Reevaluating the role of dilation and curettage in the diagnosis of pregnancy of unknown location

Reevaluating the role of dilation and curettage in the diagnosis of pregnancy of unknown location

Reevaluating the role of dilation and curettage in the diagnosis of pregnancy of unknown location Karine Chung, M.D., M.S.C.E.,a,b Uma Chandavarkar, M...

77KB Sizes 0 Downloads 8 Views

Reevaluating the role of dilation and curettage in the diagnosis of pregnancy of unknown location Karine Chung, M.D., M.S.C.E.,a,b Uma Chandavarkar, M.D.,b Neisha Opper, M.P.H.,b and Kurt Barnhart, M.D., M.S.C.E.c a

Division of Reproductive Endocrinology and Infertility, b Department of Obstetrics and Gynecology, Los Angeles County– University of Southern California Medical Center, Los Angeles, California; and c Division of Reproductive Endocrinology and Infertility, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania

Objective: To evaluate the clinical utility of dilation and curettage (D&C) in diagnosing ectopic pregnancy (EP). Design: Retrospective cohort study. Setting: University hospital. Patient(s): Clinically stable women (n ¼ 321) who underwent a diagnostic D&C with no visible intrauterine pregnancy (IUP) on transvaginal ultrasound or those with an abnormal hCG trend. Intervention(s): None. Main Outcome Measure(s): EP or IUP made by final pathologic review. Result(s): Overall, 73.2% of the patients were ultimately diagnosed with EP and 26.8% were found to have a nonviable IUP. Those with EPs had significantly lower initial hCGs than those with nonviable IUPs and were more likely to have had a history of an EP. On ultrasound, the overall impression, the presence of free fluid, and the endometrial echo complex correlated well with the final diagnoses but did not have 100% predictive value. Conclusion(s): D&C remains valuable to differentiate EP from nonviable IUP and to avoid misdiagnosis and unnecessary exposure to methotrexate. Low initial hCG values and ultrasound findings such as a thin endometrial echo complex and the presence of free fluid are associated with but are not diagnostic of an ectopic pregnancy. (Fertil Steril 2011;96:659–62. 2011 by American Society for Reproductive Medicine.) Key Words: Ectopic pregnancy, dilation and curettage, methotrexate

Ectopic pregnancy (EP) remains a leading cause of maternal death (1). Diagnostic tools, including transvaginal ultrasonography (TVUS) (2) and b-hCG curves, have been refined over the past decade to optimize early detection and reduce mortality (3–5). EP is suspected when the hCG level is >2,000 mIU/mL and no intrauterine gestation is seen or when abnormal patterns of hCG are observed. In these scenarios, definitive diagnosis can be achieved by performing dilation and curettage (D&C) (6). Clinicians often choose to abandon D&C in the diagnostic evaluation and administer methotrexate to all those suspected to have EP. This strategy avoids a surgical procedure with associated risks. However, failure to ascertain accurate diagnosis may have implications for counseling about future risk of EP and may lead to unnecessary exposure to methotrexate, a chemotherapeutic agent. The utility of D&C in the diagnosis of EP was previously studied, revealing that the presumed diagnosis of EP is incorrect in 40% of cases (6). Since then, ultrasound resolution has improved and hCG curves have been refined, possibly allowing for more accurate diagnosis of EP (3–5). The primary objective of the present study was to reevaluate the clinical utility of D&C in diagnosing EP.

MATERIALS AND METHODS This was a retrospective cohort analysis of all patients undergoing D&C as part of evaluation for EP at Los Angeles County–University of Southern Received April 18, 2010; revised May 2, 2011; accepted June 6, 2011; published online July 23, 2011. K.C. has nothing to disclose. U.C. has nothing to disclose. N.O. has nothing to disclose. K.B. has nothing to disclose. Supported by National Institutes of Health grant R01-HD036455 (K.B.) with subcontract to K.C. Reprint requests: Karine Chung, M.D., M.S.C.E., 2020 Zonal Avenue, IRD Room 532, Los Angeles, CA 90033 (E-mail: [email protected]).

0015-0282/$36.00 doi:10.1016/j.fertnstert.2011.06.017

California (USC) Medical Center between January 1, 2004, and December 31, 2007. No patients in this analysis were presumptively treated with methotrexate before pathologic review of their uterine curettings. This study was approved by the USC Institutional Review Board. During the study period, D&C was routinely performed on all women considered to be ‘‘at risk’’ for EP: 1) no visible intrauterine pregnancy (IUP) with hCG >2,000 mIU/mL; 2) abnormal rise in hCG level, defined as <50% increase in 2 days (5); or 3) abnormal fall in hCG level, defined as <20% decline in 2 days (3). Potential subjects were identified by review of a pathology database and medical records. All ultrasounds were performed at the bedside by gynecology residents, which is standard at this institution. D&Cs were performed under conscious sedation. Final diagnosis was based on the presence or absence of chorionic villi on permanent histology of the uterine curettings. When salpingostomy or salpingectomy was performed, pathologic analysis of the removed tissue confirmed the diagnosis of EP. Data collected included age, race, days of amenorrhea, pregnancy history, initial hCG levels, hCG trends, time to diagnosis, ultrasound impressions, surgical pathology, and method of treatment. If clinical suspicion of EP warranted immediate intervention, only a single hCG value was available. Ultrasound impressions were designated as ‘‘suspicious for EP,’’ in the presence of free fluid, thin endometrial echo complex (EEC), or mass without gestational sac, or ‘‘probable IUP,’’ when an intrauterine sac was seen without a yolk sac or fetal pole. In comparing the two outcome groups, statistical significance was determined using chi-square or Fisher exact test for categoric variables and independent t test or Wilcoxon rank sum test, as appropriate, for continuous variables. A P value of < .05 was considered to be statistically significant. All data were analyzed with the use of SAS 9.2 (SAS Institute).

RESULTS Three hundred eighty-seven patients at risk for EP underwent a diagnostic D&C over the study period. Sixty-three patients’ records were unavailable for review. Three patients had molar gestations and were

Fertility and Sterility Vol. 96, No. 3, September 2011 Copyright ª2011 American Society for Reproductive Medicine, Published by Elsevier Inc.

659

TABLE 1 Comparison of clinical and demographic factors between patients diagnosed with ectopic pregnancy (EP) and those with nonviable intrauterine pregnancy (IUP).

Age (y) Amenorrhea (d) hCG (mIU/mL) Gravidity Parity History of EP Race Latin American African American European Asian Other Serial hCG trendc Rise Fall Single value Plateau Time to diagnosis Immediate Delayed

All (n [ 321)

EP (n [ 235)

Nonviable IUP (n [ 86)

P value

31 (15–47) 49 (6–380) 998 (11–48,110) 3 (1–12) 1 (0–8) 36 (11.2)

32 (15–45) 49 (6–380) 731 (11–39,836) 3 (1–12) 1 (1–8) 33 (14.0)

31 (17–47) 51.5 (18–184) 2806.5 (47–48,810) 3 (1–10) 1 (0–7) 3 (3.5)

.527a .165a < .001a .650a .230a .008b .289b

250 (77.9) 26 (8.1) 16 (5.0) 23 (7.2) 6 (1.9)

181 (77.0) 19 (8.1) 10 (4.3) 21 (8.9) 4 (1.7)

69 (80.2) 7 (8.1) 6 (7.0) 2 (2.3) 2 (2.3)

56 (23.8) 23 (9.8) 126 (53.6) 30 (12.8)

16 (18.6) 7 (8.1) 53 (61.6) 10 (11.6)

120 (51.1) 115 (48.9)

49 (57) 37 (43)

.628b

.347b

Note: Values are presented as median (range) or n (%). a Wilcoxon rank sum statistical test. b Chi-square analysis or Fisher exact test. c Patterns of serial hCG levels were defined as a rise if the hCG increased by >50% in 2 days (5), a fall if it decreased by >20% in 2 days (3), and a plateau if it failed to rise or fall. Chung. D&C in ectopic pregnancy. Fertil Steril 2011.

The association of initial ultrasound impression and final outcome is presented in Table 3. In the entire cohort, 10.6% had an ultrasound impression of ‘‘probable IUP,’’ 28.7% were ‘‘suspicious for EP,’’ and 60.7% were ‘‘nondiagnostic.’’ The number of women diagnosed with EP differed based on the ultrasound impression (P<.001). As expected, the odds of diagnosis of EP were significantly decreased when the ultrasound impression was ‘‘probable IUP’’ and significantly increased when the impression was ‘‘suspicious for EP.’’ Specific characteristics of the ultrasound examinations were found to be associated with final outcome (Table 3). Women in this study whose ultrasounds showed free fluid in the cul-de-sac had a greater chance of having EP (odds ratio 3.78, 95% confidence interval 1.80–7.95). In addition, median preoperative EEC was significantly lower in the EP group compared with the nonviable

excluded from the analysis. We chose to exclude molar pregnancies from this study, because our aim was to focus on the use of ultrasound in diagnosis of EP. Table 1 compares clinical factors between those diagnosed with EP and those with nonviable IUP. There was a statistically significant difference in hCG at initial evaluation and in the proportion of women with a history of EP between the two groups. There was no difference in initial hCG value or days of amenorrhea between the two groups. Of the 321 women, 235 (73.2%) had a final diagnosis of an EP and 86 (26.8%) a final diagnosis of nonviable IUP (Table 2). The odds of EP were 3.82 times higher when the initial hCG was <2,000 mIU/mL (69.4% EP) with an abnormal hCG trend than when the initial hCG was >2,000 mIU/mL (30.6% EP). When salpingostomy or salpingectomy was performed, pathologic analysis confirmed the diagnosis of EP in all cases.

TABLE 2 Final diagnosis of clinically stable women with a presumed EP.

All patients hCG <2,000 mIU/mL hCG >2,000 mIU/mL

EP (n [ 235)

Nonviable IUP (n [ 86)

OR (95% CI) for EP

P value

235 (73.2) 163 (69.4) 72 (30.6)

86 (26.8) 32 (37.2) 54 (62.8)

3.82 (2.28–6.41)

< .001a

Note: Values are presented as n (%). CI ¼ confidence interval; OR ¼ odds ratio; other abbreviations as in Table 1. a Chi-square analysis. Chung. D&C in ectopic pregnancy. Fertil Steril 2011.

660

Chung et al.

D&C in ectopic pregnancy

Vol. 96, No. 3, September 2011

TABLE 3 Comparison of initial ultrasound evaluations between patients ultimately diagnosed with EP and nonviable IUP.

Ultrasound impression Nondiagnostic (n ¼ 195) Probable IUP (n ¼ 34) Suspicious for EP (n ¼ 92) Presence of free fluid EEC (mm) hCG %2,000 mIU/mL EEC %5 mm EEC >5 mm hCG >2,000 mIU/mL EEC %5 mm

EP (n [ 235)

Nonviable IUP (n [ 86)

147 (75.4) 3 (8.8) 85 (92.4) 9 (11.1) 8.35 (2.6–20)

48 (24.6) 31 (91.2) 7 (7.6) 72 (88.9) 7.20 (2–26)

2 (5) 28 (18.1)

38 (95) 127 (81.9)

4.19 (0.95–18.39)

.041a

11 (39.3)

17 (60.7)

1.21 (0.51–2.85)

.665a

OR (95% CI) for EP

P value < .001a

0.03 (0.009–0.108) 3.97 (1.72–9.15) 3.78 (1.80–7.95)

< .001a .03b

Note: Values are presented as median (range) or n (%). EEC ¼ endometrial echo complex; other abbreviations as in Tables 1 and 2. a Chi-square analysis or Fisher exact test. b Wilcoxon rank sum test. Chung. D&C in ectopic pregnancy. Fertil Steril 2011.

IUP groups (7.20 mm vs. 8.35 mm; P<.03). Table 3 also demonstrates the association of EEC with final outcome, stratified by initial hCG. EEC %5 mm was statistically significantly associated with EP (P¼.041) when the initial hCG was below but not when it was above 2,000 mIU/mL.

DISCUSSION Despite improvements in diagnostic tools, this study demonstrates that the presumptive diagnosis of EP in clinically stable women with a pregnancy of an unknown location is inaccurate in nearly 27% of cases. This false diagnosis rate is substantial and would entail treating a significant proportion of patients with methotrexate unnecessarily. This is especially notable when the initial hCG is above the discriminatory zone and the ‘‘true positive’’ rate is just over 57%. Therefore, when hCG levels are >2,000 mIU/mL, the use of D&C to obtain the final diagnosis is particularly important. Below the discriminatory zone, the ‘‘true positive’’ rate improves to 84%, but this remains unacceptably low. Our data suggest that ultrasound can be a powerful adjunct to hCG. When a clinician commits to a diagnosis of EP based on ultrasound findings other than an extrauterine gestational sac, the diagnosis is correct 92.4% of the time. The EEC is statistically associated with diagnosis, and the ‘‘thinnest’’ EEC (2 mm) was found in a patient ultimately diagnosed with EP. However, we were unable to detect a cutoff point for EEC that would be clinically meaningful, because the ‘‘next-to-thinnest’’ EEC was found in a patient ultimately diagnosed with IUP. Therefore, EEC should not be used as a diagnostic test and D&C continues to be important in achieving a definitive diagnosis. Some authors prefer to treat all patients presumed to have EP with methotrexate, bypassing D&C. They advocate that it may save time and expense and avoids an invasive procedure that is not 100% specific; false negatives can occur with D&C (7). Because all subjects in the present study were treated for EP in the absence of chorionic villi on final pathology, we were not able to determine the false negative

Fertility and Sterility

rate of D&C. One study reported that 20% of uterine curettage specimens from elective terminations of pregnancies fail to demonstrate chorionic villi on histopathology (8). Relatively little is known about the cost-effectiveness of empiric methotrexate treatment versus proceeding with a D&C first. One study found that empirically treating women at risk for EP with methotrexate does not reduce complications or significantly save on cost (9). That cost analysis reported that diagnostic D&C cost $173 to $223 more per patient but had nearly 14% fewer complications with about 6% fewer hospitalizations. Although rare, methotrexate administration may result in stomatitis and conjunctivitis, and very rarely bone marrow suppression. Approximately 30% of patients who receive single-dose therapy and 40% of those who receive multidose therapy will experience some type of side effect (10). Additionally, dilation and curettage may be required in cases of failed treatment to clarify the diagnosis and terminate the abnormal pregnancy. We acknowledge that there are weaknesses in the present study. Ultrasound exams were performed by residents at all levels of training, and we relied on written documentation of ultrasound characteristics rather than ultrasound images. These factors may have contributed to some degree of misclassification, biasing our results against ultrasound diagnosis and in favor of D&C. However, because ultrasound exams are commonly performed by gynecology residents, our data may be generalizable to standard clinical practices in training institutions. Additionally, owing to the limited sample size, we did not pursue multivariable modeling to determine whether consideration of multiple risk factors could improve the sensitivity for detection of EP. As part of an ongoing multicenter project, we are currently evaluating a much larger dataset in that manner. We conclude that D&C remains an important tool in diagnosing EP and that ultrasound can be used as a powerful adjunct in making a presumptive preliminary diagnosis. Based on our results, we recommend that clinically stable women undergo diagnostic D&C to reduce the chance of unnecessary exposure to methotrexate.

661

REFERENCES 1. Centers for Disease Control and Prevention. Ectopic pregnancy—United States, 1990–1992. MMWR Morb Mortal Wkly Rep 1995;44:46–8. 2. Barnhart KT, Simhan H, Kamelle SA. Diagnostic accuracy of ultrasound above and below the beta-hCG discriminatory zone. Obstet Gynecol 1999;94:583–7. 3. Barnhart K, Sammel MD, Chung K, Zhou L, Hummel AC, Guo W. Decline of serum human chorionic gonadotropin and spontaneous complete abortion: defining the normal curve. Obstet Gynecol 2004;104:975–81. 4. Silva C, Sammel MD, Zhou L, Gracia C, Hummel AC, Barnhart K. Human chorionic gonadotropin profile for women with ectopic pregnancy. Obstet Gynecol 2006;107:605–10.

662

Chung et al.

5. Barnhart KT, Sammel MD, Rinaudo PF, Zhou L, Hummel AC, Guo W. Symptomatic patients with an early viable intrauterine pregnancy. hCG curves redefined. Obstet Gynecol 2004;104:50–5. 6. Barnhart KT, Katz I, Hummel A, Gracia C. Presumed diagnosis of ectopic pregnancy. Obstet Gynecol 2002;100:505–10. 7. Condous G, Kurk E, Lu C, van Calster B, Van Huffel S, Timmerman D, et al. There is no role for uterine curettage in the contemporary diagnostic workup of women with a pregnancy of unknown location. Hum Reprod 2006;21:2706–10. 8. Lindahl B, Ahlgren M. Identification of chorion villi in abortion specimens. Obstet Gynecol 1986;67: 79–81.

D&C in ectopic pregnancy

9. Barnhart KT, Gosman G, Ashby R, Sammel M. The medical management of ectopic pregnancy: a metaanalysis comparing ‘‘single dose’’ and ‘‘multidose’’ regimens. Obstet Gynecol 2003;101:778–84. 10. Ailawadi M, Lorch SA, Barnhart KT. Cost-effectiveness of presumptively medically treating women at risk for ectopic pregnancy compared with first performing a dilatation and curettage. Fertil Steril 2005;83:376–82.

Vol. 96, No. 3, September 2011