Predictive and prognostic values of preoperative platelet parameters in patients with gynecological tumors

Abstract Background Platelets play a role in tumor cell growth, metastasis, and angiogenesis, and the present study aimed to evaluate diagnostic and prognostic values of platelet parameters in patients with gynecological tumors. Methods A total of 1062 women were included. Differences of platelet parameters (platelet count [PLT], plateletcrit [PCT], mean platelet volume [MPV], platelet‐large cell rate [P‐LCR], and platelet distribution width [PDW]) between different categories were analyzed by nonparametric test. The optimal cutoff value was calculated with receiver operating characteristic analysis. Overall survivals were analyzed with Kaplan‐Meier method and log‐rank tests for univariate analysis. Results Platelet count and PCT were significantly increased, and MPV and P‐LCR were significantly reduced in malign and benign gynecological tumor groups compared with the controls (P < .001); PDW had no significant differences. There were no significant differences in PLT, PCT, MPV, P‐LCR, and PDW between different tumor locations and pathologic types. The optimal cutoff values of PLT, PCT, MPV and P‐LCR were 274, 0.26, 10.08, and 24.8 (AUC: 0.661, 0.643, 0.593, 0.562), and PCT had preferable sensibility and specificity (50.84% and 70.42%) in predicting the presence of gynecological tumors. According to survival analysis, increased PLT (≥274 × 109/L) and PCT (≥0.26), and induced MPV (<10.08 fL) and P‐LCR (<24.8%) were associated with shorter overall survival. Conclusions Platelet count, PCT, MPV, and P‐LCR can be used as preferable auxiliary parameters for predicting the presence of gynecological tumors. Increased PLT and PCT, or decreased MPV and P‐LCR indicated a heavier tumor burden and shorter overall survival.


| INTRODUC TI ON
With rapid population growth and aging worldwide, cancer incidence and mortality are rapidly growing, thus making cancer as the leading cause of death worldwide. 1 According to the data released by the global cancer database (GLOBOCAN) in 2018, there were more than 1 309 165 new gynecologic cancer patients and 609 377 women died from gynecological cancer, including cervical cancer, uterine cancer, ovarian cancer, vaginal cancer and vulvar cancer, and cervical cancer ranked fourth for both incidence and mortality in female cancer. 2 Data also showed that almost 90% of cervical cancer occurred in developing country. 3 So, easily accessible and inexpensive indicators need to be considered for early diagnosis or prognosis of gynecologic tumors.
Platelets are small (2-4 um), hematopoietic, and anucleate cells released by bone marrow megakaryocytes in the bloodstream, 4 and were described as the major effectors of several physiological and pathophysiological processes, such as hemostasis, thrombosis, immunological defense mechanisms, and the development of inflammation. 5 In addition, a growing body of evidence has found that platelets also play a role in tumor cell growth, metastasis, and angiogenesis. 6,7 Decades of studies indicated that platelet indicators are important prognostic factors in patients with different types of cancer, such as gynecologic cancers, [8][9][10][11][12][13] lung cancer, 14,15 gastric cancer, 16 colorectal cancer, 5 pancreatic cancer, 17,18 laryngeal cancer, 19 and rectal cancer. 20 Thrombocytosis may precede the diagnosis of malignancy. 21 Some studies 6,22 have tried to explore the complex relationship between platelets and tumors, in order to provide a new basis and way for anti-tumor therapy. However, few studies have explored the relationship between platelet parameters other than blood platelet count (PLT) and gynecologic cancer, or investigate the differences of platelet parameters between malignant and benign tumors. Herein, we intended to assess the distribution of platelet parameters in patients with malignant and benign gynecological tumors, which were compared to healthy controls, in order to provide some basis for diagnosis and prognosis for gynecological tumors.

| Patients
Anamnesis, demographic, preoperative laboratory, and histopathological data of women who underwent gynecological tumor sur-

| Data analysis and definitions
The level of PLT in adults was normally ranging from 100 × 10 9 /L to 300 × 10 9 /L. 11 Thrombocytosis was defined as PLT greater than 400 × 10 9 /L. 22 The range of plateletcrit (PCT) was normally between 0.11 and 0.28, and PCT more than 0.28 was defined as high PCT. mean platelet volume (MPV) normally ranged from 6.5 to 11.0 fL. Platelet distribution width (PDW) ranged normally between 9% and 17%, and PDW more than 17% was defined as high PDW. For FIGO stages, I and II were defined as early stage, while III and IV were defined as advanced stage. 10,12

| Laboratory assay and pathological examination
Patients' platelet parameters were obtained from the data of routine preoperative examination after admission. Patients were drawn venous blood after admission which would transport to the laboratory within 1 hour and measured by Sysmex XE5000 hematology analyzer (Sysmex). All microscopic slides and immunohistochemistry of tissue excised intraoperatively were reviewed by two pathologists.
Computed tomography (CT) scan and lymph node metastasis sites were used to detect metastatic status.

| Statistical analysis
The statistical analysis was performed with SPSS statistical software (version 23.0; SPSS, IBM, Inc.). The median and the interquartile ranges (IQRs) were used for descriptive statistics of indexes in non-normal distribution. The variables that did not show a normal distribution were compared using the Kruskal-Wallis H test or the Mann-Whitney U test.
The optimal cutoff values of PLT, PCT, MPV, and P-LCR were analyzed by the ROC curves. Overall survivals (OS) were analyzed with the Kaplan-Meier method and log-rank tests for univariate analysis. ROC curves and survival analysis were constructed by GraphPad Prism 7 (GraphPad Software, Inc.). All reported P values were two-sided, and a P-value < .05 was considered as statistical significance.  Table 1).

| Differences of platelet parameters in groups, tumor locations, pathologic types, or FIGO stages
Platelet count and PCT in malign group and benign group were significantly increased than those in normal group, while MPV and P-LCR were significantly decreased than those in normal group, but there was no significant difference of PDW among three groups (Table 2). Furthermore, there was no significant difference between disparate tumor locations (cervix uteri, corpus uteri, and ovary), or showed that the PLT in the malign group and benign group was increased and the volume was smaller in comparison with the normal group, and PLT, PCT, MPV, P-LCR, and PDW were not associated with tumor locations or pathologic types.
Platelet count and PCT in advanced stage group were significantly increased than those in early stage group, while MPV and P-LCR were notably decreased than those in early stage group.
There was no difference in PDW among the groups ( Table 7). The results showed that increased PLT and PCT, or decreased MPV and P-LCR indicated a heavier tumor burden.

| Correlation between platelet parameters and survival of patients with malignancy
In survival analysis, the optimal cutoff value of PLT, PCT, MPV, and

| D ISCUSS I ON
Since the first study on platelets and tumors a century ago, the re- Temur, et al supported that there was no difference of MPV between patients with endometrial cancer and controls, 27 while others showed that MPV was higher in patients with endometrial cancer compared with controls. 28 It was unclear whether the heterogeneous size and structure of the platelet translated into differences in platelet function. 29 The potential value of MPV in gynecological cancer needs to be further investigated. PCT and P-LCR were found significantly different between patients and controls. Nevertheless, the association between PCT, P-LCR, PDW, and gynecological tumor had rarely been investigated in previous research.

-LCR Reference
A large amount of literature showed that platelets played a certain role in the growth and metastasis of malignant tumors, 6,7,23,25 accompanied by changes in number and size of platelets. 9,10,28,30 We found that there were similar changes in the number and size of platelets in patients with gynecological malignant and benign tumors (Table 2), and there was no significant difference between different pathological types and different organs ( Table 3). According to the ROC curves (Figure 1), PLT, PCT, MPV, and P-LCR had preferable specificity in predicting the presence of gynecological tumor. They could be useful parameters for clinical assistant diagnosis. However, the mechanism of changes in the numbers and size of platelet in gynecological tumors is unclear and needs to be further investigated.
It was found that patients with advanced disease had a higher mean preoperative PLT than patients with localized disease, and preoperative thrombocytosis was an independent prognostic indicator for high-risk patients with stages III to IV endometrial cancer. 31 Katrin et al supported that thrombocytopenia suggested advanced tumor, higher tumor grade, and higher incidence of serous ovarian cancer. 32 Our present data (Table 7) agreed with previous studies, and we found that MPV, PCT, and P-LCR were associated with FIGO stages.
High PCT, low MPV, and low P-LCR indicated worse tumor staging.
The association between the change of platelet parameters and OS of patients with gynecological cancer has been demonstrated in a few studies. Li   In conclusion, the increased PLT and the decreased platelet volume were not only common in gynecological cancer, but also usual in gynecological benign tumors. PLT, PCT, MPV, and P-LCR were useful auxiliary parameters for predicting the presence of gynecological tumor. In addition, increased PLT and PCT, or decreased MPV and P-LCR indicated a heavier tumor burden and shorter overall survival.

E TH I C A L A PPROVA L
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards.