Safety, tolerability, and antiviral effect of RG-101 in patients with chronic hepatitis C: a phase 1B, double-blind, randomised controlled trial.

BACKGROUND: miR-122 is an important host factor for hepatitis C virus (HCV) replication. The aim of this study was to assess the safety and tolerability, pharmacokinetics, and antiviral effect of a single dose of RG-101, a hepatocyte targeted N-acetylgalactosamine conjugated oligonucleotide that antagonises miR-122, in patients with chronic HCV infection with various genotypes. METHODS: In this randomised, double-blind, placebo-controlled, multicentre, phase 1B study, patients were randomly assigned to RG-101 or placebo (7:1). We enrolled men and postmenopausal or hysterectomised women (aged 18-65 years) with chronic HCV genotype 1, 3, or 4 infection diagnosed at least 24 weeks before screening who were either treatment naive to or relapsed after interferon-α based therapy. Patients with co-infection (hepatitis B virus or HIV infection), evidence of decompensated liver disease, or a history of hepatocellular carcinoma were excluded. Randomisation was done by an independent, unblinded, statistician using the SAS procedure Proc Plan. The first cohort received one subcutaneous injection of 2 mg/kg RG-101 or placebo; the second cohort received one subcutaneous injection of 4 mg/kg or placebo. Patients were followed up for 8 weeks (all patients) and up to 76 weeks (patients with no viral rebound and excluding those who were randomised to the placebo group) after randomisation. The primary objective was safety and tolerability of RG-101. This trial was registered with EudraCT, number 2013-002978-49. FINDINGS: Between June 4, 2014, and Oct 27, 2014, we enrolled 32 patients with chronic HCV genotype 1 (n=16), 3 (n=10), or 4 (n=6) infections. In the first cohort, 14 patients were randomly assigned to receive 2 mg/kg RG-101 and two patients were randomly assigned to receive placebo, and in the second cohort, 14 patients were randomly assigned to receive 4 mg/kg RG-101 and two patients were randomly assigned to receive placebo. Overall, 26 of the 28 patients dosed with RG-101 reported at least one treatment-related adverse event. At week 4, the median viral load reduction from baseline was 4·42 (IQR 3·23-5·00) and 5·07 (4·19-5·35) log10 IU/mL in patients dosed with 2 mg/kg RG-101 or 4 mg/kg RG-101. Three patients had undetectable HCV RNA levels 76 weeks after a single dose of RG-101. Viral rebound at or before week 12 was associated with the appearance of resistance associated substitutions in miR-122 binding regions in the 5' UTR of the HCV genome. INTERPRETATION: This study showed that one administration of 2 mg/kg or 4 mg/kg RG-101, a hepatocyte targeted N-acetylgalactosamine conjugated anti-miR-122 oligonucleotide, was well tolerated and resulted in substantial viral load reduction in all treated patients within 4 weeks, and sustained virological response in three patients for 76 weeks. FUNDING: Regulus Therapeutics, Inc.

Conceptually Funded Usability Evaluation of an Application for Leveraging Descriptive Data Analysis Models for Cardiovascular Research.

The focus of this study, and the subject of this article, resides in the conceptually funded usability evaluation of an application of descriptive models to a specific dataset obtained from the East Slovak Institute of Heart and Vascular Diseases targeting cardiovascular patients. Delving into the current state-of-the-art practices, we examine the extent of cardiovascular diseases, descriptive data analysis models, and their practical applications. Most importantly, our inquiry focuses on exploration of usability, encompassing its application and evaluation methodologies, including Van Welie's layered model of usability and its inherent advantages and limitations. The primary objective of our research was to conceptualize, develop, and validate the usability of an application tailored to supporting cardiologists' research through descriptive modeling. Using the R programming language, we engineered a Shiny dashboard application named DESSFOCA (Decision Support System For Cardiologists) that is structured around three core functionalities: discovering association rules, applying clustering methods, and identifying association rules within predefined clusters. To assess the usability of DESSFOCA, we employed the System Usability Scale (SUS) and conducted a comprehensive evaluation. Additionally, we proposed an extension to Van Welie's layered model of usability, incorporating several crucial aspects deemed essential. Subsequently, we rigorously evaluated the proposed extension within the DESSFOCA application with respect to the extended usability model, drawing insightful conclusions from our findings.

Exclusion of the 750-kb genetically unstable region at Xq27 as a candidate locus for prostate malignancy in HPCX1-linked families.

Several linkage studies provided evidence for the presence of the hereditary prostate cancer locus, HPCX1, at Xq27-q28. The strongest linkage peak of prostate cancer overlies a variable region of ~750 kb at Xq27 enriched by segmental duplications (SDs), suggesting that the predisposition to prostate cancer may be a genomic disorder caused by recombinational interaction between SDs. The large size of SDs and their sequence similarity make it difficult to examine this region for possible rearrangements using standard methods. To overcome this problem, direct isolation of a set of genomic segments by in vivo recombination in yeast (a TAR cloning technique) was used to perform a mutational analysis of the 750 kb region in X-linked families. We did not detect disease-specific rearrangements within this region. In addition, transcriptome and computational analyses were performed to search for nonannotated genes within the Xq27 region, which may be associated with genetic predisposition to prostate cancer. Two candidate genes were identified, one of which is a novel gene termed SPANXL that represents a highly diverged member of the SPANX gene family, and the previously described CDR1 gene that is expressed at a high level in both normal and malignant prostate cells, and mapped 210 kb of upstream the SPANX gene cluster. No disease-specific alterations were identified in these genes. Our results exclude the 750-kb genetically unstable region at Xq27 as a candidate locus for prostate malignancy. Adjacent regions appear to be the most likely candidates to identify the elusive HPCX1 locus.

Anti-tumor efficacy of a potent and selective non-covalent KRASG12D inhibitor.

Recent progress in targeting KRASG12C has provided both insight and inspiration for targeting alternative KRAS mutants. In this study, we evaluated the mechanism of action and anti-tumor efficacy of MRTX1133, a potent, selective and non-covalent KRASG12D inhibitor. MRTX1133 demonstrated a high-affinity interaction with GDP-loaded KRASG12D with KD and IC50 values of ~0.2 pM and <2 nM, respectively, and ~700-fold selectivity for binding to KRASG12D as compared to KRASWT. MRTX1133 also demonstrated potent inhibition of activated KRASG12D based on biochemical and co-crystal structural analyses. MRTX1133 inhibited ERK1/2 phosphorylation and cell viability in KRASG12D-mutant cell lines, with median IC50 values of ~5 nM, and demonstrated >1,000-fold selectivity compared to KRASWT cell lines. MRTX1133 exhibited dose-dependent inhibition of KRAS-mediated signal transduction and marked tumor regression (≥30%) in a subset of KRASG12D-mutant cell-line-derived and patient-derived xenograft models, including eight of 11 (73%) pancreatic ductal adenocarcinoma (PDAC) models. Pharmacological and CRISPR-based screens demonstrated that co-targeting KRASG12D with putative feedback or bypass pathways, including EGFR or PI3Kα, led to enhanced anti-tumor activity. Together, these data indicate the feasibility of selectively targeting KRAS mutants with non-covalent, high-affinity small molecules and illustrate the therapeutic susceptibility and broad dependence of KRASG12D mutation-positive tumors on mutant KRAS for tumor cell growth and survival.

The KRASG12C Inhibitor MRTX849 Reconditions the Tumor Immune Microenvironment and Sensitizes Tumors to Checkpoint Inhibitor Therapy.

KRASG12C inhibitors, including MRTX849, are promising treatment options for KRAS-mutant non-small cell lung cancer (NSCLC). PD-1 inhibitors are approved in NSCLC; however, strategies to enhance checkpoint inhibitor therapy (CIT) are needed. KRASG12C mutations are smoking-associated transversion mutations associated with high tumor mutation burden, PD-L1 positivity, and an immunosuppressive tumor microenvironment. To evaluate the potential of MRTX849 to augment CIT, its impact on immune signaling and response to CIT was evaluated. In human tumor xenograft models, MRTX849 increased MHC class I protein expression and decreased RNA and/or plasma protein levels of immunosuppressive factors. In a KrasG12C -mutant CT26 syngeneic mouse model, MRTX849 decreased intratumoral myeloid-derived suppressor cells and increased M1-polarized macrophages, dendritic cells, CD4+, and CD8+ T cells. Similar results were observed in lung KrasG12C -mutant syngeneic and a genetically engineered mouse (GEM) model. In the CT26 KrasG12C model, MRTX849 demonstrated marked tumor regression when tumors were established in immune-competent BALB/c mice; however, the effect was diminished when tumors were grown in T-cell-deficient nu/nu mice. Tumors progressed following anti-PD-1 or MRTX849 single-agent treatment in immune-competent mice; however, combination treatment demonstrated durable, complete responses (CRs). Tumors did not reestablish in the same mice that exhibited durable CRs when rechallenged with tumor cell inoculum, demonstrating these mice developed adaptive antitumor immunity. In a GEM model, treatment with MRTX849 plus anti-PD-1 led to increased progression-free survival compared with either single agent alone. These data demonstrate KRAS inhibition reverses an immunosuppressive tumor microenvironment and sensitizes tumors to CIT through multiple mechanisms.

Statistical method on nonrandom clustering with application to somatic mutations in cancer.

BACKGROUND: Human cancer is caused by the accumulation of tumor-specific mutations in oncogenes and tumor suppressors that confer a selective growth advantage to cells. As a consequence of genomic instability and high levels of proliferation, many passenger mutations that do not contribute to the cancer phenotype arise alongside mutations that drive oncogenesis. While several approaches have been developed to separate driver mutations from passengers, few approaches can specifically identify activating driver mutations in oncogenes, which are more amenable for pharmacological intervention. RESULTS: We propose a new statistical method for detecting activating mutations in cancer by identifying nonrandom clusters of amino acid mutations in protein sequences. A probability model is derived using order statistics assuming that the location of amino acid mutations on a protein follows a uniform distribution. Our statistical measure is the differences between pair-wise order statistics, which is equivalent to the size of an amino acid mutation cluster, and the probabilities are derived from exact and approximate distributions of the statistical measure. Using data in the Catalog of Somatic Mutations in Cancer (COSMIC) database, we have demonstrated that our method detects well-known clusters of activating mutations in KRAS, BRAF, PI3K, and beta-catenin. The method can also identify new cancer targets as well as gain-of-function mutations in tumor suppressors. CONCLUSIONS: Our proposed method is useful to discover activating driver mutations in cancer by identifying nonrandom clusters of somatic amino acid mutations in protein sequences.

The KRASG12C Inhibitor MRTX849 Provides Insight toward Therapeutic Susceptibility of KRAS-Mutant Cancers in Mouse Models and Patients.

Despite decades of research, efforts to directly target KRAS have been challenging. MRTX849 was identified as a potent, selective, and covalent KRASG12C inhibitor that exhibits favorable drug-like properties, selectively modifies mutant cysteine 12 in GDP-bound KRASG12C, and inhibits KRAS-dependent signaling. MRTX849 demonstrated pronounced tumor regression in 17 of 26 (65%) KRASG12C-positive cell line- and patient-derived xenograft models from multiple tumor types, and objective responses have been observed in patients with KRASG12C-positive lung and colon adenocarcinomas. Comprehensive pharmacodynamic and pharmacogenomic profiling in sensitive and partially resistant nonclinical models identified mechanisms implicated in limiting antitumor activity including KRAS nucleotide cycling and pathways that induce feedback reactivation and/or bypass KRAS dependence. These factors included activation of receptor tyrosine kinases (RTK), bypass of KRAS dependence, and genetic dysregulation of cell cycle. Combinations of MRTX849 with agents that target RTKs, mTOR, or cell cycle demonstrated enhanced response and marked tumor regression in several tumor models, including MRTX849-refractory models. SIGNIFICANCE: The discovery of MRTX849 provides a long-awaited opportunity to selectively target KRASG12C in patients. The in-depth characterization of MRTX849 activity, elucidation of response and resistance mechanisms, and identification of effective combinations provide new insight toward KRAS dependence and the rational development of this class of agents.See related commentary by Klempner and Hata, p. 20.This article is highlighted in the In This Issue feature, p. 1.

Retroposition of processed pseudogenes: the impact of RNA stability and translational control.

Human processed pseudogenes are copies of cellular RNAs reverse transcribed and inserted into the nuclear genome by the enzymatic machinery of L1 (LINE1) non-LTR retrotransposons. Although it is generally accepted that germline expression is crucial for the heritable retroposition of cellular mRNAs, little is known about the influences of RNA stability, mRNA quality control and compartmentalization of translation on the retroposition of processed pseudogenes. We found that frequently retroposed human mRNAs are derived from stable transcripts with translation-competent functional reading frames that are resistant to nonsense-mediated RNA decay. They are preferentially translated on free cytoplasmic ribosomes and encode soluble proteins. Our results indicate that interactions between mRNAs and L1 proteins seem to occur at free cytoplasmic ribosomes.

Discovery and preclinical evaluation of anti-miR-17 oligonucleotide RGLS4326 for the treatment of polycystic kidney disease.

Autosomal dominant polycystic kidney disease (ADPKD), caused by mutations in either PKD1 or PKD2 genes, is one of the most common human monogenetic disorders and the leading genetic cause of end-stage renal disease. Unfortunately, treatment options for ADPKD are limited. Here we report the discovery and characterization of RGLS4326, a first-in-class, short oligonucleotide inhibitor of microRNA-17 (miR-17), as a potential treatment for ADPKD. RGLS4326 is discovered by screening a chemically diverse and rationally designed library of anti-miR-17 oligonucleotides for optimal pharmaceutical properties. RGLS4326 preferentially distributes to kidney and collecting duct-derived cysts, displaces miR-17 from translationally active polysomes, and de-represses multiple miR-17 mRNA targets including Pkd1 and Pkd2. Importantly, RGLS4326 demonstrates a favorable preclinical safety profile and attenuates cyst growth in human in vitro ADPKD models and multiple PKD mouse models after subcutaneous administration. The preclinical characteristics of RGLS4326 support its clinical development as a disease-modifying treatment for ADPKD.

Rapid identification of somatic mutations in colorectal and breast cancer tissues using mismatch repair detection (MRD).

Mismatch repair detection (MRD) was used to screen 93 matched tumor-normal sample pairs and 22 cell lines for somatic mutations in 30 cancer relevant genes. Using a starting amount of only 150 ng of genomic DNA, we screened 102 kb of sequence for somatic mutations in colon and breast cancer. A total of 152 somatic mutations were discovered, encompassing previously reported mutations, such as BRAF V600E and KRAS G12S, G12V, and G13D, as well as novel mutations, including some in genes in which somatic mutations have not previously been reported, such as MAP2K1 and MAP2K2. The distribution of mutations ranged widely within and across tumor types. The functional significance of many of these mutations is not understood, with patterns of selection only evident in KRAS and BRAF in colon cancer. These results present a novel approach to high-throughput mutation screening using small amounts of starting material and reveal a mutation spectrum across 30 genes in a large cohort of breast and colorectal cancers.

Accelerated evolution of the ASPM gene controlling brain size begins prior to human brain expansion.

Primary microcephaly (MCPH) is a neurodevelopmental disorder characterized by global reduction in cerebral cortical volume. The microcephalic brain has a volume comparable to that of early hominids, raising the possibility that some MCPH genes may have been evolutionary targets in the expansion of the cerebral cortex in mammals and especially primates. Mutations in ASPM, which encodes the human homologue of a fly protein essential for spindle function, are the most common known cause of MCPH. Here we have isolated large genomic clones containing the complete ASPM gene, including promoter regions and introns, from chimpanzee, gorilla, orangutan, and rhesus macaque by transformation-associated recombination cloning in yeast. We have sequenced these clones and show that whereas much of the sequence of ASPM is substantially conserved among primates, specific segments are subject to high Ka/Ks ratios (nonsynonymous/synonymous DNA changes) consistent with strong positive selection for evolutionary change. The ASPM gene sequence shows accelerated evolution in the African hominoid clade, and this precedes hominid brain expansion by several million years. Gorilla and human lineages show particularly accelerated evolution in the IQ domain of ASPM. Moreover, ASPM regions under positive selection in primates are also the most highly diverged regions between primates and nonprimate mammals. We report the first direct application of TAR cloning technology to the study of human evolution. Our data suggest that evolutionary selection of specific segments of the ASPM sequence strongly relates to differences in cerebral cortical size.

Lipid Nanoparticle-Mediated Delivery of Anti-miR-17 Family Oligonucleotide Suppresses Hepatocellular Carcinoma Growth.

Hepatocellular carcinoma (HCC) is one of the most common human malignancies with poor prognosis and urgent unmet medical need. Aberrant expression of multiple members of the miR-17 family are frequently observed in HCC, and their overexpression promotes tumorigenic properties of HCC cells. However, whether pharmacologic inhibition of the miR-17 family inhibits HCC growth remains unknown. In this study, we validated that the miR-17 family was upregulated in a subset of HCC tumors and cell lines and its inhibition by a tough decoy inhibitor suppressed the growth of Hep3B and HepG2 cells, which overexpress the miR-17 family. Furthermore, inhibition of the miR-17 family led to a global derepression of direct targets of the family in all three HCC cell lines tested. Pathway analysis of the deregulated genes indicated that the genes associated with TGFß signaling pathway were highly enriched in Hep3B and HepG2 cells. A miR-17 family target gene signature was established and used to identify RL01-17(5), a lipid nanoparticle encapsulating a potent anti-miR-17 family oligonucleotide. To address whether pharmacologic modulation of the miR-17 family can inhibit HCC growth, RL01-17(5) was systemically administrated to orthotopic Hep3B xenografts. Suppression of Hep3B tumor growth in vivo was observed and tumor growth inhibition correlated with induction of miR-17 family target genes. Together, this study provides proof-of-concept for targeting the miR-17 family in HCC therapy. Mol Cancer Ther; 16(5); 905-13. ©2017 AACR.

HERVd: database of human endogenous retroviruses.

The human endogenous retroviruses database (HERVd) is maintained at the Institute of Molecular Genetics, Academy of Sciences of the Czech Republic, and is accessible via the World Wide Web at http://herv.img.cas.cz. The HERVd provides complex information on and analysis of retroviral elements found in the human genome. It can be used for searches of individual HERV families, identification of HERV parts, graphical output of HERV structures, comparison of HERVs and identification of retrovirus integration sites.

WAViS server for handling, visualization and presentation of multiple alignments of nucleotide or amino acids sequences.

Web Alignment Visualization Server contains a set of web-tools designed for quick generation of publication-quality color figures of multiple alignments of nucleotide or amino acids sequences. It can be used for identification of conserved regions and gaps within many sequences using only common web browsers. The server is accessible at http://wavis.img.cas.cz.

HERVd: the Human Endogenous RetroViruses Database: update.

An elaboration of HERVd (http://herv.img.cas.cz) is being carried out in two directions. One of them is the integration and better classification of families that diverge considerably from typical retroviral genomes. This leads to a more precise identification of members with individual families. The second improvement is better accessibility of the database and connection with human genome annotation.

Anti-miR-21 Suppresses Hepatocellular Carcinoma Growth via Broad Transcriptional Network Deregulation.

UNLABELLED: Hepatocellular carcinoma (HCC) remains a significant clinical challenge with few therapeutic options available to cancer patients. MicroRNA 21-5p (miR-21) has been shown to be upregulated in HCC, but the contribution of this oncomiR to the maintenance of tumorigenic phenotype in liver cancer remains poorly understood. We have developed potent and specific single-stranded oligonucleotide inhibitors of miR-21 (anti-miRNAs) and used them to interrogate dependency on miR-21 in a panel of liver cancer cell lines. Treatment with anti-miR-21, but not with a mismatch control anti-miRNA, resulted in significant derepression of direct targets of miR-21 and led to loss of viability in the majority of HCC cell lines tested. Robust induction of caspase activity, apoptosis, and necrosis was noted in anti-miR-21-treated HCC cells. Furthermore, ablation of miR-21 activity resulted in inhibition of HCC cell migration and suppression of clonogenic growth. To better understand the consequences of miR-21 suppression, global gene expression profiling was performed on anti-miR-21-treated liver cancer cells, which revealed striking enrichment in miR-21 target genes and deregulation of multiple growth-promoting pathways. Finally, in vivo dependency on miR-21 was observed in two separate HCC tumor xenograft models. In summary, these data establish a clear role for miR-21 in the maintenance of tumorigenic phenotype in HCC in vitro and in vivo. IMPLICATIONS: miR-21 is important for the maintenance of the tumorigenic phenotype of HCC and represents a target for pharmacologic intervention.

Length distribution of long interspersed nucleotide elements (LINEs) and processed pseudogenes of human endogenous retroviruses: implications for retrotransposition and pseudogene detection.

Deciphering the human genome includes reliable identification and structural characterization of individual retrotransposon elements. The most active group of autonomous transposable elements, the long interspersed nuclear elements (LINE), transpose themselves as well as other RNAs, including those of human endogenous retroviruses (HERV). During this transposition, however, the LINE-encoded reverse transcriptase (RT) often abortively dissociates from the RNA template, leaving a prematurely terminated, 5' truncated copy. We have analyzed the length distributions of LINEs and of processed pseudogenes derived from HERV-W. As expected, we have found that the majority of 5' truncated LINEs and HERV-W processed pseudogenes show a prevalence of very short elements terminated close to the 3' end. On the other hand, the number of complete elements is far above the expectation. The characteristic distribution in both cases indicates two important conclusions: (i) dissociation of LINE RT from the template cannot be fully explained by low processivity of RT modelled as a stochastic, Poisson-type process. (ii) Currently cited numbers of pseudogenes within the human genome are underestimated, since a large percentage of pseudogenes are terminated in the 3' untranslated region and remain undetectable in translated homology searches of protein databases against the human genome.

Representing GC variation along eukaryotic chromosomes.

Genome sequencing now permits direct visual representation, at any scale, of GC heterogeneity along the chromosomes of several higher eukaryotes. Plots can be easily obtained from the chromosomal sequences, yet sequence releases of mammalian or plant chromosomes still tend to use small scales or window sizes that obscure important large-scale compositional features. To faithfully reveal, at one glance, the compositional variation at a given scale, we have devised a simple scheme that combines line plots with color-coded shading of the regions underneath the plots. The scheme can be applied to different eukaryotic genomes to facilitate their comparison, as illustrated here for a sample of chromosomes chosen from seven selected species. As a complement to a previously published compact view of isochores in the human genome sequence, we include here an analogous map for the recently sequenced mouse genome, and discuss the contribution of repetitive DNA to the GC variation along the plots. Supplementary information, including a database of color-coded GC profiles for all recently sequenced eukaryotes and the program draw_chromosomes_gc.pl used to obtain them, are available at.

A compact view of isochores in the draft human genome sequence.

Prior to genome sequencing, information on base composition (GC level) and its variation in mammalian genomes could be obtained using density gradient ultracentrifugation. Analyses using this approach led to the conclusion that mammalian genomes are organized into mosaics of fairly homogeneous regions, called isochores. We present an initial compositional overview of the chromosomes of the recently available draft human genome sequence, in the form of color-coded moving window plots and corresponding GC level histograms. Results obtained from the draft human genome sequence agree well with those obtained or deduced earlier from CsCl experiments. The draft sequence now permits the visualization of the mosaic organization of the human genome at the DNA sequence level.

Test of four colon cancer risk-scores in formalin fixed paraffin embedded microarray gene expression data.

BACKGROUND: Prognosis prediction for resected primary colon cancer is based on the T-stage Node Metastasis (TNM) staging system. We investigated if four well-documented gene expression risk scores can improve patient stratification. METHODS: Microarray-based versions of risk-scores were applied to a large independent cohort of 688 stage II/III tumors from the PETACC-3 trial. Prognostic value for relapse-free survival (RFS), survival after relapse (SAR), and overall survival (OS) was assessed by regression analysis. To assess improvement over a reference, prognostic model was assessed with the area under curve (AUC) of receiver operating characteristic (ROC) curves. All statistical tests were two-sided, except the AUC increase. RESULTS: All four risk scores (RSs) showed a statistically significant association (single-test, P < .0167) with OS or RFS in univariate models, but with HRs below 1.38 per interquartile range. Three scores were predictors of shorter RFS, one of shorter SAR. Each RS could only marginally improve an RFS or OS model with the known factors T-stage, N-stage, and microsatellite instability (MSI) status (AUC gains < 0.025 units). The pairwise interscore discordance was never high (maximal Spearman correlation = 0.563) A combined score showed a trend to higher prognostic value and higher AUC increase for OS (HR = 1.74, 95% confidence interval [CI] = 1.44 to 2.10, P < .001, AUC from 0.6918 to 0.7321) and RFS (HR = 1.56, 95% CI = 1.33 to 1.84, P < .001, AUC from 0.6723 to 0.6945) than any single score. CONCLUSIONS: The four tested gene expression-based risk scores provide prognostic information but contribute only marginally to improving models based on established risk factors. A combination of the risk scores might provide more robust information. Predictors of RFS and SAR might need to be different.