Prognostic and Predictive Model for Stage II Colon Cancer Patients With Nonemergent Surgery

Abstract

No ideal prognostic model has been applied to clearly identify which suitable high-risk stage II colon cancer patients with negative margins undergoing nonemergent surgery should receive adjuvant chemotherapy routinely. Clinicopathologic and prognostic data of 333 stage II colon cancer patients who underwent D2 or D3 lymphadenectomy during nonemergent surgery were retrospectively analyzed. Four pathologically determined factors, including adjacent organ involvement (RR 2.831, P = 0.001), histologic differentiation (RR 2.151, P = 0.009), lymphovascular invasion (RR 4.043, P < 0.001), and number of lymph nodes retrieved (RR 2.161, P = 0.011), were identified as independent prognostic factors on multivariate analysis. Importantly, a simple cumulative scoring system clearly categorizing prognostic risk groups was generated: risk score = ∑ coefficient' × status (AOI + histological differentiated + lymphovascular invasion + LNs retrieved). Our new prognostic model may provide valuable information on the impact of lymphovascular invasion, as well as powerfully and reliably predicting prognosis and recurrence for this particular cohort of patients. This model may identify suitable patients with an R0 resection who should receive routine postoperative adjuvant therapy and may help clinicians to facilitate individualized treatment. In this study, we aim to provide an ideal and quantifiable method for clinical decision making in the nonemergent surgical treatment of stage II colon cancer. Our prognostic and predictive model should be applied in multicenter, prospective studies with large sample sizes, in order to obtain a more reliable clinical recommendation.