OBJECTIVE: Our purpose was to determine the utility of measuring serum leuk
emia inhibitory factor, a cytokine expressed in the process of pregnancy im
plantation, for the diagnosis of ectopic pregnancy.
STUDY DESIGN: Serum samples from 40 patients with positive serum quantitati
ve beta -human chorionic gonadotropin levels were used for leukemia inhibit
ory factor determination. The serum leukemia inhibitory factor concentratio
n was determined by enzyme-linked immunosorbent assay in the following 4 gr
oups: (1) normal intrauterine pregnancies, (2) threatened abortions, (3) sp
ontaneous abortions, and (4) ectopic pregnancies.
RESULTS: Ail patients had detectable concentrations of leukemia inhibitory
factor in serum, ranging from 2.44 to 8.25 pg/mL. Mean leukemia inhibitory
factor concentrations for ectopic pregnancy were significantly lower (P < .
05) than those of both the spontaneous abortion and threatened abortion gro
ups by 1-way analysis of variance. When a cutoff point of serum leukemia in
hibitory factor < 6.2 pg/mL is assigned as diagnostic of ectopic pregnancy,
leukemia inhibitory factor in patients with ectopic pregnancies versus all
other groups predicted ectopic pregnancy with a sensitivity of 73%, specif
icity of 72%, positive predictive value of 50%, and negative predictive val
ue of 88%.
CONCLUSION: Serum leukemia inhibitory factor concentration is lowest in pat
ients with ectopic pregnancy. A cutoff point of 6.2 pg/mL maximizes the sen
sitivity and specificity of the lest; however, it is not sufficiently discr
iminatory to be used clinically for the diagnosis of ectopic pregnancy.