Leukocyte alkaline phosphatase and carcinoembryonic antigen in lung cancer patients.

Abstract

Leukocyte alkaline phosphatase (LAP) scores in peripheral blood and plasma carcinoembryonic antigen (CEA) levels were determined in 58 lung cancer patients and compared with those of 22 healthy persons, who served as controls. In the control group we found LAP scores of 54 +/- 28 and CEA levels of 4 +/- 3 ng/ml. The 97.7% confidence intervals are: LAP score < 110 and CEA level < 10. In the patients with limited lung cancer we found LAP scores of 152.4 +/- 36.7 and CEA levels of 22.7 +/- 48 ng/ml. The 97.7% confidence intervals are: LAP score > 78.6 and CEA level > 0. In the patients with extensive lung cancer we found LAP scores of 272 +/- 49 and CEA levels of 47.5 +/- 80.4 ng/ml. The 97.7% confidence intervals are: LAP score > 174 and CEA level > 0. The sensitivity of each marker is discussed, here as percent of false-negative results, denoted PFN, and meaning that with a chosen threshold for the marker level, only PFN% of the patients with a given condition (e.g. extensive lung cancer) might have a marker level less than the chosen threshold and could therefore be assumed healthy. The sensitivity of the LAP score is 0.05% PFN for the group of patients with extensive lung cancer, and we conclude that indeed LAP score < 110 indicates a very low probability of having extensive malignancy. The sensitivity of the CEA level is 32% PFN. The sensitivity of the LAP score to limited lung cancer is 12% PFN for the group of patients with that condition compared with 40% of CEA. The specificities of both LAP score and CEA levels, expressed as percent of false-positive results in the control group, are fair (2.3%), provided that the common conditions, other than cancer, that elevate LAP score and/or CEA levels are well known and can be eliminated by the physician. It is concluded that the LAP score is considerably more useful than the CEA level as a marker for extensive lung cancer.