1<4 (<9)7 (3)10 (3)1PTEN<4 (<9)12 (4)14 (4)1CNAs (most common in all pts), n (%)bERBB227 (60)115 (43)142 (45)0.047MYC9 (20)61 (23)70 (22)0.84FGF196 (13)45 (17)51 (16)0.72aGroup stage is documented at the time of initial diagnosis in the pt health record or determined based on T, N, and M stage based on AJCC version at the time of the patient’s diagnosis, and does not include staging collected after systemic/radiation therapy, progression, and/or neoadjuvant treatment, if any.bp-values were not adjusted for multiple comparisons. Data not shown: additional univariate analyses of 129 genes on SNVs, CNAs and rearrangements yielded the following genes with the highest prevalence in pts with CNS mets: TP53 (82%), PIK3CA (29%), ERBB2 (<9%), CDH1 (<9%), ESR1 (<9%), BRCA1 (<9%), MLL2 (<9%), ATRX (<9%), and GATA3 (<9%) for SNVs; ERBB2 (60%), MYC (20%), CCND1 (13%), FGF19 (13%), FGF3 (13%), FGF4 (13%), and CCNE1 (13%) for CNAs; and CDK12 (<9%) and NF1 (<9%) for rearrangements. Citation Format: Preet K. Dhillon, Peter Lambert, Thibaut Sanglier, Adam Knott, Eleonora Restuccia, Sanne L. de Haas, Chiara Lambertini. Clinical features and genetic alterations in patients (pts) with HER2-positive breast cancer (BC) and central nervous system (CNS) metastases (mets) in the real world [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P3-13-04.