Cytokeratin-18 fragments predict treatment response and overall survival in gastric cancer in a randomized controlled trial

Abstract

Background: Gastric cancer is common malignancy and exhibits a poor prognosis. At the time of diagnosis, the majority of patients present with metastatic disease which precludes curative treatment. Non-invasive biomarkers which discriminate early from advanced stages or predict the response to treatment are urgently required. This study explored the cytokeratin-18 fragment M30 and full-length cytokeratin-18 M65 in predicting treatment response and survival in a randomized, placebo-controlled trial of advanced gastric cancer. Methods: Patients enrolled in the SUN-CASE study received sunitinib or placebo as an adjunct to standard therapy with leucovorin (Ca-folinate), 5-fluorouracil, and irinotecan in second or third line. Treatment response rates, progression-free survival and overall survival were assessed during a follow-up period of 12 months. Cytokeratin-18 fragments were analyzed in 52 patients at baseline and day 14 of therapy. Results: Levels of M30 correlated with the presence of metastasis and lymph node involvement and decreased significantly during chemotherapy. Importantly, baseline levels of M30 were significantly higher in patients who failed therapy. In addition, patients who did not respond to treatment were also identifiable at day 14 based on elevated M30 levels. By stepwise regression analysis, M30 at day 14 was identified as independent predictor of treatment response. Likewise, serum levels of full-length cytokeratin-18 M65 at baseline also correlated with treatment failure and progression-free survival. The addition of sunitinib did not exert any effects on serum levels of M30 or M65. Conclusion: The cytokeratin-18 fragment M30 at day 14 identifies patients that fail to second- or third-line therapy for advanced gastric cancer. Validation of this non-invasive biomarker in gastric cancer is warranted.