Prognostic significance of prospectively detected bone marrow micrometastases in esophagogastric cancer: 10-year follow-up confirms prognostic significance.

Abstract

We have previously reported that most patients with esophagogastric cancer (EGC) undergoing potentially curative resections have bone marrow micrometastases (BMM). We present 10-year outcome data of patients with EGC whose rib marrow was examined for micrometastases and correlate the findings with treatment and conventional pathologic tumor staging. A total of 88 patients with localized esophagogastric tumors had radical en-bloc esophagectomy, with 47 patients receiving neoadjuvant (5-fluorouracil/cisplatin based) chemoradiotherapy (CRT) and the remainder being treated with surgery alone. Rib marrow was examined for cytokeratin-18-positive cells. Standard demographic and pathologic features were recorded and patients were followed for a mean 10.04 years. Disease recurrences and all deaths in the follow-up period were recorded. No patients were lost to follow-up. 46 EGC-related and 10 non-EGC-related deaths occurred. Multivariate Cox analysis of interaction of neoadjuvant chemotherapy, nodal status, and BMM positivity showed that the contribution of BMM to disease-specific and overall survival is significant (P = 0.014). There is significant interaction with neoadjvant CRT (P < 0.005), and lymph node positivity (P < 0.001) but BMM positivity contributes to increase in risk of cancer-related death in patients treated with either CRT or surgery alone. Bone marrow micrometastases detected at the time of surgery for EGC is a long-term prognostic marker. Detection is a readily available, technically noncomplex test which offers a window on the metastatic process and a refinement of pathologic staging and is worthy of routine consideration.