Serum CYFRA 21-1 in advanced stage non-small cell lung cancer: An early measure of response

• Published in: Clinical cancer research 2003-05, Vol.9 (5), p.1728-1733
• Main Authors: VOLLMER, Robin T, GOVINDAN, Ramaswamy, GRAZIANO, Stephen L, GAMBLE, Gary, GARST, Jennifer, KELLEY, Michael J, CHRISTENSON, Robert H
• Format: Article
• Language: English
• Subjects: Adenocarcinoma - blood; Adenocarcinoma - drug therapy; Adenocarcinoma - secondary; Adult; Aged; Aged, 80 and over; Antigens, Neoplasm - blood; Antineoplastic Combined Chemotherapy Protocols - therapeutic use; Biological and medical sciences; Biomarkers, Tumor - blood; Carcinoma, Non-Small-Cell Lung - blood; Carcinoma, Non-Small-Cell Lung - drug therapy; Carcinoma, Squamous Cell - blood; Carcinoma, Squamous Cell - drug therapy; Carcinoma, Squamous Cell - secondary; Female; Humans; Keratin-19; Keratins; Lung Neoplasms - blood; Lung Neoplasms - drug therapy; Male; Medical sciences; Middle Aged; Neoplasm Staging; Pilot Projects; Pneumology; Prognosis; Sensitivity and Specificity; Survival Rate; Tumors of the respiratory system and mediastinum
• ISSN: 1078-0432; 1557-3265

Our objective was to test the prognostic importance of both the pretreatment level and change in serum CYFRA 21-1 after one cycle of chemotherapy in patients with advanced non-small cell lung cancer (NSCLC) and to compare these two CYFRA variables to routine clinical stage and response as measured by imaging. Our patients consisted of 58 with advanced NSCLC who were treated with chemotherapy. Fourteen were stage IIIa, 8 stage IIIb, and 36 stage IV, and none had received previous treatment. The choice of chemotherapy was left to the discretion of the treating physicians. We collected two serum samples, one before the first cycle of chemotherapy and the second before the second cycle, and analyzed these for serum CYFRA 21-1 using an electrochemiluminescence immunoassay and the ElecSys 2010 system (Roche Diagnostics Corp., Indianapolis, IN). We expressed changes in CYFRA in terms of the natural ratio logarithm of post-treatment to pretreatment CYFRA, and we used the Cox proportional hazards model to analyze survival time. Patients experienced an average drop of 27% in serum CYFRA after the first cycle of chemotherapy. Furthermore, the Cox model demonstrated that both the initial natural logarithm of serum CYFRA and presence of >27% drop in CYFRA were significantly related to subsequent survival (model P < 0.0006), but neither clinical stage nor clinical response related to survival (P > 0.1). In advanced stage NSCLC, the initial level of serum CYFRA appears to provide more prognostic information than clinical stage. Furthermore, a drop of >27% in CYFRA after one cycle of therapy adds prognostic information, so that this threshold appears to be an early measure of response to chemotherapy.