Do Leukocyte and Platelet Counts Have Benefit for \Preoperative Evaluation of Endometrial Cancer?

Endometrial cancer is the most common gynecologic malignancy in developed countries (Ferlay et al., 2012). Staging surgery which consists of peritoneal washing, total hysterectomy, bilateral salpingo-oophorectomy, and pelvic lymph node (PLN) dissection with or without paraaortic lymph node (PALN) dissection remains the cornerstone in primary treatment for endometrial cancer (Pecorelli, 2009). Prognostic factors have been investigated by several studies, as a result of which, FIGO (The International Federation of Gynecology and Obstetrics) stage, myometrial invasion, histologic subtype and grade and lymphovascular involvement were found to be the most important ones (Prat,ß 2004; Amant et al., 2005). Furthermore, alterations in preoperative hematologic parameters such as leukocytosis and thrombocytosis have been suggested to be prognostic biomarkers and independently associated with advanced disease (Gorelick et al., 2009; Worley et al., 2012; Njolstad et al., 2013). The aim of this study was to evaluate the significance of preoperative leukocytosis and thrombocytosis among patients with endometrial cancer.


Introduction
Endometrial cancer is the most common gynecologic malignancy in developed countries (Ferlay et al., 2012). Staging surgery which consists of peritoneal washing, total hysterectomy, bilateral salpingo-oophorectomy, and pelvic lymph node (PLN) dissection with or without paraaortic lymph node (PALN) dissection remains the cornerstone in primary treatment for endometrial cancer (Pecorelli, 2009). Prognostic factors have been investigated by several studies, as a result of which, FIGO (The International Federation of Gynecology and Obstetrics) stage, myometrial invasion, histologic subtype and grade and lymphovascular involvement were found to be the most important ones (Prat,ß 2004;Amant et al., 2005). Furthermore, alterations in preoperative hematologic parameters such as leukocytosis and thrombocytosis have been suggested to be prognostic biomarkers and independently associated with advanced disease (Gorelick et al., 2009;Worley et al., 2012;Njolstad et al., 2013).
The aim of this study was to evaluate the significance of preoperative leukocytosis and thrombocytosis among patients with endometrial cancer.

Materials and Methods
The study included the patients (cancer group) surgically staged for endometrial cancer according to FIGO 2009 guidelines which consisted of peritoneal cytology, total abdominal hysterectomy, bilateral salpingo-oophorectomy, systemic pelvic and para-aortic lymphadenectomy (Pecorelli, 2009), and the patients (control group) underwent total abdominal or vaginal hysterectomy for benign uterine diseases such as endometrial polip, fibroid, uterine prolapsus at Ondokuz Mayis University, Department of Gynecology and Obstetrics between 2005 and 2013.
Patients were excluded from this study if any of the following were present: incompletely staged surgery, second malignancies, hematological disease, inflammatory disease, otoimmune disease, recombinant granulocyte colony-stimulating factor use, prior chemotherapy or radiotherapy, hypertension, diabetes mellitus, metabolic syndrome, nephropathy, renal or hepatic dysfunction, left ventricular dysfunction, valvular heart disease, abnormal thyroid function tests, previous history of local or systemic infection, HIV infection, any medication that is related to patients' inflammatory condition such as corticosteroids and missing preoperative complete blood cell count or complete blood count drawn more than two weeks prior to surgery.
Patients' preoperative and postoperative data, including demographic features, complete blood count in the week prior to surgery with differentials including WBC, platelet count, histopathologic evaluations for both benign and malign endometrium lesions, grade of endometrial cancer, tumor stage, tumor size, presence of lymphovascular space invasion (LVI) and overall survival were retrospectively analyzed. Leukocytosis was defined as white blood cell count of >10.000 cells/mm3, and thrombocytosis was defined as thrombocyte count of >450 X 10 3 platelets/mm 3 . Data analysis was performed by using SPSS for Windows, version 11.5. Whether the distributions of continuous and intermittent variables were normally or not was determined by Kolmogorov Smirnov test. Homogenity of variances was analyzed by Levene test. Data for continuous and intermittent numeric variables were shown as mean ± SD or median (min-max), data for categorical variables were shown as percentage (%) and case number where applicable.
The significance of differences between the groups were evaluated by one way variant analysis (One-Way ANOVA). Mann Whitney U test was applied for comparisons of the median values between two independent groups, whereas Kruskal Wallis test was used in case of the presence of more than two groups to compare. Conover's multiple comparison test was applied in order to determine the factors causing the difference, if there was a significance in Kruskal Wallis test statistics. Categorical variables were evaluated by Pearson's Chisquare, Fisher's exact or Likelihood Ratio test, where applicable.
To determinate whether preoperative WBC and PLT were significant to discriminate the groups in terms of clinical features, ROC analysis was applied and area under curve and 95% confidence interval was calculated. The maximum sum of sensitivity and specificity for the significant test result. Sensitivity, specificity, positive and negative predictive values were also calculated at the best cut-off points.
Determining whether the cut-off points of WBC and PLT were significant to discriminate stage 1 and stage 2, 3, 4 in order from each other was evaluated by Multiple Logistic Regression analysis. Any variable whose univariable test had a p value <0.25 was accepted as a candidate for the multivariable model along with all variables of known clinical importance. Adjusted odds ratios, 95% confidence intervals and wald statistics also were calculated.
A p value less than 0.05 was considered statistically significant.

Results
The study included 177 patients diagnosed with endometrial cancer and underwent surgical staging, and 100 patients with hysterectomy and bilateral salpingooophorectomy for benign gynecological indications. The mean age of the patients with endometrial cancer was 59.6±10.8. A hundred and fourty five (81.9%) of the endometrial cancers was endometrioid type, 32 (18.1%) of them were non-endometrioid. Eighty eight (49.7%) of the cancer cases was classified as grade I, 67 (37.9%) of them was grade II, and the rest of 22 (12.4%) was grade III. When myometrial invasion was evaluated, it was <50% in 94 (53.1%) patients and >50% in 83 (46.9%) of cases. Cervical involvement was present in 42 cases (23.7%) whereas adnexal involvement was in only 17 patients (9.6%). Peritoneal cytology was positive in only 13 (7.3%) patients while negative in 84 (47.5%) cases and nondiagnostic in 80 (45.2%) cases. Lymphovascular invasion was detected in 29 (16.4%) patients. Pelvic lymph node metastasis was present in 22 (12.4%) cases while paraaortic metastasis was present in 16 (9%) patients. When stage of endometrial cancers was determined, there were 118 (66.7%) patients with stage I, 24 (13.6%) cases with stage II, 26 (14.7%) patients with stage III and only nine (5.1%) cases with advanced stage IV.
When clinicopathologic features were compared between two groups, preoperative leukocyte count was significantly higher in patients with endometrial cancer ( preoperative WBC count in cancer group: 8100±2668 cell/ mm3, in control group:7039±1474 cell/mm3, p< 0.001), while there was no significant difference in preoperative platelet count between the groups; preoperative PLT count in cancer group: 302±91x10 3 /mm 3 , in control group: 295±75x103/mm 3 , p<0.05).
When patients' clinicopathologic features were evaluated in terms of correlation with preoperative mean leukocyte and platelet counts; mean leukocyte count was significantly higher in the presence of grade 3 disease, non-endometrioid histology, adnexal involvement, lymphovascular invasion, pelvic and paraaortic metastasis and advanced stage (p<0.05) ( Table 1). ROC analyses revealed significant area under curve for WBC count in differentiation of increased grade, adnexal involvement, positive peritoneal cytology, lymphovascular invasion, pelvic and paraaortic lymph node metastasis and advanced stage ( Table 2). The optimal cut-off points were calculated for preoperative WBC count to discriminate the clinical features of endometrial cancer cases (Table 3). It was found as 6915 to differentiate stage 1 and advanced stages Platelet count was found to be significantly increased in the presence of grade 3 disease, cervical involvement, lymphovascular invasion and pelvic lymph node metastasis (p<0.05). There was no significant association between preoperative platelet counts and stage of endometrial cancer (Table 1).
When most definitive risk factors in differentiation of advanced stages (stage II,III and IV), only myometrial invasion was identified as an indepenent risk factor in dicrimination of stage II from stage I (p<0.001). Increased histopathologic grade (grade 3), myometrial invasion and pre-op WBC count >10500 (cell/mm 3 ) were found to be independent risk factors in discrimination of stage III from stage I (p=0.026, p=0.019, p=0.029). Endometrioid histology and pre-op WBC count >10500 (cell/mm 3 ) were identified as independent risk factors in discrimination of stage IV from stage I (p=0.041) ( Table 4).

Discussion
Preoperative hematological alterations such as anemia, leukocytosis and thrombocytosis have been associated with poor prognosis in different malignancies (Hefler et al., 2000;Pedersen and Milman., 2003;Chen et al., 2005;Brockmann et al., 2007;Lu et al., 2007). Although the exact mechanism remains unclear, studies suggested that  tumors can affect hematologic parameters by significant tumor bleed, infiltrating the bone marrow or by producing proinflammatory cytokines (Weiss and Goodnough, 2005). Prognostic value of these parameters have been recently investigated in gynecological malignancies and findings have shed new light to such mechanisms, Pretreatment leukocytosis was associated with advanced disease in cervical cancer when compared to patients without leukocytosis (Garcia-Arias et al., 2007;Mabuchi et al., 2011). Up-regulation of hematopoietic growth factors, such as G-CSF, GM-CSF, IL-1, IL-6 and TNFalpha were suggested to play a critical role in tumorrelated leukocytosis (Sato et al., 1987;Chen et al., 1995;Watanabe et al., 1998;Mabuchi et al., 2011). G-CSF, which was suspected to be the most influential, was found significantly elevated in cervical cancer patients with pretreatment leukocytosis (Mabuchi et al., 2011). Not only in cervical cancer, but also in ovarian cancer, preoperative leukocytosis has been recently associated with poor prognosis (So et al., 2014).
With respect to endometrial cancer, when factors such as stage, histopathologic type, lymphovascular invasion, myometrial invasion and lymph node metastasis, affecting prognosis of the disease were evaluated with multivariate analysis, studies demonstrated preoperative leukocytosis as an independent risk factor for malignant endometrial lesions, advanced stage endometrial cancer and decreased survival (Worley et al., 2012;Luomaranta et al., 2013;Worley et al., 2013;Acmaz et al., 2014). In the present study, preoperative leukocytosis was found to be significantly associated with increased grade, nonendometrioid histology, cervical involvement, positive peritoneal cytology, lymphovascular invasion, pelvic and paraaortic lymph node metastasis and advanced stage (Table 2). Furthermore, optimal cut-off for pre-op WBC level was determined to differentiate advanced stage disease (Table 3).
Thrombocytosis, which is an other prevalent hematological alteration, has been estimated to occur in about 10%-57% of all patients diagnosed with malignant disease (Sierko and Wojtukiewicz, 2004). Since the responsible mechanism for thrombocytosis in malignancy is poorly understood, some factors are accepted as contributers to this process. One of them is production of cytokines such as IL-6 and IL-1 and stimulation of megakaryocytes and their precursors (Burstein and Harker, 1983;Blay et al., 1992). An other one is that elevated platelets may enhance the aggressive behaviour of tumor and shield it from immune system detection. In addition, platelets may promote angiogenesis through some platelet-derived factors such as vascular growth factor, thrombospondin and endostatin (Karpatkin S, Pearlstein, 1981;Verheul and Pinedo, 1998). Recently, studies have investigated the relationship between thrombocytosis and gynecological cancers (Hernandez et al., 1992;Hefler et al., 2000;Lerner et al., 2007;Gorelick et al., 2009). With respect to endometrial cancer, although preoperative elevated platelet count was associated with poor differentiation in grade, advanced stage, lymph node metastasis, adnexal and cervical involvement in, thrombocytosis was not found an independent prognostic factor in the multivariate studies (Gucer et al., 1998;Ayhan et al., 2006;Metindir and Bilir, 2009;Heng and Benjapibal, 2014). In contrast, there have been studies reporting preoperative thrombocytosis as an independent prognostic factor in endometrial cancer (Scholz et al., 2000;Tamussino et al., 2001;Gorelick et al., 2009). In this study, there was no significant difference in preoperative platelet count between the patients with endometrial cancer and the control group.
Despite the studies aimed to determine whether or not preoperative leukocytosis and thrombocytosis are independent risk factors on the stage of endometrial cancer, the topic has been controversial. In the present study, pre-op WBC count >10500 (cell/mm 3 ) were found to be independent risk factors in discrimination of stage III from stage I (p=0.026, p=0.019, p=0.029) and stage IV from stage I (p=0.041) (Table 4). However, thrombocytosis was not determined as an independent risk factor on the stage of the disease.
The current study has some limitations inherit to its retrospective design. First, surgical procedures were not standardized with respect to lymphadenectomy. Second, overall and disease free survival were not reported due to the lack of communication with the patients.
In conclusion, preoperative leukocytosis (WBC >10500 cell/mm 3 ) was independently associated with advanced endometrial cancer. More detailed studies could help to clarify the exact mechanism behind pretreatment leukocytosis and the impact of it on clinical outcomes.