Building personalized treatment plans for early-stage colorectal cancer patients

Adjuvant chemotherapy in stage 1 colon cancer: Patient characteristics and survival analysis from the national cancer database

BACKGROUND

A subset of patients in ACS-NCDB with stage-1 colon cancer received adjuvant chemotherapy (AC), in contrast to national guidelines. This study aimed to define this population and evaluate associations between AC and survival.

METHODS

Patients with T1-2N0 colon cancer from 2004 to 2016 were separated into AC and non-AC groups. Adverse pathological features (APF) included T2, poor differentiation, lymphovascular invasion, positive margin, and inadequate lymph nodes (<12). Cox proportional hazard models were used to estimate prognostic factors for overall survival (OS).

RESULTS

A total of 1745 of 139,857 patients (1.2 %) received AC. Receiving AC was associated with male sex (p = 0.02), uninsured (p < 0.01), low income (p = 0.02), or having ≥2 APFs (p < 0.001). In the total cohort, AC was associated with increased mortality (HR 1.14 [1.04-1.24] P < 0.01). On subset analysis, AC was associated with improved OS for patients with ≥2 APFs (log-rank P=<0.001), and decreased mortality when adjusted for covariates (HR 0.81 [0.69-0.95] P=<0.01). The most significant predictor of mortality was old age (HR 3.78 [3.67, 3.89] p ≤ 0.01), followed by higher Charlson Comorbidity Index (HR 1.73 [1.69, 1.76] (p ≤ 0.01), and higher APF score (HR 1.46 [1.42, 15.2] p ≤ 0.01).

CONCLUSION

AC was associated with decreased survival in the total cohort of stage 1 colon cancer patients, but was associated with improved survival for patients with multiple APFs.

Molecular Subclassification Based on Crosstalk Analysis Improves Prediction of Prognosis in Colorectal Cancer

The poor performance of single-gene lists for prognostic predictions in independent cohorts has limited their clinical use. Here, we employed a pathway-based approach using embedded biological features to identify reproducible prognostic markers as an alternative. We used pathway activity score, sure independence screening, and K-means clustering analyses to identify and cluster colorectal cancer patients into two distinct subgroups, G2 (aggressive) and G1 (moderate). The differences between these two groups with respect to survival, somatic mutation, pathway activity, and tumor-infiltration by immunocytes were compared. These comparisons revealed that the survival rates in the G2 subgroup were significantly reduced compared to that in the G1 subgroup; further, the mutational burden rates in several oncogenes, including KRAS, DCLK1, and EPHA5, were significantly higher in the G2 subgroup than in the G1 subgroup. The enhanced activity of the critical pathways such as MYC and epithelial-mesenchymal transition may also lead to the progression of colorectal cancer. Taken together, we established a novel prognostic classification system that offers meritorious insights into the hallmarks of colorectal cancer.

Suppression MGP inhibits tumor proliferation and reverses oxaliplatin resistance in colorectal cancer

Matrix Gla protein (MGP), an extracellular matrix protein, has been widely reported to participate in the tumorigenic process and is abnormally expressed in several tumors. However, the role of MGP in colorectal cancer (CRC) remains unknown. Chemotherapy resistance represents a significant limitation in the treatment of CRC. Here, a comprehensive bioinformatics analysis revealed that MGP, which is overexpressed in CRC, might act as one of the critical genes conferring resistance to oxaliplatin (OXA). Furthermore, we found that MGP overexpression in tumor tissue might be correlated with cancer stage and patient prognosis, consistent with the bioinformatics analysis. The upregulation of MGP may act as an independent risk factor for CRC. The knockdown of MGP or inhibition of MGP expression significantly increased the sensitivity of the CRC cell lines to OXA. Suppression of MGP may reverse OXA resistance by upregulating copper transporter 1 (CTR1) and downregulating ATP7A and ATP7B. When used in combination with OXA, the inhibition of MGP reduced cancer cell proliferation, invasion, and migration and increased cell apoptosis in vitro. Suppression of MGP or OXA treatment alone significantly inhibited tumor growth in the CRC mouse model. Additionally, we found that OXA might promote the antitumor immune response in vivo. In summary, our study is the first to provide evidence that MGP expression confers OXA chemotherapy resistance in CRC and provides novel strategies to overcome chemotherapy resistance in CRC.

Automated Generation of Personalized Shock Wave Lithotripsy Protocols: Treatment Planning Using Deep Learning

Background

Though shock wave lithotripsy (SWL) has developed to be one of the most common treatment approaches for nephrolithiasis in recent decades, its treatment planning is often a trial-and-error process based on physicians’ subjective judgement. Physicians’ inexperience with this modality can lead to low-quality treatment and unnecessary risks to patients.

Objective

To improve the quality and consistency of shock wave lithotripsy treatment, we aimed to develop a deep learning model for generating the next treatment step by previous steps and preoperative patient characteristics and to produce personalized SWL treatment plans in a step-by-step protocol based on the deep learning model.

Methods

We developed a deep learning model to generate the optimal power level, shock rate, and number of shocks in the next step, given previous treatment steps encoded by long short-term memory neural networks and preoperative patient characteristics. We constructed a next-step data set (N=8583) from top practices of renal SWL treatments recorded in the International Stone Registry. Then, we trained the deep learning model and baseline models (linear regression, logistic regression, random forest, and support vector machine) with 90% of the samples and validated them with the remaining samples.

Results

The deep learning models for generating the next treatment steps outperformed the baseline models (accuracy = 98.8%, F1 = 98.0% for power levels; accuracy = 98.1%, F1 = 96.0% for shock rates; root mean squared error = 207, mean absolute error = 121 for numbers of shocks). The hypothesis testing showed no significant difference between steps generated by our model and the top practices (P=.480 for power levels; P=.782 for shock rates; P=.727 for numbers of shocks).

Conclusions

The high performance of our deep learning approach shows its treatment planning capability on par with top physicians. To the best of our knowledge, our framework is the first effort to implement automated planning of SWL treatment via deep learning. It is a promising technique in assisting treatment planning and physician training at low cost.

Identification of Two Subgroups of FOLFOX Resistance Patterns and Prediction of FOLFOX Response in Colorectal Cancer Patients

To dissect gene expression subgroups of FOLFOX resistance colorectal cancer(CRC) and predict FOLFOX response, gene expression data of 83 stage IV CRC tumor samples (FOLFOX responder n = 42, non-responder n = 41) are used to develop a novel iterative supervised learning method IML. IML identified two mutually exclusive subgroups of CRC patients that rely on different DNA damage repair proteins and resist FOLFOX. IML was validated in two validation sets (HR = 2.6, p Value = 0.02; HR = 2.36, p value = 0.02). A subgroup of mesenchymal subtype patients benefit from FOLFOX. Different subgroups of FOLFOX nonresponders may need to be treated differently.

A novel CpG-methylation-based nomogram predicts survival in colorectal cancer

Aberrant DNA methylation is significantly associated with the prognosis of patients with colorectal cancer (CRC). Therefore, the aim of this study was to develop a CpG-methylation-based nomogram for prognostic prediction in CRC. First, 378 CRC patients with methylation data from The Cancer Genome Atlas were randomly divided into training cohort (n = 249) and test cohort (n = 129). A multistep screening strategy was performed to identify six CpG sites that were significantly associated with overall survival in the training cohort. Then, Cox regression modelling was performed to construct a prognostic signature based on the candidate CpG sites. The six-CpG signature successfully separated patients into high-risk and low-risk groups in both training and test cohorts, and its performance was superior to that of previously published methylation markers (P < 0.05). Furthermore, we established a prognostic nomogram incorporating this signature, TNM stage, and age. The nomogram exhibited better prediction for overall survival in comparison with the three independent prognostic factors in the training cohort (C-index: 0.798 vs 0.620 to 0.737; P < 0.001). In the test cohort, the performance of nomogram was also superior to that of the three independent prognostic factors (C-index: 0.715 vs 0.590 to 0.665; P < 0.05). Meanwhile, the calibration curves for survival probability showed good agreement between prediction by nomogram and actual observation in both training and test cohorts. Together, the present study provides a novel CpG-methylation-based nomogram as a promising predictor for overall survival of CRC patients, which may help improve decision-making regarding the personalized treatments of patients with CRC.

FOLFOX treatment response prediction in metastatic or recurrent colorectal cancer patients via machine learning algorithms

Early identification of metastatic or recurrent colorectal cancer (CRC) patients who will be sensitive to FOLFOX (5-FU, leucovorin and oxaliplatin) therapy is very important. We performed microarray meta-analysis to identify differentially expressed genes (DEGs) between FOLFOX responders and nonresponders in metastatic or recurrent CRC patients, and found that the expression levels of WASHC4, HELZ, ERN1, RPS6KB1, and APPBP2 were downregulated, while the expression levels of IRF7, EML3, LYPLA2, DRAP1, RNH1, PKP3, TSPAN17, LSS, MLKL, PPP1R7, GCDH, C19ORF24, and CCDC124 were upregulated in FOLFOX responders compared with nonresponders. Subsequent functional annotation showed that DEGs were significantly enriched in autophagy, ErbB signaling pathway, mitophagy, endocytosis, FoxO signaling pathway, apoptosis, and antifolate resistance pathways. Based on those candidate genes, several machine learning algorithms were applied to the training set, then performances of models were assessed via the cross validation method. Candidate models with the best tuning parameters were applied to the test set and the final model showed satisfactory performance. In addition, we also reported that MLKL and CCDC124 gene expression were independent prognostic factors for metastatic CRC patients undergoing FOLFOX therapy.

A network-based predictive gene expression signature for recurrence risks in stage II colorectal cancer

The current criteria for defining the recurrence risks of stage II colorectal cancer (CRC) are not robust; therefore, we aimed to explore novel gene signatures to predict recurrence risks and to reveal the underlying mechanisms of stage II CRC. First, the gene expression profiles of 124 patients with stage II CRC from The Cancer Genome Atlas (TCGA) database were obtained to screen differentially expressed genes (DEGs). A total of 202 DEGs, including 128 upregulated and 74 downregulated, were identified in the recurrence group (n = 24) compared to the nonrecurrence group (n = 100). Furthermore, the top 5 DEGs (ZNF561, WFS1, SLC2A1, MFI2, and PTGR1) were identified by random forest variable hunting, and four (ZNF561, WFS1, SLC2A1, and PTGR1) were selected to create a four‐gene recurrent model (GRM), with an area under the curve (AUC) of 0.882 according to the receiver operating characteristic curve, and the robust diagnostic effectiveness of the GRM was further validated with another gene expression profiling dataset (GSE12032), with an AUC of 0.943. The diagnostic effectiveness of the GRM regarding recurrence was associated with poor disease‐free survival in all stages of CRC. In addition, gene ontology functional annotation and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses revealed 18 enriched functions and 6 enriched pathways. Four genes, ABCG2, CACNA1F, CYP19A1, and TF, were identified as hub genes by the protein‐protein interaction network, which further validated that these genes were correlated with a poor pathologic stage and overall survival in all stages of CRC. In conclusion, the GRM can effectively classify stage II CRC into groups of high and low risks of recurrence, thereby making up for the prognostic value of the traditional clinicopathological risk factors defined by the National Comprehensive Cancer Network guidelines. The hub genes may be useful therapeutic targets for recurrence. Thus, the GRM and hub genes could offer clinical value in directing individualized and precision therapeutic regimens for stage II CRC patients.

Chemotherapy-generated cell debris stimulates colon carcinoma tumor growth via osteopontin

Colon cancer recurrence after therapy, such as 5-fluorouracil (5-FU), remains a challenge in the clinical setting. Chemotherapy reduces tumor burden by inducing cell death; however, the resulting dead tumor cells, or debris, may paradoxically stimulate angiogenesis, inflammation, and tumor growth. Here, we demonstrate that 5-FU–generated colon carcinoma debris stimulates the growth of a subthreshold inoculum of living tumor cells in subcutaneous and orthotopic models. Debris triggered the release of osteopontin (OPN) by tumor cells and host macrophages. Both coinjection of debris and systemic treatment with 5-FU increased plasma OPN levels in tumor-bearing mice. RNA expression levels of secreted phosphoprotein 1, the gene that encodes OPN, correlate with poor prognosis in patients with colorectal cancer and are elevated in chemotherapy-treated patients who experience tumor recurrence vs. no recurrence. Pharmacologic and genetic ablation of OPN inhibited debris-stimulated tumor growth. Systemic treatment with a combination of a neutralizing OPN antibody and 5-FU dramatically inhibited tumor growth. These results demonstrate a novel mechanism of tumor progression mediated by OPN released in response to chemotherapy-generated tumor cell debris. Neutralization of debris-stimulated OPN represents a potential therapeutic strategy to overcome the inherent limitation of cytotoxic therapies as a result of the generation of cell debris.—Chang, J., Bhasin, S. S., Bielenberg, D. R., Sukhatme, V. P., Bhasin, M., Huang, S., Kieran, M. W., Panigrahy, D. Chemotherapy-generated cell debris stimulates colon carcinoma tumor growth via osteopontin.