Production of placental alkaline phosphatase (PLAP) and PLAP-like material by epithelial germ cell and non-germ cell tumours in vitro.

Placental and placental-like alkaline phosphatase (PLAP) levels in the culture media of 87 cell lines of neoplastic and 'normal' origin were measured by a conventional immunosorbent enzymatic assay (IAEA) and by a new immunoradiometric assay (IRMA). The IRMA detected immunoreactive PLAP in 37 of 80 (46%) human epithelial and germ cell cultures, while the IAEA detected PLAP in only 25 (33%). Of the 52 non-germ cell tumour cultures, the IRMA detected expression in 24 (46%) and the IAEA in only 16 (31%). In 17 cases (21%) the IRMA recorded levels double that of the IAEA, while in five cultures (6%) the reverse was true. The IRMA was much more robust than the IAEA and had considerably lower inter- and intra-assay coefficients of variation (3.75-8.5% vs 5.2-46%). Detection of PLAP(-like) expression by IAEA is dependent on neoplastic expression of enzymatically functional molecules and quantification assumes constant enzyme kinetics. PLAP-like material has a higher catalytic rate constant than PLAP and thus will give higher values on a stoichiometric basis in an IAEA. The higher detection rate and levels of PLAP-like material in neoplastic cultures when measured by the IRMA clearly demonstrate ectopic expression of non-enzymatic PLAP and PLAP-like genes. The incidence of PLAP(-like) expression by non-germ cell and possible germ cell tumours has been underestimated and its utility as a tumour marker should be re-examined using assays which measure antigen mass rather than phosphatase activity.