Prognostic Value of Circulating Tumor Cells in Ovarian Cancer: A Meta-Analysis

open access
  
Introduction
Ovarian cancer is the second leading cause of death among gynecologic malignancies in the world [1] owing to the fact that a majority of patients are diagnosed in late stages of the disease [2]. In such a setting, identifying prognostic indicators for patients with ovarian cancer is crucial. CA-125 is a frequently used biomarker in ovarian cancer, but some non-malignant conditions also cause elevated serum CA-125 concentrations [3]. Thus additional prognostic markers are urgently needed for ovarian cancer.

Discussion

Many studies have reported the prognostic value of CTCs in patients with ovarian cancer, and this is the first meta-analysis evaluating the value of CTCs in these patients. Overall, our results demonstrated that patients in CTC-positive group had a worse OS and PFS/DFS compared with CTC- negative group; moreover, the presence of CTCs was associated with elevated CA-125.

Subgroup analysis showed that “RT-PCR” subgroup presented significant association between CTCs and OS, whereas it was not significant in the "CellSearch" subgroup and “Other ICC” subgroup. But the detection method didn’t influence HR estimates in the meta-regression (P = 0.118). Though the CellSearch system which uses EpCAM for cell isolation is the only CTC test approved by US FDA for clinical use currently, it has some limitations. EpCAM can be downregulated by cancer stem cells in the procedure of epithelial mesenchymal transition [24]. Thus this system suffers from relatively low sensitivity, additional innovative detection methods may reveal more tumor cells [25]. On the other hand, CellSearch system may also present poor specificity because of biologic nonspecificity [26]. Therefore the methodologies which do not use epithelial markers alone for cell isolation may be more likely to associate with OS according to our subgroup analysis.
In addition, subgroup analysis based on treatment methods showed the prognostic value of CTCs for OS was significant in the "Surgery" subgroup, while it was not significant in the "Chemotherapy" subgroup. However, it is difficult to determine whether treatment methods influence the prognostic value of CTCs. The reason for the drawback is that there was a notable difference in timing for CTCs detection, response to treatment, chemotherapy regimens and operative plans. Further studies are required to investigate the prognostic relevant factor.
To explore potential causes of heterogeneity, we performed meta-regression to evaluate whether percentage of advanced stage patients (stage IV or stage III) would influence the prognostic value of CTCs. No association could be observed (p = 0.645), which meant the prognostic value of CTCs was not associated with disease stage. Moreover, no significant association could be observed in publication year (0.619) treatment type (0.469) and sample size (0.059).
Our meta-analysis also confirmed the presence of CTC was closely associated with elevated CA-125, both of which are known to be prognostic tool in ovarian cancer.
Our study has some limitations. First, CTC detection methods were different among included studies, which may partly influence the significance in survival analyses. Second, there is no consensus on the optimal cutoff of CTCs for predicting the clinical outcome in ovarian cancer. The low cutoff of ≥2 CTC/7.5 ml of blood with the CellSearch system was used in three studies [14,16,19], and other studies used different cutoffs. Though Fan et al. showed higher CTC counts could reflect later stage disease and higher CA-125 levels [20], few trials had evaluated by the prognostic value of different numbers of enumerating CTCs in patients with ovarian cancer. Further studies are required to assess prognostic relevant CTC cutoff levels. Third, some studies included in the meta-analysis were small case-series, thus more large prospective studies should be performed.
In conclusion, this meta-analysis indicates that CTC status is associated with OS and PFS/DFS in patients with ovarian cancer. To achieve clinical utility of CTCs in ovarian cancer, more large prospective studies should be conducted before CTC detection can be applied to clinical use as a prognostic indication.