- Email: [email protected]
Rapid measurement of urinary pregnanediol glucuronide to diagnose ectopic pregnancy
CA 125 in endometriosis
Volume 159 Number 6
3. Kivinen S, Kuoppala T, Leppilampi M, VuoriJ, Kauppila A. Tumor-associated antigen CA125 before and during the treatment of ovarian carcinoma. Obstet Gynecol 1986;67:468-72. 4. Barbieri RL, Niloff JM, Bast RC Jr, Schaetzl E, Kistner RW, Knapp RC. Elevated serum concentration of CAl25 in patients with advanced endometriosis. Fertil Steril I 986;45:630-4. 5. Guidice LC, Jacobs A, Pineda J, Bell CE, Lippman L. Serum levels of CA 125 in patients with endometriosis: a preliminary report. Fertil Steril I 986;45:876-8. 6. Patton EP, Field CS, Harms RW, Coulam CB. CA125 levels in endometriosis. Fertil Steril 1986;45: 770-3. 7. Pittaway DE, Fayez JA. The use of CAl25 in the diagnosis and management of endometriosis. Fertil Steril I 986;46:790-5. 8. Kabawat SE, Bast RC Jr, Bhan AK, Welch WR, Knapp RC, Colvin RB. Tissue distribution of a coelomicepithelium-related antigen recognized by the monoclonal antibody OC 125. Int J Gynecol Pathol I 983;2:275-85. 9. Pittaway DE, Fayez JA. Serum CA-125 antigen levels increased during menses. AM J OBSTET GYNECOL 1987; 156:75-6. 10. Kauppila A, Telimaa S, Ronnberg L, Vuori J. Placebo-
11. 12.
13.
14. 15.
controlled study on serum concentrations of CA 125 before and after treatment of endometriosis with danazol or high-dose medroxyprogesterone acetate alone or after surgery. Fertil Steril 1988;49:37-41. Barbieri RL, Canick JA, Makris A, Todd RB, Davies IJ, Ryan KJ. Danazol inhibits steroidogenesis. Fertil Steril 1977;28:809-13. Menon M, Azhar S, Menon KMJ. Evidence that danazol inhibits gonadotropin-induced ovarian steroidogenesis at a point distal to gonadotropin-receptor interaction and adenosine 3' ,5' cyclic monophosphate formation. AM .J 0BSTET GYNECOL 1980; 136:524-30. Musich JR, Behrman SJ, Menon KMJ. Estrogenic and antiestrogenic effects of danazol administration in studies of estradiol receptor binding. AM J OBSTET Gvr-
Rapid measurement of urinary pregnanediol glucuronide to diagnose ectopic pregnancy Mark V. Sauer, MD, Michael Vermesh, MD, Robert E. Anderson, MD, Ariel G. Vijod, BS, Frank Z. Stanczyk, PhD, and Rogerio A. Lobo, MD Los Angeles, California We investigated the ability of a single, random, urinary pregnanediol-3a-glucuronide level to differentiate early intrauterine from ectopic pregnancy. Thirty-four patients with intrauterine gestations were compared with 60 patients with ectopic pregnancies. Urinary pregnanediol-3a-glucuronide was measured by radioimmunoassay and enzyme immunoassay. Compared with intrauterine gestations, results demonstrate that urinary pregnanediol-3a-glucuronide is significantly depressed in ectopic pregnancies: 24.5 ± 2.2 versus 4.8 ± 0.7 µg/ml (p = 0.0001 ). Urinary pregnanediol-3a-glucuronide levels obtained by conventional radioimmunoassay correlated closely with values measured in minutes with enzyme immunoassay (r = 0.95, p = 0.0001), and with serum progesterone (r = 0.74, p = 0.0001). Urinary pregnanediol-3a-glucuronide measured by enzyme immunoassay exhibitea predictive values for detecting ectopic gestations comparable with random serum progesterone or serum 13-human chorionic gonadotropin values. We conclude that ectopic gestations demonstrate a reduced level of urinary pregnanediol-3a-glucuronide (55/60 cases) detectable with a rapid enzyme immunoassay, which makes this assay a practical screening test in early pregnancy. (AM J 0BSTET GYNECOL 1988;159:1531-5.)
Key words: Pregnanediol-3a-glucuronide, progesterone, ectopic pregnancy, enzyme immunoassay
From the Division of Reproductive Endocrinology, University of Southern California. Presented in part at the Thirty-fifth Annual Meeting of the Society For Gynecologic Investigation, Baltimore, Maryland, March 1720, 1988. Reprint requests: Mark V. Sauer, MD, Women's Hospital, 1240 North Mission Rd., Room L-946, Los Angeles, CA 90033.
Concentrations of serum progesterone and its urinary metabolite, pregnanediol-3a-glucuronide, are reported to be lower in abnormal pregnancies than in normal intrauterine gestations. 1• 2 However, these assays are laborious and therefore not widely used. Thus hormonal monitoring of early pregnancy continues to con-
1531
1532
Sauer et al.
December 1988 Am J Obstet Gynecol
% POSITIVE TESTS
100 80
60 40
20 o__IL_~~~L_~--'~~-.-~-L--~~~L-------,~~~,-~~
9
3
12
15
18
PDG (ug/ml) POSITIVE TEST •
>
ECTOPIC
-
INTRAUTERINE
SPECIFIED LEVEL OF PDG
Fig. I. Urinary pregnanediol-3u-glucur onide (PDG) values measured by enzyme immunoassay for both intrauterine (n = 34) and ectopic gestations (n = 60).
Table I. Hormonal profile of patients with intrauterine (n = 34) and ectopic (n = 60) gestations at time of presentation Intrauterine pregnancy
13-hCG (mIU/ml) Mean± SEM Range Serum progesterone (ng/ml) Mean± SEM Range Urinary PDG (µg/ml) Mean± SEM Range
I
Ectopic pregnancy
18,624 ± 2544 196 - 43,796
2,726 ± 810* 17 - 44,280
28.0 ± 2.2 7.1 ± 53.9
8.3 ± 1.1* 0.2 ± 42.2
24.5 ± 2.2 7.9 ± 73.4
4.8 ± 0.7* 0.2 ± 25.2
*P.;; 0.0001.
sist of quantitative measurement of 13-human chorionic gonadotropin (13-hCG). In most cases, serial serum sampling over several days is required to discriminate abnormal from normal gestations, which often delays definitive diagnosis.' Recently the clinical use of measuring serum progesterone by rapid, direct radioimmunoassay (RIA) to differentiate normal from abnormal pregnancy has been reported.'· 4 The new double antibody assays are reported to be more precise and technically simpler than the serum extraction plus column-type methods, with results obtained in less than 4 hours. 5 • 6 Similarly, urinary pregnanediol-3cx-glu curonide can now be measured in less than 2 hours with the enzyme-multiplied immunoassay technique and· has predicted early preg-
nancy outcomes on the basis of a single random sample. 7 A pregnanediol-3cx-glu curonide enzyme immunoassay, which is available commercially, measures pregnanediol-3cx-glu curonide in less than 10 minutes. Unlike the previously mentioned assays, this assay requires no special equipment or technical training and can be readily run in the office or emergency room. Therefore the purpose of this study was to determine if a single, random, urinary pregnanediol-3cxglucuronide value measured by enzyme immunoassay could reliably differentiate an early intrauterine pregnancy from an ectopic gestation. Furthermore, we compared the predictive value of urinary pregnanediol-3cxglucuronide against that of a single serum progesterone or serum 13-hCG in detecting ectopic pregnancies.
Material and methods Subjects. Between Aug. l, l 987, and Feb. l, l 988, we prospectively studied 60 women who came to the Los Angeles County-University of Southern California Medical Center with suspected ectopic pregnancies. Patients were 5 to 8 weeks from the last menstrual period and hemodynamically stable. Ectopic gestations were later confirmed at surgery. During this time period, 34 women with known intrauterine pregnancies of 5 to 8 weeks' gestation, who attended the family planning clinic, served as normal pregnant control subjects. Intrauterine gestations were confirmed either by serial ultrasound examinations or histologic evidence of products of conception obtained after elective termination of pregnancy. Serum and urine samples. Antecubital blood was
Pregnanediol in ectopic pregnancy
Volume 159 Number 6
1533
Table II. Applicability of a single random test for detecting ectopic gestations based on normal pregnancies representing positive pregnancy tests (serum progesterone ~ 10 ng I ml, urinary pregnanediol-3a-gluc uronide ~9 µg/ml, and serum 13-hCG within the 95% confidence interval of normal values for a specific gestational age)
Sensitivity Specificity Predictive value Predictive value Efficiency
(TPITP + FN)
(TN/TN+ FP) ( + ) test (TP/TP + FP) ( - ) test (TN/TN+ FN) (TP +TN) Total
Serum /3-hCG (%)
Serum progesterone (%)
Urinary pregnanediol-3aglucuronide (%)
100 93 87
94 68 63
100 75 69
100
95
100
96
78
84
TP, True positive; TN, true negative; FP, false positive; FN, false negative.
drawn at the time of patient presentation and was allowed to clot for 1 hour. A 50 ml aliquot of urine was obtained concomitantly and tested for specific gravity. To avoid a dilutional effect, only subjects with urine specific gravities ~ 1.010 were included in the study. All samples were immediately stored at - 20° C until assayed. Radioimmunoassays . Direct radioimmunoassay (RIA) for 13-hCG and progesterone were performed with double antibody RIA kits (Radioassay Systems Laboratories, Carson, Calif.). Interassay and intrassay coefficients of variation were 7% and 4% for progesterone and 11 % and 10% for 13-hCG, respectively. Direct RIA of urinary pregnanediol-3a-gluc uronide was performed as previously described. 8 Briefly, the urine was diluted with 0.1 moll L phosphate-buffered saline solution, pH 7.4, and an aliquot of the diluted urine was subjected to RIA. Antipregnanediol3a-glucuronide-bovi ne serum albumin serum, which was obtained from Dr. Robert M. Nakamura, Eiken, Tokyo, Japan, was used in conjunction with tritiated pregnanediol-3a-gluc uronide (Courtauld Institute Steroid Biochemistry Group, University College, London, England). After an overnight incubation, antibody-bound and unbound pregnanediol-3a-gluc uronide were separated with dextran-coated charcoal. The assay sensitivity was 0.31 ng/ml, and the intraassay and interassay coefficients of variation were 3% and 11 %, respectively. Pregnanediol-3a-glu curonide enzyme immunoassay. Semiquantitative analysis of urinary pregnanediol-3a-glucuron ide was performed with the PROGESTurine pregnanediol-3a-gluc uronide assay kit (Monoclonal Antibodies, Inc., Mountain View, Calif.) with a sensitivity of 3 µg/ml. To investigate the assay's ability to accurately detect higher levels of hormone, serial dilutions from 1: 1 to 1: 5 volumes were performed on each specimen with the use of sterile
water. Enzyme immunoassay results were available in less than 10 minutes, and all runs were completed on the same day. Statistical analysis. Data were analyzed with the Mann-Whitney test. Confidence limits (95%) were established for 13-hCG, pregnanediol-3a-gluc uronide, and progesterone after log transformation of the data. Correlations were performed by means of univariate regression analysis. For calculating predictive values, we assigned 10 ng/ml of progesterone in serum and 9 µg/ml of pregnanediol-3a-gluc uronide in urine as indicative of a positive normal pregnancy value in accordance with normograms constructed from our control patients and consistent with previously published standards. 2 • 4 · 7 Levels below these values represented negative tests, suggestive of abnormal gestations. The normogram for 13-hCG obtained from control patients provided data for calculating the predictive value and test efficiency of this assay. Results
Table I lists serum 13-hCG, serum progesterone, and urinary pregnanediol-3a-gluc uronide RIA results for intrauterine and ectopic pregnancies. All hormone levels were significantly lower in ectopic gestations. Patients were not significantly different in gestational age; intrauterine 6.4 7 ± 0.18 versus ectopic 6.70 ± 0.67 weeks. Urine specific gravities, averaging l.018 ± 0.001, were also similar in both groups. Regression analysis demonstrated a strong positive correlation between RIA values for urinary pregnanediol-3a-gluc uronide and serum progesterone (r = 0.74, p = 0.0001). Urinary pregnanediol-3aglucuronide measured by RIA also correlated highly with enzyme immunoassay measurements (r = 0.95, p = 0.001). By means of enzyme immunoassay, 45 of 60 (75%) patients with ectopic pregnancies had levels <9 µg/ml of pregnanediol-3a-gluc uronide, and 54 of
1534
December 1988 Am J Obstet Gynecol
Sauer et al.
PDG (ug/ml)
50
~-----------------------------
------ - -
40 30 20
..
10
••
•*
0
8
7
6
5
GESTATIONAL AGE (WKS) Fig. 2. Urinary pregnanediol-3cx-glucuronide (PDG) values (n = 60) measured by RIA of ectopic pregnancies in relation to 95% confidence limits for values of normal control subjects.
HCG (mlU/ML)
100000 3 - - - - - - -
~ 1000011000 -
1
·:1 __~-----~------~-----~--5
6
7
8
GESTATIONAL AGE (WKS)
Fig. 3. Serum J'.1-hCG values (n = 60) measured by RIA of ectopic pregnancies in relation to 95% confidence limits for values of normal control subjects.
60 (90%) were undetectable at 15 µg/ml. In contrast, 100% of the intrauterine pregnancies produced at least 9 µg/ml of pregnanediol-3a-gluc uronide and 32 of 34 (91%) produced in excess of 15 µg/ml (Fig. I). Table II lists sensitivity, specificity, and the predictive values of serum 13-hCG, serum progesterone, and urinary pregnanediol-3a-gluc uronide assays in the detection of ectopic gestations. The assays were comparably efficient, although 13-hCG data could only be reliably interpreted by means of a normogram adjusted for gestational age. Fig. 2 depicts the normogram for pregnanediol-3a-gluc uronide, which demonstrates 55of 60 cases below the 95% confidence limit of normal pregnancy values. Similarly, Fig. 3 demonstrates that 56 of 60 ectopic pregnancies produced less 13-hCG than normal for gestational age. However, although the 95% confidence limits for normal pregnancy values of uri-
nary pregnanediol-3a-gluc uronide ranged between 10 and 40 µg/ml regardless of gestational age, serum 13hCG levels rose logarithmically with increasing gestational age.
Comment The corpus luteum is the principal site of progesterone synthesis through the initial 8 weeks of pregnancy.9 Throughout this period, progesterone levels remain relatively stable after a luteal phase rise. Therefore unlike 13-hCG, serial progesterone determinations are not required to monitor the progress of normal pregnancy and the precise knowledge of gestational age is unnecessary. Furthermore, the relatively short halflife of progesterone allows pathologic changes to be noted well in advance of alterations in 13-hCG production.'· 10
Pregnanediol in ectopic pregnancy
Volume 159 Number 6
Our results agree with other investigators' profiles of steroid production in abnormal pregnancies. Using more complex methods for measuring progesterone, Milwidsky et al. 11 and later Hertz et al.'° demonstrated that levels in spontaneous abortions and ectopic pregnancies were well below those of normal intrauterine gestations. More recently, Yeko et al. 3 and Matthews et al., 4 who used a rapid direct RIA, reported a discriminatory zone of 10 to 15 ng/ml for serum progesterone to differentiate normal from abnormal gestations. Although our mean values and 95% confidence limits agree with these recent reports, we noted a wider range of absolute values. This may be attributed to differences in assay technique or the very early gestational ages of pregnancies used as controls. Also, since many of the intrauterine pregnancies were electively terminated, we speculate that an occasional low value may represent intrauterine gestations destined to abort spontaneously. It has been previously reported that low levels of progesterone characterize abnormal intrauterine pregnancies in advance of such an event.'° The pregnanediol-3o:-glucuronide enzyme immunoassay has considerable diagnostic advantage over serum measurements of 13-hCG. First, it is not dependent on accurate gestational age determination for interpretation. Although when using a normogram a single serum 13-hCG level is highly efficient at discerning ectopic gestations, an absolute lower limit for normal pregnancies is not possible without adjusting for gestational age. Second, a single pregnanediol-3o:glucuronide measurement allows the detection of abnormality well in advance of the 36- to 48-hour doubling time required for 13-hCG. Therefore low levels of pregnanediol-3o:-glucuronide alert the physician to problem pregnancies within minutes of testing, even when 13-hCG levels are below the discriminatory zone of ultrasound examination. In this series 28 of60 (47%) ectopic pregnancies had 13-hCG < 1000 mIU /ml, the level reported to be useful as a discriminatory zone for ultrasonographic diagnosis of pregnancy with a 5.0 MHz vaginal probe. 12 Furthermore, the pregnanediol3o:-glucuronide assay has advantages over ultrasound examination, since it is very inexpensive and involves no maintenance or special training. The clinical use of the enzyme immunoassay is especially attractive. The method is convenient, nonradioactive, and avoids 24-hour urine collections. These characteristics allow office or emergency room use without special equipment or trained personnel. Enzyme immunoassay is noninvasive and rapid, with results obtained in less than 10 minutes. In nondilute specimens,
1535
simultaneous creatinine measurements are unnecessary, since values maintain a close correlation with serum progesterone levels.2· 13 We conclude that a single, random, urinary pregnanediol-3o:-glucuronide measurement may differentiate a normal from an ectopic pregnancy in most cases. Although slightly less efficient than serum 13hCG, pregnanediol-3o:-glucuronide enzyme immunoassay is more versatile, does not require precise knowledge of gestational age for interpretation, and provides results in minutes. Because of the simplicity and efficiency of pregnanediol-3o:-glucuronide enzyme immunoassay, we recommend its use in early pregnancy surveillance. REFERENCES I. Acevedo HF, Vela BA, Campbell EA, et al. Urinary steroid
2.
3.
4. 5. 6.
7.
8.
9.
10.
11. 12. 13.
profile in threatened abortion. AM J OBSTET GYNECOL 1969; I 04:964-72. Mendizabal AF, Quiroga S, Farinati Z, Lahoz M, Nagle C. Hormonal monitoring of early pregnancy by a direct radioimmunoassay of steroid glucuronides in first morning urine. Fertil Steril l 984;42:737-40. Yeko TR, Gorrill MJ, Hughes LH, Rodi IA, Buster JE, Sauer MV. Timely diagnosis of early ectopic pregnancy using a single blood progesterone measurement. Fertil Steril 1987;48:1048-50. Matthews CP, Coulson PB, Wild RA. Serum progesterone levels as an aid in the diagnosis of ectopic pregnancy. Obstet Gynecol ! 986;68:390-4. Haynes SP, Corcoran JM, Eastman CJ, Doy FA. Radioimmunoassay of progesterone in unextracted serum. Cl in Chem 1980;26:1067-9. Ratcliffe WA, Corrie JET, Dalziel AH, Macpherson JS. Direct 125 1-radioligand assays for serum progesterone compared with assays involving extraction of serum. Clin Chem !982;28:1314-8. Davidson BJ. Urinary pregnanediol-3-glucuronide and estrone conjugates to creatinine ratios in early pregnancies complicated by vaginal bleeding. Fertil Steril 1986; 46:610-14. Stanczyk FZ, Miyakawa I, Goebelsmann U. Direct radioimmunoassay of urinary estrogen and pregnanediol glucuronides during the menstrual cycle. AM J OBSTET GvNECOL 1980; 117:443-50. Tulchinsky D, Hobel CJ. Plasma human chorionic gonadotropin, estrone, estradiol, progesterone, and 17 ~-hydroxyprogesterone in human pregnancy. AM J OBSTET GYNECOL 1973; 177:884-93. Hertz JB, Larsen JF, Arends J, Nielsen J. Progesterone and human chorionic gonadotropin in serum and pregnanediol in urine in threatened abortion. Acta Obstet Gynecol Scand l 980;59:23-7. Milwidsky A, Adoni A, Segal S, Palti Z. Chorionic gonadotropin and progesterone levels in ectopic pregnancy. Obstet Gynecol 1977;50:145-7. Timor-Tritsch IE, Rottem S. High frequency transvaginal sonography: new diagnostic boon. Contemp Obstet Gynecol 1988;31: 111-3. Chatterton RT, Haan JN, Jenco JM, Cheesman KL. Radioimmunoassay of pregnanediol concentrations in early morning urine specimens for assessment of luteal function in women. Fertil Steril !982;37:361-6.
Volume 159 Number 6
3. Kivinen S, Kuoppala T, Leppilampi M, VuoriJ, Kauppila A. Tumor-associated antigen CA125 before and during the treatment of ovarian carcinoma. Obstet Gynecol 1986;67:468-72. 4. Barbieri RL, Niloff JM, Bast RC Jr, Schaetzl E, Kistner RW, Knapp RC. Elevated serum concentration of CAl25 in patients with advanced endometriosis. Fertil Steril I 986;45:630-4. 5. Guidice LC, Jacobs A, Pineda J, Bell CE, Lippman L. Serum levels of CA 125 in patients with endometriosis: a preliminary report. Fertil Steril I 986;45:876-8. 6. Patton EP, Field CS, Harms RW, Coulam CB. CA125 levels in endometriosis. Fertil Steril 1986;45: 770-3. 7. Pittaway DE, Fayez JA. The use of CAl25 in the diagnosis and management of endometriosis. Fertil Steril I 986;46:790-5. 8. Kabawat SE, Bast RC Jr, Bhan AK, Welch WR, Knapp RC, Colvin RB. Tissue distribution of a coelomicepithelium-related antigen recognized by the monoclonal antibody OC 125. Int J Gynecol Pathol I 983;2:275-85. 9. Pittaway DE, Fayez JA. Serum CA-125 antigen levels increased during menses. AM J OBSTET GYNECOL 1987; 156:75-6. 10. Kauppila A, Telimaa S, Ronnberg L, Vuori J. Placebo-
11. 12.
13.
14. 15.
controlled study on serum concentrations of CA 125 before and after treatment of endometriosis with danazol or high-dose medroxyprogesterone acetate alone or after surgery. Fertil Steril 1988;49:37-41. Barbieri RL, Canick JA, Makris A, Todd RB, Davies IJ, Ryan KJ. Danazol inhibits steroidogenesis. Fertil Steril 1977;28:809-13. Menon M, Azhar S, Menon KMJ. Evidence that danazol inhibits gonadotropin-induced ovarian steroidogenesis at a point distal to gonadotropin-receptor interaction and adenosine 3' ,5' cyclic monophosphate formation. AM .J 0BSTET GYNECOL 1980; 136:524-30. Musich JR, Behrman SJ, Menon KMJ. Estrogenic and antiestrogenic effects of danazol administration in studies of estradiol receptor binding. AM J OBSTET Gvr-
Rapid measurement of urinary pregnanediol glucuronide to diagnose ectopic pregnancy Mark V. Sauer, MD, Michael Vermesh, MD, Robert E. Anderson, MD, Ariel G. Vijod, BS, Frank Z. Stanczyk, PhD, and Rogerio A. Lobo, MD Los Angeles, California We investigated the ability of a single, random, urinary pregnanediol-3a-glucuronide level to differentiate early intrauterine from ectopic pregnancy. Thirty-four patients with intrauterine gestations were compared with 60 patients with ectopic pregnancies. Urinary pregnanediol-3a-glucuronide was measured by radioimmunoassay and enzyme immunoassay. Compared with intrauterine gestations, results demonstrate that urinary pregnanediol-3a-glucuronide is significantly depressed in ectopic pregnancies: 24.5 ± 2.2 versus 4.8 ± 0.7 µg/ml (p = 0.0001 ). Urinary pregnanediol-3a-glucuronide levels obtained by conventional radioimmunoassay correlated closely with values measured in minutes with enzyme immunoassay (r = 0.95, p = 0.0001), and with serum progesterone (r = 0.74, p = 0.0001). Urinary pregnanediol-3a-glucuronide measured by enzyme immunoassay exhibitea predictive values for detecting ectopic gestations comparable with random serum progesterone or serum 13-human chorionic gonadotropin values. We conclude that ectopic gestations demonstrate a reduced level of urinary pregnanediol-3a-glucuronide (55/60 cases) detectable with a rapid enzyme immunoassay, which makes this assay a practical screening test in early pregnancy. (AM J 0BSTET GYNECOL 1988;159:1531-5.)
Key words: Pregnanediol-3a-glucuronide, progesterone, ectopic pregnancy, enzyme immunoassay
From the Division of Reproductive Endocrinology, University of Southern California. Presented in part at the Thirty-fifth Annual Meeting of the Society For Gynecologic Investigation, Baltimore, Maryland, March 1720, 1988. Reprint requests: Mark V. Sauer, MD, Women's Hospital, 1240 North Mission Rd., Room L-946, Los Angeles, CA 90033.
Concentrations of serum progesterone and its urinary metabolite, pregnanediol-3a-glucuronide, are reported to be lower in abnormal pregnancies than in normal intrauterine gestations. 1• 2 However, these assays are laborious and therefore not widely used. Thus hormonal monitoring of early pregnancy continues to con-
1531
1532
Sauer et al.
December 1988 Am J Obstet Gynecol
% POSITIVE TESTS
100 80
60 40
20 o__IL_~~~L_~--'~~-.-~-L--~~~L-------,~~~,-~~
9
3
12
15
18
PDG (ug/ml) POSITIVE TEST •
>
ECTOPIC
-
INTRAUTERINE
SPECIFIED LEVEL OF PDG
Fig. I. Urinary pregnanediol-3u-glucur onide (PDG) values measured by enzyme immunoassay for both intrauterine (n = 34) and ectopic gestations (n = 60).
Table I. Hormonal profile of patients with intrauterine (n = 34) and ectopic (n = 60) gestations at time of presentation Intrauterine pregnancy
13-hCG (mIU/ml) Mean± SEM Range Serum progesterone (ng/ml) Mean± SEM Range Urinary PDG (µg/ml) Mean± SEM Range
I
Ectopic pregnancy
18,624 ± 2544 196 - 43,796
2,726 ± 810* 17 - 44,280
28.0 ± 2.2 7.1 ± 53.9
8.3 ± 1.1* 0.2 ± 42.2
24.5 ± 2.2 7.9 ± 73.4
4.8 ± 0.7* 0.2 ± 25.2
*P.;; 0.0001.
sist of quantitative measurement of 13-human chorionic gonadotropin (13-hCG). In most cases, serial serum sampling over several days is required to discriminate abnormal from normal gestations, which often delays definitive diagnosis.' Recently the clinical use of measuring serum progesterone by rapid, direct radioimmunoassay (RIA) to differentiate normal from abnormal pregnancy has been reported.'· 4 The new double antibody assays are reported to be more precise and technically simpler than the serum extraction plus column-type methods, with results obtained in less than 4 hours. 5 • 6 Similarly, urinary pregnanediol-3cx-glu curonide can now be measured in less than 2 hours with the enzyme-multiplied immunoassay technique and· has predicted early preg-
nancy outcomes on the basis of a single random sample. 7 A pregnanediol-3cx-glu curonide enzyme immunoassay, which is available commercially, measures pregnanediol-3cx-glu curonide in less than 10 minutes. Unlike the previously mentioned assays, this assay requires no special equipment or technical training and can be readily run in the office or emergency room. Therefore the purpose of this study was to determine if a single, random, urinary pregnanediol-3cxglucuronide value measured by enzyme immunoassay could reliably differentiate an early intrauterine pregnancy from an ectopic gestation. Furthermore, we compared the predictive value of urinary pregnanediol-3cxglucuronide against that of a single serum progesterone or serum 13-hCG in detecting ectopic pregnancies.
Material and methods Subjects. Between Aug. l, l 987, and Feb. l, l 988, we prospectively studied 60 women who came to the Los Angeles County-University of Southern California Medical Center with suspected ectopic pregnancies. Patients were 5 to 8 weeks from the last menstrual period and hemodynamically stable. Ectopic gestations were later confirmed at surgery. During this time period, 34 women with known intrauterine pregnancies of 5 to 8 weeks' gestation, who attended the family planning clinic, served as normal pregnant control subjects. Intrauterine gestations were confirmed either by serial ultrasound examinations or histologic evidence of products of conception obtained after elective termination of pregnancy. Serum and urine samples. Antecubital blood was
Pregnanediol in ectopic pregnancy
Volume 159 Number 6
1533
Table II. Applicability of a single random test for detecting ectopic gestations based on normal pregnancies representing positive pregnancy tests (serum progesterone ~ 10 ng I ml, urinary pregnanediol-3a-gluc uronide ~9 µg/ml, and serum 13-hCG within the 95% confidence interval of normal values for a specific gestational age)
Sensitivity Specificity Predictive value Predictive value Efficiency
(TPITP + FN)
(TN/TN+ FP) ( + ) test (TP/TP + FP) ( - ) test (TN/TN+ FN) (TP +TN) Total
Serum /3-hCG (%)
Serum progesterone (%)
Urinary pregnanediol-3aglucuronide (%)
100 93 87
94 68 63
100 75 69
100
95
100
96
78
84
TP, True positive; TN, true negative; FP, false positive; FN, false negative.
drawn at the time of patient presentation and was allowed to clot for 1 hour. A 50 ml aliquot of urine was obtained concomitantly and tested for specific gravity. To avoid a dilutional effect, only subjects with urine specific gravities ~ 1.010 were included in the study. All samples were immediately stored at - 20° C until assayed. Radioimmunoassays . Direct radioimmunoassay (RIA) for 13-hCG and progesterone were performed with double antibody RIA kits (Radioassay Systems Laboratories, Carson, Calif.). Interassay and intrassay coefficients of variation were 7% and 4% for progesterone and 11 % and 10% for 13-hCG, respectively. Direct RIA of urinary pregnanediol-3a-gluc uronide was performed as previously described. 8 Briefly, the urine was diluted with 0.1 moll L phosphate-buffered saline solution, pH 7.4, and an aliquot of the diluted urine was subjected to RIA. Antipregnanediol3a-glucuronide-bovi ne serum albumin serum, which was obtained from Dr. Robert M. Nakamura, Eiken, Tokyo, Japan, was used in conjunction with tritiated pregnanediol-3a-gluc uronide (Courtauld Institute Steroid Biochemistry Group, University College, London, England). After an overnight incubation, antibody-bound and unbound pregnanediol-3a-gluc uronide were separated with dextran-coated charcoal. The assay sensitivity was 0.31 ng/ml, and the intraassay and interassay coefficients of variation were 3% and 11 %, respectively. Pregnanediol-3a-glu curonide enzyme immunoassay. Semiquantitative analysis of urinary pregnanediol-3a-glucuron ide was performed with the PROGESTurine pregnanediol-3a-gluc uronide assay kit (Monoclonal Antibodies, Inc., Mountain View, Calif.) with a sensitivity of 3 µg/ml. To investigate the assay's ability to accurately detect higher levels of hormone, serial dilutions from 1: 1 to 1: 5 volumes were performed on each specimen with the use of sterile
water. Enzyme immunoassay results were available in less than 10 minutes, and all runs were completed on the same day. Statistical analysis. Data were analyzed with the Mann-Whitney test. Confidence limits (95%) were established for 13-hCG, pregnanediol-3a-gluc uronide, and progesterone after log transformation of the data. Correlations were performed by means of univariate regression analysis. For calculating predictive values, we assigned 10 ng/ml of progesterone in serum and 9 µg/ml of pregnanediol-3a-gluc uronide in urine as indicative of a positive normal pregnancy value in accordance with normograms constructed from our control patients and consistent with previously published standards. 2 • 4 · 7 Levels below these values represented negative tests, suggestive of abnormal gestations. The normogram for 13-hCG obtained from control patients provided data for calculating the predictive value and test efficiency of this assay. Results
Table I lists serum 13-hCG, serum progesterone, and urinary pregnanediol-3a-gluc uronide RIA results for intrauterine and ectopic pregnancies. All hormone levels were significantly lower in ectopic gestations. Patients were not significantly different in gestational age; intrauterine 6.4 7 ± 0.18 versus ectopic 6.70 ± 0.67 weeks. Urine specific gravities, averaging l.018 ± 0.001, were also similar in both groups. Regression analysis demonstrated a strong positive correlation between RIA values for urinary pregnanediol-3a-gluc uronide and serum progesterone (r = 0.74, p = 0.0001). Urinary pregnanediol-3aglucuronide measured by RIA also correlated highly with enzyme immunoassay measurements (r = 0.95, p = 0.001). By means of enzyme immunoassay, 45 of 60 (75%) patients with ectopic pregnancies had levels <9 µg/ml of pregnanediol-3a-gluc uronide, and 54 of
1534
December 1988 Am J Obstet Gynecol
Sauer et al.
PDG (ug/ml)
50
~-----------------------------
------ - -
40 30 20
..
10
••
•*
0
8
7
6
5
GESTATIONAL AGE (WKS) Fig. 2. Urinary pregnanediol-3cx-glucuronide (PDG) values (n = 60) measured by RIA of ectopic pregnancies in relation to 95% confidence limits for values of normal control subjects.
HCG (mlU/ML)
100000 3 - - - - - - -
~ 1000011000 -
1
·:1 __~-----~------~-----~--5
6
7
8
GESTATIONAL AGE (WKS)
Fig. 3. Serum J'.1-hCG values (n = 60) measured by RIA of ectopic pregnancies in relation to 95% confidence limits for values of normal control subjects.
60 (90%) were undetectable at 15 µg/ml. In contrast, 100% of the intrauterine pregnancies produced at least 9 µg/ml of pregnanediol-3a-gluc uronide and 32 of 34 (91%) produced in excess of 15 µg/ml (Fig. I). Table II lists sensitivity, specificity, and the predictive values of serum 13-hCG, serum progesterone, and urinary pregnanediol-3a-gluc uronide assays in the detection of ectopic gestations. The assays were comparably efficient, although 13-hCG data could only be reliably interpreted by means of a normogram adjusted for gestational age. Fig. 2 depicts the normogram for pregnanediol-3a-gluc uronide, which demonstrates 55of 60 cases below the 95% confidence limit of normal pregnancy values. Similarly, Fig. 3 demonstrates that 56 of 60 ectopic pregnancies produced less 13-hCG than normal for gestational age. However, although the 95% confidence limits for normal pregnancy values of uri-
nary pregnanediol-3a-gluc uronide ranged between 10 and 40 µg/ml regardless of gestational age, serum 13hCG levels rose logarithmically with increasing gestational age.
Comment The corpus luteum is the principal site of progesterone synthesis through the initial 8 weeks of pregnancy.9 Throughout this period, progesterone levels remain relatively stable after a luteal phase rise. Therefore unlike 13-hCG, serial progesterone determinations are not required to monitor the progress of normal pregnancy and the precise knowledge of gestational age is unnecessary. Furthermore, the relatively short halflife of progesterone allows pathologic changes to be noted well in advance of alterations in 13-hCG production.'· 10
Pregnanediol in ectopic pregnancy
Volume 159 Number 6
Our results agree with other investigators' profiles of steroid production in abnormal pregnancies. Using more complex methods for measuring progesterone, Milwidsky et al. 11 and later Hertz et al.'° demonstrated that levels in spontaneous abortions and ectopic pregnancies were well below those of normal intrauterine gestations. More recently, Yeko et al. 3 and Matthews et al., 4 who used a rapid direct RIA, reported a discriminatory zone of 10 to 15 ng/ml for serum progesterone to differentiate normal from abnormal gestations. Although our mean values and 95% confidence limits agree with these recent reports, we noted a wider range of absolute values. This may be attributed to differences in assay technique or the very early gestational ages of pregnancies used as controls. Also, since many of the intrauterine pregnancies were electively terminated, we speculate that an occasional low value may represent intrauterine gestations destined to abort spontaneously. It has been previously reported that low levels of progesterone characterize abnormal intrauterine pregnancies in advance of such an event.'° The pregnanediol-3o:-glucuronide enzyme immunoassay has considerable diagnostic advantage over serum measurements of 13-hCG. First, it is not dependent on accurate gestational age determination for interpretation. Although when using a normogram a single serum 13-hCG level is highly efficient at discerning ectopic gestations, an absolute lower limit for normal pregnancies is not possible without adjusting for gestational age. Second, a single pregnanediol-3o:glucuronide measurement allows the detection of abnormality well in advance of the 36- to 48-hour doubling time required for 13-hCG. Therefore low levels of pregnanediol-3o:-glucuronide alert the physician to problem pregnancies within minutes of testing, even when 13-hCG levels are below the discriminatory zone of ultrasound examination. In this series 28 of60 (47%) ectopic pregnancies had 13-hCG < 1000 mIU /ml, the level reported to be useful as a discriminatory zone for ultrasonographic diagnosis of pregnancy with a 5.0 MHz vaginal probe. 12 Furthermore, the pregnanediol3o:-glucuronide assay has advantages over ultrasound examination, since it is very inexpensive and involves no maintenance or special training. The clinical use of the enzyme immunoassay is especially attractive. The method is convenient, nonradioactive, and avoids 24-hour urine collections. These characteristics allow office or emergency room use without special equipment or trained personnel. Enzyme immunoassay is noninvasive and rapid, with results obtained in less than 10 minutes. In nondilute specimens,
1535
simultaneous creatinine measurements are unnecessary, since values maintain a close correlation with serum progesterone levels.2· 13 We conclude that a single, random, urinary pregnanediol-3o:-glucuronide measurement may differentiate a normal from an ectopic pregnancy in most cases. Although slightly less efficient than serum 13hCG, pregnanediol-3o:-glucuronide enzyme immunoassay is more versatile, does not require precise knowledge of gestational age for interpretation, and provides results in minutes. Because of the simplicity and efficiency of pregnanediol-3o:-glucuronide enzyme immunoassay, we recommend its use in early pregnancy surveillance. REFERENCES I. Acevedo HF, Vela BA, Campbell EA, et al. Urinary steroid
2.
3.
4. 5. 6.
7.
8.
9.
10.
11. 12. 13.
profile in threatened abortion. AM J OBSTET GYNECOL 1969; I 04:964-72. Mendizabal AF, Quiroga S, Farinati Z, Lahoz M, Nagle C. Hormonal monitoring of early pregnancy by a direct radioimmunoassay of steroid glucuronides in first morning urine. Fertil Steril l 984;42:737-40. Yeko TR, Gorrill MJ, Hughes LH, Rodi IA, Buster JE, Sauer MV. Timely diagnosis of early ectopic pregnancy using a single blood progesterone measurement. Fertil Steril 1987;48:1048-50. Matthews CP, Coulson PB, Wild RA. Serum progesterone levels as an aid in the diagnosis of ectopic pregnancy. Obstet Gynecol ! 986;68:390-4. Haynes SP, Corcoran JM, Eastman CJ, Doy FA. Radioimmunoassay of progesterone in unextracted serum. Cl in Chem 1980;26:1067-9. Ratcliffe WA, Corrie JET, Dalziel AH, Macpherson JS. Direct 125 1-radioligand assays for serum progesterone compared with assays involving extraction of serum. Clin Chem !982;28:1314-8. Davidson BJ. Urinary pregnanediol-3-glucuronide and estrone conjugates to creatinine ratios in early pregnancies complicated by vaginal bleeding. Fertil Steril 1986; 46:610-14. Stanczyk FZ, Miyakawa I, Goebelsmann U. Direct radioimmunoassay of urinary estrogen and pregnanediol glucuronides during the menstrual cycle. AM J OBSTET GvNECOL 1980; 117:443-50. Tulchinsky D, Hobel CJ. Plasma human chorionic gonadotropin, estrone, estradiol, progesterone, and 17 ~-hydroxyprogesterone in human pregnancy. AM J OBSTET GYNECOL 1973; 177:884-93. Hertz JB, Larsen JF, Arends J, Nielsen J. Progesterone and human chorionic gonadotropin in serum and pregnanediol in urine in threatened abortion. Acta Obstet Gynecol Scand l 980;59:23-7. Milwidsky A, Adoni A, Segal S, Palti Z. Chorionic gonadotropin and progesterone levels in ectopic pregnancy. Obstet Gynecol 1977;50:145-7. Timor-Tritsch IE, Rottem S. High frequency transvaginal sonography: new diagnostic boon. Contemp Obstet Gynecol 1988;31: 111-3. Chatterton RT, Haan JN, Jenco JM, Cheesman KL. Radioimmunoassay of pregnanediol concentrations in early morning urine specimens for assessment of luteal function in women. Fertil Steril !982;37:361-6.