Pregnancy of unknown location is a temporary physiologic state of early pregnancy with a positive but no ultrasound findings of pregnancy, while pregnancy of uncertain prognosis is an early intrauterine pregnancy without cardiac activity. A broad differential and swift diagnosis are important to limit the morbidity and mortality associated with ectopic pregnancy. Knowledge of available management options along with patient-centered approaches is important to improve clinical provision of care, although this may be limited by state and institutional policies.
Key points
- •
Pregnancy of unknown location (PUL) is a positive hCG and no ultrasonographic evidence of pregnancy while pregnancy of uncertain prognosis (PUP) is an intrauterine pregnancy without cardiac activity.
- •
The differential for PUL includes intrauterine pregnancy, ectopic pregnancy, expelled early pregnancy loss, and causes of elevated hCG including molar pregnancy or hormone secreting tumors.
- •
Management strategies for PUL and PUP include expectant and active, with expectant consisting of serial hCG and ultrasound and active including medications and/or procedures.
- •
Patient-centered care is important in early pregnancy care, and clinicians should pay particular attention to pregnancy desiredness and risk tolerance to help guide management.
- •
Clinician and systems factors which impact patient-centered care, including state laws and institutional policies restricting access to induced abortion.
| ACEP | American College of Emergency Physicians |
| ACOG | American College of Obstetricians and Gynecologists |
| AIUM | American Institute of Ultrasound in Medicine |
| ART | assisted reproductive technology |
| EP | ectopic pregnancy |
| EPAC | early pregnancy assessment clinics |
| EPL | early pregnancy loss |
| GS | gestational sac |
| hCG | human chorionic gonadotropin |
| IUD | intrauterine device |
| IUP | intrauterine pregnancy |
| MSD | mean sac diameter |
| NPV | negative predictive value |
| PPV | positive predictive value |
| PUL | pregnancy of unknown location |
| PUP | pregnancy of uncertain prognosis |
| SAB | spontaneous abortion |
| TVUS | transvaginal ultrasound |
Overview, terminology, and prevalence
Pregnancy of Unknown Location
The development of highly sensitive and accurate human chorionic gonadotropin (hCG) assays as well as advancements in ultrasound technology allow for early detection of pregnancy, and also an increased incidence of pregnancy of unknown location (PUL). PUL is a temporary pathologic or physiologic phenomenon of early pregnancy, which encompasses all pregnancy outcomes and is not a final diagnosis. PUL is defined as a positive pregnancy test with a nondiagnostic transvaginal ultrasound, or transvaginal ultrasound without evidence of intrauterine or ectopic pregnancy (EP). , Some authors have argued to alter the definition of PUL to specify that there are no sonographic findings of definite intrauterine pregnancy (IUP) or EP; however this creates unnecessary confusion when a likely diagnosis is favored. Therefore, any signs of an IUP or EP, whether probable or definite, do not qualify as a PUL. , PUL is a critical topic for practitioners to understand well due to the potential implications and need for swift diagnosis of EP. EP constitutes 2% to 3% of all pregnancies, and complications from EP contribute significantly to pregnancy-related morbidity and mortality, with 2.7% of US pregnancy-related deaths between 2011 and 2017 attributed to EP.
In order to discuss pregnancy outcomes, early pregnancy terminology must be defined. Although there has been multidisciplinary agreement on first-trimester imaging guidelines, until recently there was a lack of consensus on terminology to describe early pregnancy, both in the medical record and to patients. To address this, the Society of Radiologists in Ultrasound recently convened a multisociety panel to create a lexicon for first trimester ultrasound, with the goal of identifying terms that were (a) clear, specific, scientifically based, and medically appropriate; (b) acceptable to sonographers, clinicians, and patients; and (c) minimize bias and harm. The panel included the Society of Abdominal Radiology, the American College of Obstetricians and Gynecologists (ACOG), the American Institute of Ultrasound in Medicine (AIUM), the Society for Maternal-Fetal Medicine, the American Society for Reproductive Medicine, the Society of Family Planning, and the American College of Emergency Physicians (ACEP). The resultant lexicon is organized into 4 categories: general first-trimester terms, early development, pregnancy location, and early pregnancy loss (EPL). Table 1 outlines the updated consensus terminology which will be used throughout this article.
| Term (s) Alternate Term (s) | Application/Definition | Term(s) to Avoid | Comments |
|---|---|---|---|
| PUL | No findings of probably or definite IUP or EP on TVUS |
| |
| IUP | Pregnancy implanted in a normal location |
|
|
| IUP of unknown prognosis Concerning for EPL Concerning for miscarriage Concerning for SAB | Normally located GS with findings that suggest a pregnancy may not progress |
|
|
| EP | Pregnancy implanted in an abnormal location |
|
|
| EPL | First trimester pregnancy that will not progress |
|
|
The prevalence of PUL is affected by many factors, such as institutional or country definitions, ultrasound quality, and access to early pregnancy care. Using consensus guidelines and terminology helps to more accurately reflect PUL prevalence. Historically, PUL prevalence rates varied from 8% to 31%, with some studies demonstrating rates as high as 42%. These older data note a generally accepted rate of PUL of 15% in patients undergoing transvaginal ultrasound in the first trimester. A 2005 statement by the International Society of Ultrasound in Obstetrics and Gynecology determined that units specialized in early gestation should aim for a PUL rate below 15% and found that the prevalence of PUL is largely determined by the quality of ultrasonography, encompassing both the technology and the expertise of the sonographer and reading physician. , A more modern cohort found a PUL rate of only 4.5%, showing variation across ultrasound setting between 2.5% (outpatient) and 8.4% (inpatient), with the lowest rate noted in an outpatient setting staffed with maternal-fetal medicine subspecialists.
Intrauterine Pregnancy of Uncertain Prognosis
Before the recent development of the consensus lexicon for first trimester ultrasound, the term pregnancy of uncertain viability was used to describe an IUP with an uncertain prognosis, including findings concerning for EPL as well as IUP without development of internal structures that suggested likelihood of continued development, namely an embryo with cardiac activity. As viability is understood to denote the possibility for a fetus to survive in the extrauterine environment, the multisociety consensus panel recommended avoiding the term viability in the first trimester pertaining to the appearance of an embryo with cardiac activity. The new preferred terminology in early pregnancy for an intrauterine gestational sac (GS) with or without a yolk sac or embryo without cardiac activity is IUP of uncertain prognosis. This term may be accompanied by other descriptors including concerning for EPL, concerning for miscarriage, or concerning for spontaneous abortion (SAB). Broadly, IUP of uncertain prognosis is defined as transvaginal ultrasound visualization (TSUV) of an intrauterine GS without demonstration of embryonic cardiac activity.
Natural history and differential diagnosis
Pregnancy of Unknown Location
The hCG levels in all stages of early pregnancy have significant overlap, and therefore a single hCG measurement is not reliable to distinguish gestational age. There has been substantial research since the 1980s to determine the appropriate discriminatory hCG level. The discriminatory hCG level or zone is the hCG value at which an intrauterine GS should be visualized on ultrasound, prompting further evaluation if it is not seen. Early studies at a time of lower ultrasound quality placed the level as high as 6500 mIU/mL, which was decreased as low as 1500 mIU/mL, and commonly positioned between 2000 and 3000 mIU/mL. The concept of the discriminatory zone has been further refined to reflect more modern patient populations with higher rates of medical comorbidities, multiple gestations and uterine abnormalities, the known harms of misdiagnosis, and improved imaging techniques. Currently, ACOG recommends a discriminatory level that is conservatively high, up to 3500 mIU/mL, which avoids potential erroneous intervention against an IUP, though both IUPs and EPs may be visible at lower levels. , Because PUL is not a final diagnosis but rather a transient state used to indicate patient risk, it is imperative that repeated diagnostic testing be completed until definitive diagnosis is reached. Earlier detection of pregnancy and more accurate diagnosis of its complications have drastically reduced the morbidity and mortality associated with EP.
The differential diagnosis for PUL includes (1) nonvisualized IUP, (2) nonvisualized EP, and (3) an early pregnancy loss (EPL) that has completely passed. Nonvisualized IUP or EP can be further subdivided into multiple possible outcomes. , , , Fig. 1 outlines the differential diagnosis of PUL, with final diagnostic possibilities and criteria. Approximately 34% to 40% of patients with PUL are ultimately diagnosed with an IUP, of which approximately 18% progress normally and the remainder end in EPL. , , Up to 14% of PUL are found to be EP, with the identification of EP and its potential morbidity and mortality being the reason why following patients with PUL is important. The majority of PUL will self-resolve, with 44% to 69% of PUL resolving without intervention, with the location of the pregnancy often never identified. , Persistent PUL comprises a small number of PUL (2%) and may include nonvisualized EP, thus requiring intervention, which may ultimately lead to refined diagnoses of nonvisualized EP, treated persisted PUL, resolution of persistent PUL, or pathologic confirmation of IUP. It is important to mention that in patients with persistently elevated hCG levels, the differential diagnosis must be expanded to include nonpregnancy diagnoses such as an hCG-secreting ovarian mass, gestational trophoblastic disease, or pituitary secretion of hCG.
Pregnancy of Uncertain Prognosis
The differential diagnosis for pregnancy of uncertain prognosis (PUP) is more limited than PUL given its confirmed intrauterine location. Diagnostic possibilities include IUP or EPL, with recommendation for repeat ultrasound in 7 to 14 days to determine the final diagnosis. Table 2 lists the findings diagnostic of and concerning for EPL as defined by the Society of Radiologists in Ultrasound Multispecialty Consensus Conference on early first trimester diagnosis of miscarriage and exclusion of a viable IUP, with the most recent terminology updates noted.
| Findings Diagnostic of EPL a | Finding Concerning for, but Not Diagnostic of EPL b |
|---|---|
| Crown-rump length of ≥7 mm and no cardiac activity c | Crown-rump length of <7 mm and no cardiac activity c |
| MSD of ≥25 mm and no embryo | MSD of 16–24 mm and no embryo |
| Absence of embryo with cardiac activity c ≥2 wk after a scan that showed a GS without a yolk sac | Absence of embryo with cardiac activity c 7–13 d after a scan that showed a GS without a yolk sac |
| Absence of embryo with cardiac activity c ≥11 d after a scan that showed a GS with a yolk sac | Absence of embryo with cardiac activity c 7–10 d after a scan that showed a GS with a yolk sac |
| Absence of embryo ≥6 wk after last menstrual period | |
| Empty amnion (amnion seen adjacent to yolk sac, with no visible embryo) | |
| Enlarged yolk sac (>7 mm) | |
| Small GS in relation to the size of the embryo (<5 mm difference between MSD and crown-rump length) |
a Note change in terminology from pregnancy failure.
b Note change in terminology from suspicious for and pregnancy failure.
c Note change in terminology from heartbeat to cardiac activity.
Beta human chorionic gonadotropin—trend and predictive models
Pregnancy of Unknown Location
The most used tools to make a final diagnosis in PUL include serial serum beta-hCG levels 48 hours apart and focused ultrasonography. Classically, it has been suggested hCG rise of 53% after 2 days will identify 99% of normal IUPs when the initial hCG level is less than 5000 mIU/mL. However, Barnhart and colleagues reassessed the minimal expected rise in hCG in normal IUPs with a retrospective cohort study of patients with PUL in whom normal IUP was ultimately confirmed. The analyzed hCG measurements found that the estimate of 53% minimum bHCG rise over 48 hours may have been too high, though the rate of rise was greater when initial values were lower. Table 3 shows the minimum percentage bHCG increase over 48 hours for 99% of normal IUPs, based on the starting hCG level determined from that study’s model. Conversely, the rate of hCG decrease in completed EPL demonstrates a greater decline with a higher starting level, and ranges from 21% to 35% at 2 days to 60% to 84% at 7 days. Rates of decline lower than these levels suggest residual products of conception or EP.
| Initial hCG Level (mIU/mL) | Minimum % Increase 2 d Later |
|---|---|
| <1500 | 49% |
| 1500–3000 | 40% |
| > 3000 | 33% |
Using only hCG trend to identify IUP or EPL is problematic, given that EP can mimic the hCG trend of a normal IUP or an EPL. The predictors of EP study found that in the 60% of EPs with initial hCG increase, the median slope was 32% over 2 days, while in the 40% of EPs with initial hCG decrease, the median decline was 15%, with 27% of overall patients having hCG trends that mimicked IUP or EPL. Therefore, while an abnormal trend supports an EP or EPL, an hCG trend within the established norms for IUP or EPL does not exclude EP. ,
Beyond hCG alone, there are predictive models, which assist in risk stratification of patients with PUL. These models use the hCG ratio and other biomarkers to predict pregnancy outcomes. Fig. 2 outlines how the hCG ratio, also known as the M1 model, is calculated and details predicted pregnancy outcomes based on results. In the United Kingdom, Condous and colleagues found that for the detection of EP, Model M1 had a sensitivity of 91.7%, a specificity of 84.2%, and positive predictive value (PPV) 27.5% and a negative predictive value (NPV) of 99.4%. , Later iterations include the Model M4, using multicategorical logistic regression with hCG levels, which was found to be clinically superior to Model M1 in the study population. The M4 model’s external validity and application to a population in the United States has been questioned, with sensitivity dropping to 49%. A different United Kingdom-based group developed Model M5, which is a two-step model, and then Model M6, which uses hCG levels with or without initial progesterone level to determine risk stratification. , Validation studies of this model found a sensitivity of 96% in the United Kingdom. Recently, an updated Model M6 has been published to include clinical factors such as patient age, pain score, and bleeding, which also showed good sensitivity for detecting EP. Fig. 3 describes each model, with increasing complexity of the models noted over time. Despite the promise of these predictive models, the lack of 100% PPV or NPV limit clinical utility. Additionally, there are benefits to collecting serial hCG values beyond the initial 2 over 48 hours, as subsequent values and trend adjustments can improve accuracy.
Full access? Get Clinical Tree
