Transformer-based biomarker prediction from colorectal cancer histology: A large-scale multicentric study.

Deep learning (DL) can accelerate the prediction of prognostic biomarkers from routine pathology slides in colorectal cancer (CRC). However, current approaches rely on convolutional neural networks (CNNs) and have mostly been validated on small patient cohorts. Here, we develop a new transformer-based pipeline for end-to-end biomarker prediction from pathology slides by combining a pre-trained transformer encoder with a transformer network for patch aggregation. Our transformer-based approach substantially improves the performance, generalizability, data efficiency, and interpretability as compared with current state-of-the-art algorithms. After training and evaluating on a large multicenter cohort of over 13,000 patients from 16 colorectal cancer cohorts, we achieve a sensitivity of 0.99 with a negative predictive value of over 0.99 for prediction of microsatellite instability (MSI) on surgical resection specimens. We demonstrate that resection specimen-only training reaches clinical-grade performance on endoscopic biopsy tissue, solving a long-standing diagnostic problem.

Oncomicrobial Community Profiling Identifies Clinicomolecular and Prognostic Subtypes of Colorectal Cancer.

BACKGROUND & AIMS: Dysbiosis of gut microbiota is linked to the development of colorectal cancer (CRC). However, microbiota-based stratification of CRC tissue and how this relates to clinicomolecular characteristics and prognosis remains to be clarified. METHODS: Tumor and normal mucosa from 423 patients with stage I to IV CRC were profiled by bacterial 16S rRNA gene sequencing. Tumors were characterized for microsatellite instability (MSI), CpG island methylator phenotype (CIMP), APC, BRAF, KRAS, PIK3CA, FBXW7, SMAD4, and TP53 mutations, subsets for chromosome instability (CIN), mutation signatures, and consensus molecular subtypes (CMS). Microbial clusters were validated in an independent cohort of 293 stage II/III tumors. RESULTS: Tumors reproducibly stratified into 3 oncomicrobial community subtypes (OCSs) with distinguishing features: OCS1 (Fusobacterium/oral pathogens, proteolytic, 21%), right-sided, high-grade, MSI-high, CIMP-positive, CMS1, BRAF V600E, and FBXW7 mutated; OCS2 (Firmicutes/Bacteroidetes, saccharolytic, 44%), and OCS3 (Escherichia/Pseudescherichia/Shigella, fatty acid ß-oxidation, 35%) both left-sided and exhibiting CIN. OCS1 was associated with MSI-related mutation signatures (SBS15, SBS20, ID2, and ID7) and OCS2 and OCS3 with SBS18 related to damage by reactive oxygen species. Among stage II/III patients, OCS1 and OCS3 both had poorer overall survival compared with OCS2 for microsatellite stable tumors (multivariate hazard ratio [HR], 1.85; 95% confidence interval [CI], 1.15-2.99; P = .012; and HR, 1.52; 95% CI 1.01-2.29; P = .044, respectively) and left-sided tumors (multivariate HR, 2.66; 95% CI, 1.45-4.86; P = .002; and HR, 1.76; 95% CI, 1.03-3.02; P = .039, respectively). CONCLUSIONS: OCS classification stratified CRCs into 3 distinct subgroups with different clinicomolecular features and outcomes. Our findings provide a framework for a microbiota-based stratification of CRC to refine prognostication and to inform the development of microbiota-targeted interventions.

Tumour infiltrating lymphocyte status is superior to histological grade, DNA mismatch repair and BRAF mutation for prognosis of colorectal adenocarcinomas with mucinous differentiation.

Mucinous colorectal adenocarcinoma (CRC) is conventionally defined by extracellular mucin comprising >50% of the tumour area, while tumours with ≤50% mucin are designated as having a mucinous component. However, these definitions are largely arbitrary and comparisons of clinico-molecular features and outcomes by proportion of mucinous component are limited. A cohort of 1643 patients with stage II/III cancer was examined for tumour mucinous component, DNA mismatch repair (MMR) status, BRAF mutation and tumour infiltrating lymphocytes (TILs). Tumours with ≤50% mucinous component exhibited similar characteristics as mucinous tumours, including association with female gender, proximal location, high grade, TIL-high, defective MMR (dMMR) and BRAF mutation. Proportion of mucinous component did not stratify disease-free survival (DFS). In univariate analysis dMMR status, but not histological grade, stratified survival for mucinous and mucinous component tumours; however, in multivariate analysis dMMR status was not an independent predictor. BRAF mutation prognostic value depended on mucinous differentiation and MMR status, with poor prognosis limited to non-mucinous pMMR tumours (HR 2.61, 95% CI 1.69-4.03; p < 0.001). TIL status was a strong independent predictor of DFS in mucinous/mucinous component tumours (HR 0.40, 95% CI 0.23-0.67; p < 0.001), and a superior predictor of prognosis compared with histological grade, MMR and BRAF mutation. Mucinous component and mucinous stage II/III CRCs exhibit clinico-molecular resemblances, with histological grade and BRAF mutation lacking prognostic value. Prognosis for these tumours was instead strongly associated with TIL status, with the most favourable outcomes in TIL-high dMMR tumours, whilst TIL-low tumours had poor outcomes irrespective of MMR status.

Evaluation of the transferability of survival calculators for stage II/III colon cancer across healthcare systems.

Adjuvant! Online Inc (A!O), the Memorial Sloan Kettering Cancer Center (MSKCC), MD Anderson (MDA) and Mayo Clinic (MC) provide calculators to predict survival probabilities for patients with resected early-stage colon cancer, trained on data from United States (US) patient cohorts or patients enrolled in international clinical trials. Limited data exist on the transferability of calculators across healthcare systems. Calculator transferability to Australian community practice was evaluated for 1,401 stage II/III patients. Calibration and discrimination were assessed for overall (OS), cancer-specific (CSS) or recurrence-free survival (RFS). The US patient cohort-based calculators, A!O, MSKCC and MDA, significantly overestimated risks of recurrence and death in Australian patients, with 5-year OS, CSS and RFS prediction differences of -6.5% to -9.9%, -9.1% to -14.4% and - 3.8% to -6.8%, respectively (p < 0.001). Significant heterogeneity in calibration was observed for subgroups by tumor stage and treatment, age, gender, tumor location, ECOG and ASA score. Calibration appeared acceptable for the clinical trial patient-based MC calculator, but restricted tool applicability (stage III patients, ≥12 examined lymph nodes, receiving adjuvant treatment) limited the sample size. Compared to AJCC 7th edition tumor staging, calculators showed improved discrimination for OS, but no improvement for CSS and RFS. In conclusion, deficiencies in calibration limited transferability of US patient cohort-based survival calculators for early-stage colon cancer to the setting of Australian community practice. Our results demonstrate the utility for multi-feature survival calculators to improve OS predictions but highlight the importance for performance assessment of tools prior to implementation in an external health care setting.

Lymphocytic response to tumour and deficient DNA mismatch repair identify subtypes of stage II/III colorectal cancer associated with patient outcomes.

OBJECTIVE: Tumour-infiltrating lymphocyte (TIL) response and deficient DNA mismatch repair (dMMR) are determinants of prognosis in colorectal cancer. Although highly correlated, evidence suggests that these are independent predictors of outcome. However, the prognostic significance of combined TIL/MMR classification and how this compares to the major genomic and transcriptomic subtypes remain unclear. DESIGN: A prospective cohort of 1265 patients with stage II/III cancer was examined for TIL/MMR status and BRAF/KRAS mutations. Consensus molecular subtype (CMS) status was determined for 142 cases. Associations with 5-year disease-free survival (DFS) were evaluated and validated in an independent cohort of 602 patients. RESULTS: Tumours were categorised into four subtypes based on TIL and MMR status: TIL-low/proficient-MMR (pMMR) (61.3% of cases), TIL-high/pMMR (14.8%), TIL-low/dMMR (8.6%) and TIL-high/dMMR (15.2%). Compared with TIL-high/dMMR tumours with the most favourable prognosis, both TIL-low/dMMR (HR=3.53; 95% CI=1.88 to 6.64; Pmultivariate<0.001) and TIL-low/pMMR tumours (HR=2.67; 95% CI=1.47 to 4.84; Pmultivariate=0.001) showed poor DFS. Outcomes of patients with TIL-low/dMMR and TIL-low/pMMR tumours were similar. TIL-high/pMMR tumours showed intermediate survival rates. These findings were validated in an independent cohort. TIL/MMR status was a more significant predictor of prognosis than National Comprehensive Cancer Network high-risk features and was a superior predictor of prognosis compared with genomic (dMMR, pMMR/BRAFwt /KRASwt , pMMR/BRAFmut /KRASwt , pMMR/BRAFwt /KRASmut ) and transcriptomic (CMS 1-4) subtypes. CONCLUSION: TIL/MMR classification identified subtypes of stage II/III colorectal cancer associated with different outcomes. Although dMMR status is generally considered a marker of good prognosis, we found this to be dependent on the presence of TILs. Prognostication based on TIL/MMR subtypes was superior compared with histopathological, genomic and transcriptomic subtypes.

Mutation burden and other molecular markers of prognosis in colorectal cancer treated with curative intent: results from the QUASAR 2 clinical trial and an Australian community-based series.

BACKGROUND: Molecular indicators of colorectal cancer prognosis have been assessed in several studies, but most analyses have been restricted to a handful of markers. We aimed to identify prognostic biomarkers for colorectal cancer by sequencing panels of multiple driver genes. METHODS: In stage II or III colorectal cancers from the QUASAR 2 open-label randomised phase 3 clinical trial and an Australian community-based series, we used targeted next-generation sequencing of 82 and 113 genes, respectively, including the main colorectal cancer drivers. We investigated molecular pathways of tumorigenesis, and analysed individual driver gene mutations, combinations of mutations, or global measures such as microsatellite instability (MSI) and mutation burden (total number of non-synonymous mutations and coding indels) for associations with relapse-free survival in univariable and multivariable models, principally Cox proportional hazards models. FINDINGS: In QUASAR 2 (511 tumours), TP53, KRAS, BRAF, and GNAS mutations were independently associated with shorter relapse-free survival (p<0·035 in all cases), and total somatic mutation burden with longer survival (hazard ratio [HR] 0·81 [95% CI 0·68-0·96]; p=0·014). MSI was not independently associated with survival (HR 1·12 [95% CI 0·57-2·19]; p=0·75). We successfully validated these associations in the Australian sample set (296 tumours). In a combined analysis of both the QUASAR 2 and the Australian sample sets, mutation burden was also associated with longer survival (HR 0·84 [95% CI 0·74-0·94]; p=0·004) after exclusion of MSI-positive and POLE mutant tumours. In an extended analysis of 1732 QUASAR 2 and Australian colorectal cancers for which KRAS, BRAF, and MSI status were available, KRAS and BRAF mutations were specifically associated with poor prognosis in MSI-negative cancers. MSI-positive cancers with KRAS or BRAF mutations had better prognosis than MSI-negative cancers that were wild-type for KRAS or BRAF. Mutations in the genes NF1 and NRAS from the MAPK pathway co-occurred, and mutations in the DNA damage-response genes TP53 and ATM were mutually exclusive. We compared a prognostic model based on the gold standard of clinicopathological variables and MSI with our new model incorporating clinicopathological variables, mutation burden, and driver mutations in KRAS, BRAF, and TP53. In both QUASAR 2 and the Australian cohort, our new model was significantly better (p=0·00004 and p=0·0057, respectively, based on a likelihood ratio test). INTERPRETATION: Multigene panels identified two previously unreported prognostic associations in colorectal cancer involving TP53 mutation and total mutation burden, and confirmed associations with KRAS and BRAF. Even a modest-sized gene panel can provide important information for use in clinical practice and outperform MSI-based prognostic models. FUNDING: UK Technology Strategy Board, National Institute for Health Research Oxford Biomedical Research Centre, Cancer Australia Project, Cancer Council Victoria, Ludwig Institute for Cancer Research, Victorian Government.

MACROD2 Haploinsufficiency Impairs Catalytic Activity of PARP1 and Promotes Chromosome Instability and Growth of Intestinal Tumors.

ADP-ribosylation is an important posttranslational protein modification that regulates diverse biological processes, controlled by dedicated transferases and hydrolases. Here, we show that frequent deletions (∼30%) of the MACROD2 mono-ADP-ribosylhydrolase locus in human colorectal cancer cause impaired PARP1 transferase activity in a gene dosage-dependent manner. MACROD2 haploinsufficiency alters DNA repair and sensitivity to DNA damage and results in chromosome instability. Heterozygous and homozygous depletion of Macrod2 enhances intestinal tumorigenesis in ApcMin/+ mice and the growth of human colorectal cancer xenografts. MACROD2 deletion in sporadic colorectal cancer is associated with the extent of chromosome instability, independent of clinical parameters and other known genetic drivers. We conclude that MACROD2 acts as a haploinsufficient tumor suppressor, with loss of function promoting chromosome instability, thereby driving cancer evolution.Significance: Chromosome instability (CIN) is a hallmark of cancer. We identify MACROD2 deletion as a cause of CIN in human colorectal cancer. MACROD2 loss causes repression of PARP1 activity, impairing DNA repair. MACROD2 haploinsufficiency promotes CIN and intestinal tumor growth. Our results reveal MACROD2 as a major caretaker tumor suppressor gene. Cancer Discov; 8(8); 988-1005. ©2018 AACR.See related commentary by Jin and Burkard, p. 921This article is highlighted in the In This Issue feature, p. 899.

Disruption of a -35 kb Enhancer Impairs CTCF Binding and MLH1 Expression in Colorectal Cells.

Purpose:MLH1 is a major tumor suppressor gene involved in the pathogenesis of Lynch syndrome and various sporadic cancers. Despite their potential pathogenic importance, genomic regions capable of regulating MLH1 expression over long distances have yet to be identified.Experimental Design: Here, we use chromosome conformation capture (3C) to screen a 650-kb region flanking the MLH1 locus to identify interactions between the MLH1 promoter and distal regions in MLH1-expressing and nonexpressing cells. Putative enhancers were functionally validated using luciferase reporter assays, chromatin immunoprecipitation, and CRISPR-Cas9-mediated deletion of endogenous regions. To evaluate whether germline variants in the enhancer might contribute to impaired MLH1 expression in patients with suspected Lynch syndrome, we also screened germline DNA from a cohort of 74 patients with no known coding mutations or epimutations at the MLH1 promoter.Results: A 1.8-kb DNA fragment, 35 kb upstream of the MLH1 transcription start site enhances MLH1 gene expression in colorectal cells. The enhancer was bound by CTCF and CRISPR-Cas9-mediated deletion of a core binding region impairs endogenous MLH1 expression. A total of 5.4% of suspected Lynch syndrome patients have a rare single-nucleotide variant (G > A; rs143969848; 2.5% in gnomAD European, non-Finnish) within a highly conserved CTCF-binding motif, which disrupts enhancer activity in SW620 colorectal carcinoma cells.Conclusions: A CTCF-bound region within the MLH1-35 enhancer regulates MLH1 expression in colorectal cells and is worthy of scrutiny in future genetic screening strategies for suspected Lynch syndrome associated with loss of MLH1 expression. Clin Cancer Res; 24(18); 4602-11. ©2018 AACR.

Validation of specificity of antibodies for immunohistochemistry: the case of ROR2.

The Wnt signalling receptor receptor tyrosine kinase-like orphan receptor 2 (ROR2) is implicated in numerous human cancers. However, there have been conflicting reports regarding ROR2 expression, some studies showing upregulation and others downregulation of ROR2 in the same cancer type. The majority of these studies used immunohistochemistry (IHC) to detect ROR2 protein, without validation of the used antibodies. There appears to be currently no consensus on the antibody best suited for ROR2 detection or how ROR2 expression changes in various cancer types. We examined three commercially available ROR2 antibodies and found that only one bound specifically to ROR2. Another antibody cross-reacted with other proteins, and the third failed to detect ROR2 at all. ROR2 detection by IHC on 107 patient samples using the ROR2 specific antibody showed that the majority of colorectal cancers show loss of ROR2 protein. We found no association between ROR2 staining and poor patient survival, as had been previously reported. These results question the previously reported association between ROR2 and poor patient survival in colorectal cancer. Future studies should use fully validated antibodies when detecting ROR2 protein, as non-specific staining can lead to irrelevant observations and misinterpretations.

Pathogenic germline MCM9 variants are rare in Australian Lynch-like syndrome patients.

Lynch syndrome is a hereditary cancer syndrome caused by the autosomal dominant inheritance of loss-of-function mutations in DNA mismatch repair (MMR) genes. Approximately one quarter of clinically suspected cases have no identifiable germline mutation in any MMR gene, a condition known as Lynch-like syndrome (LLS). MCM9 was recently identified as the DNA helicase in the mammalian MMR complex and loss of helicase activity results in microsatellite instability. We hypothesized that pathogenic variants in MCM9 may account for LLS. The 5'UTR and coding region of MCM9 were sequenced in germline DNA of 109 Australian patients with LLS and variants were cross-referenced with three population-based databases (dbSNP144, 1000 Genomes, ExAC). The functional effect of variants was assessed in silico with PolyPhen-2, SIFT and CONDEL. Fifteen variants that included six common SNPs and nine variants of unknown significance (VUS) were identified. We conclude that VUS occur in MCM9 in a small proportion of LLS patients and MCM9 mutations are unlikely to explain most LLS cases.

Integrated Genetic, Epigenetic, and Transcriptional Profiling Identifies Molecular Pathways in the Development of Laterally Spreading Tumors.

Laterally spreading tumors (LST) are colorectal adenomas that develop into extremely large lesions with predominantly slow progression to cancer, depending on lesion subtype. Comparing and contrasting the molecular profiles of LSTs and colorectal cancers offers an opportunity to delineate key molecular alterations that drive malignant transformation in the colorectum. In a discovery cohort of 11 LSTs and paired normal mucosa, we performed a comprehensive and unbiased screen of the genome, epigenome, and transcriptome followed by bioinformatics integration of these data and validation in an additional 84 large, benign colorectal lesions. Mutation rates in LSTs were comparable with microsatellite-stable colorectal cancers (2.4 vs. 2.6 mutations per megabase); however, copy number alterations were infrequent (averaging only 1.5 per LST). Frequent genetic, epigenetic, and transcriptional alterations were identified in genes not previously implicated in colorectal neoplasia (ANO5, MED12L, EPB41L4A, RGMB, SLITRK1, SLITRK5, NRXN1, ANK2). Alterations to pathways commonly mutated in colorectal cancers, namely, the p53, PI3K, and TGFß pathways, were rare. Instead, LST-altered genes converged on axonal guidance, Wnt, and actin cytoskeleton signaling. These integrated omics data identify molecular features associated with noncancerous LSTs and highlight that mutation load, which is relatively high in LSTs, is a poor predictor of invasive potential. IMPLICATIONS: The novel genetic, epigenetic, and transcriptional changes associated with LST development reveal important insights into why some adenomas do not progress to cancer. The finding that LSTs exhibit a mutational load similar to colorectal carcinomas has implications for the validity of molecular biomarkers for assessing cancer risk. Mol Cancer Res; 14(12); 1217-28. ©2016 AACR.

ROR2 is epigenetically inactivated in the early stages of colorectal neoplasia and is associated with proliferation and migration.

BACKGROUND: Colorectal cancer (CRC) is closely linked to Wnt signalling, with 94 % of cases exhibiting a Wnt related mutation. ROR2 is a receptor tyrosine kinase that is thought to repress ß-catenin dependent Wnt signalling. Our study aims to determine if ROR2 is epigenetically silenced in CRC and determine if in vitro silencing of ROR2 potentiates Wnt signalling, and alters the proliferative, migratory or invasive potential of cells. METHODS: ROR2 expression was examined in CRC cell lines and patient adenomas using qRT-PCR, while COBRA and bisulphite sequencing was used to analyse ROR2 promoter methylation. 258 patient primary tumour samples from publicly available databases were also examined for ROR2 expression and methylation. In addition, the functional effects of ROR2 modulation were investigated in HCT116 cells following ROR2 siRNA knockdown and in RKO and SW620 cells following ectopic ROR2 expression. RESULTS: Reduced ROR2 expression was found to correlate with ROR2 promoter hypermethylation in colorectal cancer cell lines, carcinomas and adenomas. ROR2 expression was downregulated in 76.7 % (23/30) of CRC cell lines with increasing ROR2 promoter hypermethylation correlating with progressively lower expression. Analysis of 239 primary tumour samples from a publicly available cohort also found a significant correlation between reduced ROR2 expression and increased promoter methylation. Methylation analysis of 88 adenomas and 47 normal mucosa samples found greater percentage of adenoma samples to be methylated. Additional analysis also revealed that adenoma samples with reduced ROR2 expression also possessed ROR2 promoter hypermethylation. ROR2 knockdown in the CRC cell line HCT116 significantly decreased expression of the ß-catenin independent Wnt targets genes JNK and NFATC1, increased cellular proliferation and migration but decreased invasion. When ROR2 was ectopically expressed in RKO and SW620 cells, there was no significant change to either cellular proliferation or migration. CONCLUSION: ROR2 is frequently epigenetically inactivated by promoter hypermethylation in the early stages of colorectal neoplasia and this may contribute to colorectal cancer progression by increasing cellular proliferation and migration.

Integration and Analysis of Heterogeneous Colorectal Cancer Data for Translational Research.

Cancer is the number one cause of death in Australia with colorectal cancer being the second most common cancer type. The translation of cancer research into clinical practice is hindered by the lack of integration of heterogeneous and autonomous data from various data sources. Integration of heterogeneous data can offer researchers a comprehensive source for biospecimen identification, hypothesis formulation, hypothesis validation, cohort discovery and biomarker discovery. Alongside the increasing prominence of big data, various translational research tools such as tranSMART have emerged that can converge and analyse different types of data. In this study, we show the integration of different data types from a significant Australian colorectal cancer cohort. Additionally, colorectal cancer datasets from The Cancer Genome Atlas were also integrated for comparison. These integrated data are accessible via http://www.tcrn.unsw.edu.au/transmart. The use of translational research tools for data integration can provide a cost-effective and rapid approach to translational cancer research.

SNP rs16906252C>T Is an Expression and Methylation Quantitative Trait Locus Associated with an Increased Risk of Developing MGMT-Methylated Colorectal Cancer.

PURPOSE: Methylation of the MGMT promoter is the major cause of O6-methylguanine methyltransferase deficiency in cancer and has been associated with the T variant of the promoter enhancer SNP rs16906252C>T. We sought evidence for an association between the rs16906252C>T genotype and increased risk of developing a subtype of colorectal cancer featuring MGMT methylation, mediated by genotype-dependent epigenetic silencing within normal tissues. EXPERIMENTAL DESIGN: By applying a molecular pathologic epidemiology case-control study design, associations between rs16906252C>T and risk for colorectal cancer overall, and colorectal cancer stratified by MGMT methylation status, were estimated using multinomial logistic regression in two independent retrospective series of colorectal cancer cases and controls. The test sample comprised 1,054 colorectal cancer cases and 451 controls from Sydney, Australia. The validation sample comprised 612 colorectal cancer cases and 245 controls from the Australasian Colon Cancer Family Registry (ACCFR). To determine whether rs16906252C>T was linked to a constitutively altered epigenetic state, quantitative allelic expression and methylation analyses were performed in normal tissues. RESULTS: An association between rs16906252C>T and increased risk of developing MGMT-methylated colorectal cancer in the Sydney sample was observed [OR, 3.3; 95% confidence interval (CI), 2.0-5.3; P < 0.0001], which was replicated in the ACCFR sample (OR, 4.0; 95% CI, 2.4-6.8; P < 0.0001). The T allele demonstrated about 2.5-fold reduced transcription in normal colorectal mucosa from cases and controls and was selectively methylated in a minority of normal cells, indicating that rs16906252C>T represents an expression and methylation quantitative trait locus. CONCLUSIONS: We provide evidence that rs16906252C>T is associated with elevated risk for MGMT-methylated colorectal cancer, likely mediated by constitutive epigenetic repression of the T allele. Clin Cancer Res; 22(24); 6266-77. ©2016 AACR.

Relative telomere lengths in tumor and normal mucosa are related to disease progression and chromosome instability profiles in colorectal cancer.

Telomeric dysfunction is linked to colorectal cancer (CRC) initiation. However, the relationship of normal tissue and tumor telomere lengths with CRC progression, molecular features and prognosis is unclear. Here, we measured relative telomere length (RTL) by real-time quantitative PCR in 90 adenomas (aRTL), 419 stage I-IV CRCs (cRTL) and adjacent normal mucosa (nRTL). Age-adjusted RTL was analyzed against germline variants in telomere biology genes, chromosome instability (CIN), microsatellite instability (MSI), CpG island methylator phenotype (CIMP), TP53, KRAS, BRAF mutations and clinical outcomes. In 509 adenoma or CRC patients, nRTL decreased with advancing age. Female gender, proximal location and the TERT rs2736100 G allele were independently associated with longer age-adjusted nRTL. Adenomas and carcinomas exhibited telomere shortening in 79% and 67% and lengthening in 7% and 15% of cases. Age-adjusted nRTL and cRTL were independently associated with tumor stage, decreasing from adenoma to stage III and leveling out or increasing from stage III to IV, respectively. Cancer MSI, CIMP, TP53, KRAS and BRAF status were not related to nRTL or cRTL. Near-tetraploid CRCs exhibited significantly longer cRTLs than CIN- and aneuploidy CRCs, while cRTL was significantly shorter in CRCs with larger numbers of chromosome breaks. Age-adjusted nRTL, cRTL or cRTL:nRTL ratios were not associated with disease-free or overall survival in stage II/III CRC. Taken together, our data show that both normal mucosa and tumor RTL are independently associated with CRC progression, and highlight divergent associations of CRC telomere length with tumor CIN profiles.

The Impact of Mismatch Repair Status in Colorectal Cancer on the Decision to Treat With Adjuvant Chemotherapy: An Australian Population-Based Multicenter Study.

BACKGROUND: Testing for mismatch repair (MMR) status in colorectal cancer (CRC) may provide useful prognostic and predictive information. We evaluated the impact of such testing on real-world practice regarding adjuvant chemotherapy for patients with resected CRC. PATIENTS AND METHODS: A total of 175 patients with stage II and III mismatch repair-deficient (MMRD) CRC were identified from an Australian population-based study of incident CRCs. Their treatment decisions were compared with those for a cohort of 773 stage-matched patients with mismatch repair-proficient (MMRP) CRCs. The effect of MMR status, age, and pathologic characteristics on treatment decisions was determined using multiple regression analysis. RESULTS: Overall, 32% of patients in stage II and 71% of patients in stage III received adjuvant chemotherapy. Among the stage II patients, those with MMRD cancer were less likely to receive chemotherapy than were MMRP cases (15% vs. 38%; p < .0001). In this group, the treatment decision was influenced by age, tumor location, and T stage. MMR status influenced the treatment decision such that its impact diminished with increasing patient age. Among patients with stage III tumors, no difference was found in the chemotherapy rates between the MMRD and MMRP cases. In this group, age was the only significant predictor of the treatment decision. CONCLUSION: The findings of this study suggest that knowledge of the MMR status of sporadic CRC influences treatment decisions for stage II patients, in an era when clear recommendations as to how these findings should influence practice are lacking. IMPLICATIONS FOR PRACTICE: Microsatellite instability (MSI) is a molecular marker of defective DNA mismatch repair found in 15% of sporadic colorectal cancers. Until recently, expert guidelines on the role of MSI as a valid biomarker in the selection of stage II patients for adjuvant chemotherapy were lacking. Conducted at a time when the clinical utility of routine MSI testing was unclear, this study found that clinicians were influenced by MSI status in selecting stage II patients for chemotherapy. Furthermore, the impact of MSI on treatment decisions was greatest in younger patients and declined progressively until age 80 years, when no effect was found.

Wild-type APC predicts poor prognosis in microsatellite-stable proximal colon cancer.

BACKGROUND: APC mutations (APC-mt) occur in ∼70% of colorectal cancers (CRCs), but their relationship to prognosis is unclear. METHODS: APC prognostic value was evaluated in 746 stage I-IV CRC patients, stratifying for tumour location and microsatellite instability (MSI). Microarrays were used to identify a gene signature that could classify APC mutation status, and classifier ability to predict prognosis was examined in an independent cohort. RESULTS: Wild-type APC microsatellite stable (APC-wt/MSS) tumours from the proximal colon showed poorer overall and recurrence-free survival (OS, RFS) than APC-mt/MSS proximal, APC-wt/MSS distal and APC-mt/MSS distal tumours (OS HR⩾1.79, P⩽0.015; RFS HR⩾1.88, P⩽0.026). APC was a stronger prognostic indicator than BRAF, KRAS, PIK3CA, TP53, CpG island methylator phenotype or chromosomal instability status (P⩽0.036). Microarray analysis similarly revealed poorer survival in MSS proximal cancers with an APC-wt-like signature (P=0.019). APC status did not affect outcomes in MSI tumours. In a validation on 206 patients with proximal colon cancer, APC-wt-like signature MSS cases showed poorer survival than APC-mt-like signature MSS or MSI cases (OS HR⩾2.50, P⩽0.010; RFS HR⩾2.14, P⩽0.025). Poor prognosis APC-wt/MSS proximal tumours exhibited features of the sessile serrated neoplasia pathway (P⩽0.016). CONCLUSIONS: APC-wt status is a marker of poor prognosis in MSS proximal colon cancer.

Adaptive Tutorials Versus Web-Based Resources in Radiology: A Mixed Methods Comparison of Efficacy and Student Engagement.

RATIONALE AND OBJECTIVES: Diagnostic imaging is under-represented in medical curricula globally. Adaptive tutorials, online intelligent tutoring systems that provide a personalized learning experience, have the potential to bridge this gap. However, there is limited evidence of their effectiveness for learning about diagnostic imaging. MATERIALS AND METHODS: We performed a randomized mixed methods crossover trial to determine the impact of adaptive tutorials on perceived engagement and understanding of the appropriate use and interpretation of common diagnostic imaging investigations. Although concurrently engaged in disparate blocks of study, 99 volunteer medical students (from years 1-4 of the 6-year program) were randomly allocated to one of two groups. In the first arm of the trial on chest X-rays, one group received access to an adaptive tutorial, whereas the other received links to an existing peer-reviewed Web resource. These two groups crossed over in the second arm of the trial, which focused on computed tomography scans of the head, chest, and abdomen. At the conclusion of each arm of the trial, both groups completed an examination-style assessment, comprising questions both related and unrelated to the topics covered by the relevant adaptive tutorial. Online questionnaires were used to evaluate student perceptions of both learning resources. RESULTS: In both arms of the trial, the group using adaptive tutorials obtained significantly higher assessment scores than controls. This was because of higher assessment scores by senior students in the adaptive tutorial group when answering questions related to topics covered in those tutorials. Furthermore, students indicated significantly better engagement with adaptive tutorials than the Web resource and rated the tutorials as a significantly more valuable tool for learning. CONCLUSIONS: Medical students overwhelmingly accept adaptive tutorials for diagnostic imaging. The tutorials significantly improve the understanding of diagnostic imaging by senior students.

Single nucleotide polymorphism array profiling identifies distinct chromosomal aberration patterns across colorectal adenomas and carcinomas.

The progression of benign colorectal adenomas into cancer is associated with the accumulation of chromosomal aberrations. Even though patterns and frequencies of chromosomal aberrations have been well established in colorectal carcinomas, corresponding patterns of aberrations in adenomas are less well documented. The aim of this study was to profile chromosomal aberrations across colorectal adenomas and carcinomas to provide a better insight into key changes during tumor initiation and progression. Single nucleotide polymorphism array analysis was performed on 216 colorectal tumor/normal matched pairs, comprising 60 adenomas and 156 carcinomas. While many chromosomal aberrations were specific to carcinomas, those with the highest frequency in carcinomas (amplification of chromosome 7, 13q, and 20q; deletion of 17p and chromosome 18; LOH of 1p, chromosome 4, 5q, 8p, 17p, chromosome 18, and 20p) were also identified in adenomas. Hierarchical clustering using chromosomal aberrations revealed three distinct subtypes. Interestingly, these subtypes were only partially dependent on tumor staging. A cluster of colorectal cancer patients with frequent chromosomal deletions had the least favorable prognosis, and a number of adenomas (n = 9) were also present in the cluster suggesting that, at least in some tumors, the chromosomal aberration pattern is determined at a very early stage of tumor formation. Finally, analysis of LOH events revealed that copy-neutral/gain LOH (CN/G-LOH) is frequent (>10%) in carcinomas at 5q, 11q, 15q, 17p, chromosome 18, 20p, and 22q. Deletion of the corresponding region is sometimes present in adenomas, suggesting that LOH at these loci may play an important role in tumor initiation.