DOI 10.1007/s10147-015-0812-9 ORIGINAL ARTICLE Comparison of prognosis between patients with renal cell carcinoma on hemodialysis and those with renal cell carcinoma in the general population Yasunobu Hashimoto 1 Toshio Takagi 1 Tsunenori Kondo 1 Junpei Iizuka 1 Hirohito Kobayashi 1 Kenji Omae 1 Kazuhiko Yoshida 1 Kazunari Tanabe 1 Received: 8 January 2015 / Accepted: 25 February 2015 Japan Society of Clinical Oncology 2015 Abstract Objective We compared the clinical features and prognosis of renal cell carcinoma (RCC) between patients on hemodialysis () and those in the general population (). Methods We included a total of 1,794 patients who underwent surgery (, 408;, 1,386) and analyzed the clinical characteristics and oncological outcomes using a stage-for-stage analysis between the two groups. Results In the group, the mean duration of dialysis before surgery was 120 months. Compared to the group, the group tended to be younger (55 vs. 60 years, p < 0.0001) and more predominately male (84 % vs. 70 %, p < 0.0001), and the tumor size was smaller in this group (39 vs. 49 mm, p < 0.0001). The pathological characteristics of the group included a higher frequency of papillary tumors (22 % vs. 5 %, p < 0.0001) and stage I tumors (82 % vs. 68 %, p < 0.0001). During the follow-up period, 39 of patients (10 %) in the group and 193 patients (14 %) in the group died of cancer. The patients on hemodialysis had better cancer-specific survival (CSS) than their counterparts (p = 0.0292) in the univariable analysis, but no significance was found in the multivariable analysis. In the stage-for-stage analysis, the 5-year CSS was similar between the two groups for each stage. Conclusions CSS appeared to be better in the RCC- HD group than in the group, which may be * Toshio Takagi t.takagi1192@gmail.com 1 Department of Urology, Tokyo Women s Medical University, 8 1 Kawada cho, Shinjuku ku, Tokyo 162 8666, Japan associated with the higher incidence of stage I disease in the group. The comparable CSS between the groups in the stage-for-stage analysis supports this finding. Keywords Kidney neoplasm Prognosis Nephrectomy Hemodialysis Abbreviations RCC Renal cell carcinoma ESRD End-stage renal disease HD Hemodialysis CSS Cancer-specific survival ACDK Acquired cystic disease of kidney Introduction In patients with end-stage renal disease (ESRD), renal cell carcinoma (RCC) is well known to develop from acquired cystic disease of the kidney (ACDK) in patients on longterm hemodialysis (HD) [1]. The incidence of RCC in patients with ESRD is likely to be higher than that in the general population [2, 3]. In addition, long-term dialysis therapy has been reported to be associated with a high incidence of RCC in ESRD patients [4]. Considering the increasing number of patients with ESRD over the years [5], the management of RCC in these patients has become even more important. A multi-institutional study recently showed favorable clinical features and prognosis for patients with RCC and ESRD (ESRD-RCC) compared to those in the general population [6]. However, the mean interval between the initiation of dialysis therapy and the diagnosis of RCC was reported to be only 65 months [6]. The prognosis of patients with ESRD-RCC was also reported to be dependent on the
duration of dialysis therapy, and patients on long-term dialysis therapy for 10 years or longer showed worse survival than those with a dialysis duration of less than 10 years [7]. ESRD-RCC tended to be diagnosed at a more advanced stage in patients on long-term dialysis therapy [7]. Thus, the favorable outcome in patients with ESRD RCC in the French report [6] may have been biased by the high proportion of patients with early-stage disease in their study cohort. Moreover, the definition of patients with ESRD included both patients on dialysis and those who underwent kidney transplantation in their study. In this study, we assessed the clinical features and oncological outcomes of ESRD-RCC only in patients on hemodialysis and compared cancer-specific survival (CSS) according to stage between patients with RCC who are receiving HD () and those with RCC in the general population (). Materials and methods Patients In this study, we enrolled 1,794 patients with RCC who underwent resection at our institution between January 1979 and June 2013. We divided the RCC cases into two groups as follows: RCC cases that occurred in HD patients (, n = 408) and RCC cases that occurred in patients from the general population (, n = 1,386). All patients were pathologically confirmed to have RCC. The preoperative diagnosis of RCC utilized computed tomography (CT) or magnetic resonance imaging; however, these imaging techniques differed according to the time period. Clinicopathological factors were examined on the basis of the medical records of the patients. Tumor stage was determined according to the 2009 TNM classification [8]. The pathological diagnosis was made according to the 2004 World Health Organization classification [9]. Clinical symptoms were defined as those that required the performance of diagnostic studies, including gross hematuria, abdominal pain, and screening for metastatic disease or palpable tumors. The study protocol was approved by the institutional review board. The following variables were evaluated for each patient: age, sex, Eastern Cooperative Oncology Group performance status, tumor size, pathological diagnosis, nuclear grade, tumor stage, the presence of clinical symptoms, and CSS rates. Patient follow up Patients were evaluated for postoperative recurrence and general condition according to blood count, blood chemistry, and CT every 6 12 months. Statistical analysis The student s t test, Mann Whitney U test, and chi-squared test were used to compare continuous and categorical variables, as appropriate. CSS was calculated using the Kaplan Meier method and compared using the log-rank and Wilcoxon tests. Univariate and multivariate Cox regression analyses were performed to evaluate the prognostic factors of CSS in patients with RCC. A difference was considered statistically significant at p < 0.05. All data were analyzed using JMP 10.0.0 software (SAS Institute, Cary, NC, USA). Results Patient background The background characteristics of the patients are described in Table 1. Compared to the group, the RCC- HD group included younger patients (age: 55 vs. 60 years, p < 0.0001) and had higher incidences of symptomatic (32 % vs. 25 %, p = 0.0063), papillary subtype (22 % vs. 5 %, p < 0.0001), and stage 1 disease (82 % vs. 68 %, p < 0.0001). Mean dialysis duration before surgery was 120 months. Follow-up period, performance status at the time of surgery, and tumor grade did not differ significantly between the groups. Comparison of survival between the RCC HD and RCC general groups The 5-year overall survival rates were 81 % and 77 % in the and groups, respectively (p = 0.05). Figure 1 presents the CSS in both groups for all subjects. The group had a better 5-year survival rate than the group (89 % vs. 86 %, p = 0.0292). However, in the stage-by-stage analysis shown in Fig. 2a d, CSS did not differ significantly between the two groups by stage (stage I: 97 % vs. 99 %, p = 0.1945; stage II: 94 % vs. 92 %, p = 0.8611; stage III: 84 % vs. 75 %, p = 0.4705; stage IV: 6 % vs. 32 %, p = 0.1834). Risk factors of CSS In Table 2, we present the results of the univariable and multivariable analyses to demonstrate the predictive factors of CSS. The univariable analysis identified older age (p = 0.0022), symptoms at diagnosis (p < 0.0001), higher tumor grade (p < 0.0001), pathological T stage of 2 or higher (p < 0.0001), poorer performance status (p < 0.0001), and (p = 0.0239) as significant predictors of poor CSS. However, in the multivariable analysis, and older age lost significance to predict CSS (p = 0.2037).
Table 1 Patient characteristics Characteristic n = 1386 n = 408 p Age (year), mean (IQR) 60 (52 69) 55 (49 62) <0.0001 Sex (male), n (%) 970 (70 %) 342 (84) <0.0001 PS, n (%) 0 1 1347 (97) 401 (98) 0.2846 2 39 (3) 7 (2) Symptomatic, n (%) 351 (25) 132 (32) 0.0063 Tumor size (mm), mean (IQR) 49 (26 62) 39 (22 50) <0.0001 Pathology, n (%) <0.0001 CRCC 1,248 (90) 310 (76) PRCC 67 (5) 86 (22) Other 71 (5) 12 (4) Stage, n (%) <0.0001 I 935 (68) 333 (82) II 78 (6) 25 (6) III 205 (15) 21 (5) IV 168 (12) 29 (7) Nuclear grade, n (%) 0.605 48 (25) 108 (26) 2 872 (63) 258 (63) 3 166 (12) 42 (11) Duration of HD (months), mean (IQR) 120 (99 241) Death from cancer, n (%) 193 (14) 39 (10) 0.0209 Follow-up period (months), mean (IQR) 65 (24 87) 67 (21 101) 0.5425 RCC renal cell carcinoma, HD hemodialysis, IQR interquartile range, PS performance status, CRCC clear cell renal cell carcinoma, PRCC papillary renal cell carcinoma Fig. 1 Comparison of cancerspecific survival (CSS) time between renal cell carcinoma patients on hemodialysis (RCC- HD) and those in the general population () 1.0 0.8 0.6 0.4 p = 0.0292 89% 86% 0.2 0.0 0 12 24 36 48 60 72 84 96 108 120 Follow-up period (months) Number at risk 408 296 213 163 113 71 1386 1057 699 447 316 209 Discussion Patients with ESRD have a higher incidence of RCC than those in the general population, although CSS is better in those with ESRD, as demonstrated in several studies [6, 10, 11]. In fact, our study presented significantly better survival in the group than in the group, as presented in Fig. 1 (p = 0.0092). However, patients in the HD-RCC group exhibited different characteristics from those in the group, including younger age (55 vs. 60 years), smaller tumor size (39 vs. 49 mm), and earlier stage (stage I, 82 % vs. 68 %), suggesting that these favorable factors may contribute to prognosis in the group. In fact, patients on HD with each stage
Fig. 2 Comparison of cancerspecific survival (CSS) time between renal cell carcinoma patients on hemodialysis (RCC- HD) and those in the general population () as stratified by stage: stage I (a), stage II (b), stage III (c), stage IV (d) A Stage I p = 0.1945 97% 99% Follow-up period (months) 333 253 184 145 99 66 935 765 518 317 215 142 B Stage II 94% 92% p = 0.8611 Follow-up period(months) 24 21 15 11 9 4 78 62 40 28 22 15 of RCC had CSS similar to that of the general population (Fig. 2). Moreover, the multivariable analysis demonstrated no significant difference in CSS between the two groups (p = 0.2037), indicating that RCC in HD patients does not of itself strongly affect prognosis and instead is modified by other favorable prognostic factors, including age or stage. Favorable clinical and pathological features of RCC in patients with ESRD in comparison to those in the general population were reported in several articles. Neuzillet and colleagues demonstrated that RCC in native kidneys of ESRD patients seemed to exhibit many favorable clinical, pathological, and outcome features compared with RCC diagnosed in patients from the general population, including smaller size (3.7 vs. 7.3 cm), asymptomatic status (87 % vs. 44 %), low grade (68 % vs. 42 %), and better performance status (performance status = 0, 76 % vs. 63 %) [6]. Similarly, Breda and colleagues reported that the ESRD group, which consisted of patients indicated for and who had undergone transplantation, had smaller tumor size (p = 0.001), lower-grade, and earlier-stage tumors than the non-esrd group [10]. However, the multivariable analysis showed that RCC in patients with ESRD was not a prognostic factor of CSS in two previous studies [6, 10], suggesting that favorable clinical and pathological features resulted in better prognosis in the RCC ESRD patients, as in our results. The present study demonstrated that the presence of clinical symptoms at the time of diagnosis was significantly more common in the group (32 %) than in the group (25 %, p = 0.0063). According to two similar studies, RCC patients with ESRD had a lower rate of symptoms than RCC patients without ESRD (Neuzillet et al., 13 % vs. 56 %; Breda et al., 16 % vs. 38 %) [6, 10], which is not in agreement with our results. However, many of the patients (60 70 %) in the ESRD group in their studies were transplant patients, whereas our study consisted only of patients on HD, suggesting that these differences
Fig. 2 continued C Stage III 84% 75% p = 0.4705 Follow-up period(months) 21 15 13 9 7 3 205 155 109 82 69 50 D Stage IV 6% 32% p = 0.1834 Follow-up period(months) 30 10 4 1 1 1 168 76 36 22 13 6 Table 2 Predictors of cancerspecific survival Variables Univariable Multivariable HR (95 % CI) p HR (95 % CI) p Age >60/ 60 years 1.51 (1.16 1.96) 0.0022 1.19 (0.91 1.56) 0.2087 Symptomatic vs. asymptomatic 6.97 (5.31 9.24) <0.0001 2.57 (1.92 3.47) <0.0001 Tumor grade 3/1 or 3/2 6.57 (4.97 8.62) <0.0001 2.23 (1.66 2.97) <0.0001 Stage II IV/I 16.2 (11.4 23.5) <0.0001 8.85 (6.03 13.3) <0.0001 PS 2/0 or 1 13.7 (9.02 20.1) <0.000.78 (2.46 5.59) <0.0001 vs. 1.46 (1.05 2.09) 0.0239 0.79 (0.55 1.14) 0.2037 CRCC vs. non-crcc 0.97 (0.67 1.47) 0.8808 HR hazard ratio, PS performance status, HD hemodialysis, CRCC clear cell renal cell carcinoma, CI confidence interval
in patient population may have caused the difference in the presence of symptoms. In general, patients who are indicated to receive organ transplantation should be confirmed to have no malignant diseases to prevent potential development of malignancy under immunosuppressive conditions. Therefore, many patients might be diagnosed with kidney cancer without exhibiting any symptoms. As presented previously, the presence of clinical symptoms at the time of diagnosis was significantly higher in the group (32 %) than in the group (25 %, p = 0.0063) in our study. However, the incidence of early-stage disease was significantly higher in the RCC- HD group (stage I, 82 %) than in the group (68 %, p < 0.0001), and tumor size in the group was significantly smaller than that in the group (39 vs. 49 mm, p < 0.0001). These results appear controversial. RCC in patients on HD has different potential oncogenic factors compared with those in the general population, which may result in different pathological findings, such as a lower incidence rate of clear cell RCC and a higher incidence rate of papillary RCC in the group than in the group. In the present study, approximately 80 % of patients in the group also had ACDK. Several factors have been reported to cause RCC in HD patients with ACDK. Increased expression levels of growth factors or uremic metabolites have been demonstrated in ESRD. The stimulation of tubular and cystic cells by these renotropic substances seems to promote the development of ACDK, with progressively florid degrees of papillary hyperplasia and papillary neoplasms developing either simultaneously or subsequently [12, 13]. In addition, the frequency of somatic mutation of the von Hippel Lindau disease tumor-suppressor gene is reported to be approximately 50 % in clear cell RCCs in dialysis patients, which accounts for approximately 80 % of adult sporadic RCCs [14, 15]. These differences may have influenced our discrepant results in terms of clinical symptoms and stages. The present study has some limitations. First, the present study was the retrospective in nature, and its study design was similar to two previous studies. However, our study was limited to patients with ESRD receiving HD, and the number of the patients in the group was comparatively large (n = 408), indicating that our study represented a comparison between patients on HD and those of the general population. Moreover, the mean duration of HD at the time of RCC diagnosis was 120 months, which is much longer than that in other studies (Neuzillet et al. [6]: mean, 65 months). HD duration is a significant prognostic factor of oncological outcome in HD patients with RCC [7, 16]. Our previous report demonstrated that patients with longduration HD had significantly worse CSS than those not on HD (5-year CSS: >180 months, 82 %; <180 months, 95 %; p = 0.0004) [16]. In some countries, including Japan, many patients receive long-term dialysis therapy for other reasons, such as a shortage of donated organs. Our data correspond to the present situation of patients with ESRD in such countries. Again, studies that use data specific to HD patients and long-term HD will strengthen the findings of our study. Second, although the performance status of the patient should have been specified as cancer related, the performance status of patients on HD also included renal failure-related performance status. Consequently, we could not categorize performance status as cancer related or renal failure related for HD patients from the medical records. Conclusion Our study demonstrated that patients in the group presented with smaller tumor size, lower stage, younger age, and favorable oncological outcomes. However, the stage-for-stage analysis identified similar CSS in each stage between the and groups. The multivariable analysis revealed that or status was not a significant predictor of oncological outcome. Further studies to delineate the oncogenic factors of are required to determine the cause or causes of favorable clinical features in. Conflict of interest The authors declare that they have no conflict of interest. References 1. Marple JT, MacDougall M, Chonko AM (1994) Renal cancer complicating acquired cystic kidney disease. J Am Soc Nephrol 4:1951 1956 2. Stewart JH, Buccianti G, Agodoa L et al (2003) Cancers of the kidney and urinary tract in patients on dialysis for end-stage renal disease: analysis of data from the United States, Europe and Australia and New Zealand. J Am Soc Nephrol 14:197 207 3. Ishikawa I, Honda R, Yamada Y et al (2004) Renal cell carcinoma detected by screening shows better patient survival than that detected following symptoms in dialysis patients. Ther Apher Dial 8:468 473 4. Ishikawa I (2004) Present status of renal cell carcinoma in dialysis patients in Japan: questionnaire study in 2002. Nephron Clin Pract 97:c11 c16 5. Nakai S, Iseki K, Itami N et al (2012) Overview of regular dialysis treatment in Japan (as of 31 December 2009). Ther Apher Dial 16:11 53 6. Neuzillet Y, Tillou X, Mathieu R et al (2011) Renal cell carcinoma (RCC) in patients with end-stage renal disease exhibits many favourable clinical, pathologic, and outcome features compared with RCC in the general population. Eur Urol 60:366 373 7. Sassa N, Hattori R, Tsuzuki T et al (2011) Renal cell carcinomas in haemodialysis patients: does haemodialysis duration influence pathological cell types and prognosis? Nephrol Dial Transpl 26:1677 1682 8. Sobin LH, Compton CC (2010) TNM seventh edition: what s new, what s changed: communication from the International
Union Against Cancer and the American Joint Committee on Cancer. Cancer (Phila) 116:5336 5339 9. Lopez-Beltran A, Scarpelli M, Montironi R et al (2006) 2004 WHO classification of the renal tumors of the adults. Eur Urol 49:798 805 10. Breda A, Luccarelli G, Rodriguez-Faba O et al. (2015) Clinical and pathological outcomes of renal cell carcinoma (RCC) in native kidneys of patients with end-stage renal disease: a longterm comparative retrospective study with RCC diagnosed in the general population. World J Urol 33:1 7 11. Shrewsberry AB, Osunkoya AO, Jiang K et al (2014) Renal cell carcinoma in patients with end-stage renal disease has favorable overall prognosis. Clin Transplant 28:211 216 12. Hughson MD, Schmidt L, Zbar B et al (1996) Renal cell carcinoma of end-stage renal disease: a histopathologic and molecular genetic study. J Am Soc Nephrol 7:2461 2468 13. Hughson MD, Bigler S, Dickman K et al (1999) Renal cell carcinoma of end-stage renal disease: an analysis of chromosome 3, 7 and 17 abnormalities by microsatellite amplification. Mod Pathol 12:301 309 14. Inoue H, Nonomura N, Kojima Y et al (2007) Somatic mutations of the von Hippel Lindau disease gene in renal carcinomas occurring in patients with long-term dialysis. Nephrol Dial Transplant 22:2052 2055 15. Yoshida M, Yao M, Ishikawa I et al (2002) Somatic von Hippel Lindau disease gene mutation in clear-cell renal carcinomas associated with end-stage renal disease/acquired cystic disease of the kidney. Genes Chromosomes Cancer 35(4):359 364 16. Takagi T, Kondo T, Izuka J et al (2011) Prognosis and characteristics of renal cell carcinoma in hemodialysis patients: bilateral occurrence does not influence cancer-specific survival. Int J Urol 18:806 812