https://orcid.org/0000-0003-0406-9489
Figures
Abstract
Objectives
The aim of this study was to develop and validate a nomogram model that predicts the risk of bone metastasis (BM) in a prostate cancer (PCa) population.
Methods
We retrospectively collected and analyzed the clinical data of patients with pathologic diagnosis of PCa from January 1, 2013 to December 31, 2022 in two hospitals in Yangzhou, China. Patients from the Affiliated Hospital of Yangzhou University were divided into a training set and patients from the Affiliated Clinical College of Traditional Chinese Medicine of Yangzhou University were divided into a validation set. Chi-square test, independent sample t-test, and logistic regression were used to screen key risk factors. Receiver operating characteristic (ROC) curves, c-index, calibration curves, and decision curves analysis (DCA) were used for the validation, calibration, clinical benefit assessment, and external validation of nomogram models.
Results
A total of 204 cases were collected from the Affiliated Hospital of Yangzhou University, including 64 cases diagnosed as PCa BM and 50 cases collected from the Affiliated Clinical College of Traditional Chinese Medicine of Yangzhou University, including 12 cases diagnosed as PCa BM. Results showed that history of alcohol consumption, prostate stiffness on Digital rectal examination(DRE), prostate nodules on DRE, FIB, ALP, cTx, and Gleason score were high-risk factors for BM in PCa and nomogram was established. The c-index of the final model was 0.937 (95% CI: 0.899–0.975). And the model was validated by external validation set (c-index: 0.929). The ROC curves and calibration curves showed that the nomogram had good predictive accuracy, and DCA showed that the nomogram had good clinical applicability.
Citation: Xu W, Zhu G, Wang X, Yan X, Wang F, Li S, et al. (2025) Predicting the risk of bone metastases in patients with newly diagnosed prostate cancer – Development and validation of a nomogram model. PLoS ONE 20(1): e0318051. https://doi.org/10.1371/journal.pone.0318051
Editor: Yuki Arita, Memorial Sloan Kettering Cancer Center, UNITED STATES OF AMERICA
Received: August 12, 2024; Accepted: January 8, 2025; Published: January 30, 2025
Copyright: © 2025 Xu et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and in the Supporting information file.
Funding: This project was funded by the National Natural Science Foundation of China (81973518), Talent’s start-up fund from Yangzhou University (137011474), and Postgraduate Research & Practice Innovation Program of Jiangsu Province (SJCX21_1662).
Competing interests: The authors have declared that no competing interests exist.
Abbreviations:: AAPR, albumin alkaline phosphatase ratio; ALP, alkaline phosphatase; AUC, area under curve; BM, bone metastasis; CI, confidence interval; CT, computed tomography; cTx, clinical t stage x; DCA, decision curve analysis; DNA, deoxyribo nucleic acid; DRE, digital rectal examination; ECT, emission computed tomography; FIB, fibrinogen; HIS, hospital information system; fPSA, free prostate specific antigen; MRI, nuclear magnetic resonance imaging; PCa, prostate cancer; PSA, prostate specific antigen; PSAD, prostate specific antigen density; ROC, receiver operating characteristic; SII, systemic immune-inflammation index; SIRI, systemic inflammatory response index; tPSA, total prostate specific antigen.
1. Introduction
According to GLOBOCAN 2020, prostate cancer (PCa) is the second most common cancer in men worldwide and the fifth leading cause of cancer death in men [1]. In 2020, about 1414,259 cases of PCa were newly diagnosed worldwide, including 375,304 cancer-related deaths [1]. Current epidemiological surveys generally agree that PCa incidence is lower in Asian countries than in Western countries [2,3]. According to GLOBOCAN 2020, the age-standardized incidence rate of PCa in Chinese men in 2020 was 10.2/100,000, and the mortality rate of PCa patients in China was about 4.6/100,000, which were significantly lower than those in Western countries [1]. However, due to the large population in China, new cases of PCa in China account for 8.2% of the world and 13.6% of deaths [1], both indicators suggest that PCa is becoming a killer of men ‘s health in China.
It has been shown that PCa can be metastasized to different organs, but is more likely to metastasized to bone [4]. The spine, pelvis, and ribs are the most common sites of bone metastases [5]. Although PCa has a long course and slow progression, patients often have no specific discomfort at the beginning of the disease, the onset is occult, and the clinical progression of individual patients is uncertain, the tendency to develop bone metastases remains strong. Although the primary tumor can be detected early in many patients, bone metastases have been detected in up to 10% of patients with a primary diagnosis of PCa. In addition, 20% to 30% of patients with T1-T3 PCa relapse and develop advanced disease after radical prostatectomy, and bone metastases may occur in 70% to 80% of these patients [1]. Therefore, it is particularly important to predict the risk of developing bone metastases early and intervene early.
Previous studies has indicated that the risk factors for bone metastasis (BM) in PCa include age, tumor stage, chemoradiotherapy, Gleason score, PSA, etc. [6–9]. However, as they paid little attention to the patients’ personal life history and the clinical physical examination results, those findings encounters difficulties in clinical application. In this study, several factors affecting the occurrence of BM in PCa were obtained by using the data of clinical patients to predict the risk of BM in PCa early, providing a theoretical basis for the clinical prevention of BM treatment.
2. Patients and methods
2.1. Patients and data collection
We retrospectively collected data from two hospitals through the electronic medical record system on patients diagnosed with prostate cancer and prostate cancer bone metastasis between January 1, 2013 and December 31, 2022.Among them, the clinical data obtained from the Affiliated Hospital of Yangzhou University were training datasets for analyzing and constructing nomogram models, and the data from the Affiliated Clinical College of Traditional Chinese Medicine of Yangzhou University were validation datasets for model validation. The study was conducted in accordance with the Declaration of Helsinki. It was reviewed and approved by the Ethics Committee of Medical College at Yangzhou University (Ethics No.YXYLL-2021-148) and we are allowed to carry out this study from June 1, 2021 to June 30, 2022.
2.1.1. Inclusion and exclusion criteria.
The inclusion and exclusion criteria of patients in the two groups followed the following criteria: 1. Inclusion criteria: PCa confirmed by prostate biopsy or surgery; Emission computed tomography (ECT) examination of systemic bone scan showed BM or no BM; Newly diagnosed cases; 2. Exclusion criteria: Patients with previous bone-related diseases; Patients with other malignant tumors; Patients with other malignant tumors and received radiotherapy, chemotherapy, immunotherapy and other treatments; Bone metastases caused by other malignant tumors; Patients who have experienced fracture, surgery and other related treatments within one month before the initial diagnosis of PCa.
Since patients with previous bone related diseases or tumor will affect the accuracy of hematological indicators and thus affect the construction of the final model, they were excluded from the study. The purpose of setting these exclusion criteria is to exclude those patients who have bone related diseases themselves, because patients have bone related diseases and PCa, and their bone related diseases will affect the accuracy of hematological indicators and thus affect the construction of the final model, so relevant criteria are set.
2.1.2. Data collection.
Relevant data were exported and collected from the hospital information system(HIS). 1. Pathological data: The specimen sections of all patients were made in the data collection hospital, and were diagnosed as PCa by the pathologist of the hospital, and graded according to the Gleason grading criteria to obtain the Gleason Score; 2. Laboratory parameters: The blood routine, coagulation function, liver function, prostate tumor indicators of the patients before prostate puncture or before surgery were collected. All relevant indicators are the data detected by the clinical laboratory of the hospital where the data were collected. Other relevant modeling indicators were obtained by further calculation; 3. Imaging indicators: prostate color ultrasound, abdominal computed tomography(CT), prostate magnetic resonance imaging(MRI) and whole-body bone scan were collected from the patients. All relevant reports were issued by the Department of Radiology of the data collection hospital. Bone metastases were confirmed by reading by the Radiologist of this hospital. At the same time, the clinical T stage of the patients was determined according to the imaging findings; 4. Other indicators: patient age, past medical history, personal history, physician physical examination results of the patients and other data were obtained from the patient ‘medical history data.
We retrospectively collected the clinical data of patients with PCa who were initially pathologically diagnosed by prostate biopsy or surgery obtained during a 10-year period from January 1, 2013 to December 31, 2022 in two hospitals. A total of 204 cases were collected from the Affiliated Hospital of Yangzhou University, including 64 cases diagnosed as PCa BM and 50 cases collected from the Affiliated Clinical College of Traditional Chinese Medicine of Yangzhou University, including 12 cases diagnosed as PCa BM.
2.2. Statistical analysis
Data analysis was performed using SPSS software (v.25.0) and R software (R-4.0.3) for visualization and further validation of the data. Among them, the measurement data were compared between groups and expressed as mean ± standard deviation (x̅ ± s) using the independent sample t-test. The enumeration data were expressed as rate (%) using the chi-square test. Univariate and multivariate logistic regression analysis were used to screen the predictors, and Hosmer-Lemesho tested the goodness of fit of each logistic regression model. R software was used to construct nomogram models and subsequent validation. The C-index and calibration curve were used to evaluate the accuracy of the nomogram. In addition, we also plotted the ROC curve and the area under the curve (AUC) to evaluate the accuracy of the nomogram. DCA identifies and compares clinical value between nomogram model and other clinical features by calculating the net benefit at each risk threshold probability. P-value < 0.05 was considered statistically significant.
3. Results
3.1. Analysis between groups
As shown in Table 1, there was no statistically significant difference in age (p = 0.851), smoking history (p = 0.381), hypertension history (p = 0.078), diabetes history (p = 0.270), coronary heart disease history (p = 0.856), dysuria (p = 0.11), and prostate volume (p = 0.865) between groups. However, history of alcohol consumption (p = 0.028), stiffness of prostate on digital rectal examination(DRE) (p < 0.001), presence of nodules in prostate on DRE (p = 0.004), systemic inflammatory response index (SIRI) (p = 0.011), systemic immune-inflammation index (SII) (p = 0.004), fibrinogen (FIB) (p = 0.006), alkaline phosphatase (ALP) (p < 0.001), albumin alkaline phosphatase ratio (AAPR) (p = 0.001), total prostate specific antigen(tPSA) (p < 0.001), free prostate specific antigen(fPSA) (p < 0.001), prostate specific antigen density (PSAD) (p < 0.001), clinical t stage x (cTx) (p < 0.001), and Gleason Score (p = 0.042) were statistically significant between groups. Among them, the values of FIB, ALP, tPSA, fPSA, PSAD and Gleason Score in BM group were (3.60 ± 1.11) g/L(299.11 ± 388.02) U/L(233.18 ± 285.58) ng/ml(26.14 ± 22.91) ng/ml(5.29 ± 8.21) ng/ml2 and (7.92 ± 1.22) points, respectively, which were higher than those in non-BM group, while SIRI, SII and AAPR in BM group were lower than those in non-BM group.
3.2. Logistic regression
We included factors with p < 0.05 in the between-group analysis in the logistic regression for further screening and the results are shown in Table 2 and 3. In Table 2, univariate logistic regression calculations showed a statistically significant regression of history of alcohol consumption (p = 0.033), prostate stiffness on DRE (p < 0.001), prostate nodules on DRE (p = 0.006), FIB (p < 0.001), ALP (p < 0.001), AAPR (p < 0.001), tPSA (p < 0.001), fPSA (p < 0.001), PSAD (p < 0.001), cTx (p < 0.001), and Gleason Score (p < 0.001) with PCa with or without bone metastases. Including the above indicators into multivariate logistic regression calculation (Table 3), it was found that history of alcohol consumption (p = 0.009), prostate stiffness on DRE (p = 0.025), prostate nodules on DRE (p = 0.002), FIB (p = 0.021), ALP (p = 0.007), cTx (pT3 = 0.006, pT4 = 0.034), and Gleason Score (p = 0.033) were risk factors for PCa BM.
3.3. ROC curve and nomogram
As shown in Fig 1A, the history of alcohol consumption, prostate stiffness on DRE, prostate nodules on DRE, FIB, ALP, cTx, and Gleason Score obtained in multivariate logistic regression were included in the construction of receiver operating characteristic(ROC) curves, and the predictive value of each factor for BM in newly diagnosed PCa patients was assessed according to the area under the curve. As shown in Table 4, the area under curve(AUC) values of the history of alcohol consumption, prostate stiffness on DRE, prostate nodules on DRE, FIB, ALP, cTx, and Gleason Score were 0.550, 0.701, 0.565, 0.645, 0.780, 0.758, and 0.729, respectively, which were lower than those when the seven were combined (AUC = 0.927), indicating that these seven played a synergistic role in the prediction of PCa BM and could be used as predictors of PCa BM together. Among them, we used the R package “ROCR” to calculate the Youden index, and the cutoff value corresponding to the maximum Youden index was the optimal diagnostic cut-off value. The Youden index of fibrinogen was 0.28 with a cutoff value of 3.495 g/L, and the Youden index of alkaline phosphatase was 0.579 with a cutoff value of 124.95 U/L. These two sets of cutoff values has been included in the model as stratification indicators for nomograms. As shown in Fig 1B, if history of alcohol consumption was excluded from the model, the AUC value of the combined model was 0.914, which was smaller than that of the combined seven.
(A). Seven parameters validated by multivariate logistic regression, the history of alcohol consumption, prostate stiffness on DRE, prostate nodules on DRE, FIB, ALP, cTx, and Gleason Score, were included in the calculation of ROC curves, and the AUC value of the final model was 0.927 (P < 0.05). (B). Subgroup analysis removing the factor drinking history revealed an AUC value of 0.914 for the ROC curve.
As shown in Fig 2, the history of alcohol consumption, prostate stiffness on DRE, prostate nodules on DRE, FIB, ALP, cTx, and Gleason Score obtained in multivariate logistic regression were included in the construction of the nomogram model.
Factors with p < 0.05 in multivariate logistic regression were included in the nomogram model. The c-index of the model was 0.937.
3.4. Evaluation of nomogram
As shown in Fig 3A, the calibration plot revealed good predictive accuracy between the actual probability and predicted probability. As shown in Fig 3C, the nomogram model used to predict PCa BM risk presented good agreement across the training set cohort. Meanwhile, the c-index of this nomogram model was 0.937 (95% CI: 0.899 – 0.975). The nomogram model also passed the validation of the validation set with a c-index of 0.929, which also illustrates the good predictive ability of the model.
(A). Calibration curve for nomogram model. The calibration curve of the model in this study is close to the ideal curve. (B). Decision curves for nomogram models. Intervention of patients within a threshold probability of 0.02 – 0.92 can be clinically beneficial. (C). ROC curves for the validation set validation nomogram model. The AUC value of the curve was 0.928534, indicating that the model predicted well.
3.5. Clinical benefit assessment
Fig 3B shows the decision curve for the nomogram model. The decision curve suggests that there may be more benefit in the clinical use of this nomogram model to predict prostate cancer bone metastasis risk if the threshold probability is between 0.02 and 0.92.
4. Discussion
BM is the most common site of metastasis in PCa, and early detection is important in clinical treatment. In our study, we used nomograms as modeling models, which are widely used in oncology and medicine to predict the risk or prognosis of diseases and can help better clinical decision-making. Seven clinically most common parameters, history of alcohol consumption, prostate stiffness on DRE, prostate nodules on DRE, FIB, ALP, cTx, and Gleason Score, were used in this study to participate in the construction of the nomogram model. These seven factors cover the personal history of patients, clinical examination results of doctors, blood biochemical results, coagulation function, imaging examination, pathological results, etc., which can comprehensively evaluate the physical status of PCa patients and their risk of BM, which is conducive to clinicians to apply and early predict and prevent the occurrence of BM of PCa. Bai G et.al. recently used six factors including total prostate-specific antigen, clinical tumor stage, Gleason score, prostate volume, red cell distribution width and serum alkaline phosphatase as the basis for constructing a nomogram model to predict the BM of PCa [9]. However, they did not include the analysis of the patient ‘s personal life history and the clinical physical examination results in their model, which will lose important information like the patient ‘s lifestyle habits on the disease and the clinical physical examination results, and hence sacrifice the accuracy of the prediction. In addition, we found that some articles discussed PCa and BM of PCa from the perspective of prostate MRI, which provided new ideas for the diagnosis and prognosis of prostate cancer and prostate cancer bone metastasis [10,11]. As far as the authors’ clinical practice is concerned, prostate magnetic resonance imaging is a very important test in urology, and in combination with changes in PSA levels, it is almost certain whether patients with benign prostatic hyperplasia or hypertrophy develop prostate cancer. It also plays an important role in cancer development and prognosis. Unfortunately, when reviewing the clinical data, it was found that many patients did not undergo MRI which is mainly due to the economic reason. Although MRI data were not evaluated, our model still passed the evaluation and validation of c-index, calibration curve and decision curve, demonstrating its good predictive ability. Our model also passed external validation, which demonstrates the validity of the model.
After univariate and multivariate logistic regression analysis (Table 2 and 3), it was found that uniquely, the history of alcohol consumption was included in the model as an important factor, which was less frequently identified as a key factor in previous studies. As shown in Fig 1B, if history of alcohol consumption was not included in the model, the AUC value of the model was lower than when it was included, indicating that history of alcohol consumption was an important factor in the model. Several studies have shown that long-term alcohol consumption can lead to oral cancer, pharyngeal cancer, laryngeal cancer, esophageal cancer, colorectal cancer, liver cancer, breast cancer, gastric cancer, pancreatic cancer, PCa and other cancers [12–15]. Alcohol, or acetaldehyde, can induce inflammatory responses and oxidative stress, and its metabolite acetaldehyde can lead to DNA damage as well as disrupt DNA methylation, all of which may lead to the development of cancer [16]. A study of the relationship between PCa and alcohol proposed that alcohol accelerates the growth of prostate tumors and significantly shortens the progression time of PCa metastasis [17], while alcohol reduces the effect of androgen deprivation therapy, resulting in the occurrence of PCa recurrence and metastasis [18–22]. The model included in this study may also be related to geographical reasons, and the proportion of men drinking alcohol has historically been higher in Jiangsu, so the probability of alcohol-induced diseases is also greater. Although the AUC value of history of alcohol consumption in this study was 0.550, which is insufficient to predict as an independent risk factor for BM in PCa, the effect on the model after inclusion in the model remains very large.
Hematological parameters also play an important role in PCa and PCa BM. In this study, FIB and ALP were found to be predictive factors for the diagnosis of bone metastases in PCa. FIB is highly expressed in various tumor tissues such as melanoma, lung cancer, renal cell carcinoma, and breast cancer, while it is most significantly expressed in breast cancer and is an independent risk factor for the prognosis of breast cancer patients [23]. FIB can provide structural support and scaffold for cancer cell metastasis and adhesion ligands for cell attachment and crawling [24], so FIB is equally important for the study of metastatic cancer. In the study of metastatic factors of gallbladder carcinoma, the experimenter found that FIB can promote tumor cell migration, cell adhesion, transendothelial cell migration, promote angiogenesis, increase vascular endothelial permeability, thereby promoting tumor cell metastasis and extravasation [25]. In a clinical analysis, a study of risk factors for bone metastases in patients with non-small cell lung cancer found that FIB existed as an independent risk factor [26]. In another study of the correlation between BM burden and FIB level in PCa, a positive relationship was found between FIB level and BM burden, and the higher the FIB level, the greater the probability of BM in PCa patients, and FIB was an independent risk factor for BM status [27]. Similarly, elevated expression levels of ALP are associated with the prognosis of a variety of malignancies [28–31] and ALP has prognostic value in castration-resistant PCa. High ALP levels have been shown to be associated with skeletal complications, lower survival rates, and bone metastases [32–34]. This is consistent with our findings that FIB and ALP can be included in the model as risk factors for BM in PCa. As drinking alcohol has been indicated to be related to the increase of alkaline phosphatase level [35–38], it justifies the reasonability of incuding both drinking history and alkaline phosphatase level in our nomogram model. Unfortunately, tPSA, fPSA, PSAD as indicators of PCa-specific tumor markers were not included in the model, which may be related to the small number of cases we collected, and we also need a large number of clinical samples as support for subsequent analysis. The same may exist, and important indicators commonly used to assess systemic inflammation, immunity, and nutritional status of patients, namely SIRI, SII, and AAPR, have not been included in the final model.
The cTx, Gleason Score, and DRE are extremely important factors for the diagnosis and treatment of PCa and judging the metastasis of PCa, as well as classical diagnostic factors. In our study, the three were also present as very important factors. It has been confirmed that the higher the T stage, the higher the Gleason Score, and the greater the risk of PCa BM [39], which is consistent with the results of this study. According to the nomogram model, the higher the T stage, the higher the Gleason Score, and the higher the score, the greater the risk of BM. ROC curves also showed the same characteristics. DRE has been the cornerstone of PCa detection [40]. Abnormal DRE is associated with an increased risk of PCa and is an important indication for PCa biopsy in both European and the United States guidelines for the diagnosis of PCa [41,42]. When PSA is < 3 ng/ml and there are abnormal DRE results, DRE can be used as a tool for multiple screenings [43–45]. At the same time, DRE results are one of the factors of a variety of classical clinical risk prediction calculation formulas, which include the European Randomised Study of Screening for Prostate Cancer risk calculator [46] and the Prostate Biopsy Collaborative Group calculator [47].
Overall, history of alcohol consumption may be included as the first time in PCa BM models, and it remains irreplaceable. Factors such as prostate stiffness on DRE, prostate nodules on DRE, FIB, ALP, cTx, and Gleason Score also play important roles in the construction of nomograms to predict the risk of prostate BM. Based on our nomogram model, especially through the novel key risk factors, the diagnostic accuracy of PCa BM can be improved greatly.
5. Limitations
Although our model predicts well, our study still has some limitations. First, limitations of the population. We selected PCa patients only from two of the three Class A tertiary hospitals in the local municipal area and did not include all similar patients from all hospitals in the area. Although these two tertiary hospitals are representative in this region, this model may not fully applicable to all PCa patients in China. Second, limitations of risk factors. We included PCa-related risk factors in detail, but parameters such as PSA could not be included in the final model due to population limitations, however this may also provide new ideas for the prediction of PCa BM. Third, limitations of applicability. Nomogram is only one of many prediction models. It has advantages such as easy to understand and apply clinically, however, it can still be improved in terms of prediction accuracy and multivariate processing ability. In our further studies, multiple machine learning, deep neural network, SHapley Additive ExPlanations and other methods are planed to be invloved to make improvements. Lastly, as this study is a retrospective study and we only selected the complete patient data. Missing data are therefore manageable. However, there still could exist unmeasured confounders. In the subsequent study, we will further adjust the selection strategy to strive for multi-regional, multi-center patient data in the hope of obtaining a better model.
6. Conclusion
Our study identified history of alcohol history, prostate stiffness on DRE, prostate nodules on DRE, FIB, ALP, cTx, and Gleason score as seven factors as high risk factors for PCa BM. Alcohol history was included in the model as a novel factor and played an important role in the model, and PCa patients with a previous history of alcohol consumption had a greater risk of BM. DRE of the prostate as a routine screening tool for PCa. Patients are at greater risk of later bone metastases if the prostate is firm to the touch or nodules are palpable on the prostate surface on DRE. Moreover, as shown in Fig 2, when FIB was greater than 3.495 g/L and ALP was greater than 124.95 U/L, FIB and ALP were present as high risk factors for PCa BM and were positively correlated with the degree of risk. At the same time, the more advanced the tumor stage and the higher the Gleason score, the greater the risk of secondary BM in PCa patients. The nomogram model involves four aspects: the basic situation of the patient, the physical examination of the clinician, the hematological examination results, and the imaging examination results. This allows a more comprehensive and accurate prediction of the likelihood of BM in PCa patients and is more conducive to clinical decision-making. In the future clinical application of this study model, it can be discussed in combination with prostate MRI results or biased to be applied in patients with a history of alcohol consumption, which can make the model more accurate.
Supporting information
S1 File. Raw data.
This file provided all of the information of patients diagnosed with prostate cancer and bone metastases of prostate cancer in this study.
https://doi.org/10.1371/journal.pone.0318051.s001
(XLSX)
Acknowledgments
The authors thank all members of Dr. Wenji Li’s laboratory for their strong support.
References
- 1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, et al. Global Cancer Statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71(3):209–49. Epub 2021/02/05. pmid:33538338
- 2. Center MM, Jemal A, Lortet-Tieulent J, Ward E, Ferlay J, Brawley O, et al. International variation in prostate cancer incidence and mortality rates. Eur Urol. 2012;61(6):1079–92. Epub 2012/03/20. pmid:22424666
- 3. Ito K. Prostate cancer in Asian men. Nat Rev Urol. 2014;11(4):197–212. Epub 2014/03/07. pmid:24595118
- 4. Gandaglia G, Abdollah F, Schiffmann J, Trudeau V, Shariat SF, Kim SP, et al. Distribution of metastatic sites in patients with prostate cancer: a population-based analysis. Prostate. 2014;74(2):210–6. Epub 2013/10/18. pmid:24132735
- 5. Kakhki VRD, Anvari K, Sadeghi R, Mahmoudian A-S, Torabian-Kakhki M. Pattern and distribution of bone metastases in common malignant tumors. Nucl Med Rev Cent East Eur. 2013;16(2):66–9. Epub 2013/09/27. pmid:24068635
- 6. Zhang H, Jiang X, Jiao L, Sui M. Development and external validation of a novel prognostic nomogram for overall survival in prostate cancer patients with bone metastatic: a retrospective study of the SEER-based and a single Chinese center. (1432-1335 (Electronic)).
- 7. Lu YJ, Duan WM. Establishment and validation of a novel predictive model to quantify the risk of bone metastasis in patients with prostate cancer. (2223-4691 (Print)).
- 8. Miyoshi Y, Noguchi K, Yanagisawa M, Taguri M, Morita S, Ikeda I, et al. Nomogram for overall survival of Japanese patients with bone-metastatic prostate cancer. (1471-2407 (Electronic)).
- 9. Bai G, Cai ZA-O, Zhai X, Xiong J, Zhang F, Li HA-OX. A new nomogram for the prediction of bone metastasis in patients with prostate cancer. (1473-2300 (Electronic)).
- 10. Arita Y, Takahara T, Yoshida S, Kwee TC, Yajima S, Ishii C, et al. Quantitative assessment of bone metastasis in prostate cancer using synthetic magnetic resonance imaging. (1536-0210 (Electronic)).
- 11. Arita Y, Akita H, Fujiwara H, Hashimoto M, Shigeta K, Kwee T. Synthetic magnetic resonance imaging for primary prostate cancer evaluation: diagnostic potential of a non-contrast-enhanced bi-parametric approach enhanced with relaxometry measurements. J Magn Reson Imaging. 2020;51(1):123–30.
- 12. Buykx P, Li J, Gavens L, Hooper L, Lovatt M, Gomes de Matos E, et al. Public awareness of the link between alcohol and cancer in England in 2015: a population-based survey. BMC Public Health. 2016;16(1):1194. Epub 2016/12/03. pmid:27899099
- 13. Manthey J, Shield KD, Rylett M, Hasan OSM, Probst C, Rehm J. Global alcohol exposure between 1990 and 2017 and forecasts until 2030: a modelling study. Lancet. 2019;393(10190):2493–502. Epub 2019/05/12. pmid:31076174
- 14. Bagnardi V, Rota M, Botteri E, Tramacere I, Islami F, Fedirko V, et al. Alcohol consumption and site-specific cancer risk: a comprehensive dose-response meta-analysis. Br J Cancer. 2015;112(3):580–93. Epub 2014/11/26. pmid:25422909
- 15. Rumgay H, Murphy N, Ferrari P, Soerjomataram I. Alcohol and Cancer: Epidemiology and Biological Mechanisms. Nutrients. 2021;13(9):3173. Epub 2021/09/29. pmid:34579050
- 16. Teissedre P-L, Rasines-Perea Z, Ruf J-C, Stockley C, Antoce AO, Romano R, et al. Effects of alcohol consumption in general, and wine in particular, on the risk of cancer development: a review. OENO One. 2020;54(4):813–32.
- 17. Macke AJ, Petrosyan A. Alcohol and prostate cancer: time to draw conclusions. Biomolecules. 2022;12(3):375. Epub 2022/03/26. pmid:35327568
- 18. Umathe S, Bhutada P, Dixit P, Shende V. Increased marble-burying behavior in ethanol-withdrawal state: modulation by gonadotropin-releasing hormone agonist. Eur J Pharmacol. 2008;587(1–3):175–80. Epub 2008/05/02. pmid:18448097
- 19. Tanvetyanon T. Physician practices of bone density testing and drug prescribing to prevent or treat osteoporosis during androgen deprivation therapy. Cancer. 2005;103(2):237–41. Epub 2004/12/15. pmid:15597384
- 20. Panday K, Gona A, Humphrey MB. Medication-induced osteoporosis: screening and treatment strategies. Ther Adv Musculoskelet Dis. 2014;6(5):185–202. Epub 2014/10/25. pmid:25342997
- 21. Diamond TH, Bucci J, Kersley JH, Aslan P, Lynch WB, Bryant C. Osteoporosis and spinal fractures in men with prostate cancer: risk factors and effects of androgen deprivation therapy. J Urol. 2004;172(2):529–32. Epub 2004/07/13. pmid:15247721
- 22. Owen PJ, Daly RM, Livingston PM, Fraser SF. Lifestyle guidelines for managing adverse effects on bone health and body composition in men treated with androgen deprivation therapy for prostate cancer: an update. Prostate Cancer Prostatic Dis. 2017;20(2):137–45. Epub 2017/01/25. pmid:28117386; PMCID: PMC5508230
- 23. Kwaan HC, Lindholm PF. Fibrin and Fibrinolysis in Cancer. Semin Thromb Hemost. 2019;45(4):413–22. Epub 2019/05/02. pmid:31041799
- 24. Weisel JW, Litvinov RI. Mechanisms of fibrin polymerization and clinical implications. Blood. 2013;121(10):1712–9. Epub 2013/01/12. pmid:23305734; PMCID: PMC3591795
- 25. Jiang C, Li Y, Li Y, Liu L, Wang X-A, Wu W, et al. Fibrinogen promotes gallbladder cancer cell metastasis and extravasation by inducing ICAM1 expression. Med Oncol. 2022;40(1):10. Epub 2022/11/10. pmid:36352295
- 26. Li Y, Xu C, Yu Q. Risk factor analysis of bone metastasis in patients with non-small cell lung cancer. Am J Transl Res. 2022;14(9):6696–702. Epub 2022/10/18. pmid:36247263; PMCID: PMC9556442
- 27. Xie G-S, Li G, Li Y, Pu J-X, Huang Y-H, Li J-H, et al. Clinical association between pre-treatment levels of plasma fibrinogen and bone metastatic burden in newly diagnosed prostate cancer patients. Chin Med J (Engl). 2019;132(22):2684–9. Epub 2019/11/15. pmid:31725446; PMCID: PMC6940101
- 28. Pochanugool L, Subhadharaphandou T, Dhanachai M, Hathirat P, Sangthawan D, Pirabul R, et al. Prognostic factors among 130 patients with osteosarcoma. Clin Orthop Relat Res. 1997;(345):206–14. Epub 1998/01/07. pmid:9418642
- 29. Liu F, Zhao J, Xie J, Xie L, Zhu J, Cai S, et al. Prognostic risk factors in patients with bone metastasis from colorectal cancer. Tumour Biol. 2016. Epub 2016/10/14. pmid:27734341
- 30. Lim SM, Kim YN, Park KH, Kang B, Chon HJ, Kim C, et al. Bone alkaline phosphatase as a surrogate marker of bone metastasis in gastric cancer patients. BMC Cancer. 2016;16:385. Epub 2016/07/06. pmid:27377907; PMCID: PMC4932725
- 31. Huang P, Lan M, Peng A-F, Yu Q-F, Chen W-Z, Liu Z-L, et al. Serum calcium, alkaline phosphotase and hemoglobin as risk factors for bone metastases in bladder cancer. PLoS One. 2017;12(9):e0183835. Epub 2017/09/14. pmid:28902911; PMCID: PMC5597169
- 32. Brown JE, Cook RJ, Major P, Lipton A, Saad F, Smith M, et al. Bone turnover markers as predictors of skeletal complications in prostate cancer, lung cancer, and other solid tumors. J Natl Cancer Inst. 2005;97(1):59–69. Epub 2005/01/06. pmid:15632381
- 33. Coleman RE, Major P, Lipton A, Brown JE, Lee K-A, Smith M, et al. Predictive value of bone resorption and formation markers in cancer patients with bone metastases receiving the bisphosphonate zoledronic acid. J Clin Oncol. 2005;23(22):4925–35. Epub 2005/06/29. pmid:15983391
- 34. Armstrong AJ, Febbo PG. Using surrogate biomarkers to predict clinical benefit in men with castration-resistant prostate cancer: an update and review of the literature. Oncologist. 2009;14(8):816–27. Epub 2009/08/18. pmid:19684076
- 35. Graham GJ, Rennie JS. Serum gamma glutamyl transferase and alkaline phosphatase as indicators of excess chronic alcohol consumption in the rat. Biochem Pharmacol. 1990;39(10):1615–7. pmid:1970932
- 36. Devgun MS, Dunbar JA, Hagart J, Martin BT, Ogston SA. Effects of acute and varying amounts of alcohol consumption on alkaline phosphatase, aspartate transaminase, and gamma-glutamyltransferase. Alcohol Clin Exp Res. 1985;9(3):235–7. pmid:2861758
- 37. Gluud C, Andersen I, Dietrichson O, Gluud B, Jacobsen A, Juhl E. Gamma-glutamyltransferase, aspartate aminotransferase and alkaline phosphatase as markers of alcohol consumption in out-patient alcoholics. Eur J Clin Invest. 1981;11(3):171–6. pmid:6115756
- 38. Nishmura M, Teschke R. Effect of chronic alcohol consumption on the activities of liver plasma membrane enzymes: gamma-glutamyltransferase, alkaline phosphatase and 5’-nucleotidase. Biochem Pharmacol. 1982;31(3):377–81. pmid:6122453
- 39. Swanson GP, Trevathan S, Hammonds KAP, Speights VO, Hermans MR. Gleason score evolution and the effect on prostate cancer outcomes. Am J Clin Pathol. 2021;155(5):711–7. pmid:33079976
- 40. Catalona WJ, Richie JP, Ahmann FR, Hudson MA, Scardino PT, Flanigan RC, et al. Comparison of digital rectal examination and serum prostate specific antigen in the early detection of prostate cancer: results of a multicenter clinical trial of 6,630 men. J Urol. 2017;197(2S):S200–7. Epub 2016/12/26. pmid:28012755
- 41. Loeb S, Gonzalez CM, Roehl KA, Han M, Antenor JAV, Yap RL, et al. Pathological characteristics of prostate cancer detected through prostate specific antigen based screening. J Urol. 2006;175(3 Pt 1):902–6. Epub 2006/02/14. pmid:16469576
- 42. Mottet N, van den Bergh RCN, Briers E, Van den Broeck T, Cumberbatch MG, De Santis M, et al. EAU-EANM-ESTRO-ESUR-SIOG guidelines on prostate cancer-2020 update. Part 1: screening, diagnosis, and local treatment with curative intent. Eur Urol. 2021;79(2):243–62. Epub 2020/11/07. pmid:33172724
- 43. Halpern JA, Oromendia C, Shoag JE, Mittal S, Cosiano MF, Ballman KV, et al. Use of digital rectal examination as an adjunct to prostate specific antigen in the detection of clinically significant prostate cancer. J Urol. 2018;199(4):947–53. Epub 2017/10/25. pmid:29061540; PMCID: PMC6719551
- 44. Beemsterboer PM, Kranse R, de Koning HJ, Habbema JD, Schröder FH. Changing role of 3 screening modalities in the European randomized study of screening for prostate cancer (Rotterdam). Int J Cancer. 1999;84(4):437–41. Epub 1999/07/15. pmid:10404100;
- 45. Schröder FH, Roobol-Bouts M, Vis AN, van der Kwast T, Kranse R. Prostate-specific antigen-based early detection of prostate cancer--validation of screening without rectal examination. Urology. 2001;57(1):83–90. Epub 2001/02/13. pmid:11164149
- 46. Roobol MJ, van Vugt HA, Loeb S, Zhu X, Bul M, Bangma CH, et al. Prediction of prostate cancer risk: the role of prostate volume and digital rectal examination in the ERSPC risk calculators. Eur Urol. 2012;61(3):577–83. Epub 2011/11/23. pmid:22104592
- 47. Ankerst DP, Straubinger J, Selig K, Guerrios L, De Hoedt A, Hernandez J, et al. A contemporary prostate biopsy risk calculator based on multiple heterogeneous cohorts. Eur Urol. 2018;74(2):197–203. Epub 2018/05/21. pmid:29778349; PMCID: PMC6082177