HTG EdgeSeq Oncology Biomarker Panel Research Articles

Relationship of adaptive subtyping and tumour heterogeneity of treatment response to neoadjuvant therapy in hormone receptor–positive HER2-negative early breast cancer: PENELOPE-B.

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.

Molecular characterization of growing teratoma syndrome in patients with testicular germ cell tumor.

e17017 Background: Teratomas are unique neoplastic growths that encompass a diverse array of tissue types derived from multiple germ layers, often exhibiting varying degrees of differentiation. Among the phenomena associated with teratomas, the growing teratoma syndrome (GTS) stands out as a rare yet clinically significant occurrence. GTS, characterized by the persistence or regrowth of teratomatous elements despite initial treatment, poses challenges for both diagnosis and management. In this abstract, we present the results of our expression array analysis of teratomas, aiming to uncover novel insights into their molecular composition, heterogeneity, and potential implications for clinical management. Methods: This study included 53 patients with retroperitoneal teratoma (7 chemotherapy naïve and 46 with postchemotherapy teratoma) and 20 patients with viable postchemotherapy germ cell tumor from retroperitoneum that had surgery at National Cancer Institute in Slovakia and for whom formalin-fixed paraffin-embedded (FFPE) tumor tissue was available. Study also includes samples from 5 patients with normal testis as control group. Ten patients (18.9%) in this cohort had growing teratoma syndrome. Analysis was performed by HTG EdgeSeq Oncology Biomarker panel (HTG Molecular Diagnostics, Inc., Tucson, USA). Identified candidate genes associated with GTS were further validated by immunohistochemistry on FFPE. Results: Molecular profiling indicates no differences in expression profile in retroperitoneal teratomas regardless of previous chemotherapy. Expression analysis indicates 1586 genes being differently expressed with high significance ( p < 0.05) in teratomas compared to normal testis and 1571 genes differentially expressed compared postchemotherapy viable tumors. We identified 14 genes being differently expressed in growing teratoma syndrome compared to teratoma without growth (CYP4A22, IGFBP1, CXCL6, LTF, CYP2C8, CFTR, CCL20, PFKFB3, SLC2A3, LIPA, HMOX1, GPNMB, CCL18, DLK1. Pathway analysis revealed MAPK and PI3K−Akt signaling pathways are mostly dysregulated in GTS. Conclusions: This molecular profiling identify genes differentially expressed in teratomas that are potential biomarkers for prediction of teratoma element in postchemotherapy residues in testicular germ cell tumor patients. Moreover, inhibition of identified signaling pathways associated with GTS could represent potentially novel therapeutic targets in case of unresectable disease.

Abstract PO2-24-02: Molecular characterization of primary tumors associated with various CTCs’ subpopulation

Abstract Background: Circulating tumor cells (CTCs) are prognostic in primary breast cancer (BC) patients and represent a heterogeneous population of cells with different phenotypes and biological value. However, molecular characterization of primary tumors associated with various CTCs’ subpopulation is lacking. This study aimed to identify expression of genes associated with the presence of different CTCs subpopulations in primary BC patients using a comprehensive genomics approach. Methods: This prospective ongoing study included 312 patients (pts) with non-metastatic BC enrolled from March 2012 to February 2015. CTCs were detected on day -1 to 0 before surgery. Isolated peripheral blood mononuclear cells (PBMC) were depleted of hematopoietic cells using RossetteSepTM kit negative selection with anti-CD45 antibody. RNA extracted from CD45-depleted (CD45-) PBMC was interrogated for expression of EMT-inducing transcription factors (TWIST1, SNAIL1, SLUG, ZEB1) and epithelial (CK19) gene transcripts by quantitative reverse transcription-PCR. Expressions of gene transcripts in CD45- PBMC of patients were compared to those of CD45- PBMC of 60 healthy donors. Formalin-fixed paraffin-embedded tumor tissue was subject for gene expression analysis performed by HTG EdgeSeq Oncology Biomarker panel (HTG Molecular Diagnostics, Inc., Tucson, USA). Results: CTCs with epithelial phenotype (CTC_EP) were detected in 30 (9.6%) pts, CTC with EMT (epithelial-to-mesenchymal transition) were detected in 54 (17.3%), while blood of 79 (25.3%) pts exhibit at least one type of CTC. Patients with detectable CTC in peripheral blood had inferior disease-free survival compared to patients without detectable CTC (HR = 0.62, 95% CI 0.37-1.06, P=0.048). There was no association between epithelial CTCs and analysed patients/tumour variables. Genes expression analysis identified 7 differentially expressed genes in patients with the CTC_EP phenotype (CHGA, MESP1, COL2A1, MT2A, COMP, MMP7 and ORM2), 200 genes associated with CTC_EMT phenotype (top 10 genes included CHGA, CECR6, SMAD2, HUS1, FGF10, HNF1A, SEC61G, CDK4, CD44 and F3). Analysis identified that low expression of CHGA and MMP7 was consistently associated with any CTC subpopulations. Conclusions: In this study for the first-time revealed gene expressions associated with various CTC subpopulations. We suppose that these genes could represent potential therapeutic target aimed to inhibit metastatic cascade in breast cancer. Citation Format: Michal Mego, Gabriel Minarik, Milan Hucko, Katarina Kalavska, Marian Karaba, Juraj Benca, Tatiana Sedlackova, Lucia Kucerova, Jozef Mardiak, Lubos Klucar, Zuzana Cierna. Molecular characterization of primary tumors associated with various CTCs’ subpopulation [abstract]. In: Proceedings of the 2023 San Antonio Breast Cancer Symposium; 2023 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2024;84(9 Suppl):Abstract nr PO2-24-02.

Mesenteric fibrosis in patients with small intestinal neuroendocrine tumors is associated with enrichment of alpha-smooth muscle actin-positive fibrosis and COMP-expressing stromal cells.

Neuroendocrine tumors of the small intestine (SI-NETs) often develop lymph node metastasis (LNM)-induced mesenteric fibrosis (MF). MF can cause intestinal obstruction as well as ischemia and render surgical resection technically challenging. The underlying pathomechanisms of MF are still not well understood. We examined mesenteric LNM and the surrounding stroma compartment from 24 SI-NET patients, including 11 with in situ presentation of strong MF (MF+) and 13 without MF (MF-). Differential gene expression was assessed with the HTG EdgeSeq Oncology Biomarker Panel comparing MF+ with MF- within LNM and paired stromal samples, respectively. Most interesting differentially expressed genes were validated by reverse transcription-quantitative polymerase chain reaction (RT-qPCR) in combination with validation of associated protein levels utilizing immunohistochemistry (IHC) staining of MF+ and MF- formalin-fixed, paraffin-embedded (FFPE) patient samples. Overall, 14 genes measured with a 2549-gene expression panel were differentially expressed in MF+ patients compared to MF-. Of those, nine were differentially expressed genes in LNM and five genes in the stromal tissue (>2-fold change, p < .05). The top hits included increased COMP and COL11A1 expression in the stroma of MF+ patients compared to MF-, as well as decreased HMGA2, COL6A6, and SLC22A3 expression in LNM of MF+ patients compared to LNM of MF- patients. RT-qPCR confirmed high levels of COMP and COL11A1 in stroma samples of MF+ compared to MF- patients. IHC staining confirmed the enrichment of α-smooth muscle actin-positive fibrosis in MF+ compared to MF- patients with corresponding increase of COMP-expressing stromal cells in MF+. Since COMP is associated with the known driver for fibrosis development transforming growth factor beta and with a cancer-associated fibroblasts enriched environment, it seems to be a promising new target for MF research.

Abstract PD17-05: Development and Validation of a Composite Biomarker Predictive of Palbociclib + Endocrine Treatment Benefit in Early Breast Cancer: PENELOPE-B and PALLAS Trials

Abstract Background: The PENELOPE-B (NCT01864746) and PALLAS (NCT02513394) trials are large prospective, randomized, phase III trials that evaluated adjuvant palbociclib (PAL) + endocrine treatment (ET) vs ET in patients with hormone receptor–positive/human epidermal growth factor receptor 2–negative (HER2–) early breast cancer (EBC). Both studies did not meet the primary endpoint of improving invasive disease-free survival (iDFS). We conducted biomarker analyses to identify patients who might benefit from PAL + ET in EBC. Methods: Resected tumor tissue was collected from consenting patients. Gene expression analyses were conducted using the HTG EdgeSeq Oncology Biomarker Panel including 2549 genes. Based on 91 genes from the HTG panel, the intrinsic molecular subtypes were calculated using Absolute Intrinsic Molecular Subtyping (AIMS). Potential predictive treatment biomarkers were established in PENELOPE-B (n=906 with resected tissue) as the development set using an outcome-oriented approach based on iDFS with a selection procedure that maximized the log-rank statistic to estimate a standard Z score–based optimal cutoff. Independent validation was conducted on PALLAS (n=2085; PENELOPE-B-like with resected tissue and HTG data). Hazard ratios and corresponding 95% CIs were calculated using the Cox proportional hazards model, and iDFS distributions between treatment arms were compared using the log-rank test. Interaction between treatment and biomarker status was assessed. Results: Patient baseline characteristics were well balanced, with no differences in iDFS between the intent-to-treat set and the biomarker set for both trials. Approximately 73% of patients (PENELOPE-B [n=663] and PALLAS [n=1516]) had luminal A subtypes whereas only 7.1 % (PENELOPE-B [n=64]) and 8.3 % (PALLAS [n=172]) had a luminal B subtype. AIMS subtypes showed overall similar prognostic patterns for iDFS between PENELOPE-B and PALLAS. The biomarker-defined subgroup found in PENELOPE-B with optimal cutoff demonstrated a preferential benefit from PAL + ET (n=364 [96 events]; hazard ratio [95% CI], 0.63 [0.42, 0.95]; P=0.025). Independent validation of the PALLAS subgroup using the pre-defined optimal cutoff confirmed a significant benefit from PAL + ET (n=916 [70 events]; 0.55 [0.34–0.90]; P=0.015) while not in the rest of the patients (interaction p=0.0025). Significant treatment effects remained (0.55 [0.34–0.89]; P=0.015) after adjusting for the randomization stratification factors of PALLAS. Conclusions: The composite predictive biomarker defined from PENELOPE-B was independently validated in a prospectively defined retrospective analysis of a subset of patients selected from PALLAS. The composite biomarker identified a subset of EBC patients deriving benefit from the addition of PAL to ET. This patient stratification approach can potentially be applied to future adjuvant clinical trials for treatment of hormone receptor–positive/HER2– EBC. Citation Format: Sibylle Loibl, Carsten Denkert, Yuan Liu, Erik S. Knudsen, Angela DeMichele, Zhe Zhang, Julia Teply-Szymanski, Martin Filipits, Peter A. Fasching, Michael Gnant, Shibing Deng, Marija Balic, Federico Rojo, Mark Watson, Chetan Deshpande, Nicholas Turner, Otto Metzger, Kathy Puyana Theall, Agnieszka Witkiewicz, Olga Valota, W. Fraser Symmans, Erica L. Mayer. Development and Validation of a Composite Biomarker Predictive of Palbociclib + Endocrine Treatment Benefit in Early Breast Cancer: PENELOPE-B and PALLAS Trials [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 PD17-05.

Use of distinct molecular signatures of appendiceal cancer subtypes to assess biomarker development.

189 Background: Appendiceal Neoplasms are diverse entities which have a variety of clinical presentations, histologic subtypes, biologic behavior, and patient outcomes. Advanced disease at initial diagnosis is not uncommon and therapeutic options are limited. This study aims to identify potential targets to develop epigenetics-based therapeutics/biomarkers. Methods: Archival (FFPE) specimens were collected from 25 patients with histologically confirmed and subtyped appendiceal neoplasia and 16 age-matched non-neoplastic controls. Four appendiceal neoplastic processes identified: Low grade appendiceal neoplasm (LAMN), pseudomyxoma peritonei (intra-peritoneal spread of LAMN), conventional-type adenocarcinoma and goblet cell adenocarcinoma. Gene expression profiling was performed using the HTG EdgeSeq Oncology Biomarker panel. Differentially expressed genes (DEGs) were selected with cut-off threshold FDR&lt;0.05 and analyzed using Qlucore Explorer and functional analyses via Ingenuity Pathway Analysis (IPA). Results: Four appendiceal neoplastic processes identified: pseudomyxoma peritonei (20%), low-grade appendiceal mucinous neoplasm (24%), adenocarcinoma (24%), and goblet cell adenocarcinoma (32%). Unsupervised hierarchical clustering identified 498 DEGs (424 up, 74 down) between all appendiceal cancer versus age-matched controls that showed enrichment in B-cell inflammatory signaling, senescence, cell cycle regulation, PI3K/AKT signaling, cell cycle checkpoint control pathways, and sets relating to NANOG in stem cell pluripotency (all p0.001). G1/S checkpoint, apoptosis, ATM, and PTEN signaling are downregulated (p0.001). TP53, E2F1, IL6, IFNG, TNF, and HGF (p0.001) regulated overlapping genes. Four-subgroups comparison found 67 DEGs involved in T-helper cell differentiation, immune-mediated apoptosis, T-cell fatigue, and PD1/PDL1 signaling to be dysregulated (p0.001). IFNG, FGF2, POU5F1, TNF, and MYC (p0.001) were upstream regulators. Conclusions: In this study, downregulated gene expressions in appendiceal cancers are related to cell proliferation and death control mechanisms, with a trend towards overexpression of inflammatory and cell cycle pathways. It is postulated that genes involved in these pathways are hypermethylated. More research is being undertaken on epigenetics-based biomarkers for diagnosis and prognosis. Given PI’s decades of experience in biomarker research, we aim to develop a "single blood test" to expand bedside monitoring alternatives that align with the National Cancer Moonshot Initiative to discover cancer early, when the likelihood of a cure is highest.

P-481 Characterization of the gene expression profile of breast cancer tumours in patients undergoing ovarian stimulation for fertility preservation purposes

Abstract Study question Could controlled ovarian stimulation (COS) protocols with letrozole supplementation induce changes in the molecular footprints of estrogen receptor-positive (ER+) breast cancer (BC) women? Summary answer Ovarian stimulation with letrozole and gonadotropins does not change the gene expression profile of estrogen receptor-positive (ER+) breast malignant tumours as compared to non-stimulated patients. What is known already Fertility preservation (FP) strategies in ER+ BC patients involve modifications of the classical COS protocols to minimise the impact of estrogen levels during ovarian stimulation. The most common approaches are using the aromatase inhibitor letrozole to block the conversion of testosterone into estradiol in the granulosa cells or the use of tamoxifen to block the ER in the breast tissue. Although retrospective studies suggest that these strategies will have no impact on patient’s survival and disease progression, little is known regarding the effect of the ovarian stimulation on the genomic fingerprints of the exposed cancerous tissue. Study design, size, duration Retrospective, non-randomized, comparative study. Gene expression profiles were compared in biopsies at diagnosis and during excision surgery in two groups of ER+ BC patients: Patients undergoing COS for FP before surgery (COS group, n = 10) and patients not undergoing any type of COS (Control group, n = 11). The former received letrozole (5mg/day) supplements during COS. The latter included 7 patients not doing FP and 4 undergoing ovarian cortex cryopreservation. Patients were recruited between 2009 and 2021. Participants/materials, setting, methods Single 5-µm sections of formalin-fixed, paraffin-embedded (FFPE) needle core biopsy (NCB) and breast surgery (BS) tumour samples were obtained prior to starting gonadotoxic therapies. Differential gene expression and gene set analysis (GSA) was performed in tumour samples before (NCB-sample) and after COS (BS-sample). Simultaneous, quantitative detection of 2549 genes associated with tumour biology was performed with the HTG EdgeSeq Oncology Biomarker Panel. Tumour cell proliferation was also assessed by Ki67 staining. Main results and the role of chance Patients were younger in the COS group (COS: 30.2±3.4 vs Control: 34.5±2.3; p = 0.004). None of the patients experienced any relapse during the observation period. The length of COS was 11.5±3.5 days and 10.6±6.8 MII oocytes were vitrified. The exploratory analysis using principal component analysis revealed no relationships between the two BC biopsy samples and gene expression levels in the experimental groups. The differential expression analysis just revealed 6 genes significantly over-expressed after ovarian stimulation (DUSP1, FOS, EGFR1, NR4A1, JUN and CYR61). From these, DUSP1, FOS, and NR4A1 were also significantly upregulated in the unstimulated group. The GSA showed the cytochrome P450 pathway was significantly enriched after COS. No pathways related to cell proliferation were differentially expressed between groups. However, in unstimulated patients, 6 KEGG pathways were upregulated, with the cytokine-cytokine receptor interaction and the Jak-STAT signalling pathway being the most enriched. Ki67 immunostaining showed no differences in cell proliferation after COS (22.8±7.2 vs 24.0±6.9, p = 0.723). The rate of cell proliferation also remained constant in unstimulated patients (22.9±8.4 vs 25.1±8.1, p = 0.560). Limitations, reasons for caution The main limitations of the present study are the retrospective design and the associated risk of bias, although the population is very homogeneous when it comes to clinical characteristics; and the fact that molecular footprints are only interim markers of survival/response to the treatment. Wider implications of the findings The present study provides molecular evidence supporting the safety of COS using letrozole for oocyte vitrification in young BC patients undergoing ovarian stimulation. This information is crucial to support patient guidance during discussions about fertility preservation. Trial registration number Not applicable

Abstract PD2-04: Molecular plasticity of luminal breast cancer and response to CDK 4/6 inhibition - The biomarker program of the PENELOPE-B trial investigating post-neoadjuvant palbociclib

Abstract Background: Molecular plasticity of breast cancer is crucial for the development of therapy-resistant disease. In this investigation, we studied changes in molecular signatures between pretherapeutic (pre-Tx) and post-therapeutic (post-NACT) tumor samples from patients included in the PENELOPE-B (NCT01864746) trial. The phase III PENELOPE-B study did not show a significant benefit from palbociclib in women with centrally confirmed HR+, HER2- primary breast cancer without a pathological complete response after taxane-containing neoadjuvant chemotherapy (NACT) and at high-risk of relapse (CPS-EG score ≥3 or 2 and ypN+) (Loibl et al. JCO 2021). However, first translational investigations showed that a small number of patients with a luminal-B tumor subtype, based on absolute intrinsic molecular subtyping (AIMS, Paquet &amp; Hallet, JNCI 2014) subtyping after NACT, had a numerical benefit from post-NACT palbociclib. We have therefore extended the analysis and included a cohort of paired pre-Tx and post-NACT samples. Methods: We investigated gene expression in pre-Tx (n=259) tumor tissue samples using the HTG EdgeSeq Oncology Biomarker Panel including 2549 genes (HTG Molecular Diagnostics Inc.); for the same patients the same panel on post-NACT residual tumor samples were available. The paired samples were selected based on a case-cohort approach. Based on 91 genes of this panel, the AIMS subtype was calculated. In addition, we performed exploratory biomarker analyses to identify genes and gene signatures with prognostic and predictive relevance. After completion of NACT, PENELOPE-B patients were randomized to palbociclib versus placebo in addition to standard endocrine therapy. Results: The prevalence of AIMS subtypes, in particular LumA vs LumB, changed in pre-Tx and post-NACT tumors. In the pre-Tx samples, 115 (44%) and 123 (47%) of tumors had LumA and LumB subtypes, respectively, as expected from a high-risk cohort. However, in the post-NACT samples, LumA tumors were predominant (n=183, 71%) over LumB (n=30; 12%). 78 (30%) and 6 (2%) tumors switched their subtype from LumB to LumA and LumA to LumB, respectively. For further analyses, we compared the groups of low proliferating (LumA and NormL) and high proliferating subtypes (LumB, BasalL and HER2E). In bivariable Cox regression analysis, the grouped pre-Tx and post-NACT AIMS subtypes were independent prognostic factors for iDFS: HR=1.85 (1.16-2.98, p=0.011) for pre-Tx LumB/BasalL/HER2E vs LumA/NormL and HR=2.18 (1.24-3.84, p=0.007) for post-NACT. Similar results were found when adjusted for prognostic clinical factors and for DDFS and OS endpoints although the pre-Tx subtype did not reach significance. These and further Cox models investigating interaction effects show that patients with tumors developing from high (pre-Tx) to low proliferation (post-NACT) had a higher iDFS risk compared to stable low proliferating tumors but a lower iDFS risk compared to stable high proliferating tumors. Neither in the pre-Tx LumB/BasalL/HER2E nor in the pre-Tx LumA/NormL subgroup a benefit from palbociclib was observed. Based on the results of the AIMS subtyping, we extended the exploratory analysis to identify genes that might be involved in the prognostic and predictive effects as well as genes driving the subtype switch. The analysis is ongoing and the relevant genes will be presented at the conference. Conclusions: Our findings show that the switch from high-risk molecular subtypes (in particular LumB) to low-risk subtypes (in particular LumA) is common in neoadjuvant therapy of luminal tumors. The adaptation of luminal high-risk tumors to therapy-induced stress is crucial for the clinical outcome and the results suggest that molecular defined tumor subtypes might not be as stable as originally thought. Citation Format: Carsten Denkert, Frederik Marmé, Miguel Martin, Michael Untch, Hervé Bonnefoi, Sung-Bae Kim, Harry Bear, Agnieszka Witkiewicz, Seock-Ah Im, Angela DeMichele, Laura Van’t Veer, Nicole McCarthy, Thorsten Stiewe, Karen A. Gelmon, José A. García-Sáenz, Catherine M. Kelly, Toralf Reimer, Erik Knudsen, Nicholas Turner, Federico Rojo, Peter A. Fasching, Julia Teply-Szymanski, Yuan Liu, Masakazu Toi, Hope S. Rugo, Michael Gnant, Andreas Makris, Bärbel Felder, Karsten Weber, Sibylle Loibl. Molecular plasticity of luminal breast cancer and response to CDK 4/6 inhibition - The biomarker program of the PENELOPE-B trial investigating post-neoadjuvant palbociclib [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 PD2-04.

Targeted gene expression profiling of inverted papilloma and squamous cell carcinoma

Inverted papilloma (IP) is a sinonasal tumor with a well-known potential for malignant transformation. The purpose of this study was to identify the genes and pathways associated with IP, with progression to carcinoma-in-situ and invasive carcinoma. To determine genes and molecular pathways that may indicate progression and correlate with histologic changes, we analyzed six IP without dysplasia, five IP with carcinoma-in-situ, and 13 squamous cell carcinoma ex-IP by targeted sequencing. The HTG EdgeSeq Oncology Biomarker Panel coupled with next-generation sequencing was used to evaluate 2560 transcripts associated with solid tumors. Progressive upregulation of 11 genes were observed (CALD1, COL1A1, COL3A1, COL4A2, COL5A2, FN1, ITGA5, LGALS1, MMP11, SERPINH1, SPARC) in the order of invasive carcinoma>carcinoma-in-situ>IP without dysplasia. When compared with IP without dysplasia, more genes are differentially expressed in invasive carcinoma than carcinoma-in-situ samples (341 downregulated/333 upregulated vs. 195 downregulated/156 upregulated). Gene set enrichment analysis determined three gene sets in common between the cohorts (epithelial mesenchymal transition, extracellular matrix organization, and coagulation). Progressive upregulation of genes specific to IP malignant degeneration has significant clinical implications. This panel of 11 genes will improve concordance of histologic classification, which can directly impact treatment and patient outcomes. Additionally, future studies on larger tumor sets may observe upregulation in the gene panel that preceded histologic changes, which may be useful for further risk stratification.

Subgroup of post-neoadjuvant luminal-B tumors assessed by HTG in PENELOPE-B investigating palbociclib in high risk HER2-/HR+ breast cancer with residual disease.

519 Background: About one third of patients with hormone-receptor-positive (HR+), HER2‐ primary breast cancer with residual invasive disease after neoadjuvant chemotherapy will relapse despite adjuvant endocrine therapy. Therapeutic inhibition of cyclin-dependent kinase 4 and 6 (CDK 4/6) by palbociclib combined with endocrine therapy demonstrated highly relevant efficacy in metastatic breast cancer. The phase III PENELOPE-B (NCT01864746) study did not show a significant benefit from palbociclib in women with centrally confirmed HR+, HER2- primary breast cancer without a pathological complete response after taxane‐containing neoadjuvant chemotherapy and at high-risk of relapse (CPS‐EG score ≥3 or 2 and ypN+) for the primary endpoint (Loibl et al. JCO 2021). Methods: After completion of neoadjuvant chemotherapy and locoregional therapy, PENELOPE-B patients were randomized (1:1) to receive 13 cycles (1 year) of palbociclib 125mg daily or placebo on days 1-21 in a 28d cycle in addition to standard endocrine therapy. Analysis of the primary endpoint of invasive disease-free survival (iDFS) was planned after 290 events. Secondary objective included iDFS in luminal-B group by treatment. Gene expression in post-neoadjuvant surgical residual tumor tissue samples was profiled using the HTG EdgeSeq Oncology Biomarker Panel targeting 2559 genes (HTG Molecular Diagnostics Inc.). Based on 91 genes of this panel the AIMS subtype (Paquet &amp; Hallett, JNCI 2014) was calculated. Results: Gene expressions were measured in tumors from 906 of 1250 (72%) PENELOPE-B patients; 663 had LumA subtype, 64 LumB, 135 NormL, 16 BasalL, and 28 HER2E. Compared to LumA the LumB patients were older, had higher post-neoadjuvant Ki-67, higher risk status (CPS-EG), and higher grade; no significant correlation was found for the region of participating sites, cT, ypT, and ypN. Patients with LumB tumors had an estimated 3-year iDFS of 71.9% with palbociclib vs 44.8% with placebo HR = 0.50 (0.24-1.05); outcome was similar in patients with LumA tumors (3-year iDFS 83.9% vs 79.5%, HR = 0.93 (0.68-1.28), interaction p = 0.132); this was confirmed in multivariable analyses. Ki-67 by IHC and proliferation biomarkers from the HTG panel also showed no significant interaction with treatment. Conclusions: PENELOPE-B did not show a benefit from the addition of 1 year palbociclib to endocrine therapy compared to placebo in the total enrolled high-risk primary breast cancer population. However, the small group of luminal-B tumors (n = 64) derived benefit from palbociclib, although without a statistically significant interaction. Further investigation is required in a larger cohort to validate a palbociclib benefit that might be confined to this group.

CCNE1 mRNA and cyclin E1 protein expression as predictive biomarkers for efficacy of palbociclib plus fulvestrant versus capecitabine in the phase III PEARL study.

1014 Background: The randomized PEARL trial found no superiority of palbociclib plus endocrine therapy over capecitabine in patients (pts) with metastatic HR-positive, HER2-negative breast cancer resistant to prior aromatase inhibitors (Martin M, Ann Oncol 2020). Gene expression analysis showed high CCNE1 mRNA ( CCNE1) conferring relative resistance to palbociclib in the PALOMA-3 trial (Turner N, JCO 2019), but further validation is needed. Cyclin E1 protein (cyclin E1) expression in this context has not been studied in randomized trials. We explored CCNE1 and cyclin E1 as predictive biomarkers in tumor samples from the PEARL study. Methods: Formalin-fixed paraffin-embeded tumor samples were retrieved from pts enrolled in PEARL cohort 2 (palbociclib (PAL) + fulvestrant (FUL) vs capecitabine (CAPE)). We measured CCNE1 using the HTG EdgeSeq Oncology Biomarker Panel (HTG Molecular Diagnostics). Cyclin E1 immunohistochemistry (IHC) staining was performed using specific mouse monoclonal antibody HE12 (Abcam) and scored as percentage of invasive nuclei stained (0-100%). CCNE1 and cyclin E1 correlations were explored using Pearson coefficients. Cox regression models were used for progression free-survival (PFS) analyses using expression levels split by median, to define high ( &gt; median values) vs. low expression. Site of disease and prior chemotherapy were used as confounders in multivariate models. Results: Analyses were conducted in 219 pts (47% receiving PAL + FUL and 53% CAPE) with available tumors, with the analysed patients representative of the overall study. Most samples were from the archival primary (72%), obtained &gt; 5 years before this analysisº (74%). CCNE1 and cyclin E1 were only moderately correlated (r = 0.5). Median CCNE1 was higher in metastatic vs primary (7.37 vs 6.94, p &lt; 0.01), and in luminal B and non-luminal subtypes compared to luminal A (p &lt; 0.001). In patients with high CCNE1 expression, median PFS on CAPE was 10.35 and PAL + FUL was 5.68 (HR = 1.63, 95% CI 1.02-2.59, p = 0.042). In patients with low CCNE1 expression, median PFS on CAPE was 9.43 and PAL + FUL was 8.97 (adjusted HR = 0.93, 95% CI 0.59-1.48, p = 0.762, interaction p = 0.072). Median cyclin E1 protein was higher in luminal B and non-luminal subtypes compared to luminal A (p &lt; 0.01). Cyclin E1 protein expression was not predictive of treatment effect (high cyclin E1 expression CAPE vs PAL + FUL HR = 1.17, low cyclin E1 expression CAPE vs PAL + FUL HR = 1.21, interaction p = 0.977). Conclusions: High tumor CCNE1 mRNA expression identified patients with relative resistance to palbociclib plus fulvestrant, validating prior observations although without statistical significance for interaction. Assessment of Cyclin E1 protein expression did not show predictive value. Investigation treatments to enhance CDK4/6 inhibitor efficacy in tumors with high CCNE1 expression is warranted. Clinical trial information: NCT02028507 .

Cross-platform comparison of immune-related gene expression to assess intratumor immune responses following cancer immunotherapy

Neoadjuvant immunotherapy can induce immune responses within the tumor microenvironment. Gene expression can be used to assess responses with limited amounts of conventionally-fixed patient-derived samples. We aim to assess the cross-platform concordance of immune-related gene expression data. We performed comparisons across three panels in two platforms: Nanostring nCounter® PanCancer Immune Profiling Panel (nS), HTG EdgeSeq Oncology Biomarker Panel (HTG OBP) and Precision Immuno-Oncology Panel (HTG PIP). All tissue samples of 14 neoadjuvant GM-CSF treated, 14 neoadjuvant Provenge treated, and 12 untreated prostate cancer patients were radical prostatectomy (RP) tissues, while 6 prostatitis patients and 6 non-prostatitis subjects were biopsies. For all 52 patients, more than 90% of the common genes were significantly correlated (p < 0.05) and more than 76% of the common genes were highly correlated (r > 0.5) between any two panels. Co-inertia analysis also demonstrated high overall dataset structure similarity (correlation>0.84). Although both dimensionality reduction visualization analysis and unsupervised hierarchical cluster analysis for highly correlated common genes (r > 0.9) suggested a high-level of consistency across the panels, there were subsets of genes that were differentially expressed across the panels. In addition, while the effect size of the differential testing for neoadjuvant treated vs. untreated localized prostate cancer patients across the panels were significantly correlated, some genes were only differentially expressed in the HTG panels. Finally, the HTG PIP panel had the best classification performance among the 3 panels. These differences detected may be a result of the different panels or platforms due to their technical setting and focus. Thus, researchers should be aware of those potential differences when deciding which platform and panel to use.

Abstract PS2-01: Plk1 expression &amp; efficacy of palbociclib in advanced hormonal receptor-positive breast cancer patients from PEARL study (GEICAM 2012-03)

Abstract Background: CDK 4/6 inhibitors (CDK 4/6i) with endocrine therapy (ET) combination therapy have improved outcomes in patients (pts) with hormonal receptor positive (HR+)/human epidermal growth factor receptor negative (HER2-) advanced breast cancer (ABC). However, most pts eventually develop resistance to these drugs, and one third never respond. Aside from HR positivity, predictive markers of clinical benefit from CDK 4/6i remains elusive. We aimed to identify biomarkers of response to palbociclib (PAL) and analyze potential therapeutic targets to reverse resistance. Methods: PEARL trial is a multicenter phase 3 study that assigned 601 postmenopausal HR+/HER2- ABC pts, whose disease progressed on aromatase inhibitors (AIs), to receive PAL + ET vs capecitabine (CAPE). We performed a differential gene expression analysis in pre-treatment tumors in extreme responders to PAL using the HTG EdgeSeq Oncology Biomarker Panel (HTG Molecular Diagnostics, Inc.), containing 2534 cancer related genes. Samples were subset in 2 categories: refractory (progressive disease as best response) vs sensitive (progression-free survival (PFS) within the upper quartile). Cox regression and Significance Analysis of Microarrays (SAM) analysis adjusting for multiple comparisons were performed. Results: We analyzed 455 (75.7%) pts with pre-treatment tumors available [from them, PAL + ET arm: 229 (50.3%) pts; CAPE arm: 226 (49.7%) pts]. Fifty genes (false discovery rate (FDR)&amp;lt;0.05) were differentially expressed in pts sensitive vs refractory to PAL (E2F target genes, epithelial-to-mesenchymal transition (EMT) and cell cycle genes, mainly). Unsupervised hierarchical clustering of pts based on the expression of these genes revealed two clusters. Cluster 1 is composed mostly of resistant tumors, highly proliferative (Ki67≥20%: 70%) with a great proportion of luminal B (59%) and non-luminal tumors (19%). Cluster 2 is composed of sensitive, low proliferative (Ki67&amp;lt;20%: 58%), mostly luminal A tumors (75%). There was no difference in ESR1 mutations distribution between the two clusters (Table 1). Forty genes were up-regulated and associated with resistance, including CCNE1 and PLK1 (Polo Like Kinase 1). In the whole cohort, pts with high levels (&amp;gt; median) of PLK1 (PLK1-high) treated with PAL, had a worse PFS in a multivariate model (5.7 months (m) vs 9.3 m of median PFS in PLK1-High vs -Low; HR=1.64, 95% CI (1.25-2.34), p=0.0008; adjusted model for confounders: age, site of disease, sites of metastasis, prior chemotherapy and Ki67). There were no differences in population treated with CAPE (9.9 m vs 9.4 m, PLK1-High vs -Low; HR=0.82, 95% CI (0.56-1.21), p=0.3189). In the METABRIC cohort, PLK1-High was associated with worse overall survival in HR+/HER2- BC but not in triple negative nor in HER2+ tumors. Among HR+/HER2- tumors, PLK1 expression was higher in luminal B and HER2-enriched intrinsic subtypes. We interrogated DepMap database and found that in BC cells lines there was an inverse correlation between PLK1 expression and effect on cell viability of CDK4 CRISPR knock-out (Pearson correlation r:0.54, p=0.009), but not of CDK6 knock-out. Also, HR+/HER2-/High Ki67 BC cell lines (HCC1428, EFM19 and MCF7) showed resistance to PAL on cell proliferation assays but sensitivity to the PLK1 inhibitor BI-2536. Conclusion: High expression of PLK1 is associated with intrinsic resistance to PAL and ET, this might be overcome with PLK1 inhibition. Table 1PATIENT CHARACTERISTICSCluster 1Cluster 2ALLn=57n=47n=104RespondersSensitive42 (73.68%)14 (29.79%)56 (53.85%)Refractory15 (26.32%)33 (70.21%)48 (46.15%)ESR1Mutated9 (15.79%)13 (27.66%)22 (21.15%)Wild type45 (78.95%)34 (72.34%)79 (75.96%)Unknown3 (5.26%)0 (0%)3 (2.88%)PriorQTN42 (73.68%)31 (65.96%)73 (70.19%)Y15 (26.32%)16 (34.04%)31 (29.81%)SubtypeLumA43 (75.44%)10 (21.28%)53 (50.96%)LumB14 (24.56%)28 (59.57%)42 (40.38%)Non Luminal0 (0%)9 (19.15%)9 (8.65%)MetastasisOne21 (36.84%)15 (31.91%)36 (34.62%)Multiple36 (63.16%)32 (68.09%)68 (65.38%)KI67 20%KI67&amp;lt;2033 (57.89%)7 (14.89%)40 (38.46%)KI67≥2016 (28.07%)33 (70.21%)49 (47.12%)Unknown8 (14.04%)7 (14.89%)15 (14.42%)Objective ResponseComplete1 (1.75%)0 (0%)1 (0.96%)Partial16 (28.07%)6 (12.77%)22 (21.15%)Progressive15 (26.32%)33 (70.21%)48 (46.15%)Stable25 (43.86%)8 (17.02%)33 (31.73%) Citation Format: Angel Guerrero- Zotano, Christoph Zielinski, Miguel Gil-Gil, Manuel Ruiz-Borrego, Eva M. Ciruelos, Montserrat Munoz, Begoña Bermejo, Mireia Margeli, Antonio Antón, Tibor Csöszi, Andrés García-Palomo, Ana Santaballa, Jose Luis Alonso, Antonio Fernández, Massimo Corsaro, Jesús Herranz, Paula López, Rosalia Caballero, Christiane Thallinger, Miguel Martin. Plk1 expression &amp; efficacy of palbociclib in advanced hormonal receptor-positive breast cancer patients from PEARL study (GEICAM 2012-03) [abstract]. In: Proceedings of the 2020 San Antonio Breast Cancer Virtual Symposium; 2020 Dec 8-11; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2021;81(4 Suppl):Abstract nr PS2-01.

Analysis of Adct-602 Pre-Clinical Activity in B-Cell Lymphoma Models and Identification of Potential Biomarkers for Its Activity

Analysis of Adct-602 Pre-Clinical Activity in B-Cell Lymphoma Models and Identification of Potential Biomarkers for Its Activity

Abstract 5880: Pan-cancer gene expression analysis of tissue microarray using EdgeSeq Oncology Biomarker Panel and a cross-comparison with ERBB3 immunohistochemical analysis

Abstract Background: Antibody-drug conjugate (ADC) is one of the recent break-through approaches of targeting tumors, in which chemotherapeutic agents are attached via linkers to antibodies recognizing membrane proteins on tumor cells and the agents are delivered directly into tumor cells. In ADC drug development, it is important to determine the antibody-target gene expression levels in tumor cells as well as normal tissues. Methods: We characterized an established pan-cancer tissue microarray set by gene expression analysis using HTG EdgeSeq Oncology Biomarker Panel (2,867 genes for 2,376 samples including 17 solid cancer types and 25 normal tissues) and compared it with immunohistochemistry (IHC) result of ERBB3, which is a target of anti-ERBB3 ADC, U3-1402, being investigated in clinical trials. Furthermore, these data were compared with RNA-seq and Reverse-Phase Protein Array (RPPA) publicly available data sets from The Cancer Genome Atlas (TCGA) and Cancer Cell Line Encyclopedia (CCLE). Results: Gene signature analyses indicated a similarity of cancer type characteristics between the tissue microarray cohort and TCGA data, such as richness of cancer-associated fibroblast and immune cell infiltrate. As expected, ERBB3 mRNA expression results by EdgeSeq were similar to TCGA RNA-seq data, suggesting an epithelial-biased expression pattern. The result of IHC for ERBB3 showed overall correlation with EdgeSeq mRNA data, such as high IHC scores in melanoma. Interestingly several cancer types and subtypes showed low IHC signals relatively to EdgeSeq data, such as low IHC signals in hepatocellular carcinoma despite of a certain mRNA expression levels. RNA-seq and RPPA data of CCLE did not show such discordance between mRNA and protein levels, which could be a result of the regulation due to in vivo environmental factors. In breast cancer, ERBB3 IHC score levels were generally lower in triple negative breast cancer (TNBC), while TNBC-Mesenchymal subtype showed relatively high IHC score levels comparable to non-TNBC tumor types. In stomach cancer, intestinal subtype showed higher IHC score levels than diffuse subtype despite similar mRNA expression levels. These subtype-specific protein-level regulations were also observed in TCGA RNA-seq and RPPA data comparison. Conclusion: These results suggested a unique protein-level regulation of ERBB3 in tumor cells. It would be important to investigate the significance of protein-level antibody-target expressions on tumor samples, such as IHC on tissue microarray, for antibody-based drug development. Citation Format: Koichiro Inaki, Naoyuki Maeda, Tomoko Shibutani, Serenella Eppenberger-Castori, Stefan Nicolet, Kumiko Koyama, Yui Tanaka, Yuki Kaneda, Kenichi Wakita, Yoshinobu Shiose, Masato Murakami. Pan-cancer gene expression analysis of tissue microarray using EdgeSeq Oncology Biomarker Panel and a cross-comparison with ERBB3 immunohistochemical analysis [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5880.

Abstract 2016: Comparison of HTG-edge targeted RNA sequencing platform with whole transcriptome RNA sequencing for clinical biomarker studies

Abstract Purpose HTG EdgeSeq Next Generation Sequencing (NGS) technology is a targeted RNA Sequencing (RNA-Seq) platform based on quantitative nuclease protection chemistry, which requires substantially less sample input than RNA-Seq (one formalin-fixed paraffin-embedded (FFPE) slide versus 4-6 slides). The automated library processor allows fast downstream NGS analysis, making it a promising technology for clinical application and companion diagnostics, especially in samples with limited availability. Given the differences from standard RNA-Seq protocols, it is critical to investigate the performance of HTG EdgeSeq compared to whole transcriptome RNA-Seq technology. We investigated the concordance between HTG EdgeSeq and RNA-Seq on a set of FFPE samples. Procedure The HTG EdgeSeq Oncology Biomarker Panel (OBP) is comprised of 2558 genes including genes relevant to immuno-oncology and oncology pathways. To evaluate the concordance between the two platforms, 57 FFPE samples from diverse tumor types and varying tumor content (16-100%) were analyzed. One slide was analyzed per sample with the HTG EdgeSeq OBP, while four to six slides were analyzed with whole transcriptome analysis using the Illumina TruSeq Total RNA library kit. A subset of samples and genes were selected for analysis by RT-qPCR using a customized Taqman 15 gene array. Data from the three platforms was analyzed and compared for concordance. Data Summary Fifty-three samples passed QC on the HTG EdgeSeq platform, while only 29 samples passed a distribution-based QC of the RNA-Seq. For samples that passed QC using both technologies, the median Spearman correlation was 0.73 (25th percentile was 0.70), indicating good, but not linear correlation. In the samples that failed RNA-Seq QC, the median Spearman correlation was 0.30. Probes with low correlation (&amp;lt;0.25) between HTG EdgeSeq and RNA-Seq had an average interquartile range of expression less than 2 log2(CPM) indicating low dynamic range of expression. The HTG EdgeSeq platform showed a tendency towards more detection of low abundance genes indicating either higher sensitivity or higher background. Comparison of RT-qPCR to HTG EdgeSeq showed good sample to sample correlation however, in many cases, genes with low expression in HTG EdgeSeq were undetectable in RT-qPCR leaving the low-end sensitivity of HTG EdgeSeq unclear. Conclusions HTG EdgeSeq produced fewer QC failures compared to whole transcriptome RNA-Seq while using less tissue. The resulting data showed sound correlation between RNA-Seq, HTG EdgeSeq, and RT-qPCR platforms, especially for genes with moderate to high expression, enabling relatively low risk translation between biomarker platforms. Further, these results were consistent across various cancer types. The characteristics demonstrated in this study support the potential use of HTG EdgeSeq for clinical biomarker applications. Citation Format: Dennis O'Rourke, Jorge F. Sanchez-Garcia, P. Alexander Rolfe, Alice Huang, Danyi Wang, Juergen Scheuenpflug, Zheng Feng. Comparison of HTG-edge targeted RNA sequencing platform with whole transcriptome RNA sequencing for clinical biomarker studies [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 2016.

Integrated-omics of MRI-visible and -invisible prostate cancer identifies molecular correlations with clinical outcome.

e17506 Background: The inclusion of multiparametric MRI (mpMRI) to prostate cancer (PCa) diagnostics has increased the detection rate and has improved the prediction rate of clinically significant PCa. Yet, mpMRI has a false negative rate of 12.6%, missing 6-13% of clinically significant PCa. The mechanisms underlying MRI visibility are poorly understood; therefore, to probe the molecular and cellular underpinnings of PCa MRI visibility, we profiled tissue from Gleason score and clinically matched patients with MRI-visible and MRI-invisible PCa who underwent radical prostatectomy. Methods: MRI-invisible (n = 7) and MRI-visible (n = 8) PCa tumors were evaluated with multiplex immunofluorescence (MxIF; 14 markers) and were subjected to gene expression profiling using the HTG EdgeSeq Oncology Biomarker Panel (2,549 genes). Gene expression analysis was also performed using The Cancer Genome Atlas (TCGA), including normal prostate (n = 52) and PCa (n = 387) tissue. Analyses were performed using the BostonGene automated pipeline. Results: MpMRI-visible PCa tumors (62.5%) displayed compact epithelial tumor architecture compared with mpMRI-invisible PCa tumors (28.5%). mpMRI-visible PCa had higher malignant cell density (p = 0.04) and increased neighboring malignant cells (p = 0.07), correlating with MRI-visible PCa complex tumor architecture (r = 0.49 for neighboring malignant cells vs tumor cell density). Tumor stromal organization differences were determined by measuring Wasserstein distances between distributions, and mpMRI-invisible PCa stroma appeared more similar to normal tissue. The visible group exhibited lower expression of stroma-enriched genes such as filamin C (FLNC) (FDR &lt; 0.1) and cellular adhesion-related genes (FDR &lt; 0.4), with gene expression signatures markedly different compared to normal prostate tissue. Higher malignant cell density, neighboring malignant cells, and Wasserstein distances, and low FLNC expression – all mpMRI visibility characteristics – were associated with patient relapse (p = 0.02). Low stroma signature expression in the TCGA cohort correlated with inferior PCa PFS (p = 0.005). Conclusions: This is the first integrated multi-omics analysis of clinically matched mpMRI-visible and -invisible PCa. mpMRI-invisible tumors exhibited molecular, cellular, and structural characteristics more akin to normal prostate tissue, which may render them undetectable by imaging. A stroma-associated gene signature, a mpMRI-invisible tumor feature, correlated with better PCa clinical outcomes.

Abstract GS2-07: Results from PEARL study (GEICAM/2013-02_CECOG/BC.1.3.006): A phase 3 trial of Palbociclib (PAL) in combination with endocrine therapy (ET) versus Capecitabine (CAPE) in hormonal receptor (HR)-positive/human epidermal growth factor receptor (HER) 2-negative metastatic breast cancer (MBC) patients (pts) whose disease progressed on aromatase inhibitors (AIs)

Abstract Background: PAL is approved for use in combination with ET in HR-positive, HER2-negative MBC pts in first and later lines. However, the relative value of PAL plus ET versus (vs) chemotherapy (CT) in pretreated pts is not established yet. PEARL is a multinational, open label, controlled, randomized phase 3 trial comparing the efficacy and safety of PAL in combination with ET (exemestane [EXE] or fulvestrant [FUL]) vs CAPE in postmenopausal women with HR-positive/HER2-negative MBC whose disease progressed on AIs. Methods: The study had two successive cohorts. In cohort 1 (C1) patients were randomized 1:1 to EXE+PAL vs CAPE. In 2016, after data showing that ESR1 mutations may induce resistance to AIs but not to FUL, cohort 2 (C2), with FUL+PAL vs CAPE, was added to the trial. Stratification criteria were site of disease (visceral vs non-visceral), prior sensitivity to ET (yes vs no), prior CT for MBC (yes vs no) and country. The study has two co-primary objectives: to demonstrate 1) the superiority of PAL+FUL over CAPE in progression-free survival (PFS) and 2) the superiority of PAL+ET (EXE or FUL) over CAPE in PFS in pts with ESR1 wild type (wt) tumors, measured in ctDNA, at study entry. In order to have an 80% power to detect a difference in median PFS between CAPE (6 months) and either PAL+FUL or PAL+ET ESR1 wt (9 months), for a hazard ratio (HR) of 0.667 and a significance level of 5%, a sample size of 300 pts and 193 events were required for each analysis. An exploratory analysis of intrinsic BC subtypes determined by an algorithm derived from the HTG EdgeSeq Oncology Biomarker Panel was also performed. Results: From March 2014 to July 2018, 601 pts were recruited (296 in C1 and 305 in C2) in 4 countries. Median age was 60 (C1) and 61 (C2) years; 26.4% (C1) and 28.2% (C2) had ESR1 mutations; 66.6% (C1) and 65.2% (C2) had visceral disease; 71.3% (C1) and 79.0% (C2) had prior sensitivity to hormonal treatment; 30.1% (C1) and 26.9% (C2) received one prior chemotherapy for MBC; 19.6% (C1) and 26.6% (C2) had not received any prior line for MBC.C2 results: with a median follow-up of 13.5 m (range 0.0 to 30.7), the median PFS was 7.5 m in PAL+FUL vs 10 m in CAPE (adjusted HR [HRa] [confidence interval (CI) 95%]: 1.09 [0.83, 1.44]; p=0.537). The Objective Response Rate (ORR) was 26.7% for PAL+FUL vs 33.3% for CAPE. The analysis of BC subtypes was performed for 74.4% of pts, 89.1% (PAL+ET) and 91.5% (CAPE) had luminal tumors. For luminal pts median PFS was 7.5 m in PAL+FUL vs 10 m in CAPE (HR [CI 95%]: 1.07 [0.77, 1.50]; p=0.684); for non-luminal pts it was 4.4 m in PAL+FUL vs 14.8 m in CAPE (HR [CI 95%]: 2.39 [0.81, 7.08]; p=0.116).ESR1 wt (C1 + C2) results: with a median follow-up of 19.0 m (range 0.0 to 56.3), the median PFS was 8.0 m in PAL+ET vs 10.6 m in CAPE (HRa [CI 95%]: 1.08 [0.85-1.36]; p=0.526). The ORR was 27.8% for PAL+ET vs 36.9% for CAPE.The analysis of BC subtypes was performed for 79.6% of pts, 89.2% (PAL+ET) and 91.8% (CAPE) had luminal tumors. For luminal pts median PFS was 9.3 m in PAL+ET vs 11.0 m in CAPE (HR [CI 95%]: 1.02 [0.77, 1.34]; p=0.913); for non-luminal pts it was 2.7 m in PAL+ET vs 13.7 m in CAPE (HR [CI 95%]: 3.19 [1.28, 7.95]; p=0.013). Most frequent grade 3-4 toxicities with EXE+PAL, FUL+PAL and CAPE respectively were neutropenia (57.4%, 55.7% and 5.5%) with febrile neutropenia (1.3%, 0.7% and 1.4%), hand/foot syndrome (0%, 0% and 23.5%) and diarrhea (1.3%, 1.3% and 7.6%). Conclusions: The PEARL study did not show a statistically superiority in PFS for PAL+ET vs CAPE in MBC pts progressing to AIs. No superiority of PAL+ET was observed in the luminal subgroup either. Treatment with PAL+ET was generally better tolerated than CAPE. Citation Format: Miguel Martín, Christoph Zielinski, Manuel Ruíz-Borrego, Eva Carrasco, Eva Ciruelos, Montserrat Muñoz, Begoña Bermejo, Mireia Margeli, Nicholas Turner, Maribel Casas, Antonio Antón, Tibor Csöszi, Massimo Corsaro, Laura Murillo, Serafín Morales, Emilio Alba, Cynthia H Bartlett, Maria Koehler, Angel Guerrero, Zsuzsanna Kahan, Miguel Gil-Gil. Results from PEARL study (GEICAM/2013-02_CECOG/BC.1.3.006): A phase 3 trial of Palbociclib (PAL) in combination with endocrine therapy (ET) versus Capecitabine (CAPE) in hormonal receptor (HR)-positive/human epidermal growth factor receptor (HER) 2-negative metastatic breast cancer (MBC) patients (pts) whose disease progressed on aromatase inhibitors (AIs) [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr GS2-07.

Variation in genetic polymorphisms and gene expression of HLA-E to predict outcomes in metastatic colorectal cancer (mCRC) patients (pts) treated with first-line FOLFIRI/cetuximab: Data from the phase III FIRE-3 trial.

245 Background: HLA-E is an MHC Class I antigen which inhibits NK cell activity by binding to CD94 receptor. Overexpression of HLA-E on crc cells is associated with poor prognostic outcomes and has recently been associated with microsatellite instability. As cet enhances NK cell activity which is dependent on HLAE, this study aims to assess whether differences in HLAE gene expression and polymorphisms lead to different outcomes. Methods: Genomic DNA from blood samples of pts treated with first-line FOLFIRI-cetuximab (cet, n = 129) and FOLFIRI-bevacizumab (bev, n = 107) was genotyped through the OncoArray, a custom array manufactured by Illumina. Gene expression levels were measured from 102 tumor samples of pts in the cet arm by HTG EdgeSeq Oncology Biomarker Panel. PFS and OS outcomes were investigated for an HLAE genetic polymorphism, rs1264457, where G &gt; A leads to glycine &gt; arginine, where glycine has stronger affinity for NK cell receptor. Results: FOLFIRI/cet and FOLFIRI/bev cohort characteristics: median FU (29.1/26.7mo); PFS (12.8/11.5mo); OS (49.8/31.4mo); RAS WT (64%/62%) and RAS mut (15%/16%). Overexpression of HLAE was associated with poor OS (log2 &gt; 13, 30 vs 34 mo) in cet arm (P &lt; .05). No significant association was observed with PFS. Multivariable analysis showed that HLAErs1264457 carriers with any A allele (n = 65) had better PFS than pts with G/G genotype (n = 18) in RAS WT pts treated with FOLFIRI/cet (12.9 vs 12.3 mo; HR = 2.36; 95%CI = 1.14-4.88; p = 0.021). A non-significant trend of improved OS was seen among carriers of A allele than G/G genotype (56.2 vs 40 mo). These effects were not observed in Ras mut pts. No significance was observed in FOLFIRI/bev overall or FOLFIRI/bev RAS WT pts. Conclusions: Overexpression of HLAE leads to poor OS among pts treated with FOLFIRI/cet. Poor PFS was seen among pts treated with cet who are RAS wt carrying polymorphisms allowing for increased binding to NK cell receptor, which in turn enhances HLAE inhibitory function of NK cell lysis. Addition of mAb against HLAE, which have been shown to restore NK cell mediated cell lysis, needs investigation in this pt population.

Genetic variants in R-Spondin/RNF43 complex and gene expression levels to predict efficacy of cetuximab (cet) in patients (pts) with metastatic colorectal cancer (mCRC): Data from the FIRE-3 phase III trial.

190 Background: Wnt signaling deregulation is a primary driver of colorectal carcinogenesis. RNF43 is a key suppressor of Wnt activation while R-Spodin inhibits RNF43 activity. RNF43 mutations are associated with the serrated neoplasia pathway, BRAF mutation and MSI. We hypothesized that genetic variants in the R-Spodin/RNF43 complex and corresponding genes expression levels may predict cetuximab efficacy in mCRC pts. Methods: Genomic DNA from blood samples of pts enrolled in the randomized FIRE-3 trial was genotyped through the OncoArray, a custom array manufactured by Illumina. The impact on outcome of 17 functional SNPs within RNF43/ ZNRF3, LGR4/5 and RSPO1/2/3 was analyzed in 129 pts treated with first-line FOLFIRI/cet and 107 pts treated with FOLFIRI/bevacizumab (bev). Gene expression levels were measured from tumor tissue samples from 102 pts in the cet arm by HTG EdgeSeq Oncology Biomarker Panel. False discovery rate (FDR) for gene expression analysis was computed using the Benjamini-Hochberg approach (significant Q &lt; 0.1). Results: In the cet cohort, pts with the C/C genotype of ZNRF3 rs132531 had significantly shorter overall survival compared to any T allele carriers (mOS: 20.3 vs 52 mo) in both univariable (HR 3.61, 95% CI 1.65-7.88, P &lt; .001) and multivariable analysis (adjusted P = .01). Conversely, RSPO1 rs4652964 any G allele carriers showed increased tumor response (TR) rates compared to the A/A genotype (83 vs 66 %, P = .04). These associations were not observed in bev arm. Lower gene expression levels of RNF43 were associated with shorter PFS in pts with right-sided tumors receiving FOLFIRI/cet ( P = .006, Q &lt; 0.1). RSPO1 expression levels were also associated with TR in the same subgroup (70 vs 10% in high vs low; P = .001, Q &lt; .05). RNF43 expression was associated with TR in pts with left-sided tumors (82% in high vs 58% in low, P = .014, Q = 0.1). Conclusions: Our results provide the first evidence that germline polymorphisms and tumor gene expression levels of RNF43/ ZNRF3 and RSPO1 may have a predictive value in mCRC pts receiving first-line cetuximab-based treatment and contribute to modulate anti-EGFRs activity.