The association of single nucleotide polymorphisms (SNPs) with breast density and breast cancer survival: the Malmö Diet and Cancer Study

Breast cancer prediction using different machine learning methods applying multi factors

OBJECTIVE

Breast cancer (BC) is a multifactorial disease and is one of the most common cancers globally. This study aimed to compare different machine learning (ML) techniques to develop a comprehensive breast cancer risk prediction model based on features of various factors.

METHODS

The population sample contained 810 records (115 cancer patients and 695 healthy individuals). 45 attributes out of 85 were selected based on the opinion of experts. These selected attributes are in genetic, biochemical, biomarker, gender, demographic and pathological factors. 13 Machine learning models were trained with proposed attributes and coefficient of attributes and internal relationships were calculated.

RESULT

Compared to other methods random forest (RF) has higher performance (accuracy 99.26%, precision 99%, and area under the curve (AUC) 99%). The results of assessing the impact and correlation of variables using the RF method based on PCA indicated that pathology, biomarker, biochemistry, gene, and demographic factors with a coefficient of 0.35, 0.23, 0.15, 0.14, and 0.13 respectively, affected the risk of BC (r2 = 0.54).

CONCLUSION

Breast cancer has several risk factors. Medical experts use these risk factors for early diagnosis. Therefore, identifying related risk factors and their effect can increase the accuracy of diagnosis. Considering the broad features for predicting breast cancer leads to the development of a comprehensive prediction model. In this study, using RF technique a breast cancer prediction model with 99.3% accuracy was developed based on multifactorial features.

The PANcreatic Disease ReseArch (PANDoRA) consortium: Ten years' experience of association studies to understand the genetic architecture of pancreatic cancer

Pancreatic cancer has an incidence that almost matches its mortality. Only a small number of risk factors and 33 susceptibility loci have been identified. so Moreover, the relative rarity of pancreatic cancer poses significant hurdles for research aimed at increasing our knowledge of the genetic mechanisms contributing to the disease. Additionally, the inability to adequately power research questions prevents small monocentric studies from being successful. Several consortia have been established to pursue a better understanding of the genetic architecture of pancreatic cancers. The Pancreatic disease research (PANDoRA) consortium is the largest in Europe. PANDoRA is spread across 12 European countries, Brazil and Japan, bringing together 29 basic and clinical research groups. In the last ten years, PANDoRA has contributed to the discovery of 25 susceptibility loci, a feat that will be instrumental in stratifying the population by risk and optimizing preventive strategies.

Effect of SMAD4 gene polymorphism on breast cancer risk in Bangladeshi women

Background

Breast cancer, one of the most prevalent cancer types among women worldwide as well as in Bangladesh, is the leading cause of cancer death in women throughout the globe. The risk of breast cancer development was found to be associated with genetic polymorphism according to several studies. As a convenient prognostic marker, a biomarker helps to identify disease progression, can lead to an effective therapeutic strategy, development of prognostic marker is very important for any cancer to initiate treatment strategy early to increase the possibility of the success rate of the treatment along with reduction of the treatment cost. This study aims to establish the correlation between polymorphism of SMAD4 rs10502913 and risk of breast cancer development in Bangladeshi women. This study was conducted on 70 breast cancer patients and 60 healthy volunteers through blood sample collection followed by DNA separation between the intervals of August 2019–October 2019. The collected DNA sample was arranged for the RFLP analysis of a PCR amplified fragments followed by gel electrophoresis. The obtained data was analyzed by structured multinomial logistic regression model.

Results

Obtained different fragment size after gel electrophoresis indicated different genotypes in this experiment. Our findings demonstrated that mutant homozygous A/A genotype, plays a significant role in breast cancer development among Bangladeshi women (P = 0.006, OR = 4.9626, 95% CI = 1.9980–12.3261) compared to the reference homozygous G/G genotype. Moreover, heterozygous G/A genotype was also found to be significantly associated with the risk of breast cancer development (P = 0.0252, OR = 2.6574, CI = 1.1295–6.2525). Considering the A/A genotype and G/A genotype combined, it also indicates a strong association of breast cancer development in Bangladeshi women (P = 0.008, OR = 3.5630, CI = 1.6907–7.5068).

Conclusion

Our study indicated a novel association between SMAD4 (rs10502913) polymorphism and increased risk of breast cancer development in Bangladeshi women.

Estrogen Receptor Gene Polymorphism as a Possible Genetic Risk Factor for Treatment Response in ER-Positive Breast Cancer Patients

Estrogen receptor-α (ESR1) single nucleotide polymorphisms (SNPs) have been related to breast cancer (BC) susceptibility. In this retrospective study we investigated ESR1 SNPs in association with survival and treatment response in BC patients. Seven ESR1 SNPs were genotyped using TaqMan probe assay in 100 formalin-fixed paraffin embedded blocks of Egyptian ER⁺BC patients. Log-binomial regression was used to assess the association of 5 ESR1 SNPs with relative risk of non-response to adjuvant-hormonal treatment. We compared the performance of five machine learning classification models for prediction of treatment response. Predictive models were developed using rs1801132, rs2228480, and rs9322354 that were significantly associated with increased risk for non-response along with the relevant clinical features. Survival analysis was performed to detect prognostic significance of ESR1 SNPs in ESR⁺BC patients. rs1801132 (C), rs2228480 (A), and rs9322354 (G) minor alleles significantly increased the risk of non-response to tamoxifen by more than 81, 84, and 117%, respectively, in ER⁺BC patients on anthracycline/anthracycline-taxanes-based chemotherapy. Multivariate Cox regression survival analysis revealed that rs1801132 (C) and large tumor size were independent predictors for poor survival outcome in ER⁺BC. The best response predictive model was a combination random forest, K-nearest neighbor, and decision tree having an area under the curve of 0.94 and an accuracy of 90.8%. Our proposed predictive model based on ESR1 rs1801132, rs2228480, and rs9322354 SNPs represents a promising genetic risk stratification for selection patients who could benefit from tamoxifen therapy in such a way that might facilitate personalized medicine required to improve ER⁺BC patients' outcome.

The epidemiologic factors associated with breast density: A review

In recent years, some studies have evaluated the epidemiologic factors associated with breast density. However, the variant and inconsistent results exist. In addition, breast density has been proved to be a significant risk factor associated with breast cancer. Our review summarized the published studies and emphasized the crucial factors including epidemiological factors associated with breast density. In addition, we also discussed the potential reasons for the discrepant results with risk factors. To decrease the incidence and mortality rates for breast cancer, in clinical practice, breast density should be included for clinical risk models in addition to epidemiological factors, and physicians should get more concentrate on those women with risk factors and provide risk-based breast cancer screening regimens.

"Association of mammographic features with molecular breast tumor profiles"

PURPOSE

Mammographic density and tumor appearance are breast cancer prognostic factors. Conceivably, mammographic features are macroscopic reflections of tumor´s molecular composition, but to an unknown extent. Our aim was to study associations of mammographic features with molecular tumor profiles.

METHODS

Invasive breast cancers (2007-2016) in Malmö Diet and Cancer Study (MDCS) for which there were tumor RNA-sequencing analyses within Sweden Cancerome Analysis Network - Breast (SCAN-B) (n=102) or All Breast Cancer in Malmö (ABIM) (n=50) were identified. Density (fatty vs. dense), tumor appearance (mass vs. spiculation), and intrinsic subtypes were registered. Differences in gene/metagene expression and Microenvironment Cell Population Counter were analyzed with R. Overall survival was used as endpoint.

RESULTS

No gene expression differences between density groups was observed. In one cohort (but not the other), Luminal A tumors associated with fatty breasts. For spiculation vs. mass, (p<0.01, t-test) 86 genes were differentially expressed; only one gene was differentially expressed comparing density. Gene set enrichment analysis showed genes highly expressed in spiculated tumors were enriched for extracellular matrix-associated genes whereas genes highly expressed with masses were associated with proliferation. A spiculation metagene, based on differentially expressed genes, showed association with estrogen receptor positivity, lower grade, and improved survival, but it was not an independent prognostic factor.

CONCLUSION

There are clear differences in molecular composition between breast tumors with a spiculated appearance vs. a mass as the dominant tumor appearance. However, there are no apparent molecular differences related to the density of the breast in which the tumor has arisen.