HIGH-PERFORMANCE LIQUID-CHROMATOGRAPHY TANDEM MASS-SPECTROMETRY AS A REFERENCE FOR ANALYSIS OF TACROLIMUS TO ASSESS 2 IMMUNOASSAYS IN PATIENTS WITH LIVER AND RENAL-TRANSPLANTS
P. Salm et al., HIGH-PERFORMANCE LIQUID-CHROMATOGRAPHY TANDEM MASS-SPECTROMETRY AS A REFERENCE FOR ANALYSIS OF TACROLIMUS TO ASSESS 2 IMMUNOASSAYS IN PATIENTS WITH LIVER AND RENAL-TRANSPLANTS, Therapeutic drug monitoring, 19(6), 1997, pp. 694-700
The accuracy and imprecision of three assays used for therapeutic moni
toring of tacrolimus were tested using blood-containing weighed-in amo
unts of the drug, an enzyme-linked immunosorbent assay (ELISA), a micr
oparticle enzyme immunoassay (MEIA I), and a high-performance liquid c
hromatography-tandem mass spectrometry (HPLC-MS2) assay. Accuracy was
acceptable for the HPLC-MS' assay al all concentrations tested (<10% d
eviation) and for the ELISA at 1.0 and 4.0 mu g/l. Accuracy was not ac
ceptable for the ELISA at IS,15.0 and 50.0 mu g/l or for the MEIA I at
all concentrations tested. Imprecision was acceptable for the HPLC-MS
2 assay at all concentrations tested (coefficient of variation < 10%),
for the ELISA at 15.0 and 50.0 mu g%l, and for the MEIA I al 15.0 and
50.0 mu g/l. Imprecision was not acceptable for the ELISA at 1.0 and
4.0 mu g/l or for the MEIA I at 1.0 and 4.0 mu g/l. This assessment wi
th weighed-in amounts of tacrolimus verified the HPLC-MS2 assay as a r
eference method. The performance of the two immunoassays? with HPLC-MS
2 was then compared in the clinical setting using blood from patients
with; liver(n = 30) and renal (n = 37) transplants. In the liver trans
plant group (127 samples), the range. of tacrolimus concentrations mea
sured by HPLC-MS2, ELISA, and MEIA I was 1.9 to 31.8, 2.1 to 35.0, and
less than 0.1 to 36.5, mg/l, respectively, In tile renal transplant g
roup (129 samples), the ranges were;ere 1.7 to 26.1, 1.9 to 24.4, and
0.9 to 28.5 mu g/l, respectively. Compared with the HPLC-MS2, the ELIS
A had minimal bias (0.1 to 0.2 mu g/l) but unacceptable variability in
values (SD > 13%), The MEIA I had unacceptable bias (1.7-1.8 mu g/l)
and variability (SD > 23%). These data indicated that neither the ELIS
A nor MEIA I is interchangeable with HPLC-MS2., Moreover, in view of t
he current trend to reduce the therapeutic dose of tacrolimus, quantit
ative results using the MEIA I would not be obtainable during therapeu
tic drug monitoring in some patients in whom effective therapeutic con
centrations can be less than 5.0 mu g/l.