Predict Therapy Response Research Articles

Identifying the programmed cell death index of hepatocellular carcinoma for prognosis and therapy response improvement by machine learning: a bioinformatics analysis and experimental validation.

Despite advancements in hepatocellular carcinoma (HCC) treatments, the prognosis for patients remains suboptimal. Cumulative evidence suggests that programmed cell death (PCD) exerts crucial functions in HCC. PCD-related genes are potential predictors for prognosis and therapeutic responses. A systematic analysis of 14 PCD modes was conducted to determine the correlation between PCD and HCC. A novel machine learning-based integrative framework was utilized to construct the PCD Index (PCDI) for prognosis and therapeutic response prediction. A comprehensive analysis of PCDI genes was performed, leveraging data including single-cell sequencing and proteomics. GBA was selected, and its functions were investigated in HCC cell lines by in vitro experiments. Two PCD clusters with different clinical and biological characteristics were identified in HCC. With the computational framework, the PCDI was constructed, demonstrating superior prognostic predictive efficacy and surpassing previously published prognostic models. An efficient clinical nomogram based on PCDI and clinicopathological factors was then developed. PCDI was intimately associated with immunological attributes, and PCDI could efficaciously predict immunotherapy response. Additionally, the PCDI could predict the chemotherapy sensitivity of HCC patients. A multilevel panorama of PCDI genes confirmed its stability and credibility. Finally, the knockdown of GBA could suppress both the proliferative and invasive capacities of HCC cells. This study systematically elucidated the association between PCD and HCC. A robust PCDI was constructed for prognosis and therapy response prediction, which would facilitate clinical management and personalized therapy for HCC.

Radiotranscriptomics of non-small cell lung carcinoma for assessing high-level clinical outcomes using a machine learning-derived multi-modal signature

BackgroundMulti-omics research has the potential to holistically capture intra-tumor variability, thereby improving therapeutic decisions by incorporating the key principles of precision medicine. The purpose of this study is to identify a robust method of integrating features from different sources, such as imaging, transcriptomics, and clinical data, to predict the survival and therapy response of non-small cell lung cancer patients.Methods2996 radiomics, 5268 transcriptomics, and 8 clinical features were extracted from the NSCLC Radiogenomics dataset. Radiomics and deep features were calculated based on the volume of interest in pre-treatment, routine CT examinations, and then combined with RNA-seq and clinical data. Several machine learning classifiers were used to perform survival analysis and assess the patient’s response to adjuvant chemotherapy. The proposed analysis was evaluated on an unseen testing set in a k-fold cross-validation scheme. Score- and concatenation-based multi-omics were used as feature integration techniques.ResultsSix radiomics (elongation, cluster shade, entropy, variance, gray-level non-uniformity, and maximal correlation coefficient), six deep features (NasNet-based activations), and three transcriptomics (OTUD3, SUCGL2, and RQCD1) were found to be significant for therapy response. The examined score-based multi-omic improved the AUC up to 0.10 on the unseen testing set (0.74 ± 0.06) and the balance between sensitivity and specificity for predicting therapy response for 106 patients, resulting in less biased models and improving upon the either highly sensitive or highly specific single-source models. Six radiomics (kurtosis, GLRLM- and GLSZM-based non-uniformity from images with no filtering, biorthogonal, and daubechies wavelets), seven deep features (ResNet-based activations), and seven transcriptomics (ELP3, ZZZ3, PGRMC2, TRAK1, ATIC, USP7, and PNPLA2) were found to be significant for the survival analysis. Accordingly, the survival analysis for 115 patients was also enhanced up to 0.20 by the proposed score-based multi-omics in terms of the C-index (0.79 ± 0.03).ConclusionsCompared to single-source models, multi-omics integration has the potential to improve prediction performance, increase model stability, and reduce bias for both treatment response and survival analysis.

Higher BST2 Expression Promotes the Anti-HBV Effect of IFN-α and BST2 Genetic Variant Predicts PegIFNα Treatment Response of HBeAg-Positive Chronic Hepatitis B Patients.

We previously reported that an interferon (IFN)-inducible protein, BST2, was regulated by the JAK-STAT pathway activated by CD40, and subsequently suppressing hepatitis B virus (HBV) repliaction and transcription. The current research attempted to assess the impact of BST2 on the IFN-treated anti-HBV effect, and explore BST2 variants for predicting pegylated IFN alpha (PegIFNα) therapy response of patients with hepatitis B e antigen (HBeAg)-positive chronic hepatitis B (CHB). Using an HBV-transfected cell model, the function of BST2 on HBV DNA replication and transcription driven by IFN was studied. The potentially functional BST2 variants were selected through a strategy of gene-wide screening. The associations of BST2 variants and polygenic score (PGS) model, which was used to quantify the combined influence of several genetic variants, with treatment response were examined in 2 separate PegIFNα-treated cohorts of 238 and 707 patients with CHB, respectively. We found that overexpression of BST2 improved the anti-HBV activity triggered by IFN-α. Among PegIFNα-treated patients with CHB, BST2_rs9576 was screened out to be significantly correlated with combined response (CR; i.e., HBeAg seroconversion along with HBV DNA level <3.3log10 IU/mL, P = 7.12 × 10-5 ). Additionally, there was a strong correlation between the PGS incorporating BST2_rs9576 and other 5 genetic variations (previously described predictors of therapy response to PegIFNα) and CR (P = 1.81 × 10-13 ), hepatitis B surface antigen (HBsAg) level (P = 0.004), as well as HBsAg decline (P = 0.017). In conclusion, higher BST2 expression responded better to IFN-α treatment. BST2_rs9576 is an effective indicator to forecast therapy response of PegIFNα-treated patients with CHB. The PGS possesses the potential to boost the ability of PegIFNα therapy response.

Early kinetics of C reactive protein for cancer-agnostic prediction of therapy response and mortality in patients treated with immune checkpoint inhibitors: a multicenter cohort study

BackgroundC reactive protein (CRP) kinetics have recently been suggested as predictive biomarkers for the efficacy of immune checkpoint inhibitor (ICI) therapy in selected cancer types. The aim of this study...

Abstract PR015: Evaluation of genetic ancestry as a covariable in gene expression analysis for chemotherapy response in Colombian breast cancer women

Abstract Breast cancer (BC) is a public health concern in Colombia that presents admixture of indigenous American (IA), European and African American (AA) ancestries. These differences could be affecting prognosis and neoadjuvant chemotherapy (NAC) response. This study aims to identify gene expression profiles associated with non-response to NAC in Colombian women diagnosed with invasive BC, including genetic ancestry as a covariable. An integrative transcriptome analysis was conducted on 87 samples of BC female patients belonging to Colombian NCI population and candidates for NAC. Patients either responded or did not respond to NAC. To estimate genetic ancestry, genotyping of 106 ancestry informative markers was performed from non-tumoral DNA. Molecular intrinsic subtypes were classified into four groups: All luminal, LuminalBHer2+/-, LuminalBHer2+ &amp; HER2 enriched, and TNBC+Her2-enriched. Two analyses were performed: differential gene expression comparing RNA-Seq data from different molecular subtype grouping, and enrichment analysis comparing baseline vs follow-up samples from non-responders, as well as baseline samples from non-responders vs responders. All analyses were performed comparing ancestry as a covariable in differential expression model. Responders and nonresponders showed distinct gene signatures in all categories. Sample distribution shown 8% AA, 59% IA and 32% European ancestry. Subtype grouping with ancestry as a covariable found 4 differential expressed genes (DEG) across nonresponders and baseline analyses for all samples (CPNE9, GTSF1, NORAD, PPP1R35) associated with cell differentiation pathways. 1 luminal DEG (CGB3), 20 DEG for LuminalBHer2+/-, 202 DEGs for LuminalBHer2+ &amp; HER2 enriched, and 7 DEGs for TNBC+Her2-enriched (AKR1E2, CEACAM6, IRS4,LIGO, MICU2, NPAS2 and STATH). Among these groups, enrichment pathways included positive regulation of cell migration, cell-cell adhesion, cellular response to hormone stimulus. We found a notable increase in differential gene expression as the proportion of IA genetic ancestry becomes greater. Genetic ancestry effect might indicate a connection between heterozygosity and genetic variance. We contrasted our results to other studies performed in other Latin American cohorts. Overall, our findings reveal a significant gene signature between responders and non-responders across subtype grouping. Our previous findings identified different sets of genes among this subtype grouping suggesting that genetic ancestry affects the gene signatures and as it's already reported, treatment response. Our results emphasizes the significance of assessing gene expression in heterogeneous populations with diverse ancestral backgrounds like Colombia, as a means to facilitate novel findings and alleviate disparities. Overexpression of these genes provides a potential to predict therapy response, alerting clinicians about patients who may not benefit from NAC. Citation Format: Hedda Michelle Guevara-Nieto, Rafael Santiago Parra-Medina, Juan Carlos Mejia Henao, Jone Garai, Jovanny Zabaleta, Liliana Lopez-Kleine, Alba Lucia Combita. Evaluation of genetic ancestry as a covariable in gene expression analysis for chemotherapy response in Colombian breast cancer women [abstract]. In: Proceedings of the 16th AACR Conference on the Science of Cancer Health Disparities in Racial/Ethnic Minorities and the Medically Underserved; 2023 Sep 29-Oct 2;Orlando, FL. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2023;32(12 Suppl):Abstract nr PR015.

Genomic Determinants of Therapy Response in ETV6::RUNX1 Leukemia

ETV6::RUNX1 leukemia is the second most common childhood B-cell acute lymphoblastic leukemia subtype. Although it has a low overall relapse risk, a significant proportion of relapses occur within this subtype due to its relatively high incidence. No consistent genomic biomarkers have been identified that predict therapy response beyond early therapy response as measured by minimal residual disease at the end of induction. In this study, our aim was to use multiomic data from 175 ETV6::RUNX1 leukemias treated according to the NOPHO ALL2000 and ALL2008 protocols to identify genomic features that predict therapy response at presentation. Cases were classified into three responder groups based on MRD at end-of-induction (EOI) therapy: fast responders with negative MRD, intermediate responders with positive but &amp;lt;0.1% MRD, and slow responders with &amp;gt;0.1% MRD. Response data was also collected at mid-induction (day 15) and end-of-consolidation (EOC), when available. Diagnostic and remission DNA were analyzed using array-based copy number data (n=142) or whole genome sequencing (WGS) (n=33), and diagnostic bulk RNA sequencing (n=51) was complemented with single cell multiomics (RNA, surface protein and VDJ sequencing) data from both healthy (n=8) and leukemic bone marrow cells upon chemotherapy (n=8). Overall, the frequency of somatic genetic changes was not significantly associated with treatment response suggesting that genetic diversification through elevated random mutagenesis events may not underlie the drug resistant phenotype. However, analysis of mutational signatures revealed that fast-responding leukemic blasts frequently exhibited the APOBEC mutational signature and had higher cell cycle activity (Figure 1A). Accordingly, we show that expression of A3B is highly specific to cycling cells also in healthy bone marrow hematopoietic progenitor cells, pro- and pre-B cells. We also examined RAG-mediated genetic instability that manifests in focal structural variations and has highest activity in ETV6::RUNX1leukemia (Figure 1B). We confirmed several recurrent RAG off-target sites, including significant enrichment of the known risk locus TBX1LR1 in slow responding cases. Strikingly, slow-responding leukemias had also frequent rearrangements in the immunoglobulin kappa genes and the kappa-deleting elements, despite their pro-B-like transcriptome state upon diagnosis. Based on the single cell multiomics profiles, we found that upon treatment the blast phenotype shifts towards an immature-B-like transcriptome, with concomitant expression of Ig light chains with productive sequence, thereby permitting activation of BCR signaling that provides a survival advantage. At CNV level, we found that the chromosomal alterations related to the translocation event associated with treatment response. Specifically, slow responders had more often the amplification of der(21)t(12;21) chromosome, impacting expression of USP5, TIGAR and SLC2A3, associated in earlier studies to chemoresistance and tumorigenesis. In comparison, fast responders were often missing a region of chromosome 12p with lower expression of several genes including oncogenic KRAS. In summary, combining multiomic and response data, we characterized genetic alterations and mutational processes that affect the treatment response in ETV6::RUNX1leukemia. Our findings highlight the genetic and transcriptomic heterogeneity of this common pediatric leukemia subtype and its implications for treatment sensitivity, paving the way for better classification of the disease at presentation.

Epigenetic Age Acceleration in Hematological Malignancies: Beyond Chronological Age, Clinical Implications, and Therapeutic Perspectives

Background: Age is important in prognosticating hematological malignancies. Biological based on DNA methylome changes offers new insights into the complex interplay between epigenetics, genetics, environment, and cancer. DNA methylation levels at specific CpG sites evolve in a predictable fashion with aging and can be used to develop biomarkers that estimate the epigenetic age (EA) of a tissue. Such biomarkers are called epigenetic clocks (Fig 1-B). Based on the EA, we can determine Epigenetic Age Acceleration (EAA), a tool to identify patients aging faster biologically than their chronological age. We sought to build a blood-based clock to predict epigenetic age in healthy young and old adults and subsequently assess the degree of EAA in AML patients. Finally, among multiple planned downstream applications, we propose to use EAA as a biomarker for monitoring treatment responses in targeted therapies (Fig 1-A). Methods: We used ClockBase and NCBI's Gene Expression Omnibus (GEO) repositories to collect methylation data based on Illumina's 450k array for healthy adults and patients with AML/MDS. Pre-processing was done using the minifi BioConductor package. We used a cohort of healthy young adults (YA, aged 18-30 ys.) to train our epigenetic clock and calibrate it with three well-studied epigenetic clocks: Horvath clock (HC), Hannum clock (HAC), and PhenoAge (PA). We hypothesized that EA and chronological ages (CA) should approximate each other in younger patients as they do not generally carry comorbidities or have a higher CHIP prevalence which could influence EA. We compared our clock's performance to known epigenetic clocks, and after fine-tuning, all 4 models were validated in a cohort of healthy older adults (OA, aged 70-85 ys.) with a higher risk of CHIP. Then, we tested our clock in an AML patient cohort and illustrated how EAA can predict therapy responses in AML patients treated with IDH1/2 inhibitors (Published methylation data from Wang et al 2021, Nat Comm.) Results: We designed the clock using an Elastic-Net regression scheme. We screened 130,000 patient samples from GEO and 1,813 eligible samples were divided into 3 cohorts: 800 healthy younger adults (aged 18-30 ys.), 800 healthy elderly adults (aged 70-85 ys.), and 213 patients with AML. To train and test our clock, we used the YA cohort (n=800) with a median age of 20 ys. (SD 3.168 ys) and our epigenetic clock predicted a median EA (MEA) of 19.1 ys. (SD 3.4 ys). HC, HAC and PA predicted MEA of 21.6 ys. (SD 3.9 ys.), 19.4 ys. (SD 4.1 ys.) and 21.5 ys. (SD 3.9 ys.). After fine-tuning of weights, we tested our clock in OA cohort (n=800) with a median age of 78 ys. (SD 4.8 ys.). Our clock predicted a MEA of 82.3 ys. (SD 4.9 ys.). HC, HAC and PA predicted a MEA of 83.7 ys. (SD 5.7 ys.), 85.2 ys. (SD 6.0 ys.) and 84.9 ys. (SD of 5.1 ys.) resp. The AML cohort (n=213) had a median age of 80.9 ys. (SD 2.5 ys.). We expected a significantly higher EA due to the mutation burden along with genomic and epigenomic changes induced by the disease, our clock predicted a MEA of 84.7 ys. (SD 3.0 ys.). EAA was calculated from the difference between EA and CA. YA cohort had a median EAA of 0.9 ys. (SD 1.8 ys.) and our clock predicted a median EAA of 1.1 ys. (SD 2.0 ys.). HC, HAC, and PA predicted a median EAA of 1.4 ys. (SD 2.3 ys.), 1.4 ys. (SD 2.9 ys.), and 1.5 ys. (SD 2.4 ys.). In the OA cohort, we expected higher EAA due to age-associated cardiometabolic disease and CHIP; our clock predicted a median EAA of 4.2 (SD 2.1). HC, HAC, and PA predicted a median EAA of 6.0 ys. (SD 2.7), 5.6 ys. (SD 3.0 ys.), and 5.1 ys. (SD 2.2 ys.). For the AML cohort, our clock predicted a median EAA of 8.0 ys. (SD 5.6). We then applied our clock to a cohort of 60 AML patients treated with IDH inhibitors to predict whether a patient cleared the IDH clone after therapy. We discovered that patients who failed to clear the IDH-clone (n=14) had a higher EAA (3.23 vs 7.91 ys., p=0.012) and shorter RFS compared to patients who cleared the IDH-clone (n=7) [RFS 20.6 vs 6.1 mo., p=0.0086]. Conclusion: We built an epigenetic clock that outperforms three standard clocks in predicting epigenetic age and applied it to AML patients demonstrating a significant increase in epigenetic age. We also illustrated how EAA can be used as a biomarker for targeted therapies. Ongoing work at Karmanos Cancer Center is focused on improving the underlying regression model for more sensitive epigenetic age predictions and implementing novel pace-of-aging algorithms for EAA calculations.

Characteristics of ABCC4 and ABCG2 High Expression Subpopulations in CRC—A New Opportunity to Predict Therapy Response

Background: Our previous findings proved that ABCC4 and ABCG2 proteins present much more complex roles in colorectal cancer (CRC) than typically cancer-associated functions as drug exporters. Our objective was to evaluate their predictive/diagnostic potential. Methods: CRC patients’ transcriptomic data from the Gene Expression Omnibus database (GSE18105, GSE21510 and GSE41568) were discriminated into two subpopulations presenting either high expression levels of ABCC4 (ABCC4 High) or ABCG2 (ABCG2 High). Subpopulations were analysed using various bioinformatical tools and platforms (KEEG, Gene Ontology, FunRich v3.1.3, TIMER2.0 and STRING 12.0). Results: The analysed subpopulations present different gene expression patterns. The protein–protein interaction network of subpopulation-specific genes revealed the top hub proteins in ABCC4 High: RPS27A, SRSF1, DDX3X, BPTF, RBBP7, POLR1B, HNRNPA2B1, PSMD14, NOP58 and EIF2S3 and in ABCG2 High: MAPK3, HIST2H2BE, LMNA, HIST1H2BD, HIST1H2BK, HIST1H2AC, FYN, TLR4, FLNA and HIST1H2AJ. Additionally, our multi-omics analysis proved that the ABCC4 expression correlates with substantially increased tumour-associated macrophage infiltration and sensitivity to FOLFOX treatment. Conclusions: ABCC4 and ABCG2 may be used to distinguish CRC subpopulations that present different molecular and physiological functions. The ABCC4 High subpopulation demonstrates significant EMT reprogramming, RNA metabolism and high response to DNA damage stimuli. The ABCG2 High subpopulation may resist the anti-EGFR therapy, presenting higher proteolytical activity.

Metabolomic profiles of intact tissues reflect clinically relevant prostate cancer subtypes

BackgroundProstate cancer (PC) is a heterogenous multifocal disease ranging from indolent to lethal states. For improved treatment-stratification, reliable approaches are needed to faithfully differentiate between high- and low-risk tumors and to predict therapy response at diagnosis.MethodsA metabolomic approach based on high resolution magic angle spinning nuclear magnetic resonance (HR MAS NMR) analysis was applied on intact biopsies samples (n = 111) obtained from patients (n = 31) treated by prostatectomy, and combined with advanced multi- and univariate statistical analysis methods to identify metabolomic profiles reflecting tumor differentiation (Gleason scores and the International Society of Urological Pathology (ISUP) grade) and subtypes based on tumor immunoreactivity for Ki67 (cell proliferation) and prostate specific antigen (PSA, marker for androgen receptor activity).ResultsValidated metabolic profiles were obtained that clearly distinguished cancer tissues from benign prostate tissues. Subsequently, metabolic signatures were identified that further divided cancer tissues into two clinically relevant groups, namely ISUP Grade 2 (n = 29) and ISUP Grade 3 (n = 17) tumors. Furthermore, metabolic profiles associated with different tumor subtypes were identified. Tumors with low Ki67 and high PSA (subtype A, n = 21) displayed metabolite patterns significantly different from tumors with high Ki67 and low PSA (subtype B, n = 28). In total, seven metabolites; choline, peak for combined phosphocholine/glycerophosphocholine metabolites (PC + GPC), glycine, creatine, combined signal of glutamate/glutamine (Glx), taurine and lactate, showed significant alterations between PC subtypes A and B.ConclusionsThe metabolic profiles of intact biopsies obtained by our non-invasive HR MAS NMR approach together with advanced chemometric tools reliably identified PC and specifically differentiated highly aggressive tumors from less aggressive ones. Thus, this approach has proven the potential of exploiting cancer-specific metabolites in clinical settings for obtaining personalized treatment strategies in PC.

Is System xc- a Suitable Target for Tumour Detection and Response Assessment with Imaging?

System xc- is upregulated in cancer cells and can be imaged using novel radiotracers, most commonly with (4S)-4-(3-[18F]fluoropropyl)-L-glutamic acid (18F-FSPG). The aim of this review was to summarise the use of 18F-FSPG in humans, explore the benefits and limitations of 18F-FSPG, and assess the potential for further use of 18F-FSPG in cancer patients. To date, ten papers have described the use of 18F-FSPG in human cancers. These studies involved small numbers of patients (range 1-26) and assessed the use of 18F-FSPG as a general oncological diagnostic agent across different cancer types. These clinical trials were contrasting in their findings, limiting the scope of 18F-FSPG PET/CT as a purely diagnostic agent, primarily due to heterogeneity of 18F-FSPG retention both between cancer types and patients. Despite these limitations, a potential further application for 18F-FSPG is in the assessment of early treatment response and prediction of treatment resistance. Animal models of cancer have shown that changes in 18F-FSPG retention following effective therapy precede glycolytic changes, as indicated by 18F-FDG, and changes in tumour volume, as measured by CT. If these results could be replicated in human clinical trials, imaging with 18F-FSPG PET/CT would offer an exciting route towards addressing the currently unmet clinical needs of treatment resistance prediction and early imaging assessment of therapy response.

Multistain deep learning as a prognostic and predictive biomarker in colorectal cancer

The tumor immune microenvironment (TIME) plays acrucial prognostic and predictive role in solid malignancies such as colorectal cancer (CRC). Nevertheless, scoring systems based on TIME such as the Immunoscore (IS) are rarely used in clinical practice. Among other reasons, this might be due to the additional time required for manual quantification of tumor-associated immune cells or costs associated with proprietary/commercial solutions. To address these issues, we developed amultistain deep learning model (MSDLM) and trained, validated, and tested it on immunohistochemical image data of different immune cell subtypes from over 1000 patients with CRC. Our model showed high prognostic accuracy and outperformed other clinical, molecular, and immune cell-based parameters. It might also be used for therapy response prediction in rectal cancer patients undergoing neoadjuvant therapy. Leveraging artificial intelligence interpretability/explainability methods, we ascertained that the MSDLM's predictions align with recognized antitumor immune response patterns. Consequently, the AImmunoscore (AIS) could emerge as apotential TIME-based decision-making tool for clinicians.

Genetic diversity, haplotype analysis, and prevalence of Hepatitis B virus MHR mutations among isolates from Kenyan blood donors.

The rapid spread of HBV has resulted in the emergence of new variants. These viral genotypes and variants, in addition to carcinogenic risk, can be key predictors of therapy response and outcomes. As a result, a better knowledge of these emerging HBV traits will aid in the development of a treatment for HBV infection. However, many Sub-Saharan African nations, including Kenya, have insufficient molecular data on HBV strains circulating locally. This study conducted a population-genetics analysis to evaluate the genetic diversity of HBV among Kenyan blood donors. In addition, within the same cohort, the incidence and features of immune-associated escape mutations and stop-codons in Hepatitis B surface antigen (HBsAg) were determined. In September 2015 to October 2016, 194 serum samples were obtained from HBsAg-positive blood donors residing in eleven different Kenyan counties: Kisumu, Machakos, Uasin Gishu, Nairobi, Nakuru, Embu, Garissa, Kisii, Mombasa, Nyeri, and Turkana. For the HBV surface (S) gene, HBV DNA was isolated, amplified, and sequenced. The sequences obtained were utilized to investigate the genetic and haplotype diversity within the S genes. Among the blood donors, 74.74% were male, and the overall mean age was 25.36 years. HBV genotype A1 (88.14%) was the most common, followed by genotype D (10.82%), genotype C (0.52%), and HBV genotype E (0.52%). The phylogenetic analysis revealed twelve major clades, with cluster III comprising solely of 68 blood donor isolates (68/194-35.05%). A high haplotype diversity (Hd = 0.94) and low nucleotide diversity (π = 0.02) were observed. Kisumu county had high number of haplotypes (22), but low haplotype (gene) diversity (Hd = 0.90). Generally, a total of 90 haplotypes with some consisting of more than one sequence were observed. The gene exhibited negative values for Tajima's D (-2.04, p<0.05) and Fu's Fs (-88.84). Several mutations were found in 139 isolates, either within or outside the Major Hydrophilic Area (MHR). There were 29 mutations found, with 37.9% of them situated inside the "a" determinant. The most common mutations in this research were T143M and K122R. Escape mutations linked to diagnostic failure, vaccination and immunoglobulin treatment evasion were also discovered. Also, one stop-codon, W163STP, inside the MHR, was found in one sample from genotype A. In Kenya, HBV/A1 is still the most common genotype. Despite limited genetic and nucleotide diversity, haplotype network analysis revealed haplotype variance among HBV genotypes from Kenyan blood donors. The virological properties of immune escape, which may be the source of viral replication endurance, were discovered in the viral strains studied and included immune-escape mutations and stop-codon. The discovery of HBsAg mutations in MHR in all isolates highlighted the need of monitoring MHR mutations in Kenya.

Computational single cell oncology: state of the art.

Single cell computational analysis has emerged as a powerful tool in the field of oncology, enabling researchers to decipher the complex cellular heterogeneity that characterizes cancer. By leveraging computational algorithms and bioinformatics approaches, this methodology provides insights into the underlying genetic, epigenetic and transcriptomic variations among individual cancer cells. In this paper, we present a comprehensive overview of single cell computational analysis in oncology, discussing the key computational techniques employed for data processing, analysis, and interpretation. We explore the challenges associated with single cell data, including data quality control, normalization, dimensionality reduction, clustering, and trajectory inference. Furthermore, we highlight the applications of single cell computational analysis, including the identification of novel cell states, the characterization of tumor subtypes, the discovery of biomarkers, and the prediction of therapy response. Finally, we address the future directions and potential advancements in the field, including the development of machine learning and deep learning approaches for single cell analysis. Overall, this paper aims to provide a roadmap for researchers interested in leveraging computational methods to unlock the full potential of single cell analysis in understanding cancer biology with the goal of advancing precision oncology. For this purpose, we also include a notebook that instructs on how to apply the recommended tools in the Preprocessing and Quality Control section.

Predictors of treatment response in therapy-resistant enuresis

The voiding chart is part of the initial evaluation of enuresis, since the data gathered this way are assumed to carry predictive information. However, there is little evidence that the voiding chart actually does predict therapy response. Lundmark & Nevéus performed a pilot investigation in 2020 and found that anamnestic and voiding chart data did not predict response to second-line therapies. This study aims at evaluating whether these findings could be replicated. This is an evaluation of clinical practice. All patients in a tertiary outpatient clinic with enuresis resistant to first-line therapy (i.e. the enuresis alarm and desmopressin medication) during the evaluation period were included in the study. Baseline anamnestic data focused on bladder and bowel habits, were gathered and the families were instructed to complete a voiding chart including measurements of nocturnal urine production. The children were then treated in accordance with international guidelines, which are anticholinergics and antidepressants as second- and third-line treatment, respectively. Desmopressin was added if needed. In total, 70 patients were included. At the end of the study 37 of these patients were dry, 11 patients were still wetting their beds and 22 patients were lost to follow-up. Of the dry patients 21 became dry on anticholinergics (and/or mirabegron, with or without desmopressin), five on tricyclic antidepressants (with or without desmopressin), seven after a new attempt with the alarm and five became dry spontaneously. The only statistically significant differences between responders and non-responders to the various treatments were that children responsive to anticholinergics had harder and more infrequent stools (p=0.04 and p=0.03, respectively). This study found that anamnestic and voiding chart data do not predict response to treatment in children with therapy-resistant enuresis. Because of this and the fact that we lose some children who need our help by demanding that they complete a voiding chart before initiating treatment, we question the use of this instrument in the evaluation of therapy-resistant enuresis.

ScCURE identifies cell types responding to immunotherapy and enables outcome prediction

A deep understanding of immunotherapy response/resistance mechanisms and a highly reliable therapy response prediction are vital for cancer treatment. Here, we developed scCURE (single-cell RNA sequencing [scRNA-seq] data-based Changed and Unchanged cell Recognition during immunotherapy). Based on Gaussian mixture modeling, Kullback-Leibler (KL) divergence, and mutual nearest-neighbors criteria, scCURE can faithfully discriminate between cells affected or unaffected by immunotherapy intervention. By conducting scCURE analyses in melanoma and breast cancer immunotherapy scRNA-seq data, we found that the baseline profiles of specific CD8+ T and macrophage cells (identified by scCURE) can determine the way in which tumor microenvironment immune cells respond to immunotherapy, e.g., antitumor immunity activation or de-activation; therefore, these cells could be predictive factors for treatment response. In this work, we demonstrated that the immunotherapy-associated cell-cell heterogeneities revealed by scCURE can be utilized to integrate the therapy response mechanism study and prediction model construction.

Longitudinal dynamic contrast-enhanced MRI radiomic models for early prediction of response to neoadjuvant systemic therapy in triple-negative breast cancer

Early prediction of neoadjuvant systemic therapy (NAST) response for triple-negative breast cancer (TNBC) patients could help oncologists select individualized treatment and avoid toxic effects associated with ineffective therapy in patients unlikely to achieve pathologic complete response (pCR). The objective of this study is to evaluate the performance of radiomic features of the peritumoral and tumoral regions from dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) acquired at different time points of NAST for early treatment response prediction in TNBC. This study included 163 Stage I-III patients with TNBC undergoing NAST as part of a prospective clinical trial (NCT02276443). Peritumoral and tumoral regions of interest were segmented on DCE images at baseline (BL) and after two (C2) and four (C4) cycles of NAST. Ten first-order (FO) radiomic features and 300 gray-level-co-occurrence matrix (GLCM) features were calculated. Area under the receiver operating characteristic curve (AUC) and Wilcoxon rank sum test were used to determine the most predictive features. Multivariate logistic regression models were used for performance assessment. Pearson correlation was used to assess intrareader and interreader variability. Seventy-eight patients (48%) had pCR (52 training, 26 testing), and 85 (52%) had non-pCR (57 training, 28 testing). Forty-six radiomic features had AUC at least 0.70, and 13 multivariate models had AUC at least 0.75 for training and testing sets. The Pearson correlation showed significant correlation between readers. In conclusion, Radiomic features from DCE-MRI are useful for differentiating pCR and non-pCR. Similarly, predictive radiomic models based on these features can improve early noninvasive treatment response prediction in TNBC patients undergoing NAST.

The importance of planning CT-based imaging features for machine learning-based prediction of pain response

Patients suffering from painful spinal bone metastases (PSBMs) often undergo palliative radiation therapy (RT), with an efficacy of approximately two thirds of patients. In this exploratory investigation, we assessed the effectiveness of machine learning (ML) models trained on radiomics, semantic and clinical features to estimate complete pain response. Gross tumour volumes (GTV) and clinical target volumes (CTV) of 261 PSBMs were segmented on planning computed tomography (CT) scans. Radiomics, semantic and clinical features were collected for all patients. Random forest (RFC) and support vector machine (SVM) classifiers were compared using repeated nested cross-validation. The best radiomics classifier was trained on CTV with an area under the receiver-operator curve (AUROC) of 0.62 ± 0.01 (RFC; 95% confidence interval). The semantic model achieved a comparable AUROC of 0.63 ± 0.01 (RFC), significantly below the clinical model (SVM, AUROC: 0.80 ± 0.01); and slightly lower than the spinal instability neoplastic score (SINS; LR, AUROC: 0.65 ± 0.01). A combined model did not improve performance (AUROC: 0,74 ± 0,01). We could demonstrate that radiomics and semantic analyses of planning CTs allowed for limited prediction of therapy response to palliative RT. ML predictions based on established clinical parameters achieved the best results.

LncRNA EILA promotes CDK4/6 inhibitor resistance in breast cancer by stabilizing cyclin E1 protein.

CDK4/6 inhibitors (CDK4/6i) plus endocrine therapy are now standard first-line therapy for advanced HR+/HER2- breast cancer, but developing resistance is just a matter of time in these patients. Here, we report that a cyclin E1-interacting lncRNA (EILA) is up-regulated in CDK4/6i-resistant breast cancer cells and contributes to CDK4/6i resistance by stabilizing cyclin E1 protein. EILA overexpression correlates with accelerated cell cycle progression and poor prognosis in breast cancer. Silencing EILA reduces cyclin E1 protein and restores CDK4/6i sensitivity both in vitro and in vivo. Mechanistically, hairpin A of EILA binds to the carboxyl terminus of cyclin E1 protein and hinders its binding to FBXW7, thereby blocking its ubiquitination and degradation. EILA is transcriptionally regulated by CTCF/CDK8/TFII-I complexes and can be inhibited by CDK8 inhibitors. This study unveils the role of EILA in regulating cyclin E1 stability and CDK4/6i resistance, which may serve as a biomarker to predict therapy response and a potential therapeutic target to overcome resistance.

Oncogenic dependency plays a dominant role in the immune response to cancer

Pancreatic ductal adenocarcinoma (PDAC) is one of the deadliest human malignancies. Advanced PDAC is considered incurable. Nearly 90% of pancreatic cancers are caused by oncogenic KRAS mutations. The mechanisms of primary or acquired resistance to KRAS inhibition are currently unknown. Here, we propose that oncogenic dependency, rather than KRAS mutation per se, plays a dominant role in the immune response to cancer, including late-stage PDAC. Classifying tumor samples according to KRAS activity scores allows accurate prediction of tumor immune composition and therapy response. Dual RAS/MAPK pathway blockade combining KRAS and MEK inhibitors is more effective than the selective KRAS inhibitor alone in attenuating MAPK activation and unblocking the influx of T cells into the tumor. Lowering KRAS activity in established tumors promotes immune infiltration, but with a limited antitumor effect, whereas combining KRAS/MEK inhibition with immune checkpoint blockade achieves durable regression in preclinical models. The results are directly applicable to stratifying human PDAC based on KRAS dependency values and immune cell composition to improve therapeutic design.

194P Early kinetics of C-reactive protein for cancer-agnostic prediction of therapy response and mortality in patients treated with immune checkpoint inhibitors: A multi-center cohort study

194P Early kinetics of C-reactive protein for cancer-agnostic prediction of therapy response and mortality in patients treated with immune checkpoint inhibitors: A multi-center cohort study