Mucins CA 125, CA 19.9, CA 15.3 and TAG-72.3 as tumor markers in patients with lung cancer: comparison with CYFRA 21-1, CEA, SCC and NSE

Tumour Biol. 2008;29(6):371-80. doi: 10.1159/000181180. Epub 2008 Dec 8.


Tumor marker serum levels were prospectively studied in 289 patients with suspected, but unconfirmed, lung cancer and in 513 patients with lung cancer [417 non-small cell lung cancer (NSCLC) patients and 96 small cell lung cancer (SCLC) patients]. In patients with benign disease, abnormal serum levels were found for the following tumor markers: CEA (in 6.6% of patients); CA 19.9 (6.2%); CA 125 (28.7%); NSE (0.7%); CYFRA (8.7%); TAG-72.3 (4.2%); SCC (3.5%), and CA 15.3 (3.5%). Excluding patients with renal failure or liver diseases, tumor marker specificity improved with abnormal levels in 0.5% for NSE, 0.9% for SCC, 2.8% for CEA, CA 15.3 and TAG-72.3, 3.8% for CA 19.9, 4.2% for CYFRA and 21.4% for CA 125. Excluding CA 125, one of the markers was abnormal in 15% of patients without malignancy. Tumor marker sensitivity was related to cancer histology and tumor extension. NSE had the highest sensitivity in SCLC and CYFRA and CEA in NSCLC. Significantly higher concentrations of CEA, SCC, CA 125, CA 15.3 and TAG-72.3 were found in NSCLC than in SCLC. Likewise, significantly higher CEA (p < 0.0001), TAG-72.3 (p < 0.001), CA 15.3 (p < 0.0001) and CA 125 (p < 0.01) were found in adenocarcinomas than in squamous tumors. Using a combination of 3 tumor markers (CEA, CYFRA 21-1 in all histologies, SCC in squamous tumors and CA 15.3 in adenocarcinomas), a high sensitivity may be achieved in all histological types. Tumor markers may be useful in the histological differentiation of NSCLC and SCLC. Using specific criteria for the differentiation of SCLC and NSCLC, the sensitivity was 84.2 and 68.8%, the specificity was 93.8 and 99.7%, the positive predictive value was 98.3 and 98.5% and the negative predictive value was 57.7 and 93.3%, respectively.

Publication types

  • Comparative Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Aged
  • Antigens, Neoplasm / blood*
  • Biomarkers, Tumor / blood*
  • Female
  • Humans
  • Lung Neoplasms / blood*
  • Lung Neoplasms / immunology
  • Lung Neoplasms / pathology
  • Male
  • Prospective Studies


  • Antigens, Neoplasm
  • Biomarkers, Tumor