Prediction of local recurrence in cervical cancer by a Cox model comprised of lymph node status, lymph-vascular space invasion, and intratumoral Th17 cell-infiltration.

The identification of cervical cancer patients at high risk of local recurrence is urgent to improve the selection of patients for more aggressive treatment. The immune contexture in human tumors has vital impact on clinical outcome. Our aim in the study was to establish a predictive model of local recurrence by assessing the prognostic significance of clinicopathologic features and five immune markers within the tumor microenvironment in cervical cancer. The expression of CD3, CD4, CD8, FoxP3, and IL-17 was assessed by immunohistochemistry in tumor tissue from 153 patients after radical resection for cervical cancer. Prognostic effects of these immune markers and clinicopathologic factors were evaluated by Kaplan-Meier and Cox regression analysis. Local recurrence was observed in 34 % patients (52/153). Independent predictors of tumor recurrence were lymph node status (P = 0.004), lymph-vascular space invasion (P = 0.012), and the number of intratumoral IL-17(+) cells (P = 0.003). The risk of local recurrence was the highest in patients with lymph node positivity, presence of lymph-vascular space invasion, and low prevalent of intratumoral IL-17(+) cells (probability, 73 %; 5-year DFS, 19 %). A Cox model composed of these three features provided a significant higher diagnostic accuracy of local recurrence than each feature alone (P < 0.05). Lymph node status, lymph node space invasion, and number of intratumoral IL-17(+) cells are three independent predictors for recurrence of cervical cancer. Their combination by a Cox model is highly predictive and may help to identify high-risk patients who may benefit from adjuvant chemotherapy.