Life Sciences Research Articles

Novel Polypeptide Automated Synthesis Devices: A Review

Peptides are a unique class of proteins and one-dimensional biological nanomaterials that play an exceptionally important role in life science and industrial applications. At present, peptide synthesis equipment encounters issues like low synthetic efficiency, challenges in scaling up, and limited automation. This article provides an overview of the key technologies in peptide synthesis equipment, covering aspects such as peptide information screening, peptide synthesis, and detection. It analyzes and summarizes the principles, methods, and critical challenges of traditional solid-phase peptide synthesis technology, microwave-assisted solid-phase peptide synthesis technology, and continuous-flow solid-phase peptide synthesis technology. It also discusses microfluidic solid-phase peptide synthesis technology and future research trends, offering insights into the search for an efficient, high-yield, high-purity, scalable, and intelligent peptide synthesis method, which holds significant research significance.

Establishing a Freshwater Turtle (Emydura subglobosa) Laboratory Line (FTLL) as a novel model species for research and education

<b>The Jardine River turtle (Emydura subglobosa) was selected as a potential model species for studies on freshwater turtles and general reptile physiology. Attempts to establish a freshwater turtle laboratory line were made when an adult pair of E. subglobosa was received in 2016 by the Laboratory of Inland Fisheries and Aquaculture (Poznań University of Life Sciences). The first generation of offspring was obtained in 2017, and the second generation in 2023. In each generation, unrelated specimens were added to the animal cohort to avoid inbreeding. Husbandry regimes were established, and a basal diet for nutritional experiments was developed and manufactured by two methods, producing extruded feed and a gelatine-solidified variant. The establishment of the Freshwater Turtle Laboratory Line (FTLL) provides an opportunity to improve the development of husbandry techniques, increase knowledge of reptile physiology, and use laboratory-raised animals as model species for research and education.

Fast Pure Shift NMR Spectroscopy Using Attention-Assisted Deep Neural Network.

Pure shift NMR spectroscopy enables the robust probing on molecular structure and dynamics, benefiting from great resolution enhancements. Despite extensive application landscapes in various branches of chemistry, the long experimental times induced by the additional time dimension generally hinder its further developments and practical deployments, especially for multi-dimensional pure shift NMR. Herein, this study proposes and implements the fast, reliable, and robust reconstruction for accelerated pure shift NMR spectroscopy with lightweight attention-assisted deep neural network. This deep learning protocol allows one to regain high-resolution signals and suppress undersampling artifacts, as well as furnish high-fidelity signal intensities along with the accelerated pure shift acquisition, benefitting from the introduction of the attention mechanism to highlight the spectral feature and information of interest. Extensive results of simulated and experimental NMR data demonstrate that this attention-assisted deep learning protocol enables the effective recovery of weak signals that are almost drown in the serious undersampling artifacts, and the distinction and recognition of close chemical shifts even though using merely 5.4% data, highlighting its huge potentials on fast pure shift NMR spectroscopy. As a result, this study affords a promising paradigm for the AI-assisted NMR protocols toward broader applications in chemistry, biology, materials, and life sciences, and among others.

A miRNA-disease association prediction model based on tree-path global feature extraction and fully connected artificial neural network with multi-head self-attention mechanism

BackgroundMicroRNAs (miRNAs) emerge in various organisms, ranging from viruses to humans, and play crucial regulatory roles within cells, participating in a variety of biological processes. In numerous prediction methods for miRNA-disease associations, the issue of over-dependence on both similarity measurement data and the association matrix still hasn’t been improved. In this paper, a miRNA-Disease association prediction model (called TP-MDA) based on tree path global feature extraction and fully connected artificial neural network (FANN) with multi-head self-attention mechanism is proposed. The TP-MDA model utilizes an association tree structure to represent the data relationships, multi-head self-attention mechanism for extracting feature vectors, and fully connected artificial neural network with 5-fold cross-validation for model training.ResultsThe experimental results indicate that the TP-MDA model outperforms the other comparative models, AUC is 0.9714. In the case studies of miRNAs associated with colorectal cancer and lung cancer, among the top 15 miRNAs predicted by the model, 12 in colorectal cancer and 15 in lung cancer were validated respectively, the accuracy is as high as 0.9227.ConclusionsThe model proposed in this paper can accurately predict the miRNA-disease association, and can serve as a valuable reference for data mining and association prediction in the fields of life sciences, biology, and disease genetics, among others.Graphical

Intelligent bell facial paralysis assessment: a facial recognition model using improved SSD network

With the continuous progress of technology, the subject of life science plays an increasingly important role, among which the application of artificial intelligence in the medical field has attracted more and more attention. Bell facial palsy, a neurological ailment characterized by facial muscle weakness or paralysis, exerts a profound impact on patients’ facial expressions and masticatory abilities, thereby inflicting considerable distress upon their overall quality of life and mental well-being. In this study, we designed a facial attribute recognition model specifically for individuals with Bell’s facial palsy. The model utilizes an enhanced SSD network and scientific computing to perform a graded assessment of the patients’ condition. By replacing the VGG network with a more efficient backbone, we improved the model’s accuracy and significantly reduced its computational burden. The results show that the improved SSD network has an average precision of 87.9% in the classification of light, middle and severe facial palsy, and effectively performs the classification of patients with facial palsy, where scientific calculations also increase the precision of the classification. This is also one of the most significant contributions of this article, which provides intelligent means and objective data for future research on intelligent diagnosis and treatment as well as progressive rehabilitation.

Systematic evaluation with practical guidelines for single-cell and spatially resolved transcriptomics data simulation under multiple scenarios

BackgroundSingle-cell RNA sequencing (scRNA-seq) and spatially resolved transcriptomics (SRT) have led to groundbreaking advancements in life sciences. To develop bioinformatics tools for scRNA-seq and SRT data and perform unbiased benchmarks, data simulation has been widely adopted by providing explicit ground truth and generating customized datasets. However, the performance of simulation methods under multiple scenarios has not been comprehensively assessed, making it challenging to choose suitable methods without practical guidelines.ResultsWe systematically evaluated 49 simulation methods developed for scRNA-seq and/or SRT data in terms of accuracy, functionality, scalability, and usability using 152 reference datasets derived from 24 platforms. SRTsim, scDesign3, ZINB-WaVE, and scDesign2 have the best accuracy performance across various platforms. Unexpectedly, some methods tailored to scRNA-seq data have potential compatibility for simulating SRT data. Lun, SPARSim, and scDesign3-tree outperform other methods under corresponding simulation scenarios. Phenopath, Lun, Simple, and MFA yield high scalability scores but they cannot generate realistic simulated data. Users should consider the trade-offs between method accuracy and scalability (or functionality) when making decisions. Additionally, execution errors are mainly caused by failed parameter estimations and appearance of missing or infinite values in calculations. We provide practical guidelines for method selection, a standard pipeline Simpipe (https://github.com/duohongrui/simpipe; https://doi.org/10.5281/zenodo.11178409), and an online tool Simsite (https://www.ciblab.net/software/simshiny/) for data simulation.ConclusionsNo method performs best on all criteria, thus a good-yet-not-the-best method is recommended if it solves problems effectively and reasonably. Our comprehensive work provides crucial insights for developers on modeling gene expression data and fosters the simulation process for users.

Multimodal analysis of traction forces and the temperature dynamics of living cells with a diamond-embedded substrate

Cells and tissues are constantly exposed to chemical and physical signals that regulate physiological and pathological processes. This study explores the integration of two biophysical methods: traction force microscopy (TFM) and optically detected magnetic resonance (ODMR) to concurrently assess cellular traction forces and the local relative temperature. We present a novel elastic substrate with embedded nitrogen-vacancy microdiamonds that facilitate ODMR-TFM measurements. Optimization efforts focused on minimizing sample illumination and experiment duration to mitigate biological perturbations. Our hybrid ODMR-TFM technique yields TFM maps and achieves approximately 1 K precision in relative temperature measurements. Our setup employs a simple wide-field fluorescence microscope with standard components, demonstrating the feasibility of the proposed technique in life science laboratories. By elucidating the physical aspects of cellular behavior beyond the existing methods, this approach opens avenues for a deeper understanding of cellular processes and may inspire the development of diverse biomedical applications.

Social and natural environmental factors cause the birth-cohort phenomenon of peptic ulcers by superposition mechanism

ABSTRACT The widely believed Helicobacter pylori infection has never explained the birth-cohort phenomenonof peptic ulcers. Although numerous studies have observed that environmental factors are associated with peptic ulcers, their role in the disease has yet to be identified. A new etiological theory proposed that environmental factors cause peptic ulcers via inducing psychological stress. Starting from this etiology, an integration of the mortality rates caused by social and natural environmental factors reproduced a representative fluctuation curve in the birth-cohort phenomenon, where a causal role of environmental factors in peptic ulcers was hidden. The reproduced fluctuation curve revealed that multiple environmental factors caused the birth-cohort phenomenon by Superposition Mechanism, and the causal role of each individual environmental factor surfaced if the fluctuation curves in the birth-cohort phenomenon were properly differentiated. A full understanding of the birth-cohort phenomenon highlights the importance of environmental management in improving clinical outcomes, and suggests that the Superposition Mechanism is an indispensable methodological concept for life science and medicine.

Data Commoning in the Life Sciences

Data Commoning in the Life Sciences

Cover Picture: Engineering in Life Sciences 6'24

Cover Picture: Engineering in Life Sciences 6'24

Kinome reprogramming as a therapeutic opportunity in ESR1 fusion driven breast cancer but not in gynecologic cancers.

1045 Background: The administration of specific tyrosine-kinase inhibitors in kinase fusion positive cancers has led to significant improvement in clinical outcomes. Gene fusions involving the ESR1 gene, a non-kinase, have been previously reported in breast and gynecologic cancers including ovarian cancer and uterine cancer. However, ERa inhibitors proved to be ineffective in ESR1 fusion positive cases, we have effectively characterized ESR1 fusions and hypothesized a surrogate downstream target as an effective approach to target ESR1 fusion positive breast cancer. Methods: We retrospectively analyzed 216,176 samples submitted for clinical molecular profiling at Caris Life Sciences via Caris MI Tumor Seek (Phoenix, AZ). Gene fusion detection was performed on mRNA isolated from a FFPE tumor sample using the Illumina NovaSeq platform. All ESR1 fusions with ≥ 3 junction reads were identified for manual review and for domain prediction, breakpoints, frame retention and co-occurring alterations by NGS. WES data was also investigated to evaluate co-occurring pathogenic mutations to establish the oncogenicity of gene fusions. Results: ESR1 fusions had the highest frequency in gynecologic and breast cancer. We evaluated ESR1 expression in cases with ESR1 fusions detected and observed higher expression of ESR1 in breast cancer, ovarian cancer (P<0.0001) and uterine cancer (P<0.0001) compared to respective cancer cases as determined by TPM values. The Kaplan-Meier curve also indicated poor prognosis for breast cancer patients with ESR1 fusions (619 days [n=105]) as opposed to patients who did not harbor any ESR1 variants (1396 days [n=755]; P<0.0001). Further, pathway analysis was performed on breast cancer ESR1 fusion positive and negative cases to discern specific downstream markers that might be upregulated. It was observed that RET(p<0.0001), IGF1R(p=0.0001) and FGFR3 (p<0.0001) were upregulated in ESR1 fusion positive breast cancer cases which may be largely determined by the TP53-FOXA1-ER alpha transcriptional axis. However, no significant increase of RET, IGF1R and FGFR3 was observed in gynecologic cancers, owing to high TP53 mutation frequency observed in ovarian cancer (90%) and uterine cancer (82%) compared to breast cancer(19%). Conclusions: We highlight that ESR1 fusions can be reliably detected in patients with breast cancer, ovarian cancer, and endometrial cancers. ESR1 fusions lose the ligand binding domain and cannot be targeted using current estrogen modulators or degraders. We have discovered the upregulation of oncogenic kinase signaling in ESR1 fusion positive breast cancer, but not in gynecologic cancers. Our observations highlight enhanced oncogenic kinase signaling as a surrogate approach with clinically approved kinase inhibitors for targeting ESR1 fusion positive breast cancer cases but not in ovarian and uterine cancers.

Metacognition and Self-Efficacy in Action: How First-Year Students Monitor and Use Self-Coaching to Move Past Metacognitive Discomfort During Problem Solving.

Stronger metacognitive regulation skills and higher self-efficacy are linked to increased academic achievement. Metacognition and self-efficacy have primarily been studied using retrospective methods, but these methods limit access to students' in-the-moment metacognition and self-efficacy. We investigated first-year life science students' metacognition and self-efficacy while they solved challenging problems, and asked: 1) What metacognitive regulation skills are evident when first-year life science students solve problems on their own? and 2) What aspects of learning self-efficacy do first-year life science students reveal when they solve problems on their own? Think-aloud interviews were conducted with 52 first-year life science students across three institutions and analyzed using content analysis. Our results reveal that while first-year life science students plan, monitor, and evaluate when solving challenging problems, they monitor in a myriad of ways. One aspect of self-efficacy, which we call self-coaching, helped students move past the discomfort of monitoring a lack of understanding so they could take action. These verbalizations suggest ways we can encourage students to couple their metacognitive skills and self-efficacy to persist when faced with challenging problems. Based on our findings, we offer recommendations for helping first-year life science students develop and strengthen their metacognition to achieve improved problem-solving performance.

The prevalence of sickle cell phenotype and its association with clinical outcomes in patients with solid malignancies.

10563 Background: Sickle cell disease (SCD) is the most common inherited blood disorder in the United States and is caused by a mutation of β-globin ( HBB-Mt). The risk of developing solid malignancies in this population remains controversial, with research showing increased rates of some solid tumors (colon, kidney, and thyroid) and decreased rates of others (breast and male genitourinary). We report the prevalence of HBB mutations across a cohort of solid tumors and clinical outcomes between HBB-Wild type (-Wt) v Mutant (-Mt) tumors. Methods: Non-small cell lung cancer (NSCLC, N = 3307), breast (BC, N = 1960), colorectal (CRC, N = 2161), prostate (N = 899), and gynecologic cancers (N = 2,805) in persons of African descentwere tested at Caris Life Sciences (Phoenix, AZ) with NextGen Sequencing of DNA (whole exome) and RNA (whole transcriptome). Her2/Neu (+: ≥3+ and >10%) and HR (ER or PR+: ≥1+ and ≥1%) expression was tested by immunohistochemistry. Tumors were assessed for single nucleotide variations (SNVs) and insertions/deletions (indels) in HBB associated with a sickle cell phenotype. Mutation prevalence for all other genes was calculated for pathogenic SNVs/indels. Differentially regulated pathways were assessed by gene set enrichment analysis (GSEA). Fisher’s Exact/χ2 tests were applied as appropriate with p-values adjusted for multiple comparisons ( p < .05). Real-world overall survival (OS) and ethnicity data was obtained from insurance claims and log-rank estimates were calculated for molecularly defined subpopulations. Results: Amongst persons of African descent, uterine sarcoma had the highest prevalence of HBB-Mt (17.1%, 43/251) followed by uterine serous carcinoma (9.1, 70/768) and endometrial carcinoma (7.8, 55/708). Tumors with the highest absolute number of HBB-Mt tumors were NSCLC (7.0, 230/3307), CRC (8.1, 176/2161) and BC (6.7, 131/1960). In CRC, there was no significant difference in the prevalence of right-sided tumors between HBB-Wt v Mt (42% v 51, p = .06), nor in the prevalence of HBB-Mt by consensus molecular subtype ( p = .34). No significant enrichment of HBB-Mt across BC subtypes was observed (HR- HER2-: 7.8%, HR-HER2+: 9.7, HR+/HER2-:5.8, HR+/HER2+: 8.1, p = .34). No difference in the prevalence of pathogenic alterations was observed between HBB-Mt vs -Wt across NSCLC, CRC, and BC ( p > .05). HBB-Mt NSCLC were significantly enriched (GSEA) for genes related interferon /, inflammatory response and IL6/JAK/STAT3. No significant difference in OS was observed in CRC (median survival HBB-Wt: 957 days v -Mt: 1034, p = .43), NSCLC (-Wt: 591 v -Mt: 648, p = .59) or BC (-Wt: 924 v -Mt: 922, p = .61). Conclusions: No significant differences in the genomic landscape nor in OS of HBB-Mt vs HBB-Wt NSCLC, CRC or BC were observed. NSCLC HBB-Mt was enriched with inflammatory response genes. Future work should focus on potential role HBB-Mt plays in NSCLC.

Association of germline HSD3B1 (c.1100) genotype with tumoral and clinical outcomes in breast and endometrial cancers.

10587 Background: A germline single-nucleotide polymorphism in HSD3B1 results in adrenal-permissive (c.1100 C) and adrenal-restrictive (c.1100 A) phenotypes with respect to androgen synthesis. In pre- and post-menopausal patients with breast cancer (n=3548 and 7871) and endometrial cancer (n=1588 and 8439), we annotated samples as adrenal permissive (CC/AC) or restrictive (AA) based on HSD3B1 variant status and then examined tumoral characteristics and clinical outcomes. Methods: We utilized tumors from the Caris Life Sciences Precision Oncology Alliance database and inferred germline HSD3B1 c.1100 genotype using variant allele frequencies (VAFs): c.1100 C VAF of 0% defined the AA genotype, VAF of 40-60% defined the AC genotype, and VAF of 100% defined the CC genotype. We explored associations of each HSD3B1genotypes with other genomic (WES) and transcriptomic (RNA-seq) features, as well as with survival outcomes. To performed consider menopausal status we used the patient’s age at the time of biopsy (>55 indicating post-menopausal). Results: In breast cancer (BC) and endometrial cancer (EC), the permissive HSD3B1genotype had greater mRNA expression of the estrogen receptor ( ESR1) and co-alterations of the TP53 (p<0.0001). In post-menopausal ECs, the permissive genotype was associated with increased alterations in the PI-3-kinase pathway ( PTEN, PIK3CA, PIK3R1, p<0.0001). In pre-menopausal HER2+ BCs, the restrictive genotype was associated with worse survival (HR=1.40, 95% CI = 1.05-1.85, p=0.027) and exhibited increases in hallmark GSEA signatures including epithelial-to-mesenchymal transition. Post-menopausal basal BCs with the permissive genotype had numerically higher PD-L1 protein expression (50% vs 13%) and less PI-3-kinase alterations ( PTEN, PIK3R1, p<0.01). Post-menopausal clear cell ECs with the restrictive genotype exhibited worse survival (HR=1.60, 95% CI=1.07-2.38, p=0.03) and were 3-fold more likely to be TMB-high or MSI-high (p<0.05). Distinctly, post-menopausal endometrioid ECs of the restrictive genotype were less likely to be TMB- or MSI-high (p<0.05). Conclusions: Breast and endometrial cancers can be classified as adrenal-permissive or -restrictive based on HSD3B1 c.1100 genotype. This classification distinguishes clinical outcomes and tumoral features, particularly when stratifying by menopausal status. Importantly, these genotypes are differentially associated with druggable pathways (PI-3-kinase) and markers of immune sensitivity (PD-L1, TMB, MSI). [Table: see text]

Molecular and immunological characterization of androgen receptor expression in different breast cancer subtypes.

1114 Background: While estrogen receptor (ER) is well-studied in breast cancer (BC), the role of androgen receptor (AR) in prognosis and therapy response is less understood. Here, we aimed to characterize the clinicopathologic and molecular features of AR gene expression in BC subtypes. Methods: 10,728 BC samples were tested by NGS (592, NextSeq; WES, NovaSeq) and WTS (NovaSeq; Caris Life Sciences, Phoenix, AZ). MSI was tested by IHC and NGS. TMB-high was designated as >10 somatic mutations/MB. Tumors with AR-high(H) and AR-low(L) RNA expression were classified by top and bottom quartile, respectively. Real world overall survival (OS) and treatment-associated survival was obtained from insurance claims and calculated from treatment start to last contact using Kaplan-Meier estimates. Statistical significance was determined by chi-square and Mann-Whitney U test with p-values adjusted for multiple comparisons ( q<.05). Results: Lobular carcinoma had higher AR expression (2.37-fold) and AR positivity by IHC (93.6% vs 63.6%) compared to ductal carcinoma ( q<.05). Luminal A (LumA) had higher AR expression and IHC positivity compared to HER2-enriched, basal-like, and normal-like tumors ( q<.05). AR-H tumors were significantly ( p<.05) enriched for PIK3CA mutations in LumA (60.3% vs 29.8%), LumB (48.1% vs 23.2%), HER2-enriched (51.2% vs 39.3%), and basal-like (23.8% vs 8.3%); CDH1 mutations in LumA (26% vs 10.3%) and normal-like (55.5% vs 4.8%); MAP3K1 in LumA (17.1% vs 5.3%); ESR1 in LumB (17.9% vs 10.7%); and NF1 in HER2-enriched (19.3% vs 7.1%). AR-H had higher immune infiltration for LumA (B cells, M2 Mφ, neutrophils, NK cells, DC), LumB (B cells, M2 Mφ, neutrophils, NK cells, CD4+ T cells, DC), basal-like (NK cells, CD4+ T cells), and normal-like (NK cells, DC; q<.05) than AR-L. Conversely, AR-H HER2-enriched tumors had lower immune infiltration (B cells, M1 and M2 Mφ, NK cells, CD8+ T cells, DC) and fewer TMB-high (15.1% vs 23.2%), dMMR/MSI-high (0.3% vs 2%), and PD-L1+ tumors (15% vs 40.8%; p<.05) compared to AR-L (all q<.05). Post-doxorubicin survival was longer for patients with AR-H tumors; however, opposing trends were observed for patients with LumB (75.5 vs 56.3 m; p=.081) and HER2-enriched tumors (47.9 vs 65.2 m; p.071). AR-H was associated with longer post-dox survival in patients with LumB PIK3CA-mt tumors (81.7 vs 39.3 m; p=.036), but shorter post-dox survival in HER2-enriched PIK3CA-wt (50.4 vs 70.4 m) and PIK3CA-mt cohorts (40.8 vs 58.0 m; p=.18). Conclusions: Findings from our updated analysis based on PAM50 designation suggest a strong association between AR and PIK3CA mutations across subtypes and suggest that distinct AR-associated alterations in the immune microenvironment require further study. Exploration of specific mutations and immune-oncology markers associated with AR-H may aid in molecularly selected clinical trial design for advanced BC patients.

Real-world impact of KRAS mutations on clinical outcomes in metastatic colorectal cancer treated with trifluridine-tipiracil in the USA.

e15566 Background: Defining the impact of KRAS mutation status on outcomes to standard of care treatments in metastatic colorectal cancer (mCRC) is of increasing clinical relevance. We interrogated a large real-world genomic database to further explore the association between KRAS mutations and benefit to trifluridine-tipiracil (FTD-TPI) in patients with mCRC. Methods: 19,364 patients with mCRC and known KRAS mutation status were analyzed by next-generation sequencing (NextSeq, 592 Genes; NovaSeq, whole exome sequencing (WES)) at Caris Life Sciences. Real-world overall survival (rwOS) and time on treatment (ToT) was inferred from insurance claims data and calculated from time of start of treatment with FTD-TPI to last contact and time from first to last treatment with FTD-TPI with comparison done by the Kaplan-Meier test, respectively. Chi-square test was applied where appropriate, with p-value adjusted for multiple comparisons (p < 0.05). Results: 764 of 10,571 (7.2%) patients with wild-type (WT) KRAS, 768 of 6,984 (11%) with KRAS G12 mutation, and 170 of 1,809 (9.4%) with KRAS G13 mutation received FTD-TPI. Age distribution was similar among groups. Sidedness of primary site was reported in 692 (40.3%) patients; an increased proportion of right-sided tumors was observed in KRAS G12 (53.4% vs. 28.3%, p < 0.0001) and G13 (56% vs 28.3%, p < 0.0001) groups compared to KRAS WT. An increased proportion of TP53 mutations in KRAS WT was observed compared to KRAS G12 mutation (87% vs. 70%, p < 0.0001) and KRAS G13 mutation groups (87% vs. 75%, p < 0.0001). KRAS G12 mutation was associated with shorter median rwOS compared to KRAS WT (7.33 vs. 8.43 months (mo), HR 1.13 (95% CI 1.11-1.42), p = 0.041). KRAS G13 mutation was associated with similar rwOS to KRAS WT. Median ToT was shorter in patients with KRAS G12 mutations than KRAS WT (1.8 vs. 2 mo, HR of 1.22 (95% CI 1.11-1.42), p < 0.001). Median ToT was longer in patients with KRAS G13 versus KRAS G12 (2.3 vs. 1.8 mo, HR 0.82 (95% CI 0.68-0.98), p = 0.032). Conclusions: To our knowledge, this is the largest primarily US-based cohort of mCRC treated with FTD-TPI to be reported. KRAS G12 mutation, but not KRAS G13 mutation, was associated with shorter rwOS and ToT compared to KRAS WT, despite enrichment of TP53 and BRAF mutations in the KRAS WT group. This finding is consistent with previously published analyses of large randomized prospective trials of FTD-TPI in mCRC and further support KRAS mutation status as a predictive biomarker in this setting. [Table: see text]

Therapeutic insights for the aggressive subset of high-grade gliomas (HGG) driven by chromosome 1q32 MDM4-containing amplicon and unmethylated MGMT.

2080 Background: MDM4/MDMX amplification ( MDM4 amp) is associated with poor outcomes, including after immune checkpoint blockade therapy, in multiple tumor types (El-Deiry et al., ASCO, 2022). Glioblastoma (GBM) displays the highest rate of MDM4 amp (9.63%) among all tumor types (Arnoff and El-Deiry, 2022). MDM4 located on chromosome 1q32 can be amplified separately from contactin 2 ( CNTN2) although they can be co-amplified among 17 genes with MDM4 as main amplification target (Riemenschneider et al., 2003). There are no FDA-approved drugs targeting MDMX although there are preclinical compounds such as XI-006 (NSC208895) that can reduce cell migration (George et al., AACR, 2023; De La Cruz, AACR, 2023). We recently uncovered immune-stimulatory effects of temozolomide (TMZ) within hours of drug exposure, prior to cell division and mutation accumulation in microsatellite stable colorectal cancer and HGG (Gonzalez et al., 2023). Here we explored MDM4 amp, co-expressed genes, MGMT, clinical outcomes and response to TMZ. Methods: HGG tested at Caris Life Sciences (Phoenix, AZ) with NextGen Sequencing on DNA and RNA (NGS) were analyzed. MDM4 amp was determined with a cutoff of >/= 4.0 copies. MGMT promoter methylation (me) was tested by pyrosequencing. Patient outcome was obtained from insurance claims data and calculated from start of TMZ (tmzOS) or tissue collection to last contact (rwOS). Cox proportional hazards model, X2/Fisher-Exact was used and p values were adjusted for multiple correction (q value) when appropriate. Results: In a large cohort of 3856 HGG wild-type for IDH1/2, H3F3A and TP53, 300 (8.4%) had MDM4 amp and showed a shorter OS compared to MDM4-not amplified tumors (median rwOS: 14.9 vs 17.3 mo, HR: 1.286; 95% CI: 1.12-1.48, p<0.001). The effect is significant after accounting for age and gender (adjusted p=0.016). MDM4 amp was associated with increased PIK3C2B (fold change: 6.7) and PPP1R15B (FC: 7.9, both q<0.01) expression among other candidates for co-targeting present on Chr 1q32.1; while no significant difference was seen with MGMT-me (p=0.3). Among patients with unmethylated MGMT treated with TMZ (N=1066), MDM4 amp (N=101) was associated with shorter tmzOS (14.3m vs. 15.9m, HR: 1.27, 1.01-1.59, p=0.038) while a trend was seen in the 926 methylated patients (MDM4 Amp: 69, not amplified: 857; tmzOS: 22 m vs. 27.7m, HR: 1.33; 0.95-1.85), p=0.092.). Conclusions: MDM4 amp is negatively associated with clinical outcomes in TP53-WT HGG. The MDM4amplicon on Chr 1q32.1 has associated targets such as PIK3C2B and phosphatase PPP1R15B that may be suitable for therapeutic targeting in a subset of aggressive HGG’s with MDM4 amp. The Chr 1q32 amplicon contains therapeutic targets known to attenuate immune responses and thus their targeting may be combined with various treatments including immunotherapy, targeted therapies, chemotherapy or vaccines.

A digital pathology AI model to predict immune-oncology biomarker status and pembrolizumab response in a real-world cohort of patients with colorectal and breast cancer.

e13617 Background: Deep learning (DL) algorithms have garnered significant attention for prediction of biomarker status in digitized tissue slides. Here, we trained and validated a digital pathology algorithm to predict microsatellite instability (MSI)/mismatch repair deficiency (MMRd) in colorectal cancer (CRC) and PD-L1 status in breast cancer (BC) using a large real-world patient cohort. Methods: H&E and immunohistochemistry (IHC) slides of CRC and BC from Caris Life Sciences (Phoenix, AZ) with known biomarker status were digitally scanned (CRC: N=27,494, 6.6% MSI-high by next-generation sequencing [NGS]; N=25,788, 6.4% MMRd by IHC; BC: N=16,069, 15.8% PD-L1+ by IHC). A transformer-based DL model (Wagner, Sophia J., et al., 2023) was trained to predict biomarker status, with 5-fold cross-validation. Accuracy, F1 score, area under the curve (AUC), and other metrics were calculated comparing the model to NGS/IHC “truth”. Risk stratification was performed on holdout datasets of pembrolizumab-treated CRC (N=422-454) and triple negative breast cancer (TNBC; N=513) pts. Post-treatment overall survival (OS) was calculated from start of pembrolizumab to last contact, per insurance claims data. Hazard ratios (HR) and 95% confidence intervals (CI) were calculated by Cox-Proportional hazards and compared between model and “truth”. Results: When applied to holdout datasets for biomarker prediction, the validated model achieved AUC of 0.78-0.95. In CRC and BC, performance was better among primary specimens than metastatic. For MSI-H/MMRd risk stratification in pembrolizumab-treated CRC, HR for OS were comparable between model and “truth” for primary and metastatic cohorts. When applied to the TNBC holdout dataset, the model performed better than “truth” for PD-L1 risk stratification in pembrolizumab-treated pts. Conclusions: A digital pathology DL model predicts biomarker status with high accuracy and is non-inferior to NGS/IHC for risk stratification of pembrolizumab-treated CRC and TNBC pts. These results support further investigation of AI algorithms developed from large datasets as possible rapid screening tools for biomarker detection in molecular testing programs. [Table: see text]

Uncovering actionable genetic alterations and immune predictive biomarkers for anal squamous cell carcinomas in the era of immunotherapy: PD-L1 and beyond.

3518 Background: Squamous cell carcinoma of the anal canal (SCAC) is a rare cancer with limited treatments. Immune checkpoint inhibitors (ICIs) use is routine for recurrent/refractory disease, but predictive biomarkers remain elusive. We analyzed the largest dataset of SCAC to date, focusing on PD-L1 expression. Methods: Next-generation sequencing of DNA (592 genes or WES) and RNA (WTS) was tested at Caris Life Sciences (Phoenix, AZ). PD-L1 was tested by immunohistochemistry (IHC) (SP142) and grouped as high (2+ and ≥5%), low (1-2 and 1-5%), and negative (0) CPS. dMMR/MSI-H was tested by IHC/NGS and tumor mutation burden (TMB)-High was defined as >10 mt/MB. RNA expression data was used to estimate the tumor microenvironment (TME) using QuantiSEQ. RNA signatures predictive of ICI response (interferon gamma [IFNγ]; T-cell inflamed signature [TIS]) were tested. X2/Fisher-Exact was used with significance shown as P-value adjusted for multiple comparisons (Q<0.05). Real-world overall survival (rwOS) was from insurance claims and calculated from tissue collection to last contact; time-on-treatment (TOT) was from start to finish of ICI. Results: Of 1242 SCAC samples, 64.3% (798) expressed PD-L1, 41.7% (518) showed high PD-L1 expression, 1.2% were dMMR/MSI-H and 14.5% were TMB-high. TMB-H trended higher in PD-L1 low vs PD-L1 negative (16.3% vs 10.2%, p=0.036). No trend was seen for dMMR/MSI-H. Mutations in PIK3CA (38.3% vs 21.7%) and CASP8(5.3% vs 0.3%) were significantly higher in PD-L1 high vs negative. Similar trends were seen in NF1 (2.4% vs 0.4%) and ARID1A (2.7% vs 0.6%). The opposite was noted for JAK2 (0% vs 2.6%), BAP1 (1.1% vs 3.6%) mutations, and FGFR3 amplifications (0% vs 2.2%). For PD-L1 high vs negative, infiltration of Tregs, M1 macrophages, neutrophils, CD8+ cells, and cancer-associated fibroblasts decreased with PD-L1 expression (fold change [FC]: 0.67-0.80, Q<0.05). Expression of immune-oncology (IO) markers IDO1, PDCD1, IFNG, CD274, HVACR2, CTLA4, LAG3, CD80, CD86 and PDCD1LG2 decreased with PD-L1 expression (FC:0.35-0.73, Q<0.05). Similarly, TIS and IFNg scores decreased with PD-L1 IHC (Q<0.05). In 316 SCAC patients treated with ICIs, PD-L1 high had longer TOT compared to PD-L1 negative (N=168 vs 80, 5.5m vs 2.9m, HR 0.758, 95% CI 0.579 to 0.992, p=0.044) while PD-L1 low (N=68) showed a TOT in between (3.2m) but not significantly different compared to high or negative. No differences in rwOS were observed. Conclusions: PD-L1 is expressed in over 50% of SCACs, while dMMR/MSI-H is rare. High PD-L1-expressing tumors have higher PIK3CAand CASP8 mutations and overexpress IO markers, are inflamed, and have prolonged treatment times with ICIs. This is the largest study to date reporting SCAC genomic profiles and identifies the utility of PD-L1 as a predictive biomarker of IO efficacy. This discovery needs to be confirmed in prospective trials.

Characterization of necroptosis activators in colorectal cancer.

3522 Background: Necroptosis is a pro-inflammatory form of programmed cell death triggered by cellular stresses and extrinsic cytokines. It is a potential therapeutic target for immunotherapy (IO) in colorectal cancer (CRC). Here, we present comprehensive clinical and molecular characterization of key necroptosis regulators in CRC. Methods: 24,257 samples of CRC tested at Caris Life Sciences (Phoenix, AZ) with WTS (Illumina NovaSeq), NextGen DNA sequencing (NextSeq, 592 Genes and NovaSEQ, WES), and PD-L1 expression (22C3 clone; TPS ≥1%) were analyzed. Expression of activators in the necroptosis pathway ( RIPK3, MLKL, CYLD, and ZBP1) was categorized as either high (H) or low (L) using median expression as the cutoff. CRCs with high expression of all four activators were considered to have high propensity for necroptosis activity (NA-H, N=4595) compared to tumors with low expression in all activators (NA-L, N = 4946). RNA deconvolution analysis with QuantiSEQ estimated immune cell infiltration in the tumor microenvironment (TME). Differences in overall survival (OS) were analyzed from insurance claims data and calculated from treatment initiation using Kaplan-Meier estimates. Statistical significance was set as a P-value adjusted for multiple comparisons ( q<0.05). Results: NA-H was associated with increased PD-L1 positivity (4.0% vs 2.3%), immune cell infiltration, and T-cell inflamed score (all q<0.001). TP53 mutations were more frequent in NA-H, while SMAD2, SMAD4, KRAS, and NRAS mutations were associated with NA-L (all q<0.01). NA-H was associated with activation of KRAS signaling, allograft rejection, inflammatory/interferon gamma response, and pancreatic beta cell function pathways (all q<0.05). NA-H showed significantly improved OS in FOLFOX/FOLFIRI (F/F) (NA-L: 34.3 vs NA-H: 39.6 months [mo], P < 0.001; HR = 0.83, 95% CI 0.77 – 0.90) and those treated with IO (15.3 vs 22.4 mo, P = 0.018; HR = 0.77, 95% CI 0.62– 0.96) treated CRC patients. The benefit remained in TP53-mutated CRC treated with F/F (34.6 vs 41.0 mo, P < 0.001; HR = 0.82, 95% CI 0.74-0.90) and IO (10.0 vs 21.8 mo, P = 0.027; HR = 0.68, 95% CI 0.48-0.96) but not in the TP53-wild-type subgroup. Improved OS was observed with increased expression of CYLD and ZBP1(Table). The benefit remained in TP53-mutated CRC but not in the TP53 wildtype subgroup. Conclusions: This is the largest molecular and clinical characterization of necroptotic genes in CRC. Our data shows that activation of the necroptosis pathway is associated with immune cell infiltration/pathway activation, PD-L1 expression, and improved OS on F/F and IO. This benefit is more pronounced in TP53-mutated CRC, suggesting possible novel therapeutic targets.[Table: see text]