Differentiating tubal abortion from viableectopic pregnancy with serum CA-125 andb-human chorionic gonadotropindeterminations

Mladen Predanic, M.D.

Department of Obstetrics and Gynecology, Flushing Hospital Medical Center, Flushing, New York

Objective: To determine whether serum CA-125 and serialb-hCG levels can be used to distinguish betweentubal abortion and viable ectopic pregnancy (EP).

Design: Retrospective cohort study.

Setting: A tertiary care institution.

Patient(s): Twenty-six women with EPs of 7–12 weeks’ duration were studied retrospectively. Five hadlaparoscopically proved tubal abortions and 21 had active, viable EPs at the time of entry into the study. Allbut 3 of the latter group were managed surgically; the others were given a single dose of methotrexate.

Intervention(s): Surgical removal of EPs by means of laparoscopy or laparotomy, or medical treatment ofthe disease.

Main Outcome Measure(s): Serum CA-125 andb-hCG determinations were used to differentiate tubalabortion and viable EP. The results were compared with the findings at surgery.

Result(s): The mean (6SD) CA-125 level was 112.26 11.9 IU/mL for the patients with tubal abortion and30.16 15.3 IU/mL for the patients with viable EP. The mean (6SD) b-hCG level was 3,6436 3,718 IU/Lfor the patients with tubal abortion and 10,7556 11,465 IU/L for the patients with viable EP. Linear regressionanalysis showed a statistically insignificant inverse relation between serum CA-125 andb-hCG levels.

Conclusion(s): The use of CA-125 levels as an adjunct to serialb-hCG levels shows promise as a means fordifferentiating tubal abortion from viable EP. (Fertil Sterilt 2000;73:522–5. ©2000 by American Society forReproductive Medicine.)

Key Words: Ectopic pregnancy, tubal abortion, CA-125,b-hCG

It is difficult to distinguish clinically be-tween an ectopic pregnancy (EP) that is viableand continues to threaten the well-being of thepatient and one that is no longer viable (i.e., atubal abortion). The ability to make this differ-ential diagnosis easily would be useful to phy-sicians, particularly because it can help incounseling the patient, managing her condition,and prognosticating. Serum CA-125 levelshave been shown to correlate with adverse out-comes of intrauterine pregnancies in somestudies (1–3) but not in others (4).

Serum hCG levels reflect trophoblastic ac-tivity and rise rapidly in early pregnancy, re-gardless of whether the pregnancy is intrauter-ine or ectopic. Although they may fall if thepregnancy fails, they are not completely reli-

able in this regard. The degeneration of thedecidua, amnion, and chorion that precedesfetal demise liberates CA-125 into the maternalcirculation (5).

We hypothesized that the correlation be-tween high CA-125 levels and adverse preg-nancy outcome that is seen in intrauterine preg-nancies also may be present in EPs. Our aimwas to determine whether high CA-125 levelswould complement low or falling serumb-hCG levels in discriminating between tubalabortion and viable EP.

MATERIALS AND METHODS

A total of 39 patients were admitted toFlushing Hospital Medical Center during the15-month period between September 1997 and

Received July 7, 1999;revised and acceptedOctober 27, 1999.Reprint requests: MladenPredanic, M.D.,Department of Obstetricsand Gynecology, FlushingHospital Medical Center,4500 Parsons Boulevard,Flushing, New York 11355(FAX: 718-670-5780).

FERTILITY AND STERILITY tVOL. 73, NO. 3, MARCH 2000

Copyright ©2000 American Society for Reproductive MedicinePublished by Elsevier Science Inc.

Printed on acid-free paper in U.S.A.

0015-0282/00/$20.00PII S0015-0282(99)00554-3

522

December 1998 with a confirmed diagnosis of EP. Amongthese patients, 26 (68%) met our study criteria, which in-cluded an accurate menstrual history and a clinical constel-lation of manifestations that supported a diagnosis of unrup-tured EP, with cardiovascular stability and minimal or nointraabdominal bleeding.

The gestational age ranged from 7–12 weeks as calculatedby the date of the last menstrual period. The patients allpresented with lower quadrant abdominal pain, abdominaland pelvic tenderness, and a palpable adnexal mass. Thefindings on ultrasonographic imaging, which was done in allcases as part of the diagnostic work-up, did not have adeterminant role in patient selection. Some patients under-went dilatation and curettage as part of their work-up. Pa-tients who had positive serum hCG levels but no chorionicvilli present at dilatation and curettage and no ectopic prod-ucts of conception on diagnostic laparoscopy were excludedfrom the study. Patients with a history of endometriosis alsowere excluded.

Blood samples were drawn routinely as part of the pread-mission emergency department evaluation, obviating theneed to obtain specific patient consent or Institutional Re-view Board approval. Serum was separated from bloodsamples by centrifugation and stored at220°C after beingassayed forb-hCG. When an EP was confirmed, a serumCA-125 level was measured in the same stored serum spec-imen. The activity ofb-hCG and CA-125 was measured byRIA (Abbott Laboratories, Chicago, IL).

Definitive management was accomplished either on theday of admission (16 cases) or, if it was necessary to confirmthe diagnosis by serial hCG determinations, within 48 hoursof admission (10 cases). Primary surgery was done in 23cases, by laparoscopy in 20 cases and by laparotomy in 3cases. The remaining 3 patients were treated with a single IMinjection of methotrexate (50 mg/m2). In no instance had theEP ruptured. Follow-up showed that the management pro-gram was successful in all cases; no patient required anyadditional therapy.

In 5 of the 20 laparoscopy cases, the products of concep-tion were found floating freely in the pelvis or imbedded inorganized or fresh blood clots; these 5 cases were designatedtubal abortions after confirmation by pathologic examina-tion. Follow-up of these 5 tubal abortions with serialb-hCGdeterminations showed no persistent residual ectopic tropho-blastic tissue.

Ectopic pregnancy was clearly viable on gross inspectionand pathologic examination in the remaining 15 laparoscopycases as well as the 3 laparotomy cases. This group includedone fimbrial pregnancy, which the surgeon considered torepresent a tubal abortion by naked-eye examination; how-ever, the case was included among the viable EPs on thebasis of histologic evidence. In addition to these 18 surgi-cally managed cases, the 3 that were treated with methotrex-

ate also were deemed to have been viable EPs, for a total of21 cases. These 3 EPs were considered to be viable on thebasis of ultrasonographic findings that unequivocally showed acystic adnexal mass with structures that resembled a yolksac. The size of these cystic masses did not exceed 3 cm inthe largest diameter, and a minimal amount of free fluid wasfound in the pelvic cavity.

The relation between CA-125 andb-hCG serum levelswas examined by linear regression analysis. In addition, dataon serum CA-125 levels from patients with viable EPs werecompared with those from patients with tubal abortions.P,.05 was considered statistically significant.

RESULTS

The distributions of CA-125 levels in the two groups ofcases were impressively distinctive by virtue of the absenceof any overlap (Fig. 1), suggesting that CA-125 levels mayserve as a utilitarian discriminator. The mean (6SD) CA-125 level for the 5 patients with tubal abortion was 112.2611.9 IU/mL (range, 96–125 IU/mL). For the patients withviable EP, the mean (6SD) CA-125 level was 30.16 15.3IU/mL (range, 10–66 IU/mL). The difference was statisti-cally significant (two-tailedt-test5 11.2,df 5 24, P,.001,95% confidence interval5 93.3–67).

The distributions ofb-hCG levels were not clearly dif-ferent from each other because of their overlap (Fig. 2). Themean (6SD) b-hCG level for the patients with tubal abor-tion was 3,6436 3,718 IU/L (range, 534–9,684 IU/L),whereas that for the patients with viable EP was 10,755611,465 IU/L (range, 927–38,113 IU/L). The difference was

F I G U R E 1

Serum CA-125 levels in the patients with tubal abortion (n 55, solid squares) and viable EP (n 5 21, open circles).

Predanic. Diagnosis with CA-125. Fertil Steril 2000.

FERTILITY & STERILITY t 523

statistically significant, but not to the same degree as forCA-125 levels (two-tailedt-test 5 2.2, df 5 24, P5.04,95% confidence interval5 378–12,672).

Serum CA-125 levels were not significantly related tob-hCG levels when examined by linear regression analysis,even though there was somewhat of a trend for the CA-125level to rise as theb-hCG level fell. The slope of the curverepresenting this relation was determined to be240.846 andis expressed in the following linear regression equation:

CA-125 5 50.20 2 @0.000489 3 b-hCG]

This equation shows a negligible rise of 5 IU/mL inCA-125 levels for every 10,000-IU/L fall inb-hCG levels.Given the highP value and the broad spread of the confi-dence interval (t-test 5 0.6994,df 5 24, P5.491, 95%confidence interval5 2161.3787–79.6867), this relationcould readily have occurred by chance alone, and thereforecannot be considered clinically meaningful.

DISCUSSION

The CA-125 tumor marker is a cell-surface antigen de-rived from the surface coelomic epithelium, including themucosa of the entire female genital tract and the germinalepithelium of the ovaries (5). The fetal chorion, amnioticfluid, and maternal decidua also have been shown to containsignificant amounts of CA-125 protein (6).

Serum CA-125 levels are increased early in pregnancy

and immediately after birth, implicating disintegration of thematernal decidua (i.e., blastocyst implantation and placentalseparation) as a possible source of the tumor marker eleva-tion (7, 8). A significant increase in serum CA-125 levelsalso was reported in a group of patients with vaginal bleed-ing and impending spontaneous abortion (1–3). It was con-cluded that the extension of decidual destruction and tropho-blast separation from decidual cells was the major source ofthe maternal serum CA-125 elevation, but this has not beensupported by other studies (4, 9).

A connection between serum CA-125 levels and abnor-mal early intrauterine pregnancies has been investigated todetermine whether it might be useful in the assessment ofEPs. It was believed that destruction of fetal tissues and/oroviduct mucosa would yield different serum CA-125 valuesin patients with EPs compared with patients with normalintrauterine pregnancies. One of the cross-sectional studiesreported significantly lower serum CA-125 concentrations inwomen with EPs (9). The low CA-125 levels were explainedby the impaired interaction between the fetal trophoblast andtubal mucosa (9).

In contrast, another study demonstrated significantly ele-vated serum CA-125 levels in a group of patients withruptured and unruptured EPs (10). It was speculated thatelevations in serum CA-125 levels could be attributed to theextent of destruction of the tubal mucosa and the “leakage”of CA-125 from mucosal cells into maternal sera. However,the investigators did not find a statistical difference betweenthe serum CA-125 levels in patients with ruptured and un-ruptured EPs to confirm their hypothesis that elevated serumCA-125 levels are correlated with ruptured EPs (10).

Therefore, one may speculate that elevated serum CA-125 levels could be associated with tubal abortion; this wasreported in the group of patients with ruptured EPs but wasnot analyzed separately (10). We hypothesized that highserum levels of CA-125 might complement low or fallingserum levels ofb-hCG and help discriminate between tubalabortion and viable EP.

Our data showed statistically significant differences inserum CA-125 levels between patients with tubal abortionsand patients with viable EPs, confirming our hypothesis.However, because of the small sample size, low serumCA-125 levels were more prognostic of active and viableproducts of conception rather than high serum CA-125 levelsbeing useful for detection of tubal abortion. In addition, datacollection bias and bias in diagnosing an EP as viable ornonviable further compromised the validity of our data.These biases could be avoided by studying a larger group ofpatients and performing pathologic examination of removedtubes to discriminate viable EP from tubal abortion on thebasis of trophoblast penetration into the tubal mucosa. Wealso observed a trend toward lower serumb-hCG levels andhigher serum CA-125 levels in relation to nonviability anddisintegration of the ectopic products of conception.

F I G U R E 2

Serum b-hCG levels in the patients with tubal abortion (n 5 5,solid squares) and viable EP (n 5 21, open circles).

Predanic. Diagnosis with CA-125. Fertil Steril 2000.

524 Predanic Diagnosis of tubal abortion with CA-125 Vol. 73, No. 3, March 2000

The ability to determine that ectopic trophoblastic or fetaltissue is disintegrated would be of great value in the man-agement of hemodynamically stable patients. Patients withassumed tubal abortion could be candidates for expectantand/or medical management, and clinicians could be reas-sured by high serum CA-125 levels, in combination withfalling serialb-hCG levels, that pharmacologic managementof an EP was successful, and vice versa. Our data showedpromising results in the differentiation of tubal abortionsfrom viable EPs. However, our conclusions require substan-tiation in larger patient populations.

Acknowledgments:The author thanks Emenuel A. Friedman, M.D., D.Sc.,for his help, support, and guidance during this study.

References1. Azogui G, Yaronovski A, Zohar S, Ben-Shlomo I. CA-125 is elevated

in viable pregnancies destined to be miscarried: a prospective longitu-dinal study. Fertil Steril 1996;65:1059–61.

2. Noci I, Biagiotti R, Periti E, Baronci D, Torricelli F, Cefala L, et al.Maternal serum CA 125 levels in first trimester abortion. Eur J ObstetGynecol Reprod Biol 1995;60:35–6.

3. Check HJ, Nowroozi K, Winkel CA, Johnson T, Seefried L. SerumCA-125 levels in early pregnancy and subsequent spontaneous abor-tions. Obstet Gynecol 1990;75:742–4.

4. Hornstein MD, Check JH, Hill AJ. Serum CA 125 levels and sponta-neous abortion. Am J Obstet Gynecol 1995;172:695–9.

5. O’Brien TJ, Hardin JW, Bannon GA, Norris JS, Quirk JG. CA 125antigen in human amniotic fluid and fetal membranes. Am J ObstetGynecol 1986;155:50–5.

6. Kabawat SE, Bast RC Jr, Bhan AK, Welch WR, Knapp RC, Colvin RB.Tissue distribution of a coelomic-epithelium related antigen recognizedby the monoclonal antibody OC-125. Int J Gynecol Pathol 1983;2:275–85.

7. Spitzer M, Kaushal N, Benjamin F. Maternal CA-125 levels in preg-nancy and puerperium. J Reprod Med 1998;43:387–92.

8. Jacobs IJ, Fay TN, Yovich J, Stabile I, Frost C, Turner J, et al. Serumlevels of CA-125 during the first trimester of normal outcome, ectopicand anembryonic pregnancies. Hum Reprod 1990;5:116–22.

9. Brumsted JR, Nakajima ST, Badger G, Riddick DH, Gibson M. Serumconcentration of CA-125 during the first trimester of normal andabnormal pregnancies. J Reprod Med 1990;35:499–502.

10. Sadovsky Y, Pineda J, Collins JL. Serum CA-125 levels in women withectopic and intrauterine pregnancies. J Reprod Med 1991;36:875–8.

FERTILITY & STERILITY t 525