The association between elevated platelet counts and malignant neoplasia has been known since first reported by Reiss in 1872.1 Although the exact mechanism is not known, it is thought to be a humorally mediated response.2 Previously, thrombocytosis was reported with various pelvic and extrapelvic solid tumors.3
Gynecologic malignancies shown to be associated with thrombocytosis were ovarian cancer, vulvar carcinoma,4 cervical cancer,5,6 and endometrial cancer.7 Studies on the prognostic significance of thrombocytosis in gynecologic malignancies reported conflicting results. For ovarian cancer, Menczer et al8 reported poor prognosis associated with increased platelet counts. On the other hand, Zeimet et al9 found no effect on survival.
In this study, we evaluated the correlation of pretreatment platelet counts on disease progression and second-look laparotomy results in patients with advanced epithelial carcinoma of ovary. Additionally by using receiver operating characteristic curve analysis, different cutoff levels of platelet counts were also evaluated with sensitivity and specificity values for predicting disease progression and second-look laparotomy result.
MATERIALS AND METHODS
Between 1999 and 2001 in Hacettepe University Faculty of Medicine, Department of Obstetrics and Gynecology optimally debulked (defined as the largest diameter of residual tumor less than 1 cm) 37 patients with advanced epithelial carcinoma of ovary were evaluated. All patients were treated with primary surgery followed by adjuvant chemotherapy. After 6 cycles of platin-paclitaxel combination, second-look laparotomy was offered to all patients with good performance status. Second-look laparotomy was performed in patients suitable for and willing the undergo procedure (n = 20). Patients with poor performance and/or refusing second-look laparotomy (n = 12) were prospectively followed for disease progression with physical examination, ultrasonography, and serum CA 125 determinations. Five patients were found to be not eligible for the second-look laparotomy procedure (Figure 1).
Patients’ age, gravida, parity, operative findings, stage, grade, serum CA 125 levels, and platelet count were prospectively recorded variables. Platelet counts were performed within 5 days before the operation. Thirty-seven patients with advanced epithelial ovarian cancer underwent primary surgical treatment. Thirty-six patients had stage III disease, and one had stage IV. Optimal debulking was performed in all patients who subsequently received chemotherapy in the form of platin-paclitaxel combination. According to second-look laparotomy or follow-up results patients were divided into 2 groups. The first group had negative second-look laparotomy results or no evidence of disease during follow-up (n = 20), and the second group had positive second-look laparotomy results or progressive disease during follow-up (n = 17). Progressive disease was defined as failure to respond to chemotherapy or progression during the first 6 months after completion of chemotherapy. Student t test and Wilcoxon signed rank test were used. Sensitivity and specificity values for different cutoff levels of platelet counts were calculated with receiver operator characteristic curve method. Statistical significance level was set to P < .05.
Age, gravida, and parity were not significantly different as compared with controls (P > .05) (Table 1). Mean platelet counts (± standard deviation) were 371 × 109/L ± 94 × 109/L and 446 × 109/L ± 100 × 109/L in the first and second groups, respectively (P = .03).
Using receiver operator characteristic curve analysis, the following sensitivity and specificity values were calculated for different cutoff points. Area under the receiver operator characteristic curve (± standard error) was 0.72± 0.08 (P = .026). The cutoff value showing the best equilibrium between sensitivity and specificity was 380 × 109/L (sensitivity of 77% and specificity of 60%). Negative and positive predictive values for this cutoff were 75% and 62%, respectively. Cutoff value of 400 × 109/L had a sensitivity of 59% with a specificity of 65%. Cutoff value of 450 × 109/L, the other frequently used threshold value for the diagnosis of thrombocytosis, yielded a lower sensitivity of 41% and a higher specificity of 75% (Figure 2).
Malignancy is a well known cause of reactive (secondary) thrombocytosis. It was found to be associated in 30–60% of patients with nongynecologic10,11 and 10–30% in gynecologic malignancies.3,5,11,12
Underlying pathophysiologic mechanisms were complex and gained importance after reports regarding the effect on prognosis. Although the exact mechanism to explain increased production is not known, it is thought to be a paraneoplastic syndrome involving interleukin-6 (IL-6), thrombopoietin.13,14 Humoral mediators stimulating platelet production might be produced by malignant cells themselves. Therefore, the degree of elevation of platelets could be a marker of tumor load. This, in turn, might be associated with poor prognosis. Probably the same humoral mediators also have a role in tumor growth, metastasis, or both. One of the best-studied mediators of increased production of platelets in cancer patients was IL-6.14 IL-6 is a well known stimulator of megakaryocytes.13 It has a therapeutic value in chemotherapy-induced thrombocytopenia in patients with ovarian cancer.15 IL-6 levels in ascites also were shown to be correlated with reactive thrombocytosis in patients with epithelial ovarian cancer.16
Thrombocytosis was found to be an independent prognostic factor in patients with cervical12 and endometrial17 cancers. Contrary to these findings, it is not found to be a prognostic factor in vulvar carcinoma.3 In epithelial ovarian cancer, Menczer et al8 reported poor prognosis in patients with thrombocytosis in their series. Contrary to these findings Zeimet et al9 reported thrombocytosis in 38% and found no association with prognosis. In this study, volume of ascites and hemoglobin concentrations were independent factors associated with thrombocytosis.
Different cutoff values between 300 × 109 /L and 450 × 109 /L for thrombocytosis were used in different studies. The two most frequently used cutoff levels to define thrombocytosis were 350 × 109 /L4,11 and 400 × 109 /L3,5 in previous studies. In this study, instead of grouping patients according to a cutoff level, we compared the mean values. In the second group the mean platelet count was significantly higher than in the first group (446 ± 100 × 109 /L versus 371 ± 94 × 109 /L, P = .03). With this finding in mind, we performed receiver operator characteristic curve analysis to determine the effect of different cutoff values on disease progression. Patients in the second group can be predicted with 59% sensitivity and 65% specificity at the cutoff value of 400 × 109/L. When the cutoff value was increased to 450 × 109/L, sensitivity and specificity were calculated as 41% and 75%, respectively. Finding of increased platelet count may reflect increased tumor burden, but also as proposed by Hernandez et al,5 might be a marker of tumor growth. Another mechanism for poor prognosis associated with thrombocytosis was the possibility that excess platelets might facilitate the vascular adhesion and distant metastasis of circulating malignant cells. Thrombospondin-1 might be involved this process.5 Although hematological metastasis is less important in the pathogenic mechanism of ovarian cancer, this might have a role in distant metastases. Another mechanism for association of aggressive behavior of tumor cells was platelets producing mediators affecting tumor growth. Also, a platelet cast around malignant cells has a role in protection from immune system.
In conclusion, high platelet count in advanced stage epithelial ovarian carcinoma is associated with increased risk of disease progression and positive second-look laparotomy results. Thrombocytosis might be an important marker of aggressive tumor behavior, and epithelial platelets might themselves produce mediators affecting tumor growth cancer.
1. Tranum BL, Haut A. Thrombocytosis: platelet kinetics in neoplasia. J Lab Clin Med 1974 Nov;84:615–9.
2. Estrov Z, Talpaz M, Mavligit G, Pazdur R, Harris D, Greenberg SM, et al. Elevated plasma thrombopoietic activity in patients with metastatic cancer-related thrombocytosis. Am J Med 1995;98:551–8.
3. Lavie O, Comerci G, Daras V, Bolger BS, Lopes A, Monaghan JM. Thrombocytosis in women with vulvar carcinoma. Gynecol Oncol 1999;72:82–6.
4. Hefler L, Mayerhofer K, Leibman B, Obermair A, Reinthaller A, Kainz C, et al. Tumor anemia and thrombocytosis in patients with vulvar cancer. Tumour Biol 2000;21:309–14.
5. Hernandez E, Donohue KA, Anderson LL, Heller PB, Stehman FB. The significance of thrombocytosis in patients with locally advanced cervical carcinoma: a Gynecologic Oncology Group study. Gynecol Oncol 2000;78:137–42.
6. Hernandez E, Heller PB, Whitney C, Diana K, Delgado G. Thrombocytosis in surgically treated stage IB squamous cell cervical carcinoma (A Gynecologic Oncology Group study). Gynecol Oncol 1994;55(3 Pt 1):328–32.
7. Gucer F, Moser F, Tamussino K, Reich O, Haas J, Arikan G, et al. Thrombocytosis as a prognostic factor in endometrial carcinoma. Gynecol Oncol 1998;70:210–4.
8. Menczer J, Schejter E, Geva D, Ginath S, Zakut H. Ovarian carcinoma associated thrombocytosis: correlation with prognostic factors and with survival. Eur J Gynaecol Oncol 1998;19:82–4.
9. Zeimet AG, Marth C, Muller-Holzner E, Daxenbichler G, Dapunt O. Significance of thrombocytosis in patients with epithelial ovarian cancer. Am J Obstet Gynecol 1994;170:549–54.
10. Costantini V, Zacharski LR, Moritz TE, Edwards RL. The platelet count in carcinoma of the lung and colon. Thromb Haemost 1990;64:501–5.
11. Rodriguez GC, Clarke-Pearson DL, Soper JT, Berchuck A, Synan I, Dodge RK. The negative prognostic implications of thrombocytosis in women with stage IB cervical cancer. Obstet Gynecol 1994;83:445–8.
12. Hernandez E, Lavine M, Dunton CJ, Gracely E, Parker J. Poor prognosis associated with thrombocytosis in patients with cervical cancer. Cancer 1992;69:2975–7.
13. Kaser A, Brandacher G, Steurer W, Kaser S, Offner FA, Zoller H, et al. Interleukin-6 stimulates thrombopoiesis through thrombopoietin: role in inflammatory thrombocytosis. Blood 2001;98:2720–5.
14. Araneda M, Krishnan V, Hall K, Kalbfleisch J, Krishnaswamy G, Krishnan K. Reactive and clonal thrombocytosis: proinflammatory and hematopoietic cytokines and acute phase proteins. South Med J 2001;94:417–20.
15. D’Hondt V, Humblet Y, Guillaume T, Baatout S, Chatelain C, Berliere M, et al. Thrombopoietic effects and toxicity of interleukin-6 in patients with ovarian cancer before and after chemotherapy: a multicentric placebo-controlled, randomized phase Ib study. Blood 1995;85:2347–53.
16. Gastl G, Plante M, Finstad CL, Wong GY, Federici MG, Bander NH, et al. High IL-6 levels in ascitic fluid correlate with reactive thrombocytosis in patients with epithelial ovarian cancer. Br J Haematol 1993;83:433–41.
17. Scholz HS, Petru E, Gucer F, Haas J, Tamussino K, Winter R. Preoperative thrombocytosis is an independent prognostic factor in stage III and IV endometrial cancer. Anticancer Res 2000;20:3983–5.