Predict Chemotherapy Response Research Articles

DCTPP1 Expression as a Predictor of Chemotherapy Response in Luminal A Breast Cancer Patients.

Breast cancer (BRCA) remains a significant global health challenge due to its prevalence and lethality, exacerbated by the development of resistance to conventional therapies. Therefore, understanding the molecular mechanisms underpinning chemoresistance is crucial for improving therapeutic outcomes. Human deoxycytidine triphosphate pyrophosphatase 1 (DCTPP1) has emerged as a key player in various cancers, including BRCA. DCTPP1, involved in nucleotide metabolism and maintenance of genomic stability, has been linked to cancer cell proliferation, survival, and drug resistance. This study evaluates the role of DCTPP1 in BRCA prognosis and chemotherapy response. Data from the Cancer Genome Atlas Program (TCGA), Genotype-Tissue Expression (GTEx), and Gene Expression Omnibus (GEO) repositories, analyzed using GEPIA and Kaplan-Meier Plotter, indicate that high DCTPP1 expression correlates with poorer overall survival and increased resistance to chemotherapy in BRCA patients. Further analysis reveals that DCTPP1 gene expression is up-regulated in non-responders to chemotherapy, particularly in estrogen receptor (ER)-positive, luminal A subtype patients, with significant predictive power. Additionally, in vitro studies show that DCTPP1 gene expression increases in response to 5-fluorouracil and doxorubicin treatments in luminal A BRCA cell lines, suggesting a hypothetical role in chemoresistance. These findings highlight DCTPP1 as a potential biomarker for predicting chemotherapy response and as a therapeutic target to enhance chemotherapy efficacy in BRCA patients.

Allysine-Targeted Molecular MRI Enables Early Prediction of Chemotherapy Response in Pancreatic Cancer.

Neoadjuvant therapy is routinely used in pancreatic ductal adenocarcinoma (PDAC), but not all tumors respond to this treatment. Current clinical imaging techniques are not able to precisely evaluate and predict the response to neoadjuvant therapies over several weeks. A strong fibrotic reaction is a hallmark of a positive response, and during fibrogenesis, allysine residues are formed on collagen proteins by the action of lysyl oxidases. Here, we report the application of an allysine-targeted molecular MRI probe, MnL3, to provide an early, noninvasive assessment of treatment response in PDAC. Allysine increased 2- to 3-fold after one dose of neoadjuvant therapy with FOLFIRINOX in sensitive human PDAC xenografts in mice. Molecular MRI with MnL3 could specifically detect and quantify fibrogenesis in PDAC xenografts. Comparing the MnL3 signal before and 3 days after one dose of FOLFIRINOX predicted subsequent treatment response. The MnL3 tumor signal increased by 70% from day 0 to day 3 in mice that responded to subsequent doses of FOLFIRINOX, whereas no signal increase was observed in FOLFIRINOX-resistant tumors. This study indicates the promise of allysine-targeted molecular MRI as a noninvasive tool to predict chemotherapy outcomes. Significance: Allysine-targeted molecular MRI can quantify fibrogenesis in pancreatic tumors and predict response to chemotherapy, which could guide rapid clinical management decisions by differentiating responders from nonresponders after treatment initiation.

A susceptibility gene signature for ERBB2-driven mammary tumour development and metastasis in collaborative cross mice

Deeper insights into ERBB2-driven cancers are essential to develop new treatment approaches for ERBB2+ breast cancers (BCs). We employed the Collaborative Cross (CC) mouse model to unearth genetic factors underpinning Erbb2-driven mammary tumour development and metastasis. 732 F1 hybrid female mice between FVB/N MMTV-Erbb2 and 30 CC strains were monitored for mammary tumour phenotypes. GWAS pinpointed SNPs that influence various tumour phenotypes. Multivariate analyses and models were used to construct the polygenic score and to develop a mouse tumour susceptibility gene signature (mTSGS), where the corresponding human ortholog was identified and designated as hTSGS. The importance and clinical value of hTSGS in human BC was evaluated using public datasets, encompassing TCGA, METABRIC, GSE96058, and I-SPY2 cohorts. The predictive power of mTSGS for response to chemotherapy was validated invivo using genetically diverse MMTV-Erbb2 mice. Distinct variances in tumour onset, multiplicity, and metastatic patterns were observed in F1-hybrid female mice between FVB/N MMTV-Erbb2 and 30 CC strains. Besides lung metastasis, liver and kidney metastases emerged in specific CC strains. GWAS identified specific SNPs significantly associated with tumour onset, multiplicity, lung metastasis, and liver metastasis. Multivariate analyses flagged SNPs in 20 genes (Stx6, Ramp1, Traf3ip1, Nckap5, Pfkfb2, Trmt1l, Rprd1b, Rer1, Sepsecs, Rhobtb1, Tsen15, Abcc3, Arid5b, Tnr, Dock2, Tti1, Fam81a, Oxr1, Plxna2, and Tbc1d31) independently tied to various tumour characteristics, designated as a mTSGS. hTSGS scores (hTSGSS) based on their transcriptional level showed prognostic values, superseding clinical factors and PAM50 subtype across multiple human BC cohorts, and predicted pathological complete response independent of and superior to MammaPrint score in I-SPY2 study. The power of mTSGS score for predicting chemotherapy response was further validated in an invivo mouse MMTV-Erbb2 model, showing that, like findings in human patients, mouse tumours with low mTSGS scores were most likely to respond to treatment. Our investigation has unveiled many new genes predisposing individuals to ERBB2-driven cancer. Translational findings indicate that hTSGS holds promise as a biomarker for refining treatment strategies for patients with BC. The U.S. Department of Defense (DoD) Breast Cancer Research Program (BCRP) (BC190820), United States; MCIN/AEI/10.13039/501100011039 (PID2020-118527RB-I00, PDC2021-121735-I00), the "European Union Next Generation EU/PRTR," the Regional Government of Castile and León (CSI144P20), European Union.

Can Ultrasound Evaluation of Lymph Node Size and Necrosis Rate Predict Chemotherapy Response in Cervical Tuberculous Lymphadenitis?

To explore the relationship between the initial size and the necrotic rate of lymph nodes evaluated by ultrasound in patients with cervical tuberculous lymphadenitis (CTL) and therapeutic response. Overall, 55 patients were included in this study. Conventional ultrasound and contrast-enhanced ultrasound examination were performed before anti-tuberculosis chemotherapy. Based on the different therapeutic outcomes, they were divided into responder groups (n = 39) and non-responder groups (n = 16). The relationship between the initial size (maximum area, length diameter, short diameter), rate of necrosis, and therapeutic response were compared and analyzed between two groups. There was a significant difference in maximum area, short diameter and rate of necrosis of lymph nodes between the responder groups and the non-responder groups (P < 0.05). The receiver-operating-characteristic (ROC) curve analysis was used to differentiate the two groups, it showed that the area under the curve was 0.746 for maximum area and 0.721 for short diameter, respectively. The cut-off value for the lymph node maximum area and short diameter based on ROC curve analysis was determined as 3.94cm2 (sensitivity 76.9%, specificity 68.7%) and 1.15cm (sensitivity 59.0%, specificity 93.7%), respectively. A negative correlation was observed between maximum area, short diameter, and therapeutic response. The initial maximum area and short diameter of lymph nodes were found to have a negative correlation with chemotherapy response in patients with CTL. The treatment outcomes are typically unsatisfactory for lymph nodes exhibiting an initial necrosis rate of 50% or higher. These findings may be helpful for evaluating therapeutic response.

A computed tomography-based nomogram for neoadjuvant chemotherapy plus immunotherapy response prediction in patients with advanced esophageal squamous cell carcinoma.

To develop a CT-based nomogram to predict the response of advanced esophageal squamous cell carcinoma (ESCC) to neoadjuvant chemotherapy plus immunotherapy. In this retrospective study, 158 consecutive patients with advanced ESCC receiving contrast-enhanced CT before neoadjuvant chemotherapy plus immunotherapy were randomized to a training cohort (TC, n = 121) and a validation cohort (VC, n = 37). Response to treatment was assessed with response evaluation criteria in solid tumors. Patients in the TC were divided into the responder (n = 69) and non-responder (n = 52) groups. For the TC, univariate analyses were performed to confirm factors associated with response prediction, and binary analyses were performed to identify independent variables to develop a nomogram. In both the TC and VC, the nomogram performance was assessed by area under the receiver operating characteristic curve (AUC), calibration slope, and decision curve analysis (DCA). In the TC, univariate analysis showed that cT stage, cN stage, gross tumor volume, gross volume of all enlarged lymph nodes, and tumor length were associated with the response (all P < 0.05). Binary analysis demonstrated that cT stage, cN stage, and tumor length were independent predictors. The independent factors were imported into the R software to construct a nomogram, showing the discriminatory ability with an AUC of 0.813 (95% confidence interval: 0.735-0.890), and the calibration curve and DCA showed that the predictive ability of the nomogram was in good agreement with the actual observation. This study provides an accurate nomogram to predict the response of advanced ESCC to neoadjuvant chemotherapy plus immunotherapy.

8P Biomarkers to predict chemotherapy response in low-grade serous ovarian carcinoma

8P Biomarkers to predict chemotherapy response in low-grade serous ovarian carcinoma

Abstract PS05-02: PET Imaging of PARP Expression as a Biomarker of Response to Chemotherapy in Breast Cancer: A Nonrandomized Clinical Trial

Abstract Introduction: Poly–(adenosine diphosphate–ribose) polymerase (PARP) proteins are involved in double-stranded break repair, mediated through homologous recombination (HR) and non-homologous end joining (NHEJ). The majority of PARP activity in this context is attributed to PARP-1 (85-90%). PARP-1 binding to damaged DNA, catalytic activity, and subsequent release to allow access for other DNA repair proteins are necessary for efficient DNA repair. The critical function of PARP-1 in these processes makes this protein a unique potential biomarker of DNA repair capability. [18F]FluorThanatrace ([18F]FTT) is an 18F-labeled PARP inhibitor analog that binds to the same site as approved PARPi drugs. We hypothesized that the biologically relevant form of PARP in untreated breast cancer might provide an indication of functional tumor DNA repair capabilities and therefore predict chemotherapy response. Thus, this study aimed to determine whether a non-invasive quantitative measure of in vivo PARP expression correlated with response to chemotherapy in breast cancer. Methods: A single-arm prospective trial enriched for subjects with triple negative breast cancer (TNBC) was conducted at the University of Pennsylvania from May 2017 to March 2022 (NCT03083288 and NCT03846167). Inclusion criteria were primary breast cancer with tumor diameter of at least 1 cm on conventional imaging and willingness to undergo an [18F]FTT-PET scan prior to neoadjuvant chemotherapy. Participants provided written informed consent. [18F]FTT uptake in breast cancer was measured pre-therapy and tested for association with pathologic complete response (pCR). Subjects were scanned on an Ingenuity TF PET/CT (Philips Healthcare) following injection of [18F]FTT using a 20-minute static acquisition. Peak 1 cm3 average uptake (SUVpeak) were recorded from a spherical region-of-interest (ROI) covering the primary malignancy, guided by CT and prior imaging studies. All SUVs were partial volume corrected and normalized using muscle (NM). A Mann-Whitney test was used to determine if [18F]FTT uptake differed by pCR. Receiver operating characteristic (ROC) curves were constructed to test the performance of [18F]FTT uptake in predicting pCR. Statistical analyses utilized STATA v15.1 with significance based on a two-sided alpha level of ≤0.05. Participants: Twenty-six women with stage I-III breast cancer met inclusion criteria with age range of 29-74 years and self-reported race as Asian (1, 4%), Black (9, 37%) or White (16, 60%). One subject had bilateral tumors (ER+ stage IIA and HER2+ stage I) that were evaluated separately (27 total tumors). Breast tumors were pathologically confirmed and included 6 (22%) ER+/HER2-, 5 (19%) HER2+, and 16 (59%) TNBC. Germline sequencing was available for 21 participants. Results: Mean tumor diameter ranged from 15 to 91 mm (median 30 mm). A considerable range of [18F]FTT uptake was seen across subjects (SUVpeak/NM 0.66-6.5). [18F]FTT uptake was independent of subtype (ER+, HER2+, TN) (P=0.35), stage (P=0.39), and between mutations in genes associated with homologous recombination deficiency (i.e., BRCA1/2, PALB2) and wild-type (P=0.73). In the TNBC cohort, pre-treatment [18F]FTT uptake was higher in subjects who went on to have a pCR (n=16, P=0.05), while a trend toward higher uptake was seen when including all tumor types (n=27, P=0.11). ROC analysis of the value of [18F]FTT uptake for predicting pCR revealed an AUC of 0.79 and showed that a threshold SUV ratio &amp;lt; 2.47 predicted pCR in TNBC patients with a 100% specificity and 64% sensitivity. Conclusions: These early clinical results suggest a relationship between [18F]FTT uptake, a measure of PARP expression, and chemotherapy sensitivity in TNBC, warranting future studies to elucidate underlying mechanistic processes and potential clinical implications. Citation Format: Sarah Gitto, Anthony Young, Ira Bleiweiss, Amy Clark, Elizabeth McDonald. PET Imaging of PARP Expression as a Biomarker of Response to Chemotherapy in Breast Cancer: A Nonrandomized Clinical Trial [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 PS05-02.

Apriori prediction of chemotherapy response in locally advanced breast cancer patients using CT imaging and deep learning: transformer versus transfer learning.

Neoadjuvant chemotherapy (NAC) is a key element of treatment for locally advanced breast cancer (LABC). Predicting the response to NAC for patients with Locally Advanced Breast Cancer (LABC) before treatment initiation could be beneficial to optimize therapy, ensuring the administration of effective treatments. The objective of the work here was to develop a predictive model to predict tumor response to NAC for LABC using deep learning networks and computed tomography (CT). Several deep learning approaches were investigated including ViT transformer and VGG16, VGG19, ResNet-50, Res-Net-101, Res-Net-152, InceptionV3 and Xception transfer learning networks. These deep learning networks were applied on CT images to assess the response to NAC. Performance was evaluated based on balanced_accuracy, accuracy, sensitivity and specificity classification metrics. A ViT transformer was applied to utilize the attention mechanism in order to increase the weight of important part image which leads to better discrimination between classes. Amongst the 117 LABC patients studied, 82 (70%) had clinical-pathological response and 35 (30%) had no response to NAC. The ViT transformer obtained the best performance range (accuracy = 71 ± 3% to accuracy = 77 ± 4%, specificity = 86 ± 6% to specificity = 76 ± 3%, sensitivity = 56 ± 4% to sensitivity = 52 ± 4%, and balanced_accuracy=69 ± 3% to balanced_accuracy=69 ± 3%) depending on the split ratio of train-data and test-data. Xception network obtained the second best results (accuracy = 72 ± 4% to accuracy = 65 ± 4, specificity = 81 ± 6% to specificity = 73 ± 3%, sensitivity = 55 ± 4% to sensitivity = 52 ± 5%, and balanced_accuracy = 66 ± 5% to balanced_accuracy = 60 ± 4%). The worst results were obtained using VGG-16 transfer learning network. Deep learning networks in conjunction with CT imaging are able to predict the tumor response to NAC for patients with LABC prior to start. A ViT transformer could obtain the best performance, which demonstrated the importance of attention mechanism.

Pathway-based signatures predict patient outcome, chemotherapy benefit and synthetic lethal dependencies in invasive lobular breast cancer.

Invasive Lobular Carcinoma (ILC) is a morphologically distinct breast cancer subtype that represents up to 15% of all breast cancers. Compared to Invasive Breast Carcinoma of No Special Type (IBC-NST), ILCs exhibit poorer long-term outcome and a unique pattern of metastasis. Despite these differences, the systematic discovery of robust prognostic biomarkers and therapeutically actionable molecular pathways in ILC remains limited. Pathway-centric multivariable models using statistical machine learning were developed and tested in seven retrospective clinico-genomic cohorts (n = 996). Further external validation was performed using a new RNA-Seq clinical cohort of aggressive ILCs (n = 48). mRNA dysregulation scores of 25 pathways were strongly prognostic in ILC (FDR-adjusted P < 0.05). Of these, three pathways including Cell-cell communication, Innate immune system and Smooth muscle contraction were also independent predictors of chemotherapy response. To aggregate these findings, a multivariable machine learning predictor called PSILC was developed and successfully validated for predicting overall and metastasis-free survival in ILC. Integration of PSILC with CRISPR-Cas9 screening data from breast cancer cell lines revealed 16 candidate therapeutic targets that were synthetic lethal with high-risk ILCs. This study provides interpretable prognostic and predictive biomarkers of ILC which could serve as the starting points for targeted drug discovery for this disease.

Abstract LB096: Telomere length as a predictor of therapy response and survival in patients diagnosed with ovarian carcinoma

Abstract Impaired telomere length (TL) maintenance in ovarian tissue may play a pivotal role in epithelial ovarian carcinoma (OvC) onset. TL in either target or surrogate tissue (blood) is currently being investigated as a predictor in anti-OvC therapy or as a biomarker of the disease progression, respectively. There is presently an urgent need for an appropriate approach to chemotherapy response prediction. We performed a monochrome multiplex qPCR measurement of TL in peripheral blood and tumor tissues of 209 OvC patients. Methylation status and gene expression of shelterin complex and telomerase catalytic subunit (hTERT) were determined within tumor tissues by High-Throughput DNA methylation profiling and RNA sequencing (RNA-Seq) analysis, respectively. The patients sensitive to cancer treatment (n=46) had shorter telomeres in peripheral blood cells compared to those treatment-resistant (n=93; P=0.037). Cancer tissue TL lower than the median predisposed to better overall survival (P=0.002). Gene expression of TPP1, which recruits telomerase to telomeres, was positively associated with TL in tumor tissue (P=0.026). TL measured in peripheral blood cells could serve as a marker of platinum therapy response in OvC patients. Additionally, TL determined in tumor tissue provides information on OvC patients' overall survival. The study was financed by the project NPO (LX22NPO05102), the Czech Health Research Council (NU21-03-00145), and the Czech Science Foundation (21-27902S). Citation Format: Kristyna Tomasova, Karolina Seborova, Michal Kroupa, Josef Horak, Ludmila Vodickova, Lukas Rob, Martin Hruda, Marcela Mrhalova, Alena Bartakova, Jiri Bouda, Thomas Fleischer, Vessela Kristensen, Pavel Vodicka, Radka Vaclavikova. Telomere length as a predictor of therapy response and survival in patients diagnosed with ovarian carcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr LB096.

Abstract LB242: Prediction of patient response to neoadjuvant chemotherapy in breast cancer from their deconvolved tumor microenvironment transcriptome

Abstract Introduction: The tumor microenvironment (TME) is a complex and dynamic ecosystem that plays critical roles in tumor development and clinical outcome. While the multifarious cellular interactions within the TME have been extensively studied with a significant emphasis on immunotherapy, understanding its role in chemotherapy outcome remains less explored. To this end, we present a novel generic computational framework named DECODEM (DEcoupling Cell-type-specific Outcomes using DEconvolution and Machine learning). DECODEM facilitates the exploration of the association of cell-type-specific gene expression profiles with the effectiveness of neoadjuvant chemotherapy (NAC) in treating breast cancer (BC). This approach holds promise for elucidating the nuanced ways in which the TME can modulate therapeutic efficacy, potentially paving the way for more personalized and effective cancer treatments. Methods: DECODEM employs cellular deconvolution of bulk transcriptomics to extract cell-type-specific gene expression profiles within the TME, and subsequently harnesses machine learning to construct cell-type-specific predictors of clinical outcomes. Utilizing DECODEM, we analyzed three publicly available cohorts of HER2-negative BC patients who received NAC. We assessed the predictive capabilities of the individual cell types as well as their collective influence, benchmarking against a state-of-the-art predictor that employs both clinical and transcriptomic features. We corroborated our findings in a single-cell (SC) transcriptomics cohort of patients with triple negative breast cancer (TNBC) undergoing chemotherapy. Further, we expanded DECODEM to DECODEMi, which accounts for cell-cell interactions (CCIs) among the predominant cell types, offering potential insights into the mechanistic processes governing clinical responses. Results: Our analysis indicated that the heterogeneous composition of the TME is crucial for chemotherapy response prediction in HER2-negative BC, beyond the influence of malignant cells alone. The key insights from our study are as follows:1. Gene expression of immune cells (myeloid, plasmablasts, B-cells) and stromal cells (endothelial, normal epithelial, cancer-associated fibroblasts) displayed comparable-to-superior predictive powers over both bulk expression and the state-of-the-art predictor.2. The predictive accuracy is further augmented by multi-cell-type ensembles, with a triad ensemble comprising endothelial, myeloid, and plasmablast cells achieving the highest predictive accuracy across multiple cohorts.3. DECODEMi can effectively capture the regulatory impact of CCIs (validated in a SC dataset sourced from the SC-TNBC cohort) and has identified novel CCIs likely to influence chemotherapy response in HER2-negative BC. Conclusion: We provide an unprecedented computational methodology to quantitatively assess the cell-type-specific influences of the TME on clinical outcome. Our analysis offers insights into the complex interactions among TME cells and their correlation with chemotherapy efficacy in HER2-negative BC, thereby enriching the knowledge base of BC treatment effectiveness. Citation Format: Saugato Rahman Dhruba, Sahil Sahni, Binbin Wang, Di Wu, Yael Schmidt, Eldad Shulman, Sanju Sinha, Stephen-John Sammut, Carlos Caldas, Kun Wang, Eytan Ruppin. Prediction of patient response to neoadjuvant chemotherapy in breast cancer from their deconvolved tumor microenvironment transcriptome [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr LB242.

Abstract 7609: Potential gene expression meta-signatures predict neoadjuvant chemotherapy response in invasive breast cancer

Abstract Early prediction of Neoadjuvant Chemotherapy (NAC) response is crucial for women with locally advanced breast cancer (BC), given their low response rates. This study aims to identify gene expression patterns and tumor microenvironment (TME) characteristics associated with resistance to NAC, which could significantly impact tumor progression and treatment outcomes. The study encompassed differential gene expression comparing non-responders and responders, enrichment analysis, tumor microenvironment estimation, and cytokines identification from a female patient's cohort with distinct molecular subtypes and candidates to NAC at the Colombian National Cancer Institute. Candidate differentially expressed genes validated with RT-PCR in a larger validation cohort and corroborated in several publicly available databases, including associated treatment drugs. Distinct gene expression profiles were observed between responders and non-responders, emphasizing immune system-related pathways such as IL-17 signaling, B cell receptor signaling, complement and coagulation cascades, natural killer cell-mediated cytotoxicity, and NF-kB signaling. Immune-related genes (AGTR1, APOD, CCL19, CGA, ECEL1, NPY1R, NTNG1, PCSK1, PFDN2, S100B, TPRG1, WDR5B, ZNF385B) were identified as biomarkers correlated with non-response and worse prognosis. Responders showed increased B cell populations, CD4+ and CD8+ effector memory T-cells in lymphoid groups, while non-responders displayed elevated levels of CD8+ naive T-cells. Additionally, responders revealed upregulated myeloid cell populations like macrophages and dendritic cells. Several cytokines like IL1B, IL-2, IL-4, IL-12, IL-21, IL-22, IL-36, and TWEAK showed elevated expression in non-responders. Dexamethasone, Vinblastine, and Streptozocin were candidate drugs that interacted with AGTR1, CGA, and S100B, respectively and are associated with progression and metastasis. Logistic regression revealed that overexpression of CGA showed a higher probability (OR:1.83, p=0.03) while ECEL1 had a lower probability (OR=0.44, p=0.04) in non-responders compared to responders who lacked expression of these genes. Immune-related genes and various subtypes of immune cells within the TME exhibit correlations with response to NAC, and several cell fractions significantly changed before and after NAC. These results suggest meta-signatures that warrant further evaluation as concordance assessment between immunohistochemistry (IHC) and gene expression profiling, in addition to an extensive validation to triage patients before NAC and target personalized interventions. Detection and evaluation of TME components are essential to predict NAC efficacy and avoid disease relapse. Finally, these findings emphasize the importance of assessing admixed populations to facilitate novel findings and alleviate disparities. Citation Format: Hedda Michelle Guevara-Nieto, Rafael S. Parra-Medina, Jovanny Zabaleta, Liliana Lopez-Kleine, Alba L. Combita. Potential gene expression meta-signatures predict neoadjuvant chemotherapy response in invasive breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7609.

Basal–epithelial subpopulations underlie and predict chemotherapy resistance in triple-negative breast cancer

Triple-negative breast cancer (TNBC) is the most aggressive breast cancer subtype, characterized by extensive intratumoral heterogeneity, high metastasis, and chemoresistance, leading to poor clinical outcomes. Despite progress, the mechanistic basis of these aggressive behaviors remains poorly understood. Using single-cell and spatial transcriptome analysis, here we discovered basal epithelial subpopulations located within the stroma that exhibit chemoresistance characteristics. The subpopulations are defined by distinct signature genes that show a frequent gain in copy number and exhibit an activated epithelial-to-mesenchymal transition program. A subset of these genes can accurately predict chemotherapy response and are associated with poor prognosis. Interestingly, among these genes, elevated ITGB1 participates in enhancing intercellular signaling while ACTN1 confers a survival advantage to foster chemoresistance. Furthermore, by subjecting the transcriptional signatures to drug repurposing analysis, we find that chemoresistant tumors may benefit from distinct inhibitors in treatment-naive versus post-NAC patients. These findings shed light on the mechanistic basis of chemoresistance while providing the best-in-class biomarker to predict chemotherapy response and alternate therapeutic avenues for improved management of TNBC patients resistant to chemotherapy.

Identification of prognostic models for glycosylation-related subtypes and tumor microenvironment infiltration characteristics in clear cell renal cell cancer

One of the most fatal forms of cancer of the urinary system, renal cell carcinoma (RCC), significantly negatively impacts human health. Recent research reveals that abnormal glycosylation contributes to the growth and spread of tumors. However, there is no information on the function of genes related to glycosylation in RCC. In this study, we created a technique that can be used to guide the choice of immunotherapy and chemotherapy regimens for RCC patients while predicting their survival prognosis. The Cancer Genome Atlas (TCGA) provided us with patient information, while the GeneCards database allowed us to collect genes involved in glycosylation. GSE29609 was used as external validation to assess the accuracy of prognostic models. The "ConsensusClusterPlus" program created molecular subtypes based on genes relevant to glycosylation discovered using differential expression analysis and univariate Cox analysis. We examined immune cell infiltration as measured by estimate, CIBERSORT, TIMER, and ssGSEA algorithms, Tumor Immune Dysfunction and Exclusion (TIDE) and exclusion of tumour stemness indices (TSIs) based on glycosylation-related molecular subtypes and risk profiles. Stratification, somatic mutation, nomogram creation, and chemotherapy response prediction were carried out based on risk factors. We built and verified 16 gene signatures associated with the prognosis of ccRCC patients, which are independent prognostic variables, and identified glycosylation-related genes by bioinformatics research. Cluster 2 is associated with lower human leukocyte antigen expression, worse overall survival, higher immunological checkpoints, and higher immune escape scores. In addition, cluster 2 had significantly better angiogenic activity, mesenchymal EMT, and stem ability scores. Higher immune checkpoint genes and human leukocyte antigens are associated with lower overall survival and a higher risk score. Higher estimated and immune scores, lesser tumor purity, lower mesenchymal EMT, and higher stem scores were all characteristics of the high-risk group. High amounts of tumor-infiltrating lymphocytes, a high mutation load, and a high copy number alteration frequency were present in the high-risk group.Discussion.According to our research, the 16-gene prognostic signature may be helpful in predicting prognosis and developing individualized treatments for patients with renal clear cell carcinoma, which may result in new personalized management options for these patients.

Stromal Type I Collagen in Breast Cancer: Correlation to Prognostic Biomarkers and Prediction of Chemotherapy Response

IntroductionFibrillar collagens accumulate in the breast cancer stroma and appear as poorly defined spiculated masses in mammography imaging. The prognostic value of tissue type I collagen remains elusive in treatment-naïve and chemotherapy-treated breast cancer patients. Here, type I collagen mRNA and protein expression were analysed in 2 large independent breast cancer cohorts. Levels were related to clinicopathological parameters, prognostic biomarkers, and outcome. MethodCOL1A1 mRNA expression was analysed in 2509 patients with breast cancer obtained from the cBioPortal database. Type I collagen protein expression was studied by immunohistochemistry in 1395 women diagnosed with early invasive breast cancer. ResultsLow COL1A1 mRNA and protein levels correlated with poor prognosis features, such as hormone receptor negativity, high histological grade, triple-negative subtype, node positivity, and tumour size. In unadjusted analysis, high stromal type I collagen protein expression was associated with improved overall survival (OS) (HR = 0.78, 95% CI = 0.61-0.99, p = .043) and trended towards improved breast cancer–specific survival (BCSS) (HR = 0.65, 95% CI = 0.42-1.01, P = 0.053), although these findings were lost after adjustment for other clinical variables. In unadjusted analysis, high expression of type I collagen was associated with better OS (HR = 0.70, 95% CI = 0.55-0.90, P = .006) and BCSS (HR = 0.55, 95% CI = 0.34-0.88, P = .014) among patients not receiving chemotherapy. Strikingly, the opposite was observed among patients receiving chemotherapy. There, high expression of type I collagen was instead associated with worse OS (HR = 1.83, 95% CI = 0.65-5.14, P = .25) and BCSS (HR = 1.72, 95% CI = 0.54-5.50, P = .357). ConclusionLow stromal type I collagen mRNA and protein expression are associated with unfavourable tumour characteristics in breast cancer. Stromal type I collagen might predict chemotherapy response.

Artificial Intelligence for Breast Ultrasound: AJR Expert Panel Narrative Review.

Breast ultrasound is used in a wide variety of clinical scenarios, including both diagnostic and screening applications. Limitations of ultrasound, however, include its low specificity and, for automated breast ultrasound screening, the time necessary to review whole-breast ultrasound images. As of this writing, four AI tools that are approved or cleared by the FDA address these limitations. Current tools, which are intended to provide decision support for lesion classification and/or detection, have been shown to increase specificity among non-specialists and to decrease interpretation times. Potential future applications include triage of patients with palpable masses in low-resource settings, preoperative prediction of axillary lymph node metastasis, and preoperative prediction of neoadjuvant chemotherapy response. Challenges in the development and clinical deployment of AI for ultrasound include: the limited availability of curated training datasets compared to mammography; the high variability in ultrasound image acquisition due to equipment- and operator-related factors (which may limit algorithm generalizability); and the lack of post-implementation evaluation studies. Furthermore, current AI tools for lesion classification were developed based on 2D data, but diagnostic accuracy could potentially be improved if multimodal ultrasound data were used, such as color Doppler, elastography, cine clips, and 3D imaging.

A multiomics analysis-assisted deep learning model identifies a macrophage-oriented module as a potential therapeutic target in colorectal cancer

Colorectal cancer (CRC) is a common malignancy involving multiple cellular components. The CRC tumor microenvironment (TME) has been characterized well at single-cell resolution. However, a spatial interaction map of the CRC TME is still elusive. Here, we integrate multiomics analyses and establish a spatial interaction map to improve the prognosis, prediction, and therapeutic development for CRC. We construct a CRC immune module (CCIM) that comprises FOLR2+ macrophages, exhausted CD8+ Tcells, tolerant CD8+ Tcells, exhausted CD4+ Tcells, and regulatory Tcells. Multiplex immunohistochemistry is performed to depict the CCIM. Based on this, we utilize advanced deep learning technology to establish a spatial interaction map and predict chemotherapy response. CCIM-Net is constructed, which demonstrates good predictive performance for chemotherapy response in both the training and testing cohorts. Lastly, targeting FOLR2+ macrophage therapeutics is used to disrupt the immunosuppressive CCIM and enhance the chemotherapy response invivo.

Digital image analysis and machine learning-assisted prediction of neoadjuvant chemotherapy response in triple-negative breast cancer

BackgroundPathological complete response (pCR) is associated with favorable prognosis in patients with triple-negative breast cancer (TNBC). However, only 30–40% of TNBC patients treated with neoadjuvant chemotherapy (NAC) show pCR, while the remaining 60–70% show residual disease (RD). The role of the tumor microenvironment in NAC response in patients with TNBC remains unclear. In this study, we developed a machine learning-based two-step pipeline to distinguish between various histological components in hematoxylin and eosin (H&E)-stained whole slide images (WSIs) of TNBC tissue biopsies and to identify histological features that can predict NAC response.MethodsH&E-stained WSIs of treatment-naïve biopsies from 85 patients (51 with pCR and 34 with RD) of the model development cohort and 79 patients (41 with pCR and 38 with RD) of the validation cohort were separated through a stratified eightfold cross-validation strategy for the first step and leave-one-out cross-validation strategy for the second step. A tile-level histology label prediction pipeline and four machine-learning classifiers were used to analyze 468,043 tiles of WSIs. The best-trained classifier used 55 texture features from each tile to produce a probability profile during testing. The predicted histology classes were used to generate a histology classification map of the spatial distributions of different tissue regions. A patient-level NAC response prediction pipeline was trained with features derived from paired histology classification maps. The top graph-based features capturing the relevant spatial information across the different histological classes were provided to the radial basis function kernel support vector machine (rbfSVM) classifier for NAC treatment response prediction.ResultsThe tile-level prediction pipeline achieved 86.72% accuracy for histology class classification, while the patient-level pipeline achieved 83.53% NAC response (pCR vs. RD) prediction accuracy of the model development cohort. The model was validated with an independent cohort with tile histology validation accuracy of 83.59% and NAC prediction accuracy of 81.01%. The histological class pairs with the strongest NAC response predictive ability were tumor and tumor tumor-infiltrating lymphocytes for pCR and microvessel density and polyploid giant cancer cells for RD.ConclusionOur machine learning pipeline can robustly identify clinically relevant histological classes that predict NAC response in TNBC patients and may help guide patient selection for NAC treatment.

Abstract C063: Keratin 17 as a predictor of chemotherapy response in pancreatic ductal adenocarcinoma

Abstract Background: Two chemotherapies based on either gemcitabine- or 5-fluorouracil (5-FU) remain standard-of-care for pancreatic ductal adenocarcinomas (PDACs). Patients, however, differ in response and thus many clinical trials for experimental drugs have failed. Therefore, there is an urgent need to establish clinically deployable and rapid biomarkers to decide the best chemotherapeutic regimen for subgroups of patients and prevent toxic side effects. We previously reported that keratin 17 (K17) expression, detected at either the level of messenger RNA or protein by immunohistochemistry (IHC), is as accurate as molecular subtyping to identify subgroups of PDAC patients with shortest survival. In this work, we determined the predictive value of K17 testing for both chemotherapeutic approaches to treat PDAC. Methods: We profiled keratin 17 (K17) expression, a hallmark biomarker of the basal molecular subtype of PDAC, using immunohistochemistry in two cohorts of formalin-fixed paraffin-embedded PDACs (total n=305). These cases were collected through the Know Your Tumor® initiative and three high-volume treating institutions in the U.S. We performed multivariate predictive analyses and determined the threshold to define positive vs negative test. Results: Patients with advance stage disease, low K17, and treated with 5-FU- chemotherapy had three-times longer survival rates compared to counterparts treated with gemcitabine-based chemotherapy. These results contrast to those found in cases with high K17, where no differences were seen with either regimen. Based on RECIST, 72% of low K17 cases responded to treatment compared to 24% of high K17 cases. Predictive value of K17 was independent of mutation status and clinical variables. Conclusions: Our findings indicate that 5-FU-, rather than gemcitabine- based therapies are more likely to extend survival for patient's tumors expressing low rather than high K17. A K17 IHC test rapid test in PDACs has been patented (US11,092,603) and it is currently being tested in prospective phase II clinical trials (NCT04469556, NCT02047474 and NCT03991962). Results from these trials will test the presented findings. Citation Format: Lyanne Oblein, Lucia Roa, Sruthi Babu, Michael Horowitz, Emanuel Petricoin, Lynn M. Matrisian, Edik Blais, Natalia Marchenko, Felicia Allard, Ali Alkalin, Wei Jiang, Brent Larson, Andrew Hendifar, Vincent Picozzi, Minsig Choi, Luisa Escobar-Hoyos, Kenneth Shroyer. Keratin 17 as a predictor of chemotherapy response in pancreatic ductal adenocarcinoma [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Pancreatic Cancer; 2023 Sep 27-30; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(2 Suppl):Abstract nr C063.

XPA tumor variant leads to defects in NER that sensitize cells to cisplatin.

Nucleotide excision repair (NER) reduces efficacy of treatment with platinum (Pt)-based chemotherapy by removing Pt lesions from DNA. Previous study has identified that missense mutation or loss of the NER genes Excision Repair Cross Complementation Group 1 and 2 (ERCC1 and ERCC2) leads to improved patient outcomes after treatment with Pt-based chemotherapies. Although most NER gene alterations found in patient tumors are missense mutations, the impact of mutations in the remaining nearly 20 NER genes is unknown. Towards this goal, we previously developed a machine learning strategy to predict genetic variants in an essential NER protein, Xeroderma Pigmentosum Complementation Group A (XPA), that disrupt repair. In this study, we report in-depth analyses of a subset of the predicted variants, including in vitro analyses of purified recombinant protein and cell-based assays to test Pt agent sensitivity in cells and determine mechanisms of NER dysfunction. The most NER deficient variant Y148D had reduced protein stability, weaker DNA binding, disrupted recruitment to damage, and degradation. Our findings demonstrate that tumor mutations in XPA impact cell survival after cisplatin treatment and provide valuable mechanistic insights to improve variant effect prediction. Broadly, these findings suggest XPA tumor variants should be considered when predicting chemotherapy response.