The trusted source for
healthcare information and
abstract & commentary
Synopsis: This investigation was designed to evaluate the use of maternal creatine phosphokinase in predicting the presence of an ectopic pregnancy in an emergency department setting.
Source: Birkhahn RH, et al. Am J Emerg Med 2000; 18(6):695-697.
Biochemical markers of early normal and abnormal pregnancy have been investigated to enhance diagnosis of ectopic pregnancy (EP) sufficiently early to increase maternal safety and decrease maternal morbidity. Birkhahn and colleagues measured serum creatine phosphokinase (CPK) levels in women presenting to an urban emergency room (ER) with a diagnosis of first-trimester pregnancy and bleeding and/or abdominal pain. A case-control design matched women with a diagnosis of EP through randomization to a cohort of women with nonEP. The mean serum CPK was 118 mIU/dL in the EP group and 64 mIU/dL in the non-EP group. A CPK more than 70 mIU/dL was 100% sensitive for detection of EP, with a specificity of 61.9%, positive predictive value of 72.4%, and negative predictive value of 100%.
Lavie and colleagues first introduced CPK as a potential marker for EP in 1993, theorizing that the invasion of tubal smooth muscle by trophoblastic tissue would produce a rise in CPK.1 CPK is elevated in the presence of damage to skeletal muscle, heart, smooth muscle, and brain; and serum levels are lower than normal or normal in pregnancy until patients enter labor. Other investigations of CPK for diagnosis of EP have used different study designs and produced variable results, with some studies finding CPK to be useful in differentiating ectopic and non-ectopic pregnancy, and others finding no difference.2-5
The use of CPK for determination of EP is limited in several clinical situations. Birkhahn et al excluded women with a recent history of chest pain, trauma, myositis, cocaine use, known renal disease, or intramuscular injection. EPs in locations where the developing trophoblast does not invade muscle will not develop CPK elevation. Birkhahn et al note that the positive predictive value obtained in this study is falsely elevated due to the case control study design with an "artificial" pretest probability of 50% for EP. In an urban ED population with a 10% prevalence of EP, the positive predictive value would be reduced to 23%, while the negative predictive value would remain 100%.
Comment by Ellen L. Sakornbut, MD
EP presents as a diagnostic concern to physicians from multiple disciplines in office or clinic settings and urgent and emergency care centers. While diagnostic goals are similar for all clinicians, resources to make the diagnosis are not. In any setting where transvaginal ultrasound is readily available, findings of normal or abnormal intrauterine pregnancy, indeterminate findings, or findings suspicious or confirmatory for EP may be obtained in a brief period of time. The use of laboratory testing in this context is greatest for patients with indeterminate findings, although quantitative HCG levels and progesterone levels may be used for patients diagnosed with EP as prognostic indicators for success of medical treatment such as methotrexate.
Birkhahn et al recommend the use of CPK when ultrasound is nondiagnostic or not immediately available and in patients with HCG levels below 1500 mIU/mL, using a CPK measurement less than 70 mIU/dL as a means of "ruling out" EP. They also recommend a large prospective study to assess further the use of CPK for diagnosis of EP. At this time, the use of CPK in conjunction with other tests to "rule out" EP appears best suited to an ER setting with rapid availability of results. The high negative predictive value of CPK is similar to a serum progesterone level less than 22 ng/mL, but CPK levels are available "STAT" in almost all emergency settings, whereas progesterone levels are not.6
1. Lavie O, et al. Am J Obstet Gynecol 1993;169: 1149-1150.
2. Spitzer M, et al. J Womens Health Gend Based Med 2000;9(5):537-544.
3. Duncan WC, et al. Br J Obstet Gynaecol 1995; 102(3):233-237.
4. Vandermolen DT, Borzelleca JF. Fertil Steril 1996; 65(5):916-921.
5. Lincoln SR, et al. J Assist Reprod Genet 1996; 13(9):702-704.
6. Buckley RG, et al. Ann Emerg Med 2000;36(2):95-100.