Clinical value of serum markers in breast neoplasm: 5-years follow-up

Several serum markers have been proposed either for diagnosis or follow-up of women with breast cancer such as CEA, CA15-3 and more recently MCA. In this retrospective analysis we evaluated the correlation between breast cancer and serum markers and their changes related to surgical treatment, with a minimum 5-year postoperative follow-up. We evaluated retrospectively 66 women who underwent curative surgery for primary breast cancer. Patients with a pT value > 2 (calculated as a mean of the two maximum diameters) and all M+ were excluded, as well as patients with a discontinuous follow-up. Mean age was 61 years (median 61, range 31-87); 52 (78.8%) were older than 50 years. Patients were further divided into 3 groups: Group 1 (pTlpN0), Group 2 (pT2pN0) and Group 3 (pT2pN+). Mean pts age in the 3 groups was respectively: 53, 65 and 63 years. Preoperative CEA serum levels (normal values 0-10 ng/ mL) were available in all patients, and CA 15.3 levels (normal values &31 kU/L) only in 45 (68.1%). Markers were measured in the same Laboratory with high sensibility techniques. Further blood samples were collected at least at 6 months and 1, 2, 3 and 5 years after operation. Mean follow-up for the 62 alive patients was 88 months (median 78, range 60-138 months) and overall survival rate was 94.0%. Three patients died respectively 26, 52 and 54 months after operation of metastatic disease; another patient died of acute miocardial infarction 13 months after surgery. Student’s t-test was used for statistical significance study. CEA mean serum concentration was 8.58±7.31 ng/mL (median 7.0, range 1-19) in Group 1; 5.55±5.44 (median 4.0, range 1-24) in Group 2; 6.28±6.66 (median 3.0, range 1-22) in Group 3. Preoperative CA 15-3 was available respectively in 16/19, 16/29 and 1308 patients in each group. Mean CA 15.3 value was 34.06±18.90 kU/L (median 34.0, range 12-82) in Group 1; 27.0±12.15 (median 26.50, range 8-53) in Group 2; 27.38 f 19.90 (median 27.38, range 7-81) in Group 3. No statistically significant difference was found among the 3 groups. Preoperative increased CEA serum levels was found in 14 (21.2%) patients and respectively in 709, l/29, 6/18 patients in each group; increased basal CA 15-3 serum levels was found in 14 (31.1%) patients and respectively in 7/16,4/16, 3/13 patients. CEA serum level at 6 months were 2.63±2.06 in Group 1; 2.72 f 1.77 in group 2; 4.22±3.28 in Group 3 and mean CA 15-3 levels were respectively 22.19±12.61, 19.19±4.93 and 15.46±9.50. Five years after operation, 62 patients were still alive. In these patients CEA levels were: 2.58±1.89 (Group l), 3.07±3.26 (Group 2), 2.29±1.07 (Group 3) and CA 15.3 levels were: 23.18±17.80 (Group l), 19.84 f 6.41 (Group 2), 16.77±5.37 (Group 3). Only in Group 1, a statistically significant difference in CEA and CA 15.3 levels (0 vs. 6 months and 0 vs. 5 years) was found. Ten (15.1%) patients showed relapse of malignancy during follow-up and mean relapse time was 40.3 months. In 7/10 and 3/10 patients respectively, CA 15-3 and CEA serum levels were increased at l-6 months before recurrence. Serum markers are not useful in diagnosis of breast cancer. In our 5 year long follow-up in a smaller group of ten pts who showed recurrence, elevation of serum CA 15.3 and CEA occurred respectively in seven and three pts with a mean of four months before evidence of relapse, while in the group of pts who had no evidence of recurrence (N=56) only in three cases was there an increase of serum CEA and CA 15.3 was discovered, in one case CEA and CA 15.3 increase was due to alcohol abuse and returned to normal. We conclude that serum markers CA 15.3 and CEA are useful during follow-up for the early detection of recurrences.