Genetic analysis of radiation-specific biomarkers in sinonasal squamous cell carcinomas

Development of a machine learning framework for radiation biomarker discovery and absorbed dose prediction

Background

Molecular radiation biomarkers are an emerging tool in radiation research with applications for cancer radiotherapy, radiation risk assessment, and even human space travel. However, biomarker screening in genome-wide expression datasets using conventional tools is time-consuming and underlies analyst (human) bias. Machine Learning (ML) methods can improve the sensitivity and specificity of biomarker identification, increase analytical speed, and avoid multicollinearity and human bias.

Aim

To develop a resource-efficient ML framework for radiation biomarker discovery using gene expression data from irradiated normal tissues. Further, to identify biomarker panels predicting radiation dose with tissue specificity.

Methods

A strategic search in the Gene Expression Omnibus database identified a transcriptomic dataset (GSE44762) for normal tissues radiation responses (murine kidney cortex and medulla) suited for biomarker discovery using an ML approach. The dataset was pre-processed in R and separated into train and test data subsets. High computational cost of Genetic Algorithm/k-Nearest Neighbor (GA/KNN) mandated optimization and 13 ML models were tested using the caret package in R. Biomarker performance was evaluated and visualized via Principal Component Analysis (PCA) and dose regression. The novelty of ML-identified biomarker panels was evaluated by literature search.

Results

Caret-based feature selection and ML methods vastly improved processing time over the GA approach. The KNN method yielded overall best performance values on train and test data and was implemented into the framework. The top-ranking genes were Cdkn1a, Gria3, Mdm2 and Plk2 in cortex, and Brf2, Ccng1, Cdkn1a, Ddit4l, and Gria3 in medulla. These candidates successfully categorized dose groups and tissues in PCA. Regression analysis showed that correlation between predicted and true dose was high with R² of 0.97 and 0.99 for cortex and medulla, respectively.

Conclusion

The caret framework is a powerful tool for radiation biomarker discovery optimizing performance with resource-efficiency for broad implementation in the field. The KNN-based approach identified Brf2, Ddit4l, and Gria3 mRNA as novel candidates that have been uncharacterized as radiation biomarkers to date. The biomarker panel showed good performance in dose and tissue separation and dose regression. Further training with larger cohorts is warranted to improve accuracy, especially for lower doses.

The MUSES∗: a prognostic study on 1360 patients with sinonasal cancer undergoing endoscopic surgery-based treatment: ∗MUlti-institutional collaborative Study on Endoscopically treated Sinonasal cancers

BACKGROUND

Over the last 2 decades, transnasal endoscopic surgery (TES) has become the most frequently employed surgical technique to treat sinonasal malignancies. The rarity and heterogeneity of sinonasal cancers have hampered large non-population-based analyses.

METHODOLOGY

All patients receiving TES-including treatment between 1995 and 2021 in 5 referral hospitals were included. A prognostic study was performed, and multivariable models were transformed into nomograms. Training and validation sets were based on results from 3 European and 2 non-European centres, respectively.

RESULTS

The training and validation set included 940 and 420 patients, respectively. The mean age at surgery, primary-versus-recurrent presentation, histology distribution, type of surgery, T category and type of adjuvant treatment were differently distributed in the training and validation set. In the training set, 5-year overall survival and recurrence-free survival with a 95%-confidence interval were 72.7% (69.5-76.0%) and 66.4% (63.1-69.8%), respectively, significantly varying with histology. At multivariable analyses, age, gender, previous treatment, the extent of resection on the cranial, lateral and posterolateral axes, grade/subtype, T category, nodal status, margin status and adjuvant treatment were all associated with different prognostic outcomes, displaying a heterogeneous significance and effect size according to histology. The internal and external validation of nomograms was satisfactory (optimism-corrected C-index >0.7 and cumulative area under curve >0.7) for all histologies but mucosal melanoma.

CONCLUSIONS

Outcomes of TES-based treatment of sinonasal cancers vary substantially with histology. This large, non-population-based study provides benchmark data on the prognosis of sinonasal cancers that are deemed suitable for treatment including TES.

Increased Collagen Type V α2 (COL5A2) in Colorectal Cancer is Associated with Poor Prognosis and Tumor Progression

Purpose

Colorectal cancer (CRC) is the third most common cancer in males and the second in females worldwide with very poor prognosis. Extracellular matrix proteins like collagens play important roles in cancer progression. Collagen type V α2 (COL5A2) is increased in several cancers but its role in cancer remains unclear.

Methods

COL5A2 expression was evaluated by interrogation of public Oncomine gene microarray datasets and immunohistochemistry (IHC) analyses of two tissue microarrays containing 180 paired CRC cases. Survival analysis was performed using Kaplan–Meier survival curve and Cox proportional hazards regression methods. COL5A2 was ectopically expressed in CRC cells, and the cell proliferation was measured using the methylthiazolyldiphenyl-tetrazolium bromide (MTT) method.

Results

COL5A2 gene was significantly upregulated in the most types of CRC comparing with the normal counterparts. The mRNA expression of COL5A2 was associated with cancer stages, gender, recurrence, microsatellite instability and KRAS status of CRC. COL5A2 protein increased in the cancer epithelial cells comparing with the normal counterpart and associated with age and T stage of CRC, whereas stromal expression of COL5A2 has no significant change between cancerous and normal tissues. COL5A2 gene and protein (epithelial expression) are independent risk factors and predict poor prognosis of CRC. Ectopic expression of COL5A2 drives colon cancer cell growth and upregulates WNT/β-catenin and PI3K/mTOR signaling via binding DDR1.

Conclusion

COL5A2 is a potential prognostic marker of CRC.