Prediction of Recurrence-associated Death from Localized Prostate Cancer with a Charlson Comorbidity Index-reinforced Machine Learning Model

Association between Charlson comorbidity index and survival outcomes in patients with prostate cancer: A meta-analysis

Objective

This meta-analysis aimed to assess the influence of comorbidity, as assessed by the Charlson comorbidity index (CCI), on survival outcomes in patients with prostate cancer (PCa).

Methods

We conducted a comprehensive search of the PubMed, Web of Science, and Embase databases to identify studies that examined the association between CCI-defined comorbidity and survival outcomes in PCa patients. We employed a random effect model to merge adjusted hazard ratios (HR) with 95 % confidence intervals (CI) for survival outcomes.

Results

Sixteen studies reporting on 17 articles, which collectively included 457,256 patients. For the presence (CCI score ≥1) versus absence (CCI score of 0) of comorbidity, the pooled HR was 1.59 (95 % CI 1.43-1.77) for all-cause mortality, 0.98 (95 % CI 0.90-1.08) for PCa-specific mortality, and 1.88 (95 % CI 1.61-2.21) for other-cause mortality. When compared to a CCI score of 0, the pooled HR of all-cause mortality was 1.30 (95 % CI 1.18-1.44) for a CCI score of 1, 1.65 (95 % CI 1.37-2.00) for a CCI score ≥2, and 1.75 (95 % CI 1.57-1.95) for a CCI score ≥3. Additionally, the pooled HR of other cause mortality was 1.53 (95 % CI 1.41-1.67) for a CCI score of 1, 1.93 (95 % CI 1.74-2.75) for a CCI score ≥2, and 3.95 (95 % CI 2.13-7.34) for a CCI score ≥3.

Conclusions

Increased comorbidity, as assessed by the CCI, significantly predicts all-cause and other-cause mortality in patients with PCa, but not PCa-specific mortality. The risk of all-cause and other-cause mortality increases with the burden of comorbidity.

Urinary tract infection after radiation therapy or radical prostatectomy on the prognosis of patients with prostate cancer: a population-based study

BACKGROUND

We aimed to assess the trends in urinary tract infections (UTIs) and prognosis of patients with prostate cancer after radical prostatectomy (RP) and radiation therapy (RT) as definitive treatment options.

METHODS

The data of patients diagnosed with prostate cancer between 2007 and 2016 were collected from the National Health Insurance Service database. The incidence of UTIs was evaluated in patients treated with RT, open/laparoscopic RP, and robot-assisted RP. The proportional hazard assumption test was performed using the scaled Schoenfeld residuals based on a multivariable Cox proportional hazard model. Kaplan-Meier analysis were performed to assess survival.

RESULTS

A total of 28,887 patients were treated with definitive treatment. In the acute phase (< 3 months), UTIs were more frequent in RP than in RT; in the chronic phase (> 12 months), UTIs were more frequent in RT than in RP. In the early follow-up period, the risk of UTIs was higher in the open/laparoscopic RP group (aHR, 1.63; 95% CI, 1.44-1.83; p < 0.001) and the robot-assisted RP group (aHR, 1.26; 95% CI, 1.11-1.43; p < 0.001), compared to the RT group. The robot-assisted RP group had a lower risk of UTIs than the open/laparoscopic RP group in the early (aHR, 0.77; 95% CI, 0.77-0.78; p < 0.001) and late (aHR, 0.90; 95% CI, 0.89-0.91; p < 0.001) follow-up periods. In patients with UTI, Charlson Comorbidity Index score, primary treatment, age at UTI diagnosis, type of UTI, hospitalization, and sepsis from UTI were risk factors for overall survival.

CONCLUSIONS

In patients treated with RP or RT, the incidence of UTIs was higher than that in the general population. RP posed a higher risk of UTIs than RT did in early follow-up period. Robot-assisted RP had a lower risk of UTIs than open/laparoscopic RP group in total period. UTI characteristics might be related to poor prognosis.

miRNAs as molecular biomarkers for prostate cancer

MicroRNAs (miRNAs) are short noncoding RNA able to regulate specific mRNA stability, thus influencing target gene expression. Disrupted levels of several miRNA have been associated with prostate cancer, the leading cause of cancer death among men and the fifth leading cause of death worldwide. Here, we investigated whether miR-145, miR-148, and miR-185 circulating levels in plasma could be used as molecular biomarkers, to allow distinguishing between individuals with benign prostatic hyperplasia, precancerous lesion, and prostate cancer. In this study, we recruited 170 urological clinic patients with suspected prostate cancer who underwent prostate biopsy. Total RNA was isolated from plasma, and TaqMan MicroRNA assays were used to analyze miR-145, miR-185, and miR-148 expression. First, differential miRNA expression among patient groups was evaluated. Then, miRNA levels were combined with clinical assessment outcomes, including results from invasive tests, using multivariate analysis to examine their ability in discriminating among the three patient groups. Our results suggest that miRNA is a promising molecular tool for clinical management of at-risk patients.

A review of AI and Data Science support for cancer management

INTRODUCTION

Thanks to improvement of care, cancer has become a chronic condition. But due to the toxicity of treatment, the importance of supporting the quality of life (QoL) of cancer patients increases. Monitoring and managing QoL relies on data collected by the patient in his/her home environment, its integration, and its analysis, which supports personalization of cancer management recommendations. We review the state-of-the-art of computerized systems that employ AI and Data Science methods to monitor the health status and provide support to cancer patients managed at home.

OBJECTIVE

Our main objective is to analyze the literature to identify open research challenges that a novel decision support system for cancer patients and clinicians will need to address, point to potential solutions, and provide a list of established best-practices to adopt.

METHODS

We designed a review study, in compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, analyzing studies retrieved from PubMed related to monitoring cancer patients in their home environments via sensors and self-reporting: what data is collected, what are the techniques used to collect data, semantically integrate it, infer the patient's state from it and deliver coaching/behavior change interventions.

RESULTS

Starting from an initial corpus of 819 unique articles, a total of 180 papers were considered in the full-text analysis and 109 were finally included in the review. Our findings are organized and presented in four main sub-topics consisting of data collection, data integration, predictive modeling and patient coaching.

CONCLUSION

Development of modern decision support systems for cancer needs to utilize best practices like the use of validated electronic questionnaires for quality-of-life assessment, adoption of appropriate information modeling standards supplemented by terminologies/ontologies, adherence to FAIR data principles, external validation, stratification of patients in subgroups for better predictive modeling, and adoption of formal behavior change theories. Open research challenges include supporting emotional and social dimensions of well-being, including PROs in predictive modeling, and providing better customization of behavioral interventions for the specific population of cancer patients.

Machine-Learning Techniques for Feature Selection and Prediction of Mortality in Elderly CABG Patients

Coronary artery bypass surgery grafting (CABG) is a commonly efficient treatment for coronary artery disease patients. Even if we know the underlying disease, and advancing age is related to survival, there is no research using the one year before surgery and operation-associated factors as predicting elements. This research used different machine-learning methods to select the features and predict older adults' survival (more than 65 years old). This nationwide population-based cohort study used the National Health Insurance Research Database (NHIRD), the largest and most complete dataset in Taiwan. We extracted the data of older patients who had received their first CABG surgery criteria between January 2008 and December 2009 (n = 3728), and we used five different machine-learning methods to select the features and predict survival rates. The results show that, without variable selection, XGBoost had the best predictive ability. Upon selecting XGBoost and adding the CHA2DS score, acute pancreatitis, and acute kidney failure for further predictive analysis, MARS had the best prediction performance, and it only needed 10 variables. This study's advantages are that it is innovative and useful for clinical decision making, and machine learning could achieve better prediction with fewer variables. If we could predict patients' survival risk before a CABG operation, early prevention and disease management would be possible.

Application of a novel machine learning framework for predicting non-metastatic prostate cancer-specific mortality in men using the Surveillance, Epidemiology, and End Results (SEER) database

BACKGROUND

Accurate prognostication is crucial in treatment decisions made for men diagnosed with non-metastatic prostate cancer. Current models rely on prespecified variables, which limits their performance. We aimed to investigate a novel machine learning approach to develop an improved prognostic model for predicting 10-year prostate cancer-specific mortality and compare its performance with existing validated models.

METHODS

We derived and tested a machine learning-based model using Survival Quilts, an algorithm that automatically selects and tunes ensembles of survival models using clinicopathological variables. Our study involved a US population-based cohort of 171 942 men diagnosed with non-metastatic prostate cancer between Jan 1, 2000, and Dec 31, 2016, from the prospectively maintained Surveillance, Epidemiology, and End Results (SEER) Program. The primary outcome was prediction of 10-year prostate cancer-specific mortality. Model discrimination was assessed using the concordance index (c-index), and calibration was assessed using Brier scores. The Survival Quilts model was compared with nine other prognostic models in clinical use, and decision curve analysis was done.

FINDINGS

647 151 men with prostate cancer were enrolled into the SEER database, of whom 171 942 were included in this study. Discrimination improved with greater granularity, and multivariable models outperformed tier-based models. The Survival Quilts model showed good discrimination (c-index 0·829, 95% CI 0·820-0·838) for 10-year prostate cancer-specific mortality, which was similar to the top-ranked multivariable models: PREDICT Prostate (0·820, 0·811-0·829) and Memorial Sloan Kettering Cancer Center (MSKCC) nomogram (0·787, 0·776-0·798). All three multivariable models showed good calibration with low Brier scores (Survival Quilts 0·036, 95% CI 0·035-0·037; PREDICT Prostate 0·036, 0·035-0·037; MSKCC 0·037, 0·035-0·039). Of the tier-based systems, the Cancer of the Prostate Risk Assessment model (c-index 0·782, 95% CI 0·771-0·793) and Cambridge Prognostic Groups model (0·779, 0·767-0·791) showed higher discrimination for predicting 10-year prostate cancer-specific mortality. c-indices for models from the National Comprehensive Cancer Care Network, Genitourinary Radiation Oncologists of Canada, American Urological Association, European Association of Urology, and National Institute for Health and Care Excellence ranged from 0·711 (0·701-0·721) to 0·761 (0·750-0·772). Discrimination for the Survival Quilts model was maintained when stratified by age and ethnicity. Decision curve analysis showed an incremental net benefit from the Survival Quilts model compared with the MSKCC and PREDICT Prostate models currently used in practice.

INTERPRETATION

A novel machine learning-based approach produced a prognostic model, Survival Quilts, with discrimination for 10-year prostate cancer-specific mortality similar to the top-ranked prognostic models, using only standard clinicopathological variables. Future integration of additional data will likely improve model performance and accuracy for personalised prognostics.

FUNDING

None.

Relation of geriatric nutritional risk index with clinical risks in elderly-onset ulcerative colitis

BACKGROUND AND AIM

Worldwide increasing aging societies have many elderlies with intractable diseases including ulcerative colitis (UC). Reportedly, each patients' frailty as well as chronological age is a clinical risk factor of elderly-onset UC (EOUC). Because malnutrition is one of the major manifestations of frailty, we aimed to investigate the effect of malnutrition on the prognosis of EOUC with geriatric nutritional risk index (GNRI), a prognostic tool for several diseases in the elderly to estimate malnutrition, and to evaluate clinical risks among EOUC patients in Japan, the world-leading aging society.

METHODS

The EOUC patients (≥ 65 years at diagnosis, n = 2778) in the previous nationwide survey were classified by age and GNRI, and odds ratios (ORs) of hospitalization and UC-related surgery were determined to evaluate the effects of malnutrition on the EOUC patients as well as aging.

RESULTS

The risks of hospitalization and surgery were elevated as age advanced. The value of GNRI, negatively correlated with disease activity (r = -0.53), could distinguish severe activity (cutoff ≤ 86.82, sensitivity = 0.79, and specificity = 0.77) and discriminate the EOUC patients suffering from surgery and hospitalization. In a multivariate analysis, GNRI ≤ 86.82 was a higher risk of hospitalization (OR: 4.0, 95% CI, 2.5-6.5) and surgery (OR: 2.7, 95% CI, 0.98-7.4) than cutoff age ≥ 75 years old (OR of hospitalization and surgery were 1.4 [95% CI, 0.99-2.0] and 2.3 [95% CI, 0.8-6.3], respectively).

CONCLUSION

Malnutrition estimated by GNRI was significantly related with poor clinical courses of the EOUC patients, suggesting that evaluation of nutritional status at the onset might be useful for predicting risks of clinical courses.

Hypogonadism and its treatment among prostate cancer survivors

Adult-onset hypogonadism (AOH) is associated with sexual dysfunction, poor bone mineralization, decreased muscle mass, metabolic syndrome disorder, and cognitive suppression. Historically, testosterone has been contraindicated in men with a history of prostate cancer. However, there has been a modern resurgence in re-evaluating this belief. Not only can testosterone be safely utilized to alleviate AOH symptoms in prostate cancer survivors, it has been also touted as a treatment option for aggressive prostatic cancer. While much work remains in understanding the relationship between testosterone and prostate cancer, those who survive this disease should not be automatically turned away from an opportunity to be treated and restored.

Predicting long-term multicategory cause of death in patients with prostate cancer: random forest versus multinomial model

The majority of patients with prostate cancer die of non-cancer causes of death (COD). It is thus important to accurately predict multi-category COD in these patients. Random forest (RF), a popular machine learning model, has been shown useful for predicting binary cancer-specific deaths. However, its accuracy for predicting multi-category COD in cancer patients is unclear. We included patients in Surveillance, Epidemiology, and End Results-18 cancer registry-program with prostate cancer diagnosed in 2004 (followed-up through 2016). They were randomly divided into training and testing sets with equal sizes. We evaluated prediction accuracies of RF and conventional statistical/multinomial models for 6-category COD by data-encoding types using the 2-fold cross-validation approach. Among 49,864 prostate cancer patients, 29,611 (59.4%) were alive at the end of follow-up, and 5,448 (10.9%) died of cardiovascular disease, 4,607 (9.2%) of prostate cancer, 3,681 (7.4%) of non-prostate cancer, 717 (1.4%) of infection, and 5,800 (11.6%) of other causes. We predicted 6-category COD among these patients with a mean accuracy of 59.1% (n=240, 95% CI, 58.7%-59.4%) in RF models with one-hot encoding, and 50.4% (95% CI, 49.7%-51.0%) in multinomial models. Tumor characteristics, prostate-specific antigen level, and diagnosis confirmation-method were important in RF and multinomial models. In RF models, no statistical differences were found between the accuracies of training versus cross-validation phases, and those of categorical versus one-hot encoding. We here report that RF models can outperform multinomial logistic models (absolute accuracy-difference, 8.7%) in predicting long-term 6-category COD among prostate cancer patients, while pathology diagnosis itself and tumor pathology remain important factors.

Data-driven translational prostate cancer research: from biomarker discovery to clinical decision

Prostate cancer (PCa) is a common malignant tumor with increasing incidence and high heterogeneity among males worldwide. In the era of big data and artificial intelligence, the paradigm of biomarker discovery is shifting from traditional experimental and small data-based identification toward big data-driven and systems-level screening. Complex interactions between genetic factors and environmental effects provide opportunities for systems modeling of PCa genesis and evolution. We hereby review the current research frontiers in informatics for PCa clinical translation. First, the heterogeneity and complexity in PCa development and clinical theranostics are introduced to raise the concern for PCa systems biology studies. Then biomarkers and risk factors ranging from molecular alternations to clinical phenotype and lifestyle changes are explicated for PCa personalized management. Methodologies and applications for multi-dimensional data integration and computational modeling are discussed. The future perspectives and challenges for PCa systems medicine and holistic healthcare are finally provided.