Giridhar, Karthik V.

Abstract Deep learning has advanced digital pathology and multi-omics data analysis. However, there is still a critical need to integrate these modalities to better understand complex diseases. Deep learning models, such as Prov-Gigapath, a state-of-the-art pathology foundation model, have been developed for analyzing whole-slide images (WSI). Our previously published deep-learning framework, OmicsFootPrint (OFP), transforms high-dimensional multi-omics data into intuitive genomically-arranged circular images and utilizes transfer learning to overcome the challenges of high-dimensional, low-sample-size data. In this study, we aimed to combine WSI and multi-omics data to develop a framework capable of distinguishing between invasive ductal carcinoma (IDC) and invasive lobular carcinoma (ILC), which is clinically significant as it may help explain differential responses to therapies, such as tamoxifen versus letrozole. We enhanced the performance of our OFP model by dividing each multi-omics circular image (1024x1024x3 pixels) into 16 equal-sized patches (256x256x3 pixels). Each patch was trained using a separate EfficientNetV2 architecture with pre-trained ImageNet weights. The probability outcomes from these 16 models were subsequently combined through a meta-learner model using the AutoGluon framework. Concurrently, WSI images corresponding to the same patients were independently trained, maintaining consistent train/validation/test splits. The Prov-Gigapath model, which was trained using a dataset from the Providence Health System, was employed to extract patch-level embeddings from the WSI data. These embeddings were fine-tuned using a LongNet-based masked autoencoder to predict outcomes. Finally, predictions from the OFP model and the WSI-based Prov-Gigapath model were integrated through a meta-learner, preserving the original train/validation/test split. Model performances (AUC) were evaluated using the held-out test set. The gold standard used for the analysis was histological identification of IDC versus ILC, against which the performance of omics data alone, WSI data alone, and the multi-modal integration approach were compared. We evaluated our integrated model as a proof-of-concept using The Cancer Genome Atlas (TCGA) breast cancer cohort, comprising 92 ILC and 391 IDC samples. Multi-omics data, including gene expression, microRNA, copy number variation, and RPPA, were analyzed using the OFP framework, while WSI data were processed using the Prov-Gigapath method. The results demonstrated that the multi-modal integration approach significantly outperformed individual modalities, achieving an AUC of 0.98 compared to 0.87 for OFP (original), 0.95 for OFP (16-patches), and 0.92 for WSI. Integrating WSI and multi-omics data improves predictive accuracy, significantly improving the classification of breast cancer subtypes over histology alone. Citation Format: Naresh Prodduturi, Xiaojia Tang, Hamid R. Tizhoosh, Kevin J. Thompson, Richard Weinshilboum, Karthik V. Giridhar, Judy C. Boughey, Eric W. Klee, Liewei Wang, Matthew P. Goetz, Vera Suman, Krishna R. Kalari. A multi-modal learning framework to integrate digital pathology image and multi-omics data in breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 2429.

Abstract Purpose: The aim of PREDICT was to confirm clinical validity and the potential for clinical utility of serial circulating tumor cell (CTC) enumeration in patients with metastatic breast cancer, focusing on its prognostic value in different breast cancer subtypes and clinical settings. Experimental Design: In total, 4,436 individual patient-level data with CTC results from both baseline and one follow-up (CellSearch; Menarini Silicon Biosystems) were analyzed to evaluate the association between CTC detection and overall survival (OS) in the full patient cohort and separately for tumor and treatment types. Results: Using the cutoff ≥1 CTC for CTC positivity, 913 (20.6%) patients had 0 CTCs at both time points (neg/neg) and 325 (7.3%) and 1,189 (26.8%) patients converted from CTC negative to CTC positive (neg/pos) or vice versa (pos/neg), whereas 2,009 (45.3%) patients had at least one CTC at both time points (pos/pos). The median OS for the neg/neg, neg/pos, pos/neg, and pos/pos group was 45.6, 26.1, 32.3, and 17.3 months, respectively (P < 0.0001, global log-rank test). CTC responders (pos/neg) showed a lower risk of death compared with CTC nonresponders (pos/pos; HR, 0.48; 95% confidence interval, 0.44–0.53). Similar results were obtained in subgroup analyses according to hormone receptor and HER2 subtype, treatment type, and with a ≥5 CTC cutoff for CTC positivity. Conclusions: Follow-up CTC assessments strongly predict OS independently from tumor subtype and treatment. New randomized trials to define the clinical utility of CTC monitoring for risk stratification and as an early response marker in metastatic breast cancer are urgently needed.

Background: Chemotherapy-induced peripheral neuropathy (CIPN) from oxaliplatin and taxane drugs is a bothersome toxicity. Palmitoylethanolamide (PEA) has been reported to improve myelinated nerve fiber function in patients experiencing painful CIPN. We conducted a double-blind, placebo-controlled, randomized trial of PEA in patients with established CIPN. Methods: Eligible patients were adults who had pain, numbness, tingling, or other symptoms of CIPN for at least three months following completion of paclitaxel, oxaliplatin, or cisplatin-based chemotherapy. Study patients were randomized to one of the two treatment groups (PEA versus placebo, both administered either once or twice daily). The CIPN20 questionnaire was assessed weekly. Results: A total of 17 males and 71 females participated in the study; most had neuropathy from paclitaxel. Most (85%) finished 8 weeks of treatment. There was no suggestion that either of the PEA arms did any better than the combined placebo arms. There was no signal of significant toxicity differences between the three study arms. Quality of life outcome measures were similar between the study arms, as were cognitive function evaluations. Discussion: PEA failed to improve established CIPN. Future trials might explore whether PEA may be effective in preventing CIPN or cognitive changes based on data that suggest it may be helpful in this situation. Conclusions: PEA failed to improve established chemotherapy-induced neuropathy.

Background: Patients have been known to use cannabinoids for treating established chemotherapy-induced peripheral neuropathy (CIPN) based on anecdotal information and retrospective reports suggesting that such might be beneficial. In response, a double-blinded, placebo-controlled, randomized, pilot clinical trial was developed to evaluate whether resultant data would support a phase III trial for testing whether a cannabidiol (CBD) cream might improve CIPN. Methods: Forty patients with established CIPN were randomized, in a double-blinded manner, to topical CBD or a placebo cream. The study product was applied for 2 weeks, followed by a crossover for 2 weeks. Neuropathy was evaluated using the European Organization of Research and Treatment of Cancer (EORTC)–CIPN20, the Chemotherapy-Induced Peripheral Neuropathy Assessment Tool, and the Global Impression of Change instruments. Side effects were recorded by symptom diaries. Results: The EORTC-CIPN20 scores were similar in the patients receiving CBD versus the placebo. Likewise, the toxicity scores were similar in patients who received the CBD versus the placebo. Conclusions: This pilot trial did not support that the studied CBD isolate cream improved painful established CIPN. It was well tolerated overall. Clinical Trial Registration Number: NCT05388058

601 Background: EOP is an I-SPY2 neoadjuvant sub-study designed to test the tolerability and impact of novel endocrine-based strategies in stage 2/3 BC pts predicted to have lower benefit from chemotherapy. Amcenestrant (A) is an oral selective estrogen receptor degrader (SERD) with activity in hormone receptor positive (HR+) HER2- metastatic BC. Methods: EOP eligibility included pts screened for the main I-SPY2 trial with HR+ HER2-, MammaPrint (MP) low risk Stage 2/3 BC. Pts with MP High risk tumor signatures were eligible if clinically node-negative (cN-). Pts were stratified by SET index and menopausal status. Pts were randomized to 3 arms: A 200 mg qd (A), A + abemaciclib 150 mg bid (AA), A + letrozole 2.5 mg qd (AL). Premenopausal pts received concurrent ovarian function suppression. Treatment was for 6 mos followed by surgery. Baseline (BL) and 3-wk biopsies were performed. Breast MRIs were conducted at BL (T0), 3 wks (T1), 12 wks (T2), 6 mos (T3). The primary endpoint was feasibility, defined as >75% of pts completing >75% study therapy. Results: Between May 2021 and Aug 2022, 74 pts were randomized: 31 A, 21 AA, 22 AL. Median age 50 years, 54% premenopausal, 38% clinically node-positive (cN+). 41% pts had ductal, 38% lobular, 11% mixed ductal/lobular histology. Mean BL tumor Ki67 (central testing) was 16.9%, 88% MP low, 76% SET index Hi. Pt and tumor characteristics were balanced across the 3 arms. Treatment in all 3 arms was feasible with 95% of pts completing >75% study therapy. Most common AEs include grade 1-2 hot flashes (69%) and fatigue (66%). Grade 3 AEs include neutropenia (2 pts) and diarrhea (1 pt) in AA. Table summarizes key response endpoints, which were similar between pre- and postmenopausal pts. Of 28 cN+ pts, 2 (7%) were pathologically node-negative (pN-). Mean MRI Functional Tumor Volume (cc) change between T0->T1, T1->T2, T2->T3, T0->T3 was -6.06 (-34.2%), -3.55 (-29.2%), -0.829 (-20.1%), -11.3 (-68.4%), respectively, similar across arms (p=0.396).ctDNA was detected in 10/59 (17%) pts at T0. ctDNA became undetectable in 6 pts, decreased in 4 pts, and went from ctDNA undetectable at baseline to detectable at T3 in 2 pts. Conclusions: Six months of neoadjuvant endocrine therapy (NET) with the oral SERD amcenestrant is feasible, well-tolerated, and demonstrates anti-proliferative and anti-tumor activity in pre- and postmenopausal pts. NET provides a rich platform for biomarker discovery and investigating response to endocrine therapy. Clinical trial information: NCT01042379 . [Table: see text]

5589 Background: Imlunestrant is a next-generation oral SERD designed to deliver continuous estrogen receptor (ER) target inhibition aimed to improve outcomes for patients (pts) with ER+ cancers. In the first-in-human Phase 1a/1b EMBER trial (NCT04188548), imlunestrant showed favorable safety, pharmacokinetics, and clinical benefit in pre-treated pts with ER+ advanced breast cancer or EEC. Here, we report updated imlunestrant monotherapy data and the first clinical data of imlunestrant + abemaciclib in ER+ EEC pts from EMBER. Methods: Eligible pts had ER+ EEC previously treated with (or deemed inappropriate for) platinum-containing chemotherapy. Prior fulvestrant or aromatase inhibitor therapy was not allowed. In Phase 1a, pts received imlunestrant alone, and in Phase 1b, pts were randomized (1:1) to imlunestrant alone or with abemaciclib. Plasma samples were collected at baseline (all pts) and at cycle 2 (monotherapy pts) for ctDNA analysis (Guardant 360 assay) to correlate with clinical benefit. Primary endpoints were RP2D of imlunestrant monotherapy and in combination. Secondary endpoints included safety, tolerability, and overall response rate (ORR), clinical benefit rate (CBR: complete or partial response, or stable disease ≥24 weeks), and progression-free survival (PFS) using RECIST v1.1. Results: Overall, 72 pts were treated; 39 pts received monotherapy (33 at 400mg [RP2D] and 6 at 800mg po daily) and 33 pts received imlunestrant (29 at 400mg and 4 at 800mg po daily) + abemaciclib (150mg po twice daily). Baseline characteristics, safety, and preliminary efficacy are presented (Table). For monotherapy, the most common TEAEs (all grade/grade 3) were nausea (36%/3%), diarrhea (26%/0), urinary tract infection (26%/3%), and abdominal pain (21%/8%). For imlunestrant + abemaciclib, the most common TEAEs (all grade/grade 3) were diarrhea (88%/3%), nausea (67%/0), fatigue (49%/9%) and anemia (46%/12%). In the monotherapy arm, pts with deeper (≥ 50%) declines in ctDNA variant allele frequency had a higher CBR (4/7 vs 4/17). Conclusions: Imlunestrant, as monotherapy or with abemaciclib, appears safe and tolerable with preliminary evidence of efficacy in pts with metastatic ER+ EEC. Clinical trial information: NCT04188548 . [Table: see text]

e13036 Background: In January 2023, the FDA approved elacestrant for ESR1-mutated, ER+/HER2- MBC based on a companion diagnostic cell free (cf)DNA test (Guardant Health). In this study, we analyze the trends of liquid biopsy utilization at Mayo Clinic (MC) over a 3 year period to understand the impact of this drug approval on testing patterns and identify emergent alterations in patients (pts) with MBC. Methods: We extracted clinically performed Guardant 360 and Guardant 360 CDx ordered at MC sites in Rochester (MN), Florida (FL), Arizona (AZ), and across 9 additional Mayo Clinic Health System (MCHS) sites. Pts with MBC were identified using the EMR and analyzing visit diagnosis codes and staging information. Shifts in testing patterns were analyzed over a 3 year period from January 1, 2021 - December 31, 2023. Data was collected on the number of tests, serial testing, frequency of pathogenic mutations, and emergent mutations in ESR1 and PIK3CA. Chi-square test was used to compare the categorical values. Results: Between 2021-2023, 2,741 pts with MBC were seen across all sites. A total of 408 cfDNA tests were performed in 370 unique patients with MBC. Most pts (90%) had 1 test performed. The proportion of pts with MBC tested was 85/1350 (6.3%) in 2021, 77/1495 (5.2%) in 2022, and 208/1558 (13.4%) in 2023. The increased proportion of pts tested in 2023 was statistically significant compared to prior years (p<0.0001), while no difference was observed between 2021 and 2022 (p 0.19). Testing patterns varied by site, with FL ordering the highest number of tests, followed by MN and AZ. Of the 34 providers practicing during the entire period seeing more than 10 MBC pts, the number of providers ordering Guardant testing increased from 15/34 (2021) and 14/34 (2022) to 20/34 (2023), though not statistically significant (p 0.15). Of the 41% with EMR-documented staging available, testing was highest in the ER+/HER2- (81.3%) subtype, followed by TNBC (13.6%) and then HER2+ (5.2%). The most commonly identified mutations identified in the first testing were PIK3CA (41.1%), TP53 (40.5%), ESR1 (30.5%), and ATM (13.2%). D538G (8.0%), Y537N (4.4%), and Y537S (3.5%) were frequent emergent mutations seen in ESR1. Conclusions: We observed an increase in cfDNA testing at the start of 2023 which aligns with the FDA approval of elacestrant. We observed variation by site and individual providers that warrants further exploration to improve cfDNA testing in MBC pts at Mayo Clinic.

1083 Background: T-DXd and SG are both FDA-approved treatments for patients (pts) with HER2-low MBC, but little data exists on the sequential use of these ADCs. Methods: In this multicenter retrospective cohort study, we identified pts with HER2-low MBC treated sequentially with both T-DXd and SG in either order, with or without intervening therapies (IntTx), at 5 academic centers between 2020-2023. We previously reported real-world progression free survival (rwPFS) and real-world overall survival (rwOS) by HR-status and ADC sequence order. Here we provide updated rwPFS and rwOS data, as well as new subgroup analyses by age, sites of disease, and use of IntTx between ADCs for HR- and HR+ disease. Results: 84 pts with HER2-low MBC treated with both T-DXd and SG were included, including 56 pts with HR+/HER2-low MBC (66.7%) and 28 pts with HR-/HER2-low MBC (33.3%). Prior to the start of ADC1, median age was 58 yrs and 22 pts were >65 yrs (26.2%), 65 pts (77.4%) had visceral disease, and 14 pts (16.7%) had central nervous system (CNS) disease. 36 pts (42.9%) had an IntTx between ADCs. Table 1 shows updated rwPFS and rwOS data by HR-status and ADC sequence order. For pts with HR+ and HR- MBC, there was no difference in ADC1 rwPFS or ADC2 rwPFS by age ≤65 vs. >65 yr or by presence vs. absence of visceral disease. For pts with HR- MBC, ADC1 rwPFS and rwOS was shorter for pts with CNS disease vs. pts without CNS disease (rwPFS 5.4 mo (n=6) vs. 8.5 mo (n=22), p=0.03; rwOS 11.6 mo vs. 21.1 mo, p<0.01 respectively). For pts with HR+ and HR- MBC, there was no difference in ADC2 rwPFS with or without receipt of IntTx. There was no difference in time to treatment failure for first IntTx vs. ADC2 if no InTx was given. Conclusions: This is one of the largest multicenter cohorts of pts treated with sequential ADCs for HER2-low MBC. Median rwPFS was longer for ADC1 than ADC2 in all subgroups, regardless of HR-status, ADC sequence order, age ≤65 or >65 yr, presence of visceral disease, or use of IntTx. Among pts with HR-/HER2-low MBC, pts with CNS disease had shorter ADC1 rwPFS and rwOS than pts without CNS disease. [Table: see text]

e12641 Background: Many promising agents or combinations regimens relevant to breast cancer lack sufficient clinical information for use in the neoadjuvant setting. To enable acceleration of drug development by reducing time and cost of clinical trial planning and conduct, the Phase I initiative PRE-ISPY Program was created. Methods: The PRE-ISPY Program is built within the I-SPY network, with the overall aim to streamline the regulatory pathway and the processes of trial design, contracting, activation, enrollment, execution, data gathering and analysis, as well as reporting of new therapeutic agents or combinations of relevance to breast cancer. Results: The PRE-ISPY trial is an open-label platform phase IB study designed to evaluate compelling single agent or combination regimens with the overall goal of moving promising drug regimens into I-SPY2 Trial and/or into other oncology-based trials in a timely manner. Graduation of a PRE-ISPY regimen to I-SPY2 is not mandatory. The primary objectives in each study Arm are to determine safety and to obtain preliminary efficacy data of the new regimen in patients with metastatic cancer, including a cohort of breast cancer patients. Each Arm may contain a Part 1 (dose finding) and a Part 2 (dose expansion). The new treatment can be tested also in patients with other solid malignancies than breast cancer, in which the treatment may be applicable, as specified in the design of the Arm. A breast cancer specific cohort is required in Part 2. Candidate treatment regimens are selected with advice from the I-SPY2 Agent Selection Working Group. The PRE-ISPY/Phase IB trial has multiple ongoing drug regimen Arms subsequently added as new protocol appendices through amendments to a Master protocol. There is no randomization; participants are enrolled, and treatment selected according to the relevance of therapy to their disease. Prospective companion biomarkers are evaluated. The current arms under study are PRE1: ALX-148 combined with trastuzumab deruxtecan; PRE2: tucatinib combined with zanidatamab; and PRE3: vidutolimod combined with cemiplimab. Selected agents or therapeutic combinations are tested through this protocol in the I-SPY Network Centers that participate in the PRE-ISPY Program. Toxicities are graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE). Clinical responses are evaluated based on RECIST v1.1 criteria. Conclusions: Once safety is established, promising regimens can be rapidly transitioned to I-SPY 2.2 or other phase 2/3 trials. The first patient was enrolled on 2/1/2023. Clinical trial information: NCT05868226 .