Effect of lyophilization on determinations of lipoprotein(a) in serum

We studied the effect of lyophilization of serum on the determination of serum lipoprotein(a) [Lp(a)], using two enzyme-linked immunosorbent assays (ELISAs) and one immunoturbidimetric assay (ITA). After adjusting for the effect of dilution due to the reconstitution of lyophilized serum, Lp(a) values obtained for lyophilized sera were consistently lower than the original values for the corresponding fresh samples, the difference ranging from 17% to 83% for the monoclonal-polyclonal antibody-based ELISAs and from 12% to 49% for the exclusively polyclonal ITA. Statistical analysis showed that differences between either ELISA and the ITA were significant but not the differences between the ELISAs. Moreover, the effect was not qualitatively or quantitatively uniform, which suggested that the effect was serum-source dependent. Addition of sucrose (600 mmol/L) to sera before lyophilization decreased but did not eliminate the effect. Apparently, serum pools in lyophilized form are not a suitable reference material for Lp(a) determination. Because different methods produce different results for Lp(a) in the same common reference material intended for standardization of Lp(a) methodology, any Lp(a) method should be tested for the effect of lyophilization if lyophilized samples are to be analyzed.