Principal Component Analysis (PCA) was performed employing Lewis y/IgM/CIC, Lewis y/IgG/CIC, MUC1/IgG/CIC and MUC1/IgM/CIC. First and second component explained 68% of data variability; normal samples and benign samples appeared grouped (PC1 (-)) and separated from cancer samples which remained spread. All variables weighed similar in the model, Lewis y/IgM/CIC, MUC1/IgG/CIC and MUC1/IgM/CIC predominated PC1 (+) while Lewis y/IgG/CIC was shared between PC1(+) and PC2(+). Rays and circles represent CIC analyzed and cases, respectively. C: cancer, B: benign, N: normal.