The prognosis of different distant metastases pattern in prostate cancer: A population based retrospective study

Transcription

1 Received: 21 September 2017 Accepted: 22 January 2018 DOI: /pros ORIGINAL ARTICLE The prognosis of different distant metastases pattern in prostate cancer: A population based retrospective study Jiafeng Shou Qi Zhang Shuai Wang Dahong Zhang Department of Urology, Zhejiang Provincial People's Hospital, Hangzhou Medical College, Hangzhou, Zhejiang Province, People's Republic of China Correspondence Department of Urology, Zhejiang Provincial People's Hospital, Hangzhou Medical College, 158 Shangtang Road, Hangzhou, Zhejiang Province , People's Republic of China @zju.edu.cn Funding information Science and Technology Agency Research Foundation of Zhejiang Province, Grant number: 2013C ; National Natural Science Foundation of China, Grant number: ; Medicine and Health Research Foundation of Zhejiang Province, Grant number: Background: The present of metastases is a poor prognostic factor in prostate cancer, but the prognostic impact of different distant metastases pattern is unclear. The aim of this study is to investigate the impact of different distant metastases pattern on the survival of patients with stage IV prostate cancer. Methods: Data queried for this study include prostate cancer ( ) from the Surveillance, Epidemiology, and End Results (SEER) program. Metastatic distribution information was provided for bone, brain, liver and lung. The overall survival was calculated by the Kaplan-Meier method. Multivariable Cox regression models were used to analyze survival outcome and risk factors. Results: A total of eligible patients were identified from SEER database. Among these patients, stage of IV prostate cancer accounted for 7.53% (20 034/ ) at diagnosis. Patients who suffered metastasis to either one of the four organs occupied 61.24% (12 268/20 034) in stage of IV patients. Comparing with other three single metastases, the patients with liver metastasis exhibited worst OS whose mean survival was months (P < 0.001). The mean survival of metastases with bong and lung was months, which was the best survival of the six forms with two metastatic sites (P < 0.001). The results of univariate survival analysis showed that metastatic forms, race, N-classification and differentiated grade did not have impact on the overall survival of patients with three metastatic sites (all, P > 0.05). Conclusions: In analysis of both one and two metastatic sites, patients with liver metastasis seemed to have worse survival outcome. On the other hand, bone metastasis had better outcome than other three visceral metastases. Knowledge of these differences in metastatic patterns may help to better guide pre-treatment evaluation of prostate cancer and make determination regarding curative-intent interventions. KEYWORDS distant metastasis, overall survival, prostate cancer 1 INTRODUCTION As the most common malignancy in male of United States, prostate cancer is expected to take up 19% of all new cancer cases and 8% of all cancer deaths in men. 1,2 Despite improvement in surgery, radiotherapy, chemotherapy, and endocrinetherapy, most deaths of patients with prostate cancer (PCa) occurred due to distant metastasis. The bone and lymph node metastases have been recognized the most typical metastatic pattern of extraprostatic tumor spread. 3 5 Nevertheless, several studies reported that up to The Prostate. 2018;1 7. wileyonlinelibrary.com/journal/pros 2018 Wiley Periodicals, Inc. 1

2 2 SHOU ET AL. 15% of patients with PCa might be affected by atypical metastases at diagnosis, including brain, liver, and lung metastases, etc. 6,7 Visceral involvement of prostate cancer might be associated with worse outcome. Several studies have focused on evaluating the prognosis of different metastases in metastatic castration-resistant prostate cancer (mcrpc) However, to date little attention has been focused on the prognostic significance of distant metastatic pattern of IV PCa patients at the first diagnosis. Since knowledge of prognosis of different distant metastatic pattern is crucial for pre-treatment evaluation of IV PCa patients, our study aimed to describe the distant metastatic site, frequency of occurrence and pattern of these metastases based on a large population using SEER database. Importantly, we investigated the prognostic significance of different distant metastatic pattern in included patients. On the other hand, surviving rates of PCa patients are rising because of earlier diagnosis and more optional treatments. As a result, likelihood of encountering distant metastases from prostate cancer was rising. 13,14 Hence, heightened awareness of prognosis of different distant metastatic pattern in PCa patients at the first diagnosis might also help make suitable medical decision for IV PCa patients developed from early stages after diverse treatments. 2 MATERIALS AND METHODS 2.1 Database and patient selection The Surveillance, Epidemiology, and End Results (SEER) program, is a population-based cancer registry covering approximately 26% of African Americans, 38% of Hispanics, 44% of American Indians, and Alaska Natives, 50% of Asians, and 67% of Hawaiian/Pacific Islanders. A total of 18 population-based cancer registries in the United States were included in the current SEER database. The specific inclusion criteria for prostate cancers were as follows: (1) the year of diagnosis from 2010 to 2014; (2) the diagnosis was confirmed microscopically with active follow-up and confirmed age; and (3) the enrolled patients should have confirmed metastatic information of bone, brain, liver, and lung. Patients with benign or borderline tumors, unknown age, and unknown survival months were excluded. We used SEER*Stat software (SEER*Stat 8.2.3) to identify eligible patients. 2.2 Statistical analysis Frequency distribution of demographic and clinicalpathogical characteristics across metastatic groups were compared using Pearson s chisquare tests. Overall survival (OS) was calculated from the date of diagnosis to the date of death from any cause. Survival curves were generated using Kaplan-Meier methods, and the log-rank test was carried out to evaluate the survival differences between groups. Comparative risks of mortality were evaluated using multivariate Cox proportional hazards regression models. All statistical tests were 2-sided, and P < 0.05 was considered statistically significant. The statistical software SPSS 22.0 (SPSS Inc., Chicago, IL) was used for all data analyses. 3 RESULTS 3.1 Patient demographics and metastasis pattern A total of eligible patients were identified from SEER database. Among these patients, stage of IV prostate cancers accounted for 7.53% (20 034/ ) at the time of diagnosis. The SEER database only offered metastatic information of bone, brain, liver, and lung metastasis. Patients who suffered metastasis to either one of the four organs occupied 61.24% (12 268/20 034) in stage of IV patients. The clinical characteristics of these patients were presented in Table 1. The patients with either one of these four metastatic sites were significantly older than not (P < 0.001) at diagnosis. The distribution of race among patients with bone metastasis and without bone metastasis was significantly distinguishing (P<0.001). Similar phenomenon was observed in liver and lung metastasis (P<0.001) while not in brain metastasis (P>0.05). As shown in Table 1, the distribution of both nodal status, histological grade and insurance were significantly associated with these four metastatic sites (all, P < 0.001). The metastatic pattern of prostate cancer was presented in Table 2. For these four metastatic sites, there are 15 possible metastatic forms, including four single metastases and 11 combinations of metastases. Among stage of IV PCa, patients with only bone metastases accounted for 53.3% (10 683/20 034). The patients with only brain metastases occupied least of all, which was 0.1% (20/20 034). As for two sites, the combination of bone and lung metastases occupied most of stage of IV patients achieving to 3.4% (672/20 034). There was only one case for both liver and brain metastases. 3.2 Univariate survival analysis of patients with four single metastases As shown in Table 3, the prognoses of patients with four single metastases were further analyzed using univariate analysis. The results have showed that metastatic sites, race, N-classification, differentiated grade, and insurance status were prognostic factors affecting overall survival (OS) (all, P < 0.05). Comparing with other three single metastases, the patients with liver metastasis exhibited worst OS whose mean of survival was months. Nevertheless, there was no significant difference between patients with lymph node metastases and without lymph node metastases (N0: months; N1: months; P = 0.199). Similarly, there was also no significant difference between white and black people (white: months; black: months; P = 0.793). Interestingly, patients with moderate differentiated grade seemed to have better OS than well differentiated grade (well: months; moderate: ; P = 0.003). As our expected, the insured patients exhibited survival advantage comparing with uninsured patients (insured: months; uninsured: months; P = 0.003). Figure 1 exhibited the survival curves generated by Kaplan-Meier analyses using univariate model.

3 SHOU ET AL. 3 TABLE 1 Clinical features and metastasis sites Features Bone metastasis (%) Brain metastasis (%) Liver metastasis (%) Lung metastasis (%) No No Yes P No Yes P No Yes P No Yes P Age Mean (years) < < < <0.001 SD Race White (95.0) 9655 (5.0) < (99.9) 140 (0.1) > (99.7) 481 (0.3) < (99.6) 841 (0.4) <0.001 Black (94.1) 2271 (5.9) (99.9) 32 (0.1) (99.6) 144 (0.4) (99.4) 234 (0.6) Other (93.7) 779 (6.3) (99.9) 10 (0.1) (99.7) 33 (0.3) (99.4) 79 (0.6) Unknown 7558 (98.8) 94 (1.2) 7642 (100) 0 (0) 7641 (99.97) 2 (0.03) 7640 (99.95) 4 (0.05) Nodal status Nx 5325 (63.4) 3074 (36.6) < (99.4) 49 (0.6) < (98.0) 161 (2.0) < (97.0) 293 (3.0) <0.001 N (97.1) 6783 (2.9) (99.96) 86 (0.04) (99.9) 271 (0.1) (99.8) 490 (0.2) N (63.0) 2934 (37.0) 7815 (99.4) 47 (0.6) 7679 (97.2) 224 (2.8) 7489 (95.4) 364 (4.6) Histological grade Well (99.6) 60 (0.4) < (100) 0 (0) < (99.99) 1 (0.01) < (99.98) 4 (0.02) <0.001 Moderate (99.5) 490 (0.5) (99.99) 9 (0.01) (99.98) 21 (0.02) (99.97) 32 (0.03) Poorly (94.1) 7169 (5.9) (99.9) 64 (0.1) (99.7) 305 (0.3) (99.6) 547 (0.4) Undifferentiatd 386 (80.4) 94 (19.6) 477 (99.6) 2 (0.4) 466 (96.9) 15 (3.1) 461 (96.0) 19 (4.0) Unknown 9641 (65.9) 4986 (34.1) (99.3) 107 (0.7) (97.8) 318 (2.2) (96.1) 556 (3.9) Insurance Insured (95.4) 7616 (4.6) < (99.9) 103 (0.1) < (99.8) 393 (0.2) < (99.6) 677 (0.4) <0.001 Uninsured 3405 (86.7) 522 (13.3) 3899 (99.6) 14 (0.4) 3879 (99.1) 36 (0.9) 3856 (98.5) 60 (1.5) Unknown (98.3) 430 (1.7) (99.98) 5 (0.02) (99.9) 21 (0.1) (99.9) 38 (0.1) SD: Standard deviation

4 4 SHOU ET AL. TABLE 2 Frequencies of combination metastasis Prostate cancer Features Number Percentage (%) One site Only bone Only brain Only liver Only lung Two sites Lung and liver Lung and bone Lung and brain Liver and bone Liver and brain Bone and brain Three sites Lung and liver and bone Lung and liver and brain Liver and bone and brain Bone and brain and lung Four sites Lung and liver and bone and brain Multivariate survival analysis of patients with four single metastases Further multivariate analysis showed that metastatic site was an independent prognostic factor of OS (P < 0.05) (Table 4). In addition, race, N-classification and differentiated grade were also independent risk factor affecting OS (P < 0.05). Patients with liver metastases had worse OS than lung metastases (HR: 1.810, 95%CI: ; P = 0.028). The minimal HR could be found in bone metastases (HR: 0.897, 95%CI: ; P = 0.535) though there was no significantly difference comparing with lung metastases. Lymph node-negative had a positive effect on OS among patients with these four single metastases (HR: 0.909, 95%CI: ; P = 0.042). However, the multivariate analysis showed that there was no statistical difference between insured and uninsured groups (HR: 1.047, 95%CI: ; P = 0.663). The moderate (HR: 0.370, 95%CI: ; P < 0.001) and poorly (HR: 0.654, 95%CI: ; P = 0.007) differentiated patients both had better survival than undifferentiated cases. However, there was no statistically difference between well and undifferentiated groups (HR: 0.812, 95%CI: ; P = 0.539). 3.4 Univariate and multivariate survival analysis of patients with different combinations of metastases Patients with two metastatic sites had six forms, including bone with brain metastasis, bone with liver metastasis, bone with lung metastasis, brain with liver metastasis, brain with lung metastasis, and liver with lung metastasis. The results showed that the mean survival months of metastases with bong and lung was months, which was the best survival of the six forms of metastases using Kaplan-Meier methods (Supplementary Table S1). Survival curves for these six forms of metastases using univariate model was shown in Figure 2. There was only one case with metastases of brain and liver, whose survival months was 11. Similarly, the count of patients with metastases of brain and lung was only 2, whose mean survival months was The results of multivariate Cox proportional hazard regression analysis were shown in (Supplementary Table S2). Metastatic form was an independent risk factor affecting OS (P < 0.05) while race, N-classification, differentiated grade, and insurance status were all not (all, P > 0.05). We also compared the survival outcome of patients with three metastatic sites. There were four metastatic forms, including bone and brain and liver metastasis, bone and liver and lung metastasis, bone and brain and lung metastasis, brain and liver, and lung metastasis. The results of univariate survival analysis showed that metastatic forms, race, N-classification, and differentiated grade did not have impact on the OS (all, P > 0.05) (Supplementary Table S3). Survival curves for these six forms of metastases using univariate model was shown in Figure 3. Similar results were exhibited in the multivariate survival analysis (all, P > 0.05) (Supplementary Table S4). 4 DISCUSSION Although PCa commonly spreads to the lymph nodes and bone, visceral involvements are occurred in 15% of patients diagnosed with metastatic PCa. 3 Knowledge of clinical features and prognosis of different distant metastatic pattern is crucial to formulate medical plan. Currently, only few studies focused on this point that were based on small institutional cohorts. 6,15,16 In the present study, we identified risks of specific combinations of metastatic sites in PCa. These differences might have promising implications for clinical decisionmaking in those patients. As for the metastatic pattern of patients with only one site, we reaffirm that PCa predominantly metastasize to bone. 4,17 On the other hand, the most common sites of visceral metastases in patients without bone involvement were lung and liver. This is in line with what Gandaglia et al. 3 reported. Referring to demographics and clinical features of patients, we found that older people seemed to suffer distant metastases more frequently at diagnosis. We also found that black people had a higher risk of bone, liver, and lung metastases than white people. Of note, we found that uninsured patients had more metastases to bone, brain, liver, and lung than insured patients. In addition, patients with lymph node positive and poorly differentiated histological grade were more inclined to suffer distant metastases. We also made some findings in survival analysis. First, in both univariate and multivariate survival analysis of patients with four single metastases, we found that the patients with liver metastasis exhibited worst OS. Armstrong et al. 11 also reported that liver metastases were associated with shorter overall survival. Pouessel et al. 18 reported that

5 SHOU ET AL. 5 TABLE 3 Univariate survival analysis of patients with four single metastases Risk Factors Mean of survival months 95%CI P Metastasis site <0.001 Bone metastasis (29.993, ) Ref. Brain metastasis (15.891, ) Liver metastasis (13.570, ) Lung metastasis (28.212, ) < Race <0.001 White (29.467, Ref ) Black (29.004, ) Other (32.555, ) Unknown (33.574, ) < N-classification <0.001 Nx (24.679, < ) N (31.389, ) N (29.810, ) Differentiated grade Well (21.776, ) Moderate (41.068, ) Poorly (33.455, ) Undifferentiated (21.386, ) Unknown (22.500, ) Ref <0.001 Ref Insurance status Insured (29.824, Ref ) Uninsured (29.552, ) Unknown (26.950, ) liver metastases are frequently associated with the neuroendocrine differentiation (NED). NED in PCa appears to be a poor prognostic factor, possibly related to the increasing degree of dedifferentiation, loss of androgen receptor, and resistance to hormonal therapy. 19,20 FIGURE 1 Kaplan-Meier curves and Log-rank test for overall survival according to different metastasis (only one site). Note:1, Bone metastasis; 2, Brain metastasis; 3, Liver metastasis; 4, Lung metastasis The multivariate analysis revealed that patients with lymph node positive plus either one of other four metastases had worse overall survival than those with only distant organ metastases. Several studies reported that tumor cells metastasizing only to lymph nodes might specific epigenetic modifications that could prevent them spread to visceral organs. 21,22 On the other hand, these tumor cells could spread into vascular circulation and metastasize to distant organ only after subsequent neoplastic transformations. 23 This indicated that the tumor cells in lymph node of patients with both lymph node and distant organ metastases had a more aggressive phenotype. 16 Hence, we should focus on both distant organ metastases and lymph node status in clinical practice. In addition, poorly differentiated grade was associated with worse outcome other than the well-differentiated grade. Our results indicated that patients with both moderate and poorly differentiated grade seemed to have better OS than welldifferentiated grade. This phenomenon needs prospective study to be confirmed and basic research is warranted to explore the biological mechanisms. Furthermore, we found that there was no significant difference of prognosis between white and black people. Similar results were exhibited among insured or uninsured patients. There are no reports about prognosis of different metastatic combination of bone, brain, liver, and lung so far. In survival analysis of patients with combination of two metastatic sites, we found that patients with bone metastases plus either one of other three organ metastases had better OS than those without bone metastases. This implied that visceral metastases resulted in shorter survival than bone metastases, which was also confirmed in univariate and multivariate survival analysis of patients with only one metastasis. Distinguished from single distant metastases, our results showed that both N-classification and differentiated grade had no association with OS

6 6 SHOU ET AL. TABLE 4 Multivariate survival analysis of patients with four single metastases Risk Factors HR 95%CI P Metastasis site Bone metastasis (0.636, 1.265) Brain metastasis (0.519, 3.427) Liver metastasis (1.068, 3.068) Lung metastasis 1 Ref. Race White (1.115, 1.654) Black (1.208, 1.849) <0.001 Other 1 Ref. N-classification N (0.829, 0.996) N1 1 Ref. Differentiated grade <0.001 Well (0.418, 1.579) Moderate (0.253, 0.541) <0.001 Poorly (0.481, 0.889) Undifferentiated 1 Ref. Insurance status Insured (0.852, 1.288) Uninsured 1 Ref. FIGURE 3 Kaplan-Meier curves and Log-rank test for overall survival according to different metastasis (three sites). Note:1, Bone and brain and liver metastasis; 2, Bone and liver and lung metastasis; 3, Bone and brain and lung metastasis; 4, Brain and liver and lung metastasis in patients with two distant metastases. As for combination of three metastatic sites, we found that metastatic forms, race, N-classification, differentiated grade, and insurance status all did not have impact on the OS. This indicated that prognosis of patients with three distant metastases was predominantly decided by the metastases themselves. Meanwhile, there was no difference of prognosis among four combinations of three metastatic sites. Despitevaluablefindingsabove,thereareseverallimitations in our study due to the retrospective nature. First, we only have information on synchronous metastasistobone,brain,liver,and lung at diagnosis, a relative minority compared to those metastases developed from primary tumor during treatment and surveillance. Second, the lack of information regarding treatments of included patients might contribute to potential bias. Third, laboratory parameters including PSA level could not be obtained from SEER database. 5 CONCLUSION FIGURE 2 Kaplan-Meier curves and Log-rank test for overall survival according to different metastasis (two sites). Note:1, Bone and brain metastasis; 2, Bone and liver metastasis; 3, Bone and lung metastasis; 4, Brain and liver metastasis; 5, Brain and lung metastasis; 6, Liver and lung metastasis In conclusion, patients with liver metastasis seemed to have worst survival outcome among single metastases. On the other hand, patients with bone metastases plus either one of other three organ metastases had better OS than other forms of two sites. Patients with lymph node positive plus either one of other four metastases had worse OS than those with only distant organ metastases. Understanding of these differences in metastatic patterns may contribute to better pre-treatment evaluation of distant metastatic PCa and make appropriate determination regarding curative-intent interventions.

7 SHOU ET AL. 7 ACKNOWLEDGMENTS The authors would like to thank the members of the research group for useful discussions. This work was supported by Science and Technology Agency Research Foundation of Zhejiang Province (Grant No. 2013C ), Medicine and Health Research Foundation of Zhejiang Province (Grant No ) and National Natural Science Foundation of China (Grant No ). CONFLICTS OF INTEREST The authors declare that they have no competing interests. ORCID Dahong Zhang REFERENCES 1. Siegel RL, Miller KD, Jemal A. Cancer statistics, CA: Cancer J Clin. 2017;67: Jemal A, Center MM, DeSantis C, Ward EM. Global patterns of cancer incidence and mortality rates and trends. Cancer Epidemiol Biomarkers Prev. 2010;19: Gandaglia G, Abdollah F, Schiffmann J, et al. Distribution of metastatic sites in patients with prostate cancer: a population-based analysis. Prostate. 2014;74: Whitmore WF, Jr. Natural history and staging of prostate cancer. Urol Clin North Am. 1984;11: Pollard ME, Moskowitz AJ, Diefenbach MA, Hall SJ. Cost-effectiveness analysis of treatments for metastatic castration resistant prostate cancer. Asian J Urol. 2017;4: Vinjamoori AH, Jagannathan JP, Shinagare AB, et al. Atypical metastases from prostate cancer: 10-year experience at a single institution. AJR Am J Roentgenol. 2012;199: Long MA, Husband JE. Features of unusual metastases from prostate cancer. Br J Radiol. 1999;72: Pond GR, Sonpavde G, de Wit R, Eisenberger MA, Tannock IF, Armstrong AJ. The prognostic importance of metastatic site in men with metastatic castration-resistant prostate cancer. Eur Urol. 2014;65: Halabi S, Small EJ, Kantoff PW, et al. Prognostic model for predicting survival in men with hormone-refractory metastatic prostate cancer. J Clin Oncol. 2003;21: Omlin A, Pezaro C, Mukherji D, et al. Improved survival in a cohort of trial participants with metastatic castration-resistant prostate cancer demonstrates the need for updated prognostic nomograms. Eur Urol. 2013;64: Armstrong AJ, Garrett-Mayer ES, Yang YC, de Wit R, Tannock IF, Eisenberger M. A contemporary prognostic nomogram for men with hormone-refractory metastatic prostate cancer: a TAX327 study analysis. Clin Cancer Res. 2007;13: Halabi S, Lin CY, Kelly WK, et al. Updated prognostic model for predicting overall survival in first-line chemotherapy for patients with metastatic castration-resistant prostate cancer. J Clin Oncol. 2014;32: Quinn M, Babb P. Patterns and trends in prostate cancer incidence, survival, prevalence and mortality. Part II: individual countries. BJU Int. 2002;90: Collin SM, Martin RM, Metcalfe C, et al. Prostate-cancer mortality in the USA and UK in : an ecological study. Lancet Oncol. 2008;9: Bubendorf L, Schopfer A, Wagner U, et al. Metastatic patterns of prostate cancer: an autopsy study of 1,589 patients. Human Pathol. 2000;31: Furuya Y, Akakura K, Akimoto S, Ito H. Prognosis of patients with prostate carcinoma presenting as nonregional lymph node metastases. Urologia Internationalis. 1998;61: Coleman RE. Skeletal complications of malignancy. Cancer. 1997;80: Pouessel D, Gallet B, Bibeau F, et al. Liver metastases in prostate carcinoma: clinical characteristics and outcome. BJU Int. 2007;99: Wang W, Epstein JI. Small cell carcinoma of the prostate. A morphologic and immunohistochemical study of 95 cases. Am J Surg Pathol. 2008;32: Berruti A, Dogliotti L, Mosca A, et al. Circulating neuroendocrine markers in patients with prostate carcinoma. Cancer. 2000;88: Briganti A, Passoni NM, Abdollah F, Nini A, Montorsi F, Karnes RJ. Treatment of lymph node-positive prostate cancer: teaching old dogmas new tricks. Eur Urol. 2014;65: discussion Barekati Z, Radpour R, Lu Q, et al. Methylation signature of lymph node metastases in breast cancer patients. BMC Cancer. 2012; 12: Arya M, Bott SR, Shergill IS, Ahmed HU, Williamson M, Patel HR. The metastatic cascade in prostate cancer. Surg Oncol. 2006;15: SUPPORTING INFORMATION Additional Supporting Information may be found online in the supporting information tab for this article. How to cite this article: Shou J, Zhang Q, Wang S, Zhang D. The prognosis of different distant metastases pattern in prostate cancer: A population based retrospective study. The Prostate. 2018;1 7.