Incorporating non-biological factors into the TNM staging system for better prognostication and decision-making in testicular cancer

A Novel Machine Learning Algorithm Combined With Multivariate Analysis for the Prognosis of Renal Collecting Duct Carcinoma

Objectives

To investigate the clinical and non-clinical characteristics that may affect the prognosis of patients with renal collecting duct carcinoma (CDC) and to develop an accurate prognostic model for this disease.

Methods

The characteristics of 215 CDC patients were obtained from the U.S. National Cancer Institute’s surveillance, epidemiology and end results database from 2004 to 2016. Univariate Cox proportional hazard model and Kaplan-Meier analysis were used to compare the impact of different factors on overall survival (OS). 10 variables were included to establish a machine learning (ML) model. Model performance was evaluated by the receiver operating characteristic curves (ROC) and calibration plots for predictive accuracy and decision curve analysis (DCA) were obtained to estimate its clinical benefits.

Results

The median follow-up and survival time was 16 months during which 164 (76.3%) patients died. 4.2, 32.1, 50.7 and 13.0% of patients were histological grade I, II, III, and IV, respectively. At diagnosis up to 61.9% of patients presented with a pT3 stage or higher tumor, and 36.7% of CDC patients had metastatic disease. 10 most clinical and non-clinical factors including M stage, tumor size, T stage, histological grade, N stage, radiotherapy, chemotherapy, age at diagnosis, surgery and the geographical region where the care delivered was either purchased or referred and these were allocated 95, 82, 78, 72, 49, 38, 36, 35, 28 and 21 points, respectively. The points were calculated by the XGBoost according to their importance. The XGBoost models showed the best predictive performance compared with other algorithms. DCA showed our models could be used to support clinical decisions in 1-3-year OS models.

Conclusions

Our ML models had the highest predictive accuracy and net benefits, which may potentially help clinicians to make clinical decisions and follow-up strategies for patients with CDC. Larger studies are needed to better understand this aggressive tumor.

Clinical Implications of Nonbiological Factors With Colorectal Cancer Patients Younger Than 45 Years

Background

To evaluate the clinical implications of non-biological factors (NBFs) with colorectal cancer (CRC) patients younger than 45 years.

Methods

In the present study, we have conducted Cox proportional hazard regression analyses to evaluate the prognosis of different prognostic factors, the hazard ratios (HRs) were shown with 95% confidence intervals (CIs). Kaplan-Meier method was utilized to compare the prognostic value of different factors with the log-rank test. NBF score was established according to the result of multivariate Cox analyses.

Results

In total, 15129 patients before 45 years with known NBFs were identified from the SEER database. Only county-level median household income, marital status and insurance status were NBFs that significantly corelated with the cause specifical survival in CRC patients aged less than 45 years old (P < 0.05). Stage NBF 1 showed 50.5% increased risk of CRC-specific mortality (HR = 1.505, 95% CI = 1.411-1.606, P < 0.001). Stage NBF 0 patients were associated with significantly increased CRC-specific survival (CCSS) when compared with the stage NBF 1 patients in different AJCC TNM stages.

Conclusions

NBF stage (defined by county-level median household income, marital status and insurance status) was strongly related to the prognosis of CRC patients. NBFs should arouse enough attention of us in clinical practice of patients younger than 45 years.

Construction and Validation of a Nomogram for Predicting Progression- Free Survival in Patients with Early-Stage Testicular Germ Cell Tumor

BACKGROUND

Testicular Germ Cell Tumor (TGCT) is the most common malignant tumor in young men, but there is a lack of a prediction model to evaluate the prognosis of patients with TGCT.

OBJECTIVE

To explore the prognostic factors for Progression-Free Survival (PFS) and construct a nomogram model for patients with early-stage TGCT after radical orchiectomy.

METHODS

Patients with TGCT from The Cancer Genome Atlas (TCGA) database were used as the training cohort; univariate and multivariate cox analysis was performed. A nomogram was constructed based on the independent prognostic factors. Patients from the Nanfang Hospital affiliated with Southern Medical University were used as the cohort to validate the predictive ability using the nomogram model. Harrell's concordance index (C-index) and calibration plots were used to evaluate the nomogram.

RESULTS

A total of 110 and 62 patients with TGCT were included in the training cohort and validation cohort, respectively. Lymphatic Vascular Invasion (LVI), American Joint Committee on Cancer (AJCC) stage and adjuvant therapy were independent prognostic factors in multivariate regression analyses and were included to establish a nomogram. The C-index in the training cohort for 1- , 3-, and 5-year PFS were 0.768, 0.74, and 0.689, respectively. While the C-index for 1-, 3-, and 5- year PFS in the external validation cohort were 0.853, 0.663 and 0.609, respectively. The calibration plots for 1-, 3-, and 5-year PFS in the training and validation cohort showed satisfactory consistency between predicted and actual outcomes. The nomogram revealed a better predictive ability for PFS than AJCC staging system.

CONCLUSION

The nomogram as a simple and visual tool to predict individual PFS in patients with TGCT could guide clinicians and clinical pharmacists in therapeutic strategy.

Lymph Node Ratio Rather Than Positive Lymph Node Counts Has Better Prognostic Value in Patients With Testicular Germ Cell Tumors

Background:

Testicular cancer represents the most common malignancy in young adult men. In the current study, we sought to analyze and compare the prognostic value of lymph node ratio (LNR) as well as positive lymph node counts (LNC) to understand its clinical significance in testicular germ cell tumors.

Methods:

We employed eligibility criteria to recruit a total of 931 patients, with testicular cancer, from 2010 to 2015 from The Surveillance, Epidemiology, and End Results (SEER) database. We then used the X-Tile program to calculate LNR and LNC cutoff values and discriminate survival. We then calculated the overall and cancer specific survival rates and analyzed the association between LNR/LNC and clinical pathological characteristics using the χ² test. Finally, we assessed the relationships between clinical pathological factors and patient survival using univariate Cox proportional hazard analysis.

Results:

Univariate analysis revealed a significant association between prognosis with age (HR, 5.169; 95% CI, 1.758-15.200; P = 0.003), AJCC stage (III vs I: HR, 9.298; 95% CI, 2.691-32.131; P < 0.001), M stage (HR, 7.897; 95% CI, 3.417-18.251; P < 0.001) and LNR (HR, 3.009; 95% CI, 1.275-7.098; P = 0.012). On the other hand, LNC (HR, 1.743; 95% CI, 0.687-4.420; P = 0.242) was not significantly associated with prognosis. Analysis of the association between LNR/LNC and clinical pathological characteristics showed that high LNR patients tended to have significantly larger tumor sizes (χ² = 7.877, P = 0.005), as well as advanced T (χ² = 13.195, P = 0.004), N ( χ² = 86.775, P < 0.001), M (χ² = 19.948, P < 0.001) and 7th AJCC (χ² = 103.074, P < 0.001) stages. In addition, high LNC patients were significantly associated with T (χ² = 8.799, P = 0.032), N (χ² = 74.390, P < 0.001) and 7th AJCC (χ² = 111.759, P < 0.001) stages.

Conclusion:

LNR was a better predictor for long-term prognosis and was closely associated with clinical pathological characteristics than LNC in patients with testicular germ cell tumors.

A nomogram for predicting survival and retroperitoneal lymph node dissection treatment in patients with resected testicular germ cell tumors

BACKGROUND AND OBJECTIVES

To build nomogram incorporating potential prognostic factors for predicting survival outcomes of testicular germ cell tumors (TGCT) patients after resection of the primary tumor.

METHODS

Data of TGCT patients from the Surveillance, Epidemiology, and End Results database (2010-2016) who underwent resection of the primary tumor were collected. Overall survival (OS) and cancer-specific survival (CSS) were analyzed by using Cox regression models, nomogram, Kaplan-Meier method, and log-rank test.

RESULTS

We identified 7272 TGCT patients. Age at diagnosis, histology, tumor size, American Joint Committee on Cancer (AJCC) staging system, and number of metastases sites were independent prognostic factors and were integrated into nomograms. The nomograms had higher C-indexes for both OS and CSS compared with the AJCC 7th staging system (0.881 vs 0.831 and 0.895 vs 0.856, respectively). Moreover, the new stratification of risk groups based on the nomograms showed a more significant distinction between Kaplan-Meier curves for survival outcomes than the AJCC staging system. Retroperitoneal lymph node dissection was associated with statistically improved survival probability in the nomogram middle-risk group in resected TGCT patients.

CONCLUSION

The novel nomogram-based staging system could provide satisfactory risk stratification and survival prediction ability beyond traditional AJCC staging systems.