Energy balance-related factors in childhood and adolescence and risk of colorectal cancer expressing different levels of proteins involved in the Warburg-effect.

Early-life (childhood to adolescence) energy balance-related factors (height, energy restriction, BMI) have been associated with adult colorectal cancer (CRC) risk. Warburg-effect activation via PI3K/Akt-signaling might explain this link. We investigated whether early-life energy balance-related factors were associated with risk of Warburg-subtypes in CRC. We used immunohistochemistry for six proteins involved in the Warburg-effect (LDHA, GLUT1, MCT4, PKM2, P53, and PTEN) on tissue microarrays of 2399 incident CRC cases from the prospective Netherlands Cohort Study (NLCS). Expression levels of all proteins were combined into a pathway-based sum score and categorized into three Warburg-subtypes (Warburg-low/-moderate/-high). Multivariable Cox-regression analyses were used to estimate associations of height, energy restriction proxies (exposure to Dutch Hunger Winter; Second World War [WWII]; Economic Depression) and adolescent BMI with Warburg-subtypes in CRC. Height was positively associated with colon cancer in men, regardless of Warburg-subtypes, and with Warburg-low colon and Warburg-moderate rectal cancer in women. Energy restriction during the Dutch Hunger Winter was inversely associated with colon cancer in men, regardless of Warburg-subtypes. In women, energy restriction during the Hunger Winter and WWII was inversely associated with Warburg-low colon cancer, whereas energy restriction during the Economic Depression was positively associated with Warburg-high colon cancer. Adolescent BMI was positively associated with Warburg-high colon cancer in men, and Warburg-moderate rectal cancer in women. In conclusion, the Warburg-effect seems to be involved in associations of adolescent BMI with colon cancer in men, and of energy restriction during the Economic Depression with colon cancer in women. Further research is needed to validate these results.

Energy balance-related factors in childhood and adolescence and risk of colorectal cancer based on KRAS, PIK3CA, and BRAF mutations and MMR status.

KRAS mutations (KRASmut ), PIK3CAmut , BRAFmut , and deficient DNA mismatch repair (dMMR) have been associated with the Warburg effect. We previously reported differential associations between early-life energy balance-related factors (height, energy restriction, body mass index [BMI]) and colorectal cancer (CRC) subtypes based on the Warburg effect. We now investigated associations of early-life energy balance-related factors and the risk of CRC subgroups based on mutation and MMR status. Data from the Netherlands Cohort Study was used. KRASmut , PIK3CAmut, BRAFmut, and MMR status were available for 2349 CRC cases, and complete covariate data for 1934 cases and 3911 subcohort members. Multivariable-adjusted Cox regression was used to estimate associations of height, energy restriction proxies (exposure to Dutch Hunger Winter, Second World War, Economic Depression), and early adult BMI (age 20 years) with risk of CRC based on individual molecular features and combinations thereof (all-wild-type+MMR-proficient [pMMR]; any-mutation/dMMR). Height was positively associated with any-mutation/dMMR CRC but not all-wild-type+pMMR CRC, with the exception of rectal cancer in men, and with heterogeneity in associations observed for colon cancer in men (p-heterogeneity = 0.049) and rectal cancer in women (p-heterogeneity = 0.014). Results on early-life energy restriction proxies in relation to the risk of CRC subgroups did not show clear patterns. Early adult BMI was positively, but not significantly, associated with KRASmut colon cancer in men and with BRAFmut and dMMR colon cancer in women. Our results suggest a role of KRASmut , PIK3CAmut , BRAFmut , and dMMR in the etiological pathway between height and CRC risk. KRASmut might potentially play a role in associations of early adult BMI with colon cancer risk in men, and BRAFmut and dMMR in women.

Deep learning detects genetic alterations in cancer histology generated by adversarial networks.

Deep learning can detect microsatellite instability (MSI) from routine histology images in colorectal cancer (CRC). However, ethical and legal barriers impede sharing of images and genetic data, hampering development of new algorithms for detection of MSI and other biomarkers. We hypothesized that histology images synthesized by conditional generative adversarial networks (CGANs) retain information about genetic alterations. To test this, we developed a 'histology CGAN' which was trained on 256 patients (training cohort 1) and 1457 patients (training cohort 2). The CGAN synthesized 10 000 synthetic MSI and non-MSI images which contained a range of tissue types and were deemed realistic by trained observers in a blinded study. Subsequently, we trained a deep learning detector of MSI on real or synthetic images and evaluated the performance of MSI detection in a held-out set of 142 patients. When trained on real images from training cohort 1, this system achieved an area under the receiver operating curve (AUROC) of 0.742 [0.681, 0.854]. Training on the larger cohort 2 only marginally improved the AUROC to 0.757 [0.707, 0.869]. Training on purely synthetic data resulted in an AUROC of 0.743 [0.658, 0.801]. Training on both real and synthetic data further increased AUROC to 0.777 [0.715, 0.821]. We conclude that synthetic histology images retain information reflecting underlying genetic alterations in colorectal cancer. Using synthetic instead of real images to train deep learning systems yields non-inferior classifiers. This approach can be used to create large shareable data sets or to augment small data sets with rare molecular features. © 2021 The Authors. The Journal of Pathology published by John Wiley & Sons, Ltd. on behalf of The Pathological Society of Great Britain and Ireland.

Clinical-Grade Detection of Microsatellite Instability in Colorectal Tumors by Deep Learning.

BACKGROUND & AIMS: Microsatellite instability (MSI) and mismatch-repair deficiency (dMMR) in colorectal tumors are used to select treatment for patients. Deep learning can detect MSI and dMMR in tumor samples on routine histology slides faster and less expensively than molecular assays. However, clinical application of this technology requires high performance and multisite validation, which have not yet been performed. METHODS: We collected H&E-stained slides and findings from molecular analyses for MSI and dMMR from 8836 colorectal tumors (of all stages) included in the MSIDETECT consortium study, from Germany, the Netherlands, the United Kingdom, and the United States. Specimens with dMMR were identified by immunohistochemistry analyses of tissue microarrays for loss of MLH1, MSH2, MSH6, and/or PMS2. Specimens with MSI were identified by genetic analyses. We trained a deep-learning detector to identify samples with MSI from these slides; performance was assessed by cross-validation (N = 6406 specimens) and validated in an external cohort (n = 771 specimens). Prespecified endpoints were area under the receiver operating characteristic (AUROC) curve and area under the precision-recall curve (AUPRC). RESULTS: The deep-learning detector identified specimens with dMMR or MSI with a mean AUROC curve of 0.92 (lower bound, 0.91; upper bound, 0.93) and an AUPRC of 0.63 (range, 0.59-0.65), or 67% specificity and 95% sensitivity, in the cross-validation development cohort. In the validation cohort, the classifier identified samples with dMMR with an AUROC of 0.95 (range, 0.92-0.96) without image preprocessing and an AUROC of 0.96 (range, 0.93-0.98) after color normalization. CONCLUSIONS: We developed a deep-learning system that detects colorectal cancer specimens with dMMR or MSI using H&E-stained slides; it detected tissues with dMMR with an AUROC of 0.96 in a large, international validation cohort. This system might be used for high-throughput, low-cost evaluation of colorectal tissue specimens.

Relationship between the Warburg effect in tumour cells and the tumour microenvironment in colorectal cancer patients: Results from a large multicentre study.

Colorectal cancer (CRC) remains one of the most prevalent and deadly cancers worldwide. The tumour-node-metastasis stage (TNM) is currently the most clinically important tool to predict prognosis for CRC patients. However, patients with the same TNM stage can have different prognoses. The metabolic status of tumour cells (Warburg-subtype) has been proposed as potential prognostic factor in CRC. However, potential biological mechanisms underlying the relationship between Warburg-subtype and prognosis have not been investigated in detail. One potential mechanism could be that the metabolic status of tumour cells affects the tumour microenvironment (TME). Our objective was to investigate the relationship between Warburg-subtypes and the TME. Haematoxylin/Eosin stained tumour tissue microarray cores from 2171 CRC patients from the Netherlands Cohort Study were semi quantitatively assessed for tumour infiltrating lymphocytes (TILs) and relative tumour stroma content. 5745 cores were assessed by putting each core in one of four categories for both TILs and stroma. The relationship between Warburg-subtype, TILs, and tumour stroma content was investigated. The frequency of CRC in the different TIL categories was (n, %): very low (2538, 44.2), low (2463, 42.9), high (722, 12.6), and very high (22, 0.4). The frequency of CRC in the different tumour stroma content categories was: ≤ 25% (2755, 47.9), > 25% ≤ 50% (1553, 27) > 50% ≤ 75% (905, 15.8), and > 75% (532, 9.3). There was neither an association between Warburg-subtype and tumour stroma content (p = 0.229) nor between Warburg-subtype and TILs (p = 0.429). This is the first study to investigate the relationship between Warburg-subtypes and the TME in a large population-based series of CRC patients. Our data suggest that the prognostic value of Warburg-subtypes cannot be directly attributed to differences in TILs or tumour stroma content. Our results require confirmation in an independent series.

Association between adjuvant therapy and survival in colorectal cancer patients according to metabolic Warburg-subtypes.

PURPOSE: Tumor location and tumor node metastasis (TNM) stage guide treatment decisions in colorectal cancer (CRC) patients. However, patients with the same disease stage do not benefit equally from adjuvant therapy. Hence, there remains an urgent clinical need to identify prognostic and/or predictive biomarker(s) to personalize treatment decisions. In this exploratory study, we investigated whether our previously defined metabolic Warburg-subtypes can predict which CRC patients might derive survival benefit from adjuvant therapy. METHODS: Information regarding treatment (surgery only: n = 1451; adjuvant radiotherapy: n = 82; or adjuvant chemotherapy: n = 260) and Warburg-subtype (Warburg-low: n = 485, -moderate: n = 641, or -high: n = 667) was available for 1793 CRC patients from the Netherlands Cohort Study (NLCS). Kaplan-Meier curves and Cox regression models were used to investigate survival benefit from adjuvant therapy compared to surgery-only for the different Warburg-subtypes. RESULTS: Patients with Warburg-moderate CRC (HRCRC-specific 0.64; 95% CI 0.47-0.86, HRoverall 0.61; 95% CI 0.47-0.80), and possibly Warburg-high CRC (HRCRC-specific 0.86; 95% CI 0.65-1.14, HRoverall 0.82; 95% CI 0.64-1.05), had survival benefit from adjuvant therapy. No survival benefit was observed for patients with Warburg-low CRC (HRCRC-specific 1.07; 95% CI 0.76-1.52, HRoverall 0.95; 95% CI 0.70-1.30). There was a significant interaction between Warburg-subtype and adjuvant therapy for CRC-specific survival (p = 0.049) and overall survival (p = 0.035). CONCLUSION: Our results suggest that Warburg-subtypes may predict survival benefit from adjuvant therapy in CRC patients. A survival benefit from adjuvant therapy was observed for patients with Warburg-moderate and possibly Warburg-high CRC, but not for patients with Warburg-low CRC. Future prospective studies are necessary to validate our findings.

Association between mutational subgroups, Warburg-subtypes, and survival in patients with colorectal cancer.

BACKGROUND: Previous research suggests that Warburg-subtypes are related to potentially important survival differences in colorectal cancer (CRC) patients. In the present study, we investigated whether mutational subgroups based on somatic mutations in RAS, BRAF, PIK3CA, and MET, which are known to promote the Warburg-effect, as well as mismatch repair (MMR) status, hold prognostic value in CRC. In addition, we investigated whether Warburg-subtypes provide additional prognostic information, independent of known prognostic factors like TNM stage. METHODS: CRC patients (n = 2344) from the prospective Netherlands Cohort Study (NLCS) were classified into eight mutually exclusive mutational subgroups, based on observed mutations in RAS, BRAF, PIK3CA, and MET, and MMR status: All-wild-type + MMRproficient , KRASmut  + MMRproficient , KRASmut  + PIK3CAmut  + MMRproficient , PIK3CAmut  + MMRproficient , BRAFmut  + MMRproficient , BRAFmut  + MMRdeficient , other + MMRproficient , and other + MMRdeficient . Kaplan-Meier curves and Cox regression models were used to investigate associations between mutational subgroups and survival, as well as associations between our previously established Warburg-subtypes and survival within these mutational subgroups. RESULTS: Compared to patients with all-wild-type + MMRproficient CRC, patients with KRASmut  + MMRproficient , KRASmut  + PIK3CAmut  + MMRproficient , BRAFmut  + MMRproficient , or other + MMRproficient CRC had a statistically significant worse survival (HRCRC-specific ranged from 1.29 to 1.88). In contrast, patients with other + MMRdeficient CRC had the most favorable survival (HRCRC-specific 0.48). No statistically significant survival differences were observed for the Warburg-subtypes within mutational subgroups. CONCLUSION: Our results highlight the prognostic potential of mutational subgroups in CRC. Warburg-subtypes did not provide additional prognostic information within these mutational subgroups. Future larger-scale prospective studies are necessary to validate our findings and to examine the potential clinical utility of CRC subtyping based on mutational subgroups.

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.

Expression of proteins associated with the Warburg-effect and survival in colorectal cancer.

Previous research has suggested that the expression of proteins related to the Warburg effect may have prognostic value in colorectal cancer (CRC), but results remain inconsistent. Our objective was to investigate the relationship between Warburg-subtypes and patient survival in a large population-based series of CRC patients. In the present study, we investigated the expression of six proteins related to the Warburg effect (LDHA, GLUT1, MCT4, PKM2, p53, PTEN) by immunohistochemistry on tissue microarrays (TMAs) from 2,399 incident CRC patients from the prospective Netherlands Cohort Study. Expression levels of the six proteins were combined into a pathway-based sum-score and patients were categorised into three Warburg-subtypes (low/moderate/high). The associations between Warburg-subtypes and CRC-specific and overall survival were investigated using Kaplan-Meier curves and Cox regression models. CRC patients were classified as Warburg-low (n = 695, 29.0%), Warburg-moderate (n = 858, 35.8%) or Warburg-high (n = 841, 35.1%). Patients with Warburg-high CRC had the poorest CRC-specific [hazard ratio (HR) 1.17; 95% CI 1.00-1.38] and overall survival (HR 1.19; 95% CI 1.05-1.35), independent of known prognostic factors. In stratified analyses, this was particularly true for patients with tumour-node-metastasis (TNM) stage III CRC (HRCRC-specific 1.45; 95% CI 1.10-1.92 and HRoverall 1.47; 95% CI 1.15-1.87), and cancers located in the rectum (HRoverall 1.56; 95% CI 1.15-2.13). To our knowledge, this is the first study to identify the prognostic value of immunohistochemistry-based Warburg-subtypes in CRC. Our data suggest that Warburg-subtypes are related to potentially important differences in CRC survival. Further research is required to validate our findings and to investigate the potential clinical utility of these Warburg-subtypes in CRC.

Energy Balance-Related Factors and Risk of Colorectal Cancer Expressing Different Levels of Proteins Involved in the Warburg Effect.

BACKGROUND: Energy balance-related factors [body mass index (BMI), waist circumference, physical activity] have been associated with colorectal cancer risk. Warburg effect activation via PI3K/Akt signaling is one of the proposed mechanisms. We investigated whether energy balance-related factors were associated with risk of Warburg subtypes in colorectal cancer. METHODS: We investigated this using immunohistochemistry for six proteins involved in the Warburg effect (LDHA, GLUT1, MCT4, PKM2, P53, PTEN) on tissue microarrays of 2,399 incident colorectal cancer cases from the prospective Netherlands Cohort Study (ntotal = 120,852; nsubcohort = 5,000; aged 55-69 in 1986; 20.3 years follow-up). Data analyses included 3,911 subcohort members and 1,972 colorectal cancer cases with complete covariate data. Expression levels of all proteins were combined into a pathway-based sum score and categorized into three "Warburg subtypes" (Warburg-low/moderate/high). Multivariable Cox regression analyses were used to estimate associations of BMI, clothing size (waist circumference proxy), and physical activity with Warburg subtypes in colorectal cancer. RESULTS: BMI and clothing size were positively associated with Warburg-moderate and Warburg-high colon cancer risk in men (Pheterogeneity = 0.192). In women, clothing size was positively associated with Warburg-low and Warburg-high colon cancer (Pheterogeneity = 0.005). Nonoccupational physical activity was inversely associated with Warburg-low and Warburg-moderate colon cancer in women (Pheterogeneity = 0.045), but positively associated with Warburg-high rectal cancer in men (Pheterogeneity = 0.089). CONCLUSIONS: The Warburg effect might be involved in associations between adiposity and colon cancer risk, though additional mechanisms could be at play in women as well. The inverse association between physical activity and colon cancer might be explained by mechanisms other than the Warburg effect. IMPACT: Further research is needed to reproduce these results and investigate possible additional mechanisms.

Energy balance-related factors and risk of colorectal cancer based on KRAS, PIK3CA, and BRAF mutations and MMR status.

INTRODUCTION: KRAS mutations (KRASmut), PIK3CAmut, BRAFmut, and mismatch repair deficiency (dMMR) have been associated with the Warburg-effect. We previously observed differential associations between energy balance-related factors (BMI, clothing-size, physical activity) and colorectal cancer (CRC) subtypes based on the Warburg-effect. We now investigated whether associations between energy balance-related factors and risk of CRC differ between subgroups based on mutation and MMR status. METHODS: Information on molecular features was available for 2349 incident CRC cases within the Netherlands Cohort Study (NLCS), with complete covariate data available for 1934 cases and 3911 subcohort members. Multivariable-adjusted Cox-regression was used to estimate associations of energy balance-related factors with risk of CRC based on individual molecular features (KRASmut; PIK3CAmut; BRAFmut; dMMR) and combinations thereof (all-wild-type + MMR-proficient (pMMR); any-mutation/dMMR). RESULTS: In men, BMI and clothing-size were positively associated with risk of colon, but not rectal cancer, regardless of molecular features subgroups; the strongest associations were observed for PIK3CAmut colon cancer. In women, however, BMI and clothing-size were only associated with risk of KRASmut colon cancer (p-heterogeneityKRASmut versus all-wild-type+pMMR = 0.008). Inverse associations of non-occupational physical activity with risk of colon cancer were strongest for any-mutation/dMMR tumors in men and women, and specifically for PIK3CAmut tumors in women. Occupational physical activity was inversely associated with both combination subgroups of colon cancer in men. CONCLUSION: In men, associations did not vary according to molecular features. In women, a role of KRAS mutations in the etiological pathway between adiposity and colon cancer is suggested, and of PIK3CA mutations between physical activity and colon cancer.

Validity and Reproducibility of Immunohistochemical Scoring by Trained Non-Pathologists on Tissue Microarrays.

BACKGROUND: Scoring of immunohistochemistry (IHC) staining is often done by non-pathologists, especially in large-scale tissue microarray (TMA)-based studies. Studies on the validity and reproducibility of scoring results from non-pathologists are limited. Therefore, our main aim was to assess interobserver agreement between trained non-pathologists and an experienced histopathologist for three IHC markers with different subcellular localization (nucleus/membrane/cytoplasm). METHODS: Three non-pathologists were trained in recognizing adenocarcinoma and IHC scoring by a senior histopathologist. Kappa statistics were used to analyze interobserver and intraobserver agreement for 6,249 TMA cores from a colorectal cancer series. RESULTS: Interobserver agreement between non-pathologists (independently scored) and the histopathologist was "substantial" for nuclear and membranous IHC markers (κrange = 0.67-0.75 and κrange = 0.61-0.69, respectively), and "moderate" for the cytoplasmic IHC marker (κrange = 0.43-0.57). Scores of the three non-pathologists were also combined into a "combination score" (if at least two non-pathologists independently assigned the same score to a core, this was the combination score). This increased agreement with the pathologist (κnuclear = 0.74; κmembranous = 0.73; κcytopasmic = 0.57). Interobserver agreement between non-pathologists was "substantial" (κnuclear = 0.78; κmembranous = 0.72; κcytopasmic = 0.61). Intraobserver agreement of non-pathologists was "substantial" to "almost perfect" (κnuclear,range = 0.83-0.87; κmembranous,range = 0.75-0.82; κcytopasmic = 0.69). Overall, agreement was lowest for the cytoplasmic IHC marker. CONCLUSIONS: This study shows that adequately trained non-pathologists are able to generate reproducible IHC scoring results, that are similar to those of an experienced histopathologist. A combination score of at least two non-pathologists yielded optimal results. IMPACT: Non-pathologists can generate reproducible IHC results after appropriate training, making analyses of large-scale molecular pathological epidemiology studies feasible within an acceptable time frame.

Mitochondrial Dysfunction Inhibits Hypoxia-Induced HIF-1α Stabilization and Expression of Its Downstream Targets.

mtDNA variations often result in bioenergetic dysfunction inducing a metabolic switch toward glycolysis resulting in an unbalanced pH homeostasis. In hypoxic cells, expression of the tumor-associated carbonic anhydrase IX (CAIX) is enhanced to maintain cellular pH homeostasis. We hypothesized that cells with a dysfunctional oxidative phosphorylation machinery display elevated CAIX expression levels. Increased glycolysis was observed for cytoplasmic 143B mutant hybrid (m.3243A>G, >94.5%) cells (p < 0.05) and 143B mitochondrial DNA (mtDNA) depleted cells (p < 0.05). Upon hypoxia (0.2%, 16 h), genetic or pharmacological oxidative phosphorylation (OXPHOS) inhibition resulted in decreased CAIX (p < 0.05), vascular endothelial growth factor (VEGF) and hypoxia-inducible factor 1-alpha (HIF-1α) expression levels. Reactive oxygen species (ROS) and prolyl-hydroxylase 2 (PHD2) levels could not explain these observations. In vivo, tumor take (>500 mm3) took longer for mutant hybrid xenografts, but growth rates were comparable with control tumors upon establishment. Previously, it has been shown that HIF-1α is responsible for tumor establishment. In agreement, we found that HIF-1α expression levels and the pimonidazole-positive hypoxic fraction were reduced for the mutant hybrid xenografts. Our results demonstrate that OXPHOS dysfunction leads to a decreased HIF-1α stabilization and subsequently to a reduced expression of its downstream targets and hypoxic fraction in vivo. In contrast, hypoxia-inducible factor 2-alpha (HIF-2α) expression levels in these xenografts were enhanced. Inhibition of mitochondrial function is therefore an interesting approach to increase therapeutic efficacy in hypoxic tumors.