The presented study investigated the relevance of mutations in 17 cancer genes and response to neoadjuvant chemotherapy in two clinical cohorts of HER2+ breast cancer. 364 samples from HER2+ tumors of the neoadjuvant studies GeparTrio (no anti-HER2 treatment, n = 71) and GeparSepto (dual HER2 blockade and randomization for paclitaxel vs. nab-paclitaxel, n = 293) were analyzed by targeted next generation sequencing of hot spot regions of 17 genes. Mutations in TP53 (47.3%) and PIK3CA (23.9%) were most prevalent. EGFR, KRAS, NRAS, HRAS were combined to the MAPK module with 2.5% harboring mutations. In GeparSepto, the pCR rate was significantly lower in PIK3CA-mutant vs wild-type (wt) tumors (47.7% vs. 66.7%; p = 0.009). In patients treated with nab-paclitaxel, pCR rates were significantly lower in PIK3CA-mutated tumors compared to wt-tumors (38.7% vs. 72.0%; p = 0.001). In the GeparTrio cohort without neoadjuvant anti-HER2 therapy the pCR rate was 27.3% in the mutant cohort compared to 16.3% in the PIK3CA-wt cohort (p = 0.339). In HER2+ breast cancer, PIK3CA mutations were significantly associated with reduced response to dual HER2 blockade with pertuzumab+trastuzumab as well as reduced response to nab-paclitaxel. This reduction was not observed in GeparTrio without anti-HER2 therapy.

G7 trial design; demographics; tissue types; gene names; displayed data from additional analyses; statistical data

We evaluate therapy-induced molecular heterogeneity in longitudinal samples from high-risk, hormone-receptor positive/HER2-negative breast cancer patients with residual tumor after neoadjuvant chemotherapy from the Penelope-B trial (NCT01864746; EudraCT 2013-001040-62). Intrinsic subtypes are prognostic in pre-therapeutic (Tx) samples (n = 629, p < 0.0001) and post-Tx residual tumors (n = 782, p < 0.0001). After neoadjuvant chemotherapy, a shift of intrinsic subtypes is observed from pre-Tx luminal (Lum) B to post-Tx LumA, with reverse transition back to LumB in metastases. In a combined analysis of 540 paired pre-Tx and post-Tx samples, we identify five adaptive clusters (AC-1-5) based on transcriptomic changes before and after neoadjuvant chemotherapy. These AC-subtypes are prognostic beyond classical intrinsic subtyping, categorizing patients into groups with excellent prognosis (AC-1 and AC-2), poor prognosis (AC-3 and AC-4), and very poor prognosis (AC-5, enriched for basal-like subtype). Our analysis provides a basis for an extended molecular classification of breast cancer patients and improved identification of high-risk patient populations.

PURPOSE:The PENELOPE-B trial (ClinicalTrials.gov identifier: NCT01864746) recruited patients with hormone receptor+/human epidermal growth factor receptor 2- early breast cancer without a pathological complete response after taxane-containing neoadjuvant chemotherapy and at a high risk of relapse. Patients were randomly assigned (1:1) to receive 13 cycles of palbociclib once daily or placebo on days 1-21 in a 28-day cycle in addition to endocrine therapy (ET). PENELOPE-B did not show improved invasive disease-free survival (iDFS) after adding palbociclib to ET. This retrospective analysis investigated the impact of germline pathogenic variant (PV) status of BRCA1/2 and non-BRCA1/2 cancer predisposition genes on the outcomes of PENELOPE-B trial patients. METHODS:In total, 445 patients were sampled following a case-cohort design and 442 were analyzed for germline PVs. Statistical analyses were performed for time-to-event end points (iDFS, distant disease-free survival [DDFS], and overall survival [OS]). RESULTS:Of the 442 patients, 42 carried PVs in any cancer predisposition gene; 15 carried BRCA1/2 PVs. Irrespective of the treatment arms, PV status was not a prognostic factor. Regarding the treatment arms in BRCA1/2 PV carriers, numerically better 3-year outcomes were observed in the palbociclib arm (iDFS, 95%; DDFS, 95%; OS, 100%) than in the placebo arm (iDFS, 72.8%; DDFS, 72.8%; OS, 87.5%; hazard ratios palbociclib v placebo 0.349 [iDFS] and 0.562 [DDFS], not calculated for OS, too few events). In patients without BRCA1/2 PVs, the differences in 3-year outcomes were negligible. PVs in non-BRCA1/2 cancer predisposition genes did not influence the efficacy of palbociclib, although gene-specific effects could not be excluded. CONCLUSION:Patients with BRCA1/2 PVs had numerically better outcomes after palbociclib. However, the number of BRCA1/2 carriers was small. Larger randomized clinical trials should consider the PV status to further evaluate whether BRCA1/2 PV carriers benefit from cyclin-dependent kinase 4 and 6 inhibitor treatment.

Supplementary Figure S3. CAV1 RNA expression is positively correlated with CAV1 protein expression in the TCGA invasive breast carcinoma database.

Supplementary Figure S1. CAV1 and CAV2 before (A, C) and after (B, D) transformation of the continuous expression values.

Background: Nearly 30% of HR+ early breast cancer (eBC) patients (pts) treated with neoadjuvant chemotherapy (NACT) and surgery will experience BC recurrence, many with incurable distant metastatic disease. There is currently no blood-based biomarker that can identify pts with residual disease and at high risk of recurrence before and during adjuvant therapy. Thymidine kinase (TK) is an enzyme that plays a key role in DNA replication during cell division. The expression of TK is strongly linked to the cell cycle and measurement of TK activity (TKa) is a validated biomarker for both disease prognosis and therapy efficacy in metastatic breast cancer (mBC). We investigated the prognostic utility of TKa in eBC using patient serum samples from the PENELOPE-B study (NCT01864746). Methods: The PENELOPE-B phase III trial explored the addition of one year of palbociclib (P) to endocrine therapy (ET), in HR+, HER2- eBC pts at high risk of relapse after NACT. Serum samples were collected at baseline (BL), 6 months after starting therapy (C7), and end of treatment (EOT ∼ 13 months after starting therapy). Samples were analyzed retrospectively for TKa with the FDA cleared DiviTum® TKa assay (Biovica) using DuA (DiviTum unit of Activity) as the measuring unit. The cutoff for defining high vs low TKa was 250 DuA. In previous mBC studies, a TKa threshold value above 250 DuA was significantly associated with likelihood of disease progression. This same threshold value was applied to the PENELOPE-B sample set. TKa association with early invasive disease-free survival (iDFS) and distant disease-free survival (dDFS) was analyzed using restricted mean survival time (RMST) at a pre-specified timepoint of 1 year. The same method was also used to analyze the association of the clinical-pathologic stage- estrogen/grade (CPS-EG) score with early iDFS and dDFS using a cut-off score of <3 or >3. Results: 1251 pts enrolled in PENELOPE-B. BL TKa levels were analyzed in a sample set of 871 pts, 444 P+ET, 427 ET alone. Median follow-up time was 84 months. Among these 871 pts, 311 pts had an iDFS event (P + ET, 156; ET alone, 155) with 45 of these pts recurring in the first year. BL TKa ranged between 22 and 1025 DuA, with a mean value of 102 DuA, and a median of 81 DuA. BL TKa levels >250 DuA were significantly associated with both invasive and distant disease recurrence within the first 12 months of adjuvant therapy (RMST estimates for BL TKa>250 DuA vs BL TKa≤250 DuA: iDFS=9.6 vs 11.6 months, p=0.01, dDFS=9.9 vs 11.7 months, p= 0.02). In early relapse pts who had a serum sample taken at the time of study discontinuation, 18/22 (82%) had an increase in TKa at the time of relapse as compared to BL. The CPS-EG score when dichotomized as <3 or =4-5 was not prognostic for early relapse within a year (RMST estimates for CPS-EG >3 vs CPS-EG≤3, iDFS=11.3 vs 11.6 months, p=0.12, dDFS=11.3 vs 11.7 months, p= 0.08). TKa dynamic changes were also analyzed in 360 pts who had both C7 and EOT samples available for testing. Updated analysis looking at increases/decreases in TKa levels during therapy and association with recurrence, as well as treatment interaction will be presented. Conclusions: In a subset of pts from PENELOPE-B, TKa was detectable in all pts at BL and levels were significantly correlated with early disease recurrence where BL TKa values >250 DuA were prognostic for early relapse within the 1st year of adjuvant therapy. CPS-EG scores in contrast were not prognostic. Pts with early disease relapse also had an increase in TKa levels at the time of recurrence. These findings warrant further investigation of serum TKa as a non-invasive dynamic biomarker that could be used to assess in “real-time” the presence of actively proliferating disease in adjuvant BC pts and to monitor response to treatment via serial serum TKa testing. Citation Format: Amy J. Williams, Harry D. Bear, Carsten Denkert, Frederik Marmé, Erik Knudsen, Masey M. Ross, Seock-Ah Im, Angela DeMichele, José Angel García Sáenz, Agnieszka Witkiewicz, Laura Van’t Veer, Sung-Bae Kim, Zhe Zhang, Nicholas Turner, Federico Rojo, Martin Filipits, Lesley-Ann Martin, Olga Valota, Peter A. Fasching, Christian Schem, Nicole Mc Carthy, Toralf Reimer, Bärbel Felder, Karsten Weber, Valentina Nekljudova, Sibylle Loibl. Evaluation of proliferation biomarker serum thymidine kinase activity and prediction of early relapse in HR positive HER2 negative high risk early breast cancer: Analysis from the PENELOPE-B trial [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr P1-09-06.

566 Background: The concept of intrinsic subtyping has been an important step for understanding of breast cancer (BC) as a heterogenous disease. However, these subtypes are not adapted to therapy-induced molecular plasticity. We have evaluated a high-risk luminal BC clinical trial cohort to identify new additional adaptive BC subtypes based on molecular alterations induced during neoadjuvant chemotherapy (NACT). Methods: A total of 1250 ER+/HER2- BC patients (pts) with residual disease after NACT and increased risk (CPS-EG score of ≥3 or 2 with ypN+) were randomized into the PENELOPE-B (NCT01864746) trial to receive palbociclib or placebo. Biomarker analysis was performed in 1411 pre- or postNACT tumor samples including 540 paired samples using the HTG EdgeSeq Oncology Biomarker Panel targeting 2549 genes (HTG Molecular Diagnostics Inc.), with assessment of the absolute assignment of breast cancer intrinsic molecular subtype (AIMS). For a subcohort of 29 pts, we evaluated triplicate samples before and after NACT as well as at the time of metastatic disease. Results: In paired biopsies, a total of 335 genes were significantly different between the pre- and the postNACT cohort. With hierarchical unsupervised clustering of these 335 genes, we characterized five different tumor subtypes with highly significant iDFS survival differences (p<0.0001, n=539). We identified two large groups of tumors with excellent prognosis (adaptive cluster AC1 and AC2, together n=284, 6 events), as well as two groups with a poor prognosis; AC-3 (n=163, 82 events) and AC-4 (n=78, 29 events). The worst prognosis was observed in a small group of tumors with a pre-NACT Basal/HER2E AIMS subtype (n=14, 11 events). AC1 and AC3 are enriched for tumors with an AIMS subtype change from LumB to LumA during NACT, with improved iDFS for AC1. AC2 and AC4 are enriched for tumors with a constant LumA AIMS subtype before and after NACT, with improved iDFS for AC2. The low-risk adaptive subtypes AC1 and AC2 cover a total of 52.7% of pts with a combined event rate of 2.1%. In contrast, the best pretherapeutic conventional AIMS-subtype, preNACT LumA, covers 51.7% of pts, but still has an event rate of 15.4%. This suggests that adaptive subtypes are superior to classical AIMS subtypes for prediction of survival after NACT in luminal BC. Conclusions: Classical approach of intrinsic subtyping relies on constant gene expression, limiting its scope. Adaptive subtyping can complement the classical approach, focusing on those genes that are changed during therapy. Comparison of paired samples before and after therapy is superior to classical subtyping of baseline samples. This approach identified large subgroups of pts with excellent prognosis after NACT despite clinical high risk. This can be an important step to focus on high-risk pts for new neoadjuvant and post-neoadjuvant therapy concepts.

Therapy-induced molecular adaptation of triple-negative breast cancer is crucial for immunotherapy response and resistance. We analyze tumor biopsies from three different time points in the randomized neoadjuvant GeparNuevo trial (NCT02685059), evaluating the combination of durvalumab with chemotherapy, for longitudinal alterations of gene expression. Durvalumab induces an activation of immune and stromal gene expression as well as a reduction of proliferation-related gene expression. Immune genes are positive prognostic factors irrespective of treatment, while proliferation genes are positive prognostic factors only in the durvalumab arm. We identify stromal-related gene expression as a contributor to immunotherapy resistance and poor therapy response. The results provide evidence from clinical trial cohorts suggesting a role for stromal reorganization in therapy resistance to immunotherapy and in the generation of an immune-suppressive microenvironment, which might be relevant for future therapy approaches targeting the tumor stroma parallel to immunotherapy, such as combinations of immunotherapy with anti-angiogenic therapy.

Background: It is well known that immunological pathways are relevant for response to classical neoadjuvant chemotherapy as well as combined chemo-immunotherapy. In addition, it has been shown that combined chemo-immunotherapy significantly improves survival, even in the context of only moderate effects on pCR. Due to the window therapy with durvalumab-alone and the option to analyze multiple consecutive biopsies, the GeparNuevo trial offers the opportunity to 1) determine gene expression patterns for pCR and DDFS endpoints 2) identify pathways most relevant for pCR and DDFS 3) identify genes specifically regulated by immunotherapy (comparison of samples pre-and post-window) 4) identify genes specifically regulated by chemotherapy (comparison of samples pre-Tx and after 4 cycles of chemotherapy 5) identify longitudinal patterns of gene expression by comparison of up to four time points and 6) identify changes in the tumor microenvironment by spatial sequencing of tumor cell and stroma areas. Methods: 292 tumor samples were evaluated by gene expression analysis: 162 pretherapeutic core biopsies, 79 post-window biopsies, 32 biopsies during chemotherapy and 19 biopsies of the residual tumor after therapy. These samples were analyzed by HTG OBP panel targeting 2549 genes which are assigned to 25 different biological mechanisms or cellular pathways. In addition, spatial profiling was compared in a subset of pre-and post-window samples using Nanostring GeoMx spatial profiling system. Endpoints were pCR and DDFS. Results: A total of more than 600 genes were significantly associated with either the pCR or the DDFS endpoint in either the complete GeparNuevo cohort or one of the two therapy arms. Interestingly, there was a large number of predictive or prognostic genes (n=247 for pCR and n=179 for DDFS) in the durvalumab arm, while the number of genes in the placebo arm was considerably lower (n=113 for pCR and n=61 for DDFS). We used existing pathway information for HTG OBP panel to analyze the contribution of different cellular processes to pCR and DDFS signatures in different therapy arms. Immune pathways were particularly relevant for durvalumab signatures (pCR and DDFS), while cell cycle related gene expression patterns were particularly involved in signatures predictive of pCR in both therapy arms. To further assign genes to the cellular response to durvalumab-alone or chemotherapy-alone, we compared gene expression patterns in durvalumab arm before and after the window phase (gene expression patterns induced by one dose of durvalumab) with gene expression patterns in placebo arm before and after 4 cycles of chemotherapy. Further longitudinal alterations were analyzed by comparison of longitudinal samples for 4 different time-points (a: before NACT, n=162; b: after window phase, n=79; c: after 4 cycles, n=31 and d: at surgery, n=19). Using the Nanostring GeoMx spatial RNA profiling system guided by cytokeratine immunofluorescence, we compared areas with high tumor cell content with stromal areas with or without TILs. In combination with the HTG gene expression data, we were able allocate the changes induced by durvalumab vs chemotherapy to the stromal cell and tumor cell compartment, indicating a re-organization of the tumor-microenvironment. Conclusions: In our analysis, we show that immune gene signatures are particularly relevant for neoadjuvant response to durvalumab as well as prognosis after durvalumab treatment, while proliferation signatures are involved in pCR-signatures after durvalumab as well as chemotherapy. The spatial analysis showed that relevant changes occur in the stromal compartment, indicating a re-organization of the tumor microenvironment. The parallel targeting of immune- and proliferation pathways might explain why a combined immunotherapy-chemotherapy approach is more successful than each single therapy strategy alone. Citation Format: Carsten Denkert, Andreas Schneeweiss, Julia Rey, Akira Hattesohl, Thomas Karn, Michael Braun, Paul Jank, Jens Huober, Hans-Peter Sinn, Dirk-Michael Zahm, Claus Hanusch, Frederik Marmé, Jenny Furlanetto, Jörg Thomalla, Jens-Uwe Blohmer, Marion van Mackelenbergh, Thorsten Stiewe, Peter Staib, Christian Jackisch, Julia Teply-Szymanski, Peter A. Fasching, Bruno V. Sinn, Michael Untch, Karsten Weber, Sibylle Loibl. PD4-02 Spatial and temporal heterogeneity of predictive and prognostic signatures in triple-negative breast cancer treated with neoadjuvant combination immune-chemotherapy [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr PD4-02.

AbstractPurpose: Caveolin-1 and -2 (CAV1/2) dysregulation are implicated in driving cancer progression and may predict response to nab-paclitaxel. We explored the prognostic and predictive potential of CAV1/2 expression for patients with early-stage HER2-negative breast cancer receiving neoadjuvant paclitaxel-based chemotherapy regimens, followed by epirubicin and cyclophosphamide. Experimental Design: We correlated tumor CAV1/2 RNA expression with pathologic complete response (pCR), disease-free survival (DFS), and overall survival (OS) in the GeparSepto trial, which randomized patients to neoadjuvant paclitaxel- versus nab-paclitaxel–based chemotherapy. Results: RNA sequencing data were available for 279 patients, of which 74 (26.5%) were hormone receptor (HR)–negative, thus triple-negative breast cancer (TNBC). Patients treated with nab-paclitaxel with high CAV1/2 had higher probability of obtaining a pCR [CAV1 OR, 4.92; 95% confidence interval (CI), 1.70–14.22; P = 0.003; CAV2 OR, 5.39; 95% CI, 1.76–16.47; P = 0.003] as compared with patients with high CAV1/2 treated with solvent-based paclitaxel (CAV1 OR, 0.33; 95% CI, 0.11–0.95; P = 0.040; CAV2 OR, 0.37; 95% CI, 0.12–1.13; P = 0.082). High CAV1 expression was significantly associated with worse DFS and OS in paclitaxel-treated patients (DFS HR, 2.29; 95% CI, 1.08–4.87; P = 0.030; OS HR, 4.97; 95% CI, 1.73–14.31; P = 0.003). High CAV2 was associated with worse DFS and OS in all patients (DFS HR, 2.12; 95% CI, 1.23–3.63; P = 0.006; OS HR, 2.51; 95% CI, 1.22–5.17; P = 0.013), in paclitaxel-treated patients (DFS HR, 2.47; 95% CI, 1.12–5.43; P = 0.025; OS HR, 4.24; 95% CI, 1.48–12.09; P = 0.007) and in patients with TNBC (DFS HR, 4.68; 95% CI, 1.48–14.85; P = 0.009; OS HR, 10.43; 95% CI, 1.22–89.28; P = 0.032). Conclusions: Our findings indicate high CAV1/2 expression is associated with worse DFS and OS in paclitaxel-treated patients. Conversely, in nab-paclitaxel–treated patients, high CAV1/2 expression is associated with increased pCR and no significant detriment to DFS or OS compared with low CAV1/2 expression.

Comparison of gene signature detection in High Quality and Low Quality Samples

Supplementary Figure S3: Kaplan Meier plots for disease-free (DFS) and overall survival (OS) according to CTC detection after NT. (A, B) DFS, analysis of patients with HER2-positive tumors (A) and of patients with triple negative tumors (B) at cut-off {greater than or equal to}2 CTCs/7.5 mL, (C, D) OS, analysis of patients with HER2-positive tumors (C) and of patients with triple negative tumors (D) at cut-off {greater than or equal to}2 CTCs/7.5 mL.

Correlation of Affymetrix microarray expression data and IHC scoring of three genes in independent TNBC dataset

Bivariate logistic regression models for the immune- and proliferation-associated gene signatures

Immunohistochemical validation of hypoxia signature