Novel prognostic prediction models for patients with stage IV colorectal cancer after concurrent curative resection

Machine learning for predicting liver and/or lung metastasis in colorectal cancer: A retrospective study based on the SEER database

OBJECTIVE

This study aims to establish a machine learning (ML) model for predicting the risk of liver and/or lung metastasis in colorectal cancer (CRC).

METHODS

Using the National Institutes of Health (NIH)'s Surveillance, Epidemiology, and End Results (SEER) database, a total of 51265 patients with pathological diagnosis of colorectal cancer from 2010 to 2015 were extracted for model development. On this basis, We have established 7 machine learning algorithm models. Evaluate the model based on accuracy, and AUC of receiver operating characteristics (ROC) and explain the relationship between clinical pathological features and target variables based on the best model. We validated the model among 196 colorectal cancer patients in Beijing Electric Power Hospital of Capital Medical University of China to evaluate its performance and universality. Finally, we have developed a network-based calculator using the best model to predict the risk of liver and/or lung metastasis in colorectal cancer patients.

RESULTS

51265 patients were enrolled in the study, of which 7864 (15.3 %) had distant liver and/or lung metastasis. RF had the best predictive ability, In the internal test set, with an accuracy of 0.895, AUC of 0.956, and AUPR of 0.896. In addition, the RF model was evaluated in the external validation set with an accuracy of 0.913, AUC of 0.912, and AUPR of 0.611.

CONCLUSION

In this study, we constructed an RF algorithm mode to predict the risk of colorectal liver and/or lung metastasis, to assist doctors in making clinical decisions.

Machine learning for predicting colon cancer recurrence

INTRODUCTION

Colorectal cancer (CRC) is a global public health concern, ranking among the most commonly diagnosed malignancies worldwide. Despite advancements in treatment modalities, the specter of CRC recurrence remains a significant challenge, demanding innovative solutions for early detection and intervention. The integration of machine learning into oncology offers a promising avenue to address this issue, providing data-driven insights and personalized care.

METHODS

This retrospective study analyzed data from 396 patients who underwent surgical procedures for colon cancer (CC) between 2010 and 2021. Machine learning algorithms were employed to predict CC recurrence, with a focus on demographic, clinicopathological, and laboratory characteristics. A range of evaluation metrics, including AUC (Area Under the Receiver Operating Characteristic), accuracy, recall, precision, and F1 scores, assessed the performance of machine learning algorithms.

RESULTS

Significant risk factors for CC recurrence were identified, including sex, carcinoembryonic antigen (CEA) levels, tumor location, depth, lymphatic and venous invasion, and lymph node involvement. The CatBoost Classifier demonstrated exceptional performance, achieving an AUC of 0.92 and an accuracy of 88 % on the test dataset. Feature importance analysis highlighted the significance of CEA levels, albumin levels, N stage, weight, platelet count, height, neutrophil count, lymphocyte count, and gender in determining recurrence risk.

DISCUSSION

The integration of machine learning into healthcare, exemplified by this study's findings, offers a pathway to personalized patient risk stratification and enhanced clinical decision-making. Early identification of individuals at risk of CC recurrence holds the potential for more effective therapeutic interventions and improved patient outcomes.

CONCLUSION

Machine learning has the potential to revolutionize our approach to CC recurrence prediction, emphasizing the synergy between medical expertise and cutting-edge technology in the fight against cancer. This study represents a vital step toward precision medicine in CC management, showcasing the transformative power of data-driven insights in oncology.

Survival Benefit of Primary Tumor Resection Combined With Chemotherapy in Patients With Unresectable Colorectal Mucinous Adenocarcinoma With Liver Metastasis

OBJECTIVE

To evaluate the survival benefit of combining primary tumor resection (PTR) and chemotherapy in patients with unresectable colorectal mucinous adenocarcinoma with liver metastasis (UCR-MAC-LM).

METHODS

We obtained data from the surveillance, epidemiology, and end results database for patients with UCR-MAC-LM from 2010 to 2017. Clinicopathological characteristics were analyzed using the χ2 test. Propensity score matching was performed to balance baseline characteristics. Kaplan-Meier analysis and log-rank tests were used to estimate and compare survival outcomes. Univariate and multivariate Cox regression analyses were conducted to identify the prognostic factors.

RESULTS

A total of 10,178 patients with unresectable colorectal adenocarcinoma with liver metastasis were included, of whom 6.01% (n=612) had UCR-MAC-LM. The UCR-MAC-LM group had a higher proportion of female patients, a greater number of elderly patients, an increased incidence of right colon localization, larger tumor size, and higher T and N staging than the unresectable colorectal non-mucinous adenocarcinoma with liver metastasis group (P<0.05). Multivariate analysis identified several independent prognostic factors (P<0.05). Patients with unresectable colorectal adenocarcinoma with liver metastasis who underwent PTR+C had superior survival rates compared with those who received PTR/C alone or no treatment (cancer-specific survival, P<0.05; overall survival, P<0.05). Subgroup analysis revealed that 17 of 22 groups of patients with UCR-MAC-LM who received PTR+C had significantly prolonged long-term survival compared with those who received PTR/C alone.

CONCLUSIONS

This surveillance, epidemiology, and end results-based study indicates that PTR+C may offer a survival advantage for a specific subgroup of patients with UCR-MAC-LM compared with PTR/C alone. Nonetheless, additional clinical trials are necessary to validate these findings.

Prognostic nomogram for colorectal cancer patients with multi-organ metastases: a Surveillance, Epidemiology, and End Results program database analysis

BACKGROUND

A nomogram that integrates risk models and clinical characteristics can accurately predict the prognosis of individual patients. We aimed to identify the prognostic factors and establish nomograms for predicting overall survival (OS) and cause-specific survival (CSS) in patients with multi-organ metastatic colorectal cancer (CRC).

METHODS

Demographic and clinical information on multi-organ metastases from 2010 to 2019 were extracted from the Surveillance, Epidemiology, and End Results (SEER) Program. Univariate and multivariate Cox analyses were used to identify independent prognostic factors that were used to develop nomograms to predict CSS and OS, and to assess the concordance index (C-index), area under the curve (AUC), and calibration curve.

RESULTS

The patients were randomly assigned to the training and validation groups at a 7:3 ratio. A Cox proportional hazards model was conducted for CRC patients to identify independent prognostic factors, including age, sex, tumor size, metastases, degree of differentiation, stage T, stage N, primary and metastasis surgery. The competing risk models employed by Fine and Gray were used to identify the risk factors for CRC. Death from other causes was treated as a competing event, and Cox models were used to identify the factors for death to identify the independent factors of CSS. By incorporating the corresponding independent prognostic factors, we established prognostic nomograms for OS and CSS. Finally, we used the C-index, ROC curve, and calibration plots to assess the utility of the nomogram.

CONCLUSIONS

Using the SEER database, we constructed a predictive model for CRC patients with multi-organ metastases. Nomograms provide clinicians with 1-, 3-, and 5-year OS and CSS predictions for CRC, allowing them to formulate appropriate treatment plans.

Predicting disease-free survival in colorectal cancer by circulating tumor DNA methylation markers

BACKGROUND

Recurrence represents a well-known poor prognostic factor for colorectal cancer (CRC) patients. This study aimed to establish an effective prognostic prediction model based on noninvasive circulating tumor DNA methylation markers for CRC patients receiving radical surgery.

RESULTS

Two methylation markers (cg11186405 and cg17296166) were identified by Cox regression and receiver operating characteristics, which could classify CRC patients into high recurrence risk and low recurrence risk group. The 3-year disease-free survival was significantly different between CRC patients with low and high recurrence risk [Training set: hazard ratio (HR) 28.776, 95% confidence interval (CI) 3.594-230.400; P = 0.002; Validation set: HR 7.796, 95% CI 1.425-42.660, P = 0.018]. The nomogram based on the above two methylation markers and TNM stage was established which demonstrated robust prognostic prediction potential, as evidenced by the decision curve analysis result.

CONCLUSIONS

A cell-free DNA methylation model consisting of two DNA methylation markers is a promising method for prognostic prediction in CRC patients.

Survival nomograms for colorectal carcinoma patients with lung metastasis and lung-only metastasis, based on the SEER database and a single-center external validation cohort

Background

We analysed the survival of colorectal cancer (CRC) patients with lung metastasis and lung-only metastasis and determined the risk factors for lung metastasis in CRC patients.

Methods

Data from colorectal cancer patients with lung metastasis diagnosed from 2010 to 2015 were obtained from the SEER database. Survival was analysed using the Kaplan–Meier method and log-rank test, the Cox proportional hazards regression model, and a competing risk model. The predictive ability of the nomgram was assessed by the concordance index (C-index) and calibration curves. The data from the SEER database for the period 2016–2019 was used as an external validation set. The characteristics of 70 CRC patients treated at Shanghai East Hospital between 2016 and 2019 were retrospectively analysed and data from China was chosen as an external validation set.

Results

The median survival time for colorectal cancer patients with lung metastasis was 12 months, while this value was 24 months in patients with lung-only metastasis. Among all CRC patients with lung metastasis, age, grade, T stage, N stage, presence of liver, brain or bone metastasis, anatomic site and surgery were related to overall survival (OS). In CRC patients with lung-only metastasis, age, T stage, marital status, chemotherapy and surgery were independent prognostic factors affecting OS. Two nomograms predicting OS were established, with great discrimination (C-index between 0.67 and 0.81) and excellent calibration. Factors including age, race, sex, tumour grade, T stage, N stage, presence of liver, brain or bone metastasis, marital status, insurance status and anatomic location were related to the occurrence of lung metastasis in CRC patients.

Conclusion

We developed two reliable clinical prediction models among CRC patients to predict the OS rates in patients with lung metastasis and lung metastasis only.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12876-022-02547-9.

A bayesian approach to model the underlying predictors of early recurrence and postoperative death in patients with colorectal Cancer

Objective

This study aimed at utilizing a Bayesian approach semi-competing risks technique to model the underlying predictors of early recurrence and postoperative Death in patients with colorectal cancer (CRC).

Methods

In this prospective cohort study, 284 patients with colorectal cancer, who underwent surgery, referred to Imam Khomeini clinic in Hamadan from 2001 to 2017. The primary outcomes were the probability of recurrence, the probability of Mortality without recurrence, and the probability of Mortality after recurrence. The patients ‘recurrence status was determined from patients’ records. The Bayesian survival modeling was carried out by semi-competing risks illness-death models, with accelerated failure time (AFT) approach, in R 4.1 software. The best model was chosen according to the lowest deviance information criterion (DIC) and highest logarithm of the pseudo marginal likelihood (LPML).

Results

The log-normal model (DIC = 1633, LPML = -811), was the optimal model. The results showed that gender(Time Ratio = 0.764: 95% Confidence Interval = 0.456–0.855), age at diagnosis (0.764: 0.538–0.935 ), T₃ stage (0601: 0.530–0.713), N₂ stage (0.714: 0.577–0.935 ), tumor size (0.709: 0.610–0.929), grade of differentiation at poor (0.856: 0.733–0.988), and moderate (0.648: 0.503–0.955) levels, and the number of chemotherapies (1.583: 1.367–1.863) were significantly related to recurrence. Also, age at diagnosis (0.396: 0.313–0.532), metastasis to other sites (0.566: 0.490–0.835), T₃ stage (0.363: 0.592 − 0.301), T₄ stage (0.434: 0.347–0.545), grade of differentiation at moderate level (0.527: 0.387–0.674), tumor size (0.595: 0.500–0.679), and the number of chemotherapies (1.541: 1.332–2.243) were the significantly predicted the death. Also, age at diagnosis (0.659: 0.559–0.803), and the number of chemotherapies (2.029: 1.792–2.191) were significantly related to mortality after recurrence.

Conclusion

According to specific results obtained from the optimal Bayesian log-normal model for terminal and non-terminal events, appropriate screening strategies and the earlier detection of CRC leads to substantial improvements in the survival of patients.

Specific causes of recurrence after surgery and mortality in patients with colorectal cancer: A competing risks survival analysis

Background

In situation where there are more than one cause of occurring the outcome such as recurrence after surgery and death, the assumption of classical survival analyses are not satisfied. To cover this issue, this study aimed at utilizing competing risks survival analysis to assess the specific risk factors of local-distance recurrence and mortality in patients with colorectal cancer (CRC) undergoing surgery.

Materials and Methods

In this retrospective cohort study, 254 patients with CRC undergoing resection surgery were studied. Data of the outcome from the available documents in the hospital were gathered. Furthermore, based on pathological report, the diagnosis of CRC was considered. We model the risk factors on the hazard of recurrence and death using competing risk survival in R3.6.1 software.

Results

A total of 114 patients had local or distant recurrence (21 local recurrences, 72 distant recurrences, and 21 local and distant recurrence). Pathological stage (adjusted hazard ratio [AHR] = 4.28 and 5.37 for stage 3 and 4, respectively), tumor site (AHR = 2.45), recurrence (AHR = 3.92) and age (AHR = 3.15 for age >70) was related to hazard of death. Also based on cause-specific hazard model, pathological stage (AHR = 7.62 for stage 4), age (AHR = 1.46 for age >70), T stage (AHR = 1.8 and 2.7 for T3 and T4, respectively), N stage (AHR = 2.59 for N2), and white blood cells (AHR = 1.95) increased the hazard of recurrence in patients with CRC.

Conclusion

This study showed that older age, higher pathological, rectum tumor site and presence of recurrence were independent risk factors for mortality among CRC patients. Also age, higher T/N stage, higher pathological stage and higher values of WBC were significantly related to higher hazard of local/distance recurrence of patients with CRC.

Prognostic and predictive values of tumour budding in stage IV colorectal cancer

BACKGROUND

Tumour budding is an important prognostic feature in early-stage colorectal cancer, but its prognostic significance in metastatic disease has not been fully investigated.

METHODS

Patients with stage IV disease who had primary colorectal tumour resection without previous chemotherapy or radiotherapy from January 2000 to December 2018 were reviewed retrospectively. Budding was evaluated at the primary site and graded according to the criteria of the International Tumor Budding Consensus Conference (ITBCC) (BD1, low; BD2, intermediate; BD3, high). Patients were categorized by metastatic (M1a, M1b) and resectional (R0/R1, R2/unresected) status. Subgroups were compared for overall (OS) and recurrence-free (RFS) survival in R0/R1 subgroups; R2/unresected patients were evaluated for the rate of tumour progression, based on change in tumour size from baseline.

RESULTS

Of 371 patients observed during the study, 362 were analysed. Patients with BD3 had a lower 5-year OS rate than those with BD1 + BD2 (18·4 versus 40·5 per cent; P < 0·001). Survival analyses according to metastatic and resection status also showed that BD3 was associated with shorter OS than BD1 + BD2. In multivariable analysis, BD3 (hazard ratio (HR) 1·51, 95 per cent c.i. 1·11 to 2·10; P = 0·009), T4 status (HR 1·39) and R2/unresected status (HR 3·50) were associated with decreased OS. In the R0/R1 subgroup, the 2-year RFS rate was similar for BD3 and BD1 + BD2 according to metastatic status. There was no significant difference between BD3 and BD1 + BD2 for change in tumour size in the R2/unresected subgroup (P = 0·094). Of 141 patients with initially unresectable metastases who had chemotherapy, 35 achieved conversion from unresectable to resectable status. The conversion rate was significantly higher for BD1 + BD2 than for BD3 (36 versus 18 per cent; P = 0·016).

CONCLUSION

Stage IV colorectal cancer with high-grade tumour budding according to ITBCC criteria correlates with poor prognosis.

Risk factors and risk prediction models for colorectal cancer metastasis and recurrence: an umbrella review of systematic reviews and meta-analyses of observational studies

BACKGROUND

There is a clear need for systematic appraisal of models/factors predicting colorectal cancer (CRC) metastasis and recurrence because clinical decisions about adjuvant treatment are taken on the basis of such variables.

METHODS

We conducted an umbrella review of all systematic reviews of observational studies (with/without meta-analysis) that evaluated risk factors of CRC metastasis and recurrence. We also generated an updated synthesis of risk prediction models for CRC metastasis and recurrence. We cross-assessed individual risk factors and risk prediction models.

RESULTS

Thirty-four risk factors for CRC metastasis and 17 for recurrence were investigated. Twelve of 34 and 4/17 risk factors with p < 0.05 were estimated to change the odds of the outcome at least 3-fold. Only one risk factor (vascular invasion for lymph node metastasis [LNM] in pT1 CRC) presented convincing evidence. We identified 24 CRC risk prediction models. Across 12 metastasis models, six out of 27 unique predictors were assessed in the umbrella review and four of them changed the odds of the outcome at least 3-fold. Across 12 recurrence models, five out of 25 unique predictors were assessed in the umbrella review and only one changed the odds of the outcome at least 3-fold.

CONCLUSIONS

This study provides an in-depth evaluation and cross-assessment of 51 risk factors and 24 prediction models. Our findings suggest that a minority of influential risk factors are employed in prediction models, which indicates the need for a more rigorous and systematic model construction process following evidence-based methods.

Development of a novel prediction model for recurrent stage II colon cancer

PURPOSE

Adjuvant chemotherapy is recommended for patients with high-risk stage II colon cancer. High-risk stage II is defined by clinicopathological factors in some guidelines. However, there is no unified definition. The aim of this study was to examine the risk factors and develop a novel model to predict the recurrence of stage II colon cancer.

METHODS

Three hundred fifty patients who underwent curative resection for stage II colon cancer at Osaka International Cancer Institute and Yao Municipal Hospital from 2004 to 2012 were included. Clinicopathological factors were assessed in a subgroup of 298 patients (Learning Set), and the relapse-free survival (RFS) rate was evaluated as the main outcome. A statistical analysis was performed using a proportional hazards model to determine the factors associated with RFS and a nomogram was developed to predict recurrence. A second subgroup of 52 independent patients who underwent curative resection in 2012 (Validation Set) was used to validate the nomogram.

RESULTS

The median RFS time was 4.96 years, and recurrence was observed in 35 patients. A univariate analysis revealed that a high serum CEA level, preoperative occlusion, tumor location (left-side colon), lymphatic invasion, and vascular invasion were significantly correlated with RFS. These variables were used to develop the nomogram. The C-index was 0.701 in the learning set and 0.585 in the validation set. Using nomogram points, the patients were classified into low-risk, middle-risk, and high-risk categories.

CONCLUSION

A recurrence prediction model was developed that integrated multiple risk factors in stage II colon cancer patients. High-risk patients were identified by the nomogram.

Development and validation of novel scoring system for the prediction of disease recurrence following resection of colorectal liver metastasis

BACKGROUND

The aim of this study was to identify predictive factors for the recurrence of colorectal cancer liver metastasis (CRLM) and then to develop a corresponding novel scoring system that should improve the sensitivity of predicting recurrence in patients with CRLM.

METHODS

A total of 295 consecutive CRLM patients were enrolled in our institution between January 2002 and December 2015. Multivariate analyses were performed to identify the variables associated with disease recurrence and established the novel scoring system based on it.

RESULTS

The scoring system considered seven variables: synchronosity, CA19-9 level, number of liver metastasis, largest size of liver metastasis, resection margin of hepatic lesion, neutrophil-to-lymphocyte ratio and prognostic nutritional index. The area under the curve of ROC was 0.824 (95% confidence interval 0.767-0.882); the sensitivity of our scoring system was 87.9%, specificity was 66.7%, positive predictive value was 20.6%, and negative predictive value was 20.9%.

CONCLUSION

For patients with CRLM undergoing curative hepatic resection, our novel scoring system would improve the sensitivity for prediction of disease recurrence in Case of CRLM patients.

Outcomes of stage IV patients with colorectal cancer treated in a single institution: What is the key to the long-term survival?

OBJECTIVES

The purpose of this study is to summarize our short- and long-term treatment results for stage IV colorectal cancer (CRC) and to clarify the factors predicting the favorable long-term survival.

METHODS

Between January 2008 and December 2015, 149 consecutive patients with stage IV CRC underwent initial treatment at Nagoya University Hospital. Their clinical and pathological characteristics, the treatment methods used, and the outcomes were retrospectively analyzed.

RESULTS

The median observation period was 23 months. All of the primary and metastatic lesions were technically resectable in 74 patients; however, the remaining 75 were judged as initially unresectable. R0/1 resection during the treatment course was achieved in 74 patients (50%). For the cohort as a whole, the 5-year overall survival (OS) rate was 35%. The 5-year OS rate in the R0/1 resection group was 57%, which was significantly better than that of the non-R0/1 resection group (6%, p < 0.001). In the R0/1 resection group, perioperative chemotherapy significantly improved the outcome (5-year OS; 62% vs. 0%, p = 0.03). In the non-R0/1 resection group, primary tumor resection was associated with a significantly higher favorable prognosis (3-year OS; 20.4% vs. 0%, p = 0.026). Moreover, the additional use of molecular targeted drugs significantly improved the survival. In multivariate analysis, the differentiated histologic type, R0/1 resection, and parallel use of molecular targeted drugs remained independent factors of a favorable outcome.

CONCLUSIONS

The present study suggested that aggressive curative resection with perioperative chemotherapy might improve survival and that primary tumor resection might improve the outcome in the non-R0/1 group.

Primary tumor resection in stage IV unresectable colorectal cancer: what has changed?

Most current guidelines do not recommend primary tumor resection in stage IV unresectable colorectal cancer. Rapid chemotherapy development over the last decade has substantially changed the decision making. However, results of recently published trials and meta-analyses suggest that primary tumor resection may in fact be beneficial, principally in terms of prolonged survival. Additional factors, such as use of minimally invasive approach or protocols of enhanced recovery after surgery, affect clinical outcomes as well, but are often neglected when discussing the state of the art in this area. There are still no randomized studies determining the legitimacy of upfront surgery in asymptomatic patients. Also, quality of life also plays an important role in choosing appropriate treatment. Having said that, there is no data that would prove whether primary tumor resection has an advantage on that issue. With all the uncertainty, currently decision making in unresectable stage IV colorectal cancer is primarily up to clinicians' knowledge, common sense and patients' preferences.