CEA and CA 19-9 measurement as a monitoring parameter in metastatic colorectal cancer (CRC) under palliative first-line chemotherapy with weekly 24-hour infusion of high-dose 5-fluorouracil (5-FU) and folinic acid (FA).

Background: There have been contradictory reports on the benefit of CEA and CA 19-9 measurements in patients with metastatic colorectal cancer using palliative chemotherapy. The object of this study was to examine the diagnostic accuracy of monitoring of palliative chemotherapy by means of CEA and CA 19-9.
Patients and Methods: The tumour markers CEA and CA 19-9 were subjected to serial measurement in patients with metastatic colorectal cancer (n = 90) using palliative first-line chemotherapy with weekly 24-hour infusion of high-dose 5-FU with FA and were compared with objective response according to WHO criteria. 85 patients could be evaluated. 43 patients (51%) initially had elevated CEA (> or = 10 ng/ml) and 33 patients (39%) elevated CA19-9 (> or = 50 IE/ml). In 24 patients (28%), both markers were initially increased. With CEA positive patients, 143 cycles of chemotherapy were carried out, which showed the following response in the various cycles: 21 episodes with progressions (ePD), 69 episodes with no change (eNC), 53 episodes with partial/complete remission (ePR/eCR). With CA 19-9 positive patients, 100 cycles of chemotherapy were carried out with the following results: 21 episodes with progressions (ePD), 48 episodes with eNC, and 31 episodes with ePR/eCR.
Results: A CEA rise by at least 50% differentiated between ePD versus eNC/ePR/eCR with a sensitivity of 76% and specificity of 90%. With CEA decreases of at least 30% in 99% of these patient episodes (78 of 79), a tumour progression could be excluded. Patients with an initial drop in CEA after the first cycle of chemotherapy of at least 50% of the initial level had a significantly higher probability of achieving an ePR/eCR in further therapy (relative risk 2.9; P = 0.002). With an CA 19-9 increase of at least 30%, a sensitivity progression of 62% and a specifity of 90% could be calculated. A CA 19-9 decrease of at least 60% excludes a progression in 95% of the patient episodes.
Conclusions: A CEA or CA 19-9 rise is only conditionally appropriate for recording progressions. A progression however, can be excluded with falling levels with high diagnostic accuracy, in which CEA offers a greater degree of certainty than CA 19-9. With a drop in CEA 79 of 143 (= 55%) of the CT scans could be saved, in which case 78 of 79 patient episodes (99%) were correctly assessed as 'no progression'. In patients with an increased CEA and CA 19-9 the CEA determination is sufficient for the further monitoring. A confirmation of these results by multicenter trials can result in a considerable decrease of monitoring costs for palliative treatment.