RCC accounts for 3% of all malignant tumors and now

Transcription

1 Prognostic Value of Thrombocytosis in Renal Cell Carcinoma Karim Bensalah, Emmanuelle Leray, Patricia Fergelot, Nathalie Rioux-Leclercq, Jacques Tostain, François Guillé and Jean-Jacques Patard* From the Department of Urology (KB, FG, JJP), Department of Biostatistics (EL), Laboratory of Molecular Biology (PF) and Department of Pathology (NRL), Hôpital Pontchaillou, Rennes, and the Department of Urology, CHU of Saint-Etienne (JT), Saint-Etienne, France Purpose: We correlated platelet count to renal tumor characteristics, and evaluated the potential prognostic value of thrombocytosis in localized and metastatic tumors. Materials and Methods: A total of 804 patients operated on for a renal tumor in 2 French centers were included in this study. In all cases TNM stage, Fuhrman grade, tumor size, nodal invasion, ECOG score, histological subtype and platelet count were recorded. Survival estimates were compared with the Kaplan-Meier method and multivariate analysis was performed using a Cox model. Results: There were 538 (66.9%) males and 266 (33%) females (ratio 2.02) in the patient group. Median age was 62 years. T stages were T1, T2, T3 and T4 in 284 (35.3%), 116 (14.4%), 387 (48.1%) and 17 (2.1%), respectively. There were 112 (13.9%) and 126 (15.7%) patients with nodal and metastatic invasion, respectively. Platelet count was strongly correlated with T stage (p 0.001), Fuhrman grade (p 0.001), tumor size (p 0.001), nodal invasion (p 0.001) and the presence of distant metastasis (p 0.01). In univariate analysis all studied factors were significantly predictive of survival (p ). Thrombocytosis had an impact on prognosis in localized (p 0.001) and metastatic (p 0.003) disease. Several factors were retained in multivariate analysis, namely TNM stage, Fuhrman grade, tumor size, ECOG score and platelet count. In cases with a platelet count of less than 450,000/mm 3, the 5-year survival rate was 70%, compared to 38% when the platelet count was 450,000/mm 3 or greater. Conclusions: Platelet count appears to be an independent prognostic factor in renal cell carcinoma. It reflects a cascade of biological events correlated with tumor aggressiveness. This observation opens new perspectives for exploring carcinogenesis mechanisms and tumor progression in renal cell carcinoma. Key Words: carcinoma, renal cell; thrombocytosis, prognosis, platelet count RCC accounts for 3% of all malignant tumors and now ranks third in frequency among all urological cancers 1 with a continuous increase in incidence in the last 30 years. Localized disease is best treated with surgery with excellent survival achieved for small tumors. Immunotherapy using interleukin-2 and/or interferon gives a 15% global response rate in metastatic selected patients despite frequent toxicity. 2 Numerous prognostic factors have been reported in RCC, most important of which are tumor stage, tumor grade and performance status. There is an increasing interest in molecular and genetic markers which could reinforce the prognostic value of clinical data. Integrated staging systems have been settled to assist the clinician in following and counseling patients with RCC, as well as in helping with inclusion in clinical trials. 3 Thrombocytosis has been observed in many malignancies. Some authors have investigated the prognostic value of platelet count in patients with localized and metastatic RCC, and found thrombocytosis to have an adverse effect on Submitted for publication May 3, * Correspondence: CHRU Pontchaillou, 2 rue Henri Le Guilloux, Rennes Cedex 9 France (telephone: ; FAX: ; jean-jacques.patard@chu-rennes.fr). See Editorial on page 813. survival. 4 In this study we assessed the prognostic value of thrombocytosis in comparison with the usual prognostic variables in localized and metastatic RCC, and correlated platelet count with usual pathological and clinical parameters. MATERIALS AND METHODS Patient characteristics. The study retrospectively reviewed the data of 804 patients who had been treated for RCC in 2 French centers (Rennes and Saint-Etienne) between 1984 and All patients had undergone partial or radical nephrectomy according to tumor size and location. There were 538 males (66.9%) and 266 females (33.1%), and median age was 62 years old. Prognostic variables. All patients were investigated with abdominal computerized tomography and chest x-ray before the operation. Radionuclide bone scan was done in presence of suggestive symptoms. Several parameters were obtained in all patients, including age, gender, tumor size, nodal invasion, presence of distant metastasis, Fuhrman grade, ECOG score, platelet count and histological subtype. Tumor stage was established according to the 1997 TNM classification of the UICC. 5 Platelet counts were obtained from patient records from the anesthesia consultation before /06/ /0 859 Vol. 175, , March 2006 THE JOURNAL OF UROLOGY Printed in U.S.A. Copyright 2006 by AMERICAN UROLOGICAL ASSOCIATION DOI: /S (05)

2 860 PROGNOSTIC VALUE OF THROMBOCYTOSIS FOR RENAL CELL CARCINOMA TABLE 1. Demographic and clinical characteristics of the 804 patients with renal cancer surgery. Thrombocytosis was defined as at least 1 platelet count greater than 450,000/mm 3. Patients were followed every 6 months with clinical examination, abdominal ultrasound and chest x-ray. Mean followup was 50 months (range 10 to 265). Statistical analysis. All patient data were entered into an electronic database (FileMaker Pro 4.1). The student and chi-square tests were used to compare mean platelet counts and thrombocytosis occurrence according to demographic and clinical characteristics. Results were presented as mean SD for quantitative variables and number of patients (%) for qualitative variables. The p value was set at 5% and all tests were 2-sided. Univariate analysis was conducted to assess the prognostic value of all variables on cancer specific survival, using Kaplan-Meier estimates and log rank tests. Independent prognostic factors were identified with multivariate analysis with a Cox proportional hazards regression model. All analyses were performed with SPSS 12.0 for Windows. RESULTS No. (%) Younger than (27.5) (22.9) (31.8) 75 or older 143 (17.8) Less than (49.1) 6.5 or Greater 409 (50.9) 1 or (49.8) 3 or (50.2) No 692 (86.1) Yes 112 (13.9) No 678 (84.3) Yes 126 (15.7) 1 or (49.4) 3 or (50.6) ECOG score: (61.4) 1 or More 310 (38.6) Histology: Clear cell Ca 711 (88.4) Other histological subtype 93 (11.6) Cancer specific survival: Alive 599 (74.5) Dead 205 (25.5) Characteristics of the population. The study included 804 patients who had been operated for RCC. Mean age SE was years. Mean tumor size was cm. 50.2% of patients had pt3 or pt4 tumors. Nodal and metastatic invasion were found in 13.9% and 15.7% of patients, respectively. Mean platelet count was ,000/mm 3 (range 44,000 to 947,000) and 63 patients (7.8%) had thrombocytosis. Mean followup was months. Demographic and clinical characteristics are summarized in table 1. Association of platelet count with clinical and pathological variables. Platelet count was higher in men compared to women (p 0.002) and in case of poor clinical performance status (p 0.001). Thrombocytosis and platelet count were significantly associated with tumor size, T stage, nodal invasion, distant metastasis, Fuhrman grade and cancer specific survival (table 2). Platelet count and thrombocytosis were not related to histological subtype. Survival and prognostic factors. A total of 205 patients (25.5%) died. Survival probabilities SD at 1, 5 and 10 years were %, % and %, respectively. Prevalence of thrombocytosis was higher among deceased patients (15.1%) compared to living patients (5.3%, p 0.001). Kaplan-Meier analysis and log rank tests showed that tumor size, ECOG score, T stage, Fuhrman grade, nodal invasion, distant metastasis and thrombocytosis were associated with cancer specific survival (all p ). Influence of gender was at the limit of statistical significance (p 0.07), women having a better survival than men. Subgroup analysis was performed to assess the impact of thrombocytosis on survival among patients with localized and metastatic disease. The impact of thrombocytosis remained significant in the 2 groups (p and p 0.003, respectively). Survival curves are shown in figures 1 and 2. Results from the Cox proportional hazards regression model are presented in table 3. The final model included recognized prognostic factors such as T stage, Fuhrman grade, ECOG score, nodal and metastatic invasion, and thrombocytosis. Our study confirmed the independent prognostic value of thrombocytosis in RCC, that when adjusted on other covariates, the risk of death in patients with thrombocytosis was twice that of patients with a normal platelet count. TABLE 2. Platelet count and prevalence of thrombocytosis according to demographic and clinical characteristics Mean Platelet Count SE p Value No. Thrombocytosis (%) p Value Gender: Male (6.5) Female (10.5) Younger than (9.0) (11.4) (5.1) 75 or Older (6.3) Less than (2.8) or Greater (12.7) ECOG score: (3.4) or More (14.8) 1or (4.0) or (11.6) G1 or G (3.5) G3 or G (12.0) No (6.6) Yes (15.2) No (6.9) Yes (12.7) Histology: Clear cell Ca (8.2) Other histology (5.4) Cancer specific survival: Alive (5.3) Dead (15.1)

3 PROGNOSTIC VALUE OF THROMBOCYTOSIS FOR RENAL CELL CARCINOMA 861 FIG. 1. Kaplan-Meier cancer specific survival estimates according to platelet count in 678 patients with localized tumor. FIG. 2. Kaplan-Meier cancer specific survival estimates according to platelet count in 126 patients with metastatic disease. DISCUSSION The incidence of RCC is increasing due to the incidental detection of small tumors and environmental factors. RCC remains a major cause of mortality with approximately 40% of patients dying of cancer progression. 6 Important prognostic factors have been established including tumor related factors (tumor size, grade and stage), clinical symptoms and laboratory findings. Pathological stage is recognized to be the most important prognostic factor in RCC. Nuclear grade, mostly represented by the Fuhrman grade, 7 also has an important prognostic value with 5-year survival decreasing from 64% to 10% in grade 1 and 4 tumors, respectively. Presenting symptoms, suboptimal performance status or significant weight loss have an adverse effect on prognosis in all stages of RCC. In terms of laboratory values, an increased erythrocyte sedimentation rate, anemia, hypercalcemia and increased alkaline phosphatase are the strongest parameters recognized to have a prognostic value in retrospective studies. 8 Thrombocytosis has been recently suggested to indicate a poor prognosis in patients with various malignancies. In a series of 369 patients with gastric cancer thrombocytosis was an independent prognostic indicator of survival. Symbas et al noted a poor outcome and a shorter life expectancy in patients with thrombocytosis and metastatic RCC. 9 They reviewed the charts of 259 patients with stage 4 RCC operated on in 2 centers who had undergone immunotherapy, chemotherapy or hormonal therapy after surgery. By using a thrombocytosis cutoff of 400,000/mm 3 platelets, they found a significantly better survival (151 vs 92 months) among patients whose platelet count returned to a normal level after treatment. Authors from the same institution performed a retrospective study of 204 patients who underwent radical nephrectomy for localized RCC (stage lower than T3N0M0). They found thrombocytosis to be associated with decreased survival. The specific death rate was 42% in 26 patients with thrombocytosis compared to 7.3% in 178 patients with a normal platelet count. 10 Our findings are con- TABLE 3. Factors affecting survival in multivariate analysis OR (95% CI) No 1 Yes 5.6 ( ) 1or2 1 3 or ( ) No 1 Yes 2.1 ( ) Less than or Greater 1.7 ( ) Platelet count: 450 or Less 1 Greater than ( ) 1or2 1 3 or ( ) Younger than or Older 1.7 ( ) ECOG: 0 1 Score 1 or More 1.5 ( )

4 862 PROGNOSTIC VALUE OF THROMBOCYTOSIS FOR RENAL CELL CARCINOMA sistent with those previously described in that thrombocytosis was predictive of cancer specific survival in univariate (p ) or in multivariate (p 0.009) analysis. Thrombocytosis had an adverse impact on prognosis in localized (fig. 1) and metastatic (fig. 2) disease. Moreover, thrombocytosis was associated with tumor aggressiveness. The platelet count was significantly increased in higher stage and grade tumors as well as in patients who died of cancer (table 2). Although our thrombocytosis cutoff (450,000/mm 3 ) was different from that used by previously cited authors, our results tend to indicate a similarly poor outcome associated with increased platelet blood count. Thrombocytosis mechanisms in renal as well as in other solid malignancies are poorly understood. Platelets may be involved in tumor induced angiogenesis because of their ability to release angiogenic growth factors when activated by angiogenic endothelium. 11 In addition to VEGF, platelets contain several other angiogenic growth factors and inhibitors that are released when activated, namely platelet derived growth factor, transforming growth factor-, hepatocyte growth factor, thrombospondin and endostatin. Studies report increased VEGF content in platelets of patients with cancer compared to healthy volunteers. 12 It may be possible that tumors stimulate megakaryocytes in the bone marrow, which in turn generate increasing amounts of platelets that contain more VEGF. We do not know which tumor released factors are responsible and the effect on tumor progression. It has been shown that platelets contain TPO and release it when activated. TPO is the specific cytokine that stimulates megakaryocytes in the bone marrow to generate platelets. It also stimulates VEGF release from Mk cells. 13 In response to an injury of the vessel wall, platelets can attach to the vessel wall and subsequently become activated and release their content, including TPO. Then, in turn, TPO stimulates the bone marrow to generate new platelets into the circulation. Verheul and Pinedo hypothesized that activated platelets in the tumor vasculature release TPO and thereby stimulate bone marrow generation of platelets. 4 As a consequence, an increased number of platelets will reach the circulation. 4 Alternatively, VEGF could have a major role at the bone marrow level. It has recently been shown by Casella et al that hypoxia may induce VEGF production by Mk cells, resulting in enhanced megakaryocytic maturation and platelet production. 14 Since the authors demonstrate that exogenous VEGF promotes Mk development, platelet production might be up regulated through an autocrine VEGF loop. Thus, VEGF may synergize with TPO to stimulate Mk cells maturation. This phenomenon could be emphasized during the development of renal tumors since VEGF over expression is a common feature in RCC. To improve the prognostic evaluation of patients with RCC, several investigators made an attempt to combine usual prognostic parameters, thus creating integrated prognostic systems. It could provide useful tools for postoperative counseling and for decision making about adjuvant therapies. Kattan et al settled a nomogram out of the data of 601 patients with a localized RCC operated at the Memorial Sloan-Kettering Cancer Center. 15 Integrated prognostic factors included the presence of symptoms at presentation, tumor histology, tumor size and TNM stage. Pantuck et al used the University of California-Los Angeles database of 661 patients with localized and metastatic RCC, and subsequently proposed the University of California-Los Angeles Integrated Staging System, 3 which combines TNM stage, nuclear grade and ECOG performance status in 3 prognostic categories. Those 2 integrated prognostic systems are built with preoperative and postoperative data. Cindolo et al recently reported a preoperative prognostic model performed with the records of 660 patients from 3 European institutions. 16 Using clinical presentation and clinical tumor size they constructed a recurrent risk formula that could predict the rate of disease recurrence after surgery. It is worth noting that none of these integrated prognostic systems included any laboratory finding. Since our study showed a strong correlation between platelet count and other usual prognostic parameters we believe that it could improve RCC prognostication particularly when used preoperatively. Our data suggest that thrombocytosis may be used as an adjunctive prognostic parameter to better stratify patient risk. Either alone or integrated in more complex prognostic systems, thrombocytosis may alert the clinician to possible metastatic or aggressive disease. Finally, a better understanding of the relationship between thrombocytosis, tumor aggressiveness and angiogenesis could provide some rationale for including selected patients in angiogenesis inhibitors based clinical trials. CONCLUSIONS Our study showed that platelet count is correlated with tumor aggressiveness and that thrombocytosis has an adverse effect on survival for patients with localized or advanced RCC. Therefore, we believe that platelet count could improve prognostication in RCC when used in conjunction with other well established clinical and pathological parameters. This new prognostic biological variable should be tested for its integration into prognostic systems besides TNM stage, ECOG score, symptoms and Fuhrman grade. Abbreviations and Acronyms ECOG Eastern Cooperative Oncology Group RCC renal cell carcinoma TPO thrombopoietin VEGF vascular endothelial growth factor REFERENCES 1. Chow, W. H., Devesa, S. S., Warren, J. L. and Fraumeni, J. F., Jr.: Rising incidence of renal cell cancer in the United States. JAMA, 281: 1628, Motzer, R. J., Mazumdar, M., Bacik, J., Russo, P., Berg, W. J. and Metz, E. M.: Effect of cytokine therapy on survival for patients with advanced renal cell carcinoma. J Clin Oncol, 18: 1928, Pantuck, A. J., Zisman, A. and Belldegrun, A. S.: The changing natural history of renal cell carcinoma. J Urol, 166: 1611, Verheul, H. M. and Pinedo, H. M.: The importance of platelet counts and their contents in cancer. Clin Cancer Res, 9: 3219, Guinan, P., Sobin, L. H., Algaba, F., Badellino, F., Kameyama, S., MacLennan, G. et al: TNM staging of renal cell carcinoma: Workgroup No. 3. Union International Contre le Cancer (UICC) and the American Joint Committee on Cancer (AJCC). Cancer, 80: 992, Kontak, J. A. and Campbell, S. C.: Prognostic factors in renal cell carcinoma. Urol Clin North Am, 30: 467, 2003

5 PROGNOSTIC VALUE OF THROMBOCYTOSIS FOR RENAL CELL CARCINOMA Fuhrman, S. A., Lasky, L. C. and Limas, C.: Prognostic significance of morphologic parameters in renal cell carcinoma. Am J Surg Pathol, 6: 655, Gelb, A. B.: Renal cell carcinoma: current prognostic factors. Union Internationale Contre le Cancer (UICC) and the American Joint Committee on Cancer (AJCC). Cancer, 80: 981, Symbas, N. P., Townsend, M. F., El-Galley, R., Keane, T. E., Graham, S. D. and Petros, J. A.: Poor prognosis associated with thrombocytosis in patients with renal cell carcinoma. BJU Int, 86: 203, O Keefe, S. C., Marshall, F. F., Issa, M. M., Harmon, M. P. and Petros, J. A.: Thrombocytosis is associated with a significant increase in the cancer specific death rate after radical nephrectomy. J Urol, 168: 1378, Verheul, H. M., Jorna, A. S., Hoekman, K., Broxterman, H. J., Gebbink, M. F. and Pinedo, H. M.: Vascular endothelial growth factor-stimulated endothelial cells promote adhesion and activation of platelets. Blood, 96: 4216, Salven, P., Orpana, A. and Joensuu, H.: Leukocytes and platelets of patients with cancer contain high levels of vascular endothelial growth factor. Clin Cancer Res, 5: 487, Bobik, R., Hong, Y., Breier, G., Martin, J. F. and Erusalimsky, J. D.: Thrombopoietin stimulates VEGF release from c-mplexpressing cell lines and haematopoietic progenitors. FEBS Lett, 423: 10, Casella, I., Feccia, T., Chelucci, C., Samoggia, P., Castelli, G., Guerriero, R. et al: Autocrine-paracrine VEGF loops potentiate the maturation of megakaryocytic precursors through Flt1 receptor. Blood, 101: 1316, Kattan, M. W., Reuter, V., Motzer, R. J., Katz, J. and Russo, P.: A postoperative prognostic nomogram for renal cell carcinoma. J Urol, 166: 63, Cindolo, L., de la Taille, A., Messina, G., Romis, L., Abbou, C. C., Altieri, V. et al: A preoperative clinical prognostic model for non-metastatic renal cell carcinoma. BJU Int, 92: 901, 2003