Amy Hummel

Symptomatic patients with an early viable intrauterine pregnancy: HCG curves redefined.

OBJECTIVE: To analyze the change in serial human chorionic gonadotropin (hCG) levels in women symptomatic with pain or bleeding who presented with nondiagnostic ultrasonography but were ultimately confirmed to have a viable intrauterine pregnancy. METHODS: The rise in serial hCG measures were modeled over time, with the start point defined in 2 ways: by last menstrual period and by date of presentation for care. Both semiparametric (spline) curves and linear random-effects models were explored. The slope and projected increase of hCG were calculated to define 99% of viable intrauterine pregnancies. RESULTS: A total of 287 subjects met inclusion criteria and contributed 861 measurements of hCG. On average, these subjects contributed 3.00 observations and were followed up for 5.25 days. A linear increase in log hCG best described the pattern of rise. Curves derived from last menstrual period and day of presentation do not differ substantially. The median slope for a rise of hCG after 1 day was 1.50, (or a 50% increase); 2.24 after 2 days (or a 124% rise), and 5.00 after 4 days. The fastest rise was 1.81 at 1 day, 3.28 at 2 days, and 10.76 at 4 days. The slowest or minimal rise for a normal viable intrauterine pregnancy was 24% at 1 day and 53% at 2 days. CONCLUSION: These data define the slowest rise in serial hCG values for a potentially viable gestation and will aid in distinguishing a viable early pregnancy from a miscarriage or ectopic pregnancy. The minimal rise in serial hCG values for women with a viable intrauterine pregnancy is “slower” than previously reported, suggesting that intervention to diagnosis and treat an abnormal gestation should be more conservative. LEVEL OF EVIDENCE: II-2

Presumed diagnosis of ectopic pregnancy

OBJECTIVE To evaluate the accuracy of the diagnosis of presumed ectopic pregnancy. METHODS This was a retrospective cohort analysis at a tertiary care medical center. The patient population was composed of 1) clinically stable pregnant women with human chorionic gonadotropin (hCG) above 2000 mIU/mL and no evidence of an intrauterine pregnancy by ultrasound, or 2) women with an abnormal rise or fall of serial hCG below 2000 mIU/mL. Outcome was determined by pathologic evidence of chorionic villi in the endometrial curettings (or fallopian tube), or complete resolution of hCG. RESULTS Overall, 38.4% (43/112) of the women were diagnosed with a miscarriage and 61.6% (69/112) were found to have an ectopic pregnancy. No significant difference was found in race, age, gravity, parity, hCG trends, or time to diagnosis between women with ectopic pregnancies and those with miscarriages. Patients were more likely to be diagnosed with an ectopic pregnancy if the initial hCG value was below the discriminatory zone (relative risk 2.44; 95% confidence interval 1.07, 5.52). Ultrasound correlated well with the final diagnosis (P = .001) but was not definitive. CONCLUSION In an effort to save time, avoid dilation and curettage (D&C), and treat with methotrexate, the presence of an ectopic pregnancy is often presumed. The presumed diagnosis of ectopic pregnancy is inaccurate in almost 40% of cases. A D&C is necessary to differentiate an ectopic pregnancy from a miscarriage before a woman is presumptively treated with methotrexate.

Human chorionic gonadotropin profile for women with ectopic pregnancy

OBJECTIVE: To analyze serial human chorionic gonadotropin (hCG) levels in women presenting to the emergency department who were ultimately confirmed to have ectopic pregnancies. METHODS: Human chorionic gonadotropin levels were obtained over time until definitive diagnosis. To be included, women had to have at least 2 hCG measurements. Human chorionic gonadotropin curves were characterized and their slopes calculated. RESULTS: Two hundred women received diagnoses of ectopic pregnancy with the help of serial hCG values and were included in the study. No curve adequately characterized the pattern of hCG values so attention was focused on the initial 2 values. The median slope of log hCG among all subjects was 0.11 (25% increase in 2 days). However, 60% of subjects had an initial rise in hCG, and 40% had an initial fall. The rise in hCG for women with ectopic pregnancies (0.28; 75% increase in 2 days) was slower than the mean increase reported for a viable intrauterine pregnancy. The decline in hCG for women with ectopic pregnancies (−0.225; 27% decline in 2 days) was slower than the mean reported for completed spontaneous abortion. However, 20.8% of women presented with a rise in hCG values similar to the minimal rise for women with a viable gestation, and 8% of women presented with a fall in hCG values similar to women with a completed spontaneous abortion. CONCLUSION: There is no single way to characterize the pattern of hCG for ectopic pregnancy. The number of women with ectopic pregnancy who experience an increase in hCG values is approximately equal to the number of those who experience a decrease. The hCG profile in women with ectopic pregnancy can mimic that of an intrauterine pregnancy or a completed spontaneous abortion in approximately 29% of cases. LEVEL OF EVIDENCE: II-2

Decline of serum human chorionic gonadotropin and spontaneous complete abortion: defining the normal curve.

OBJECTIVE: We sought to estimate a standard curve of serum human chorionic gonadotropin (hCG) decline that characterizes spontaneous abortion. METHODS: Data were extracted from a clinical database of women with symptomatic early pregnancies and nondiagnostic ultrasonography who required follow-up with serial hCG levels. The evaluation was restricted to women who had a pregnancy of unknown location, a decrease in serum levels, and who were ultimately diagnosed with miscarriage (a decrease in serum hCG to < 5 mIU/mL in the absence of surgical intervention or confirmation of products of conception after dilation and curettage). The starting point of the curve was the hCG concentration at presentation, with serial levels plotted until the time of definitive diagnosis. Semiparametric statistical techniques were used to characterize the shape of the curve. We present slopes for the decrease in log hCG levels and the projected rate of fall. RESULTS: Of the 1,543 patients, 710 were diagnosed with miscarriage. A quadratic curve for log hCG with a steeper rate of decline for large initial hCG values best described the pattern of change. One curve could not fully estimate the decline because the rate was dependent on the initial hCG level. In other words, more rapid decline was associated with a higher starting concentration. The rate of decline ranged from 21% to 35% at 2 days and 60% to 84% at 7 days, depending on initial hCG value. CONCLUSION: The rate of hCG decrease in spontaneous abortions is described by a quadratic profile, with a faster decline in hCG value with higher presentation levels. A rate of decline less than 21% at 2 days or 60% at 7 days suggests retained trophoblasts or an ectopic pregnancy.

Risk factors for spontaneous abortion in early symptomatic first-trimester pregnancies

OBJECTIVE: To evaluate the association of an ultimate diagnosis of miscarriage with various clinical symptoms and historical factors in a cohort of women presenting with pain, bleeding, or both in the first trimester of pregnancy. METHODS: This was a case–control study from a population of women presenting for care with pelvic pain or vaginal bleeding in the first trimester of pregnancy whose diagnoses were not definite upon initial evaluation. Analyses were performed in 2 ways. In one instance cases were defined as women ultimately definitively diagnosed with a miscarriage and controls were defined as women with a pregnancy that did not result in miscarriage (ectopic pregnancy or ongoing intrauterine pregnancy). The second analysis compared women with a miscarriage only to women who had an ongoing intrauterine pregnancy. RESULTS: A total of 2,026 women were evaluated, with 1,192 ultimately diagnosed with a spontaneous abortion, 367 with ectopic pregnancy, and 467 with a viable intrauterine pregnancy. Although many risk factors were individually associated with miscarriage in preliminary analysis, in the final analysis only extremes in age (< 25 and > 35) and the complaint of bleeding (odds ratio [OR] 7.35, 95% confidence interval[CI] 5.74–9.41) were associated with miscarriage. The complaint of pain (OR 0.72, 95% CI 0.57–0.92), human chorionic gonadotropin (hCG) value greater than 500 (hCG ≤ 500 IU/mL compared with hCG 501–2000: OR 0.52, 95% CI 0.39–0.69) and concurrent cervical infection (OR 0.69, 95% CI 0.55–0.88) were negatively associated with miscarriage. CONCLUSION: Few factors predict miscarriage in women who present with a symptomatic first trimester pregnancy of unknown location. Heavy bleeding was most strongly associated with miscarriage. Concurrent cervical infections should not be overlooked as a cause of bleeding in early pregnancy and were not associated with miscarriage. LEVEL OF EVIDENCE: II-2

A serum proteomics approach to the diagnosis of ectopic pregnancy.

Abstract: An ectopic pregnancy (EP) occurs when implantation of the embryo occurs outside of the uterus. If left untreated, the developing fetus will continue to grow, leading to life‐threatening consequences for the mother. A major difficulty with the diagnosis of ectopic pregnancy is that methods of detection are limited, and some, such as ultrasound, are not very reliable in the earliest days of gestation. Currently, no effective serum test exists to distinguish an ectopic pregnancy from a normal intrauterine pregnancy. The incidence of ectopic pregnancy is increasing and has doubled in the last 20 years. It is now the second most common cause of maternal death in the first trimester of pregnancy. To address this issue, we initiated a project to identify serum markers of ectopic pregnancy. The subjects for these studies presented at the Hospital of the University of Pennsylvania. We obtained over 140 serum samples from women with suspected ectopic pregnancy: women presenting with pain and/or bleeding in the first trimester of pregnancy. The approximate racial breakdown of the subjects is as follows: African American, 36%; Caucasian, 3%; Asian, 2%; Hispanic, 1%; unknown, 58%. Serum samples from 139 women (62 with ectopic pregnancy and 77 with a normal intrauterine pregnancy) were applied to WCX2 (weak ion exchange) protein chip surfaces and analyzed for serum markers using surface‐enhanced laser desorption/ionization time‐of‐flight mass spectrometry (SELDI‐TOF‐MS). Several proteins in the 7500–18,000 Da mass range were identified that may discriminate an ectopic pregnancy from an intrauterine pregnancy. The most promising markers were analyzed using classification and regression tree analysis (CART) with and without clinical variables (serum hCG value, length of amenorrhea). Two different algorithms were developed that classify the patients on the basis of sensitivity (number of EPs who screen positive/# of EPs) or specificity (# of healthy patients who screen negative/# of healthy). Our current approach is to refine these two “rule sets” to segregate patients into three groups: those who need immediate intervention for a probable ectopic pregnancy, those who appear to have a normal pregnancy, and those who need further monitoring for diagnosis.

Human Chorionic Gonadotropin Profile for Women With Ectopic Pregnancy

OBJECTIVE To analyze serial human chorionic gonadotropin (hCG) levels in women presenting to the emergency department who were ultimately confirmed to have ectopic pregnancies. METHODS Human chorionic gonadotropin levels were obtained over time until definitive diagnosis. To be included, women had to have at least 2 hCG measurements. Human chorionic gonadotropin curves were characterized and their slopes calculated. RESULTS Two hundred women received diagnoses of ectopic pregnancy with the help of serial hCG values and were included in the study. No curve adequately characterized the pattern of hCG values so attention was focused on the initial 2 values. The median slope of log hCG among all subjects was 0.11 (25% increase in 2 days). However, 60% of subjects had an initial rise in hCG, and 40% had an initial fall. The rise in hCG for women with ectopic pregnancies (0.28; 75% increase in 2 days) was slower than the mean increase reported for a viable intrauterine pregnancy. The decline in hCG for women with ectopic pregnancies (-0.225; 27% decline in 2 days) was slower than the mean reported for completed spontaneous abortion. However, 20.8% of women presented with a rise in hCG values similar to the minimal rise for women with a viable gestation, and 8% of women presented with a fall in hCG values similar to women with a completed spontaneous abortion. CONCLUSION There is no single way to characterize the pattern of hCG for ectopic pregnancy. The number of women with ectopic pregnancy who experience an increase in hCG values is approximately equal to the number of those who experience a decrease. The hCG profile in women with ectopic pregnancy can mimic that of an intrauterine pregnancy or a completed spontaneous abortion in approximately 29% of cases. LEVEL OF EVIDENCE II-2.