Artifacts, Pitfalls, and Normal Variants

An Unfortunate But Common Pitfall

I begin this chapter as the subheading suggests, with a common but unfortunate pitfall.

A 34-year-old woman presented to the emergency department complaining of mild pelvic pain with a known history of positive pregnancy test. A quantitative serum β-hCG (human chorionic gonadotropin) revealed a value of 3870 mIU/mL. (This value is nearly double most centers' “discriminatory zone.”) A diagnostic ultrasound examination revealed no evidence of an intrauterine pregnancy (IUP) or adnexal masses. As a result, the patient was presumed to have an ectopic pregnancy (EP) and was treated with intramuscular methotrexate. A repeat sonogram obtained weeks later revealed a living IUP. The patient elected to terminate the pregnancy.

When a woman of child-bearing age presents to her clinician or emergency room with pelvic pain or bleeding and a positive pregnancy test, clearly the first thought in the diagnostic tree is the possibility of an EP. Most clinicians are aware that the risk of a heterotopic pregnancy (in which both an extrauterine [ectopic] and IUP occur simultaneously) is quite low (1 in 15,000 to 1 in 30,000), although this phenomenon is more common in patients following assisted reproduction. Thus, the goal of sonography is to identify an IUP, making an EP unlikely. Two factors have helped achieve this goal:

• 1.

  In the early 1980s, Kadar and others attempted to determine the maternal serum hCG level above which a gestational sac should be consistently seen during sonograms in a patient with a normal IUP. The value was termed the “discriminatory zone.” It should be noted that the values initially reported were significantly higher prior to the advent of transvaginal sonography in use today. Although this upper limit value today is variable at differing institutions, it tends to be between 1500 and 2000 mIU/mL (first IUP).

• 2.

  With the advent of high-resolution transvaginal sonography, early pregnancies can now often be identified within the uterus when gestational sacs are as small as 2 to 3 mm.

A noninvasive method of treating EPs with methotrexate became available in the early 1980s. As a result, some clinicians, then and now, treat the patient with methotrexate when both an “empty uterus” at the time of sonography and an elevated β-hCG level (above the discriminatory zone) are present, in an attempt to treat the EP. Unfortunately, some of these patients subsequently have been found to have IUPs, not EPs, on follow-up. It is quite clear that even marked elevations of β-hCG above the discriminatory zone are not always reliable in excluding an intrauterine pregnancy when a gestational sac is not seen at sonography. In these cases, follow-up, rather than immediate treatment, is indicated. Likewise, even in the best of hands and with excellent ultrasound equipment, at 4 to 5 weeks' gestation an IUP may not be identified sonographically, though the patient may in fact have an IUP. The report should ideally read “An ectopic pregnancy, or a pregnancy too early to be seen within the uterus, cannot be excluded.”

In October 2012, the Society of Radiologists in Ultrasound convened a multispecialty consensus conference to establish criteria on early first trimester diagnosis of miscarriage and exclusion of a viable IUP when seen in the uterus, as well as to discuss diagnosing and ruling out a viable IUP in a woman with a pregnancy of unknown location. These criteria and an editorial discussion were published in the New England Journal of Medicine ( ; ). The takeaway points from this panel regarding this pitfall were as follows:

• A single hCG assessment, regardless of level, does not reliably determine a pregnancy's location or viability (this is because hCG levels in women with nonviable IUPs, viable IUPs, and EP overlap substantially).

• A single hCG level less than 3000 mIU/mL should not elicit treatment for presumed EP because there is a substantial risk that administering methotrexate will damage a normal IUP.

• A single hCG level 3000 mIU/mL or higher in the absence of a sonographically visualized IUP indicates that a viable IUP is possible but unlikely. The most likely diagnosis is failed IUP, so if methotrexate is administered, it is more likely being given unnecessarily to a woman with a failed IUP than to one with an EP. At least one additional hCG level should be measured before initiating treatment for EP.