Background: No large-scale studies have examined the use of serial measurements of serum p53 antibodies (s-p53Abs) combined with carcinoembryonic antigen (CEA) measurements during the follow-up of colorectal cancer (CRC) patients after curative resection.
Methods: A highly specific enzyme-linked immunosorbent assay was used to analyze s-p53Abs levels in 305 CRC patients before and after curative resection at a single institution. Agreement between recurrence and serial s-p53Ab and CEA measurements was evaluated by diagnostic accuracy odds ratio (DOR), kappa, and area under receiver operating characteristic curve (AUC).
Results: Among 305 patients, 76 (25%) patients had disease recurrence during follow-up. None of the 168 s-p53Ab seronegative patients (s-p53Ab < 10 U/microL) without recurrence had an abnormal s-p53Ab test during follow-up. Among the remaining low-level (10 U/microL <or= s-p53Ab <or= 76 U/microL, n = 103) and high-level (s-p53Ab titer > 76 U/microL, n = 34) seropositive patients, recurrence defined by s-p53Ab tests resulted in a DOR of 4.3 and infinity, a kappa of 0.35 and 1.00, and an AUC of 0.633 [95% confidence interval (CI), 0.495 to 0.772; P = 0.047], and 1.0 (95% CI, 1.000 to 1.000; P < 0.0001), respectively. Recurrence defined by CEA tests had an AUC of 0.781 (95% CI, 0.654 to 0.909) for low-level and 0.796 (95% CI, 0.611 to 0.982) for high-level seropositive patients.
Conclusions: Agreement between clinical recurrence and serial s-p53Ab test was dependent upon preoperative s-p53Ab level. Serial s-p53Ab testing outperformed CEA testing when predicting clinical recurrence in colorectal cancer patients with an abnormal preoperative s-p53Ab level.