Monday, July 28, 2008

Ectopic Pregnancy 2


Medical Management

Methotrexate therapy of ectopic pregnancy has been used successfully over the last 2 decades. The folic acid antagonist, methotrexate, inhibits de novo synthesis of purines and pyrimidines, interfering with DNA synthesis and cell multiplication. Rapidly proliferating trophoblasts are very dependent on folic acid and thus differentially vulnerable to the cytotoxic effect of methotrexate, and this differential sensitivity forms the basis of the therapy. When methotrexate is administered to pregnant women undergoing planned termination, a single dose of 50 mg/m2 significantly blunts the β-hCG increment over the following 7 days and has been associated with a drop in circulating progesterone and 17-α-hydroxyprogesterone concentrations prior to abortion. It appears that methotrexate directly impairs trophoblastic production of hCG with a secondary decrement of corpus luteum progestin secretion. Hemodynamically stable patients with unruptured ectopic pregnancy measuring less than or equal to 4 cm by ultrasonography are eligible for methotrexate therapy. Patients with larger masses or evidence of acute intra-abdominal bleeding should undergo immediate surgical treatment. Methotrexate treatment regimens are shown in Table 2 and include the multiple dose, single dose, and the newly introduced two-dose protocol.

TABLE 2 Comparison of Methotrexate Regimens

Single-Dose Regimen

Multiple-Dose Regimen

Two-Dose Regimen

Methotrexate dose

50 mg/m2

1 mg/kg

50 mg/m2

Leucovorin dose


0.1 mg/kg


Dose frequency

Day 0
Potential dose on day 7

Alternating-day dosing of methotrexate and leucovorin, maximum of four doses of each

Day 0 and day 4
Potential doses on day 7 and day 11

β-hCG monitoring

Day 0, day 4, day 7
Day 11 and day 14 if additional doses given

Day 0, and then odd-numbered days until success

Day 0, day 4, day 7
Day 11 and day 14 if additional doses given

Success determined by

15% drop in β-hCG day 4 to day 7
15% drop in β-hCG day 11 to day 14 if second dose given

15% drop in β-hCG between any two blood draws

15% drop in β-hCG day 4 to day 7
15% drop in hCG day 7 to day 11 or day 11 to day 14 if third and fourth dose given

Multiple-Dose Methotrexate

Multiple-dose methotrexate therapy is tailored to the patient's weight and ectopic pregnancy responsiveness. Outcomes of 12 studies comparing multiple-dose systemic methotrexate with laparoscopic salpingostomy are presented in Table 3. Between 1982 and 1997, this tabulation shows 338 cases of ectopic pregnancy treated with variable-dose methotrexate (number of medication administrations varies according to response). Of these cases, 93% were treated successfully with multiple-dose systemic methotrexate (no subsequent therapy was required), and 75% of the women tested had patent fallopian tubes; in addition, of the women desiring pregnancy, 58% had a subsequent intrauterine pregnancy and 7% developed a repeat ectopic pregnancy. These rates all compare favorably with conservative surgical management.

There is one randomized clinical trial comparing laparoscopic salpingostomy with systemic multiple dose methotrexate. In it, 100 patients with laparoscopy-confirmed ectopic pregnancy were randomly treated with systemic methotrexate or laparoscopic salpingostomy. In the 51 patients treated with methotrexate, seven (14%) required surgical intervention for active bleeding or tubal rupture. An additional course of methotrexate was required in two patients (4%) for persistent trophoblast, based on continued β-hCG secretion. Of the 49 patients in the salpingostomy group, four patients (8%) failed and required salpingectomies, and ten patients (20%) required treatment with methotrexate for persistent trophoblast. Homolateral tubal patency was present in 23/42 (55%) of the patients assessed in the methotrexate group and in 23/39 (59%) of those assessed in the salpingostomy group. This randomized study and previous meta-analysis have demonstrated the effectiveness of systemic methotrexate therapy as equal to laparoscopic salpingostomy.

TABLE 3 Outcome of Different Treatments for Ectopic Pregnancy

Subsequent Fertility Rate


Number of Studiesa

Number of Patients

Number with Successful Resolution

Tubal Patency Rate

Intrauterine Pregnancy

Ectopic Pregnancy

Conservative laparoscopic surgery



1,516 (93%)

170/223 (76%)

366/647 (57%)

87/647 (13%)

Variable-dose methotrexate



314 (93%)

136/182 (75%)

55/95 (58%)

7/95 (7%)

Single-dose methotrexate



340 (87%)

61/75 (81%)

39/64 (61%)

5/64 (8%)

Direct-injection methotrexate



502 (76%)

130/162 (80%)

87/152 (57%)

9/152 (6%)

Expectant management



425 (68%)

60/79 (76%)

12/14 (86%)

1/14 (7%)

aReferences available on the Lancet Web site ( or from the journal's London office. Accessed December 6, 2002.
From Pisarska MD, Carson SA, Buster JE, et al. Ectopic pregnancy. Lancet 1998;351:1115-1120.

Single-Dose Methotrexate

Single-dose methotrexate, although more convenient, is not as efficacious as multiple-dose methotrexate. The high success rates in the initial studies using single-dose methotrexate were most likely due to the inclusion of spontaneously aborting intrauterine pregnancies. Subsequent studies of single-dose methotrexate therapy involving 393 patients are presented in Table 3. Although overall success of treatment, measured as no surgical intervention, is 87%, 8% of patients required more than one dose of methotrexate. Of the patients considered successfully treated (with one or more doses), tubal patency was found in 81% of the women evaluated. The subsequent intrauterine pregnancy rate was 61%, and for ectopic pregnancies 8%, in the patients desiring future fertility in the same group (those treated with either one or more doses of methotrexate). Based on the clinical evidence presently available, the routine use of methotrexate as a single-dose intramuscular regimen is probably not as effective as multiple doses. However, single-dose therapy remains a standard according to publications of the American College of Obstetricians and Gynecologists.

With this background, a recent meta-analysis of 26 studies evaluating methotrexate dosing for ectopic pregnancy by Barnhart and colleagues showed an odds ratio of 1.96 higher likelihood of rupture with use of single-dose methotrexate over multidose therapy. Controlling for initial β-hCG value and the presence of cardiac activity, the failure rate with single-dose therapy was almost five times greater (odds ratio 4.75). What makes direct comparisons of these protocols even harder is that based on the data from this meta-analysis, 15% of patients under a single-dose protocol actually receive more than one dose, while 10%, 23%, and 14% of those under a multidose protocol actually need only 1, 2, or 3 doses of methotrexate, respectively.

Two-Dose Methotrexate Protocol

In recognition that the single-dose protocol has fewer visits and fewer injections but may have a higher failure rate, a two-dose protocol was introduced. This protocol uses the dosing and monitoring parameters of the single-dose protocol but gives a second dose of 50 mg/m2 on day 4, when only a serum β-hCG would have been drawn according to the single-dose protocol. No leucovorin rescue is used in this protocol. The same logic is used to determine if more methotrexate is needed based on the difference between the serum β-hCG on day 4 and day 7. In the single-dose protocol, a second dose would be given on day 7 if the β-hCG did not decline by at least 15% between day 4 and day 7 (Table 2). In the two-dose protocol, a second dose is given on day 4, and a third dose can be given on day 7 if the β-hCG did not decline by a least 15% between days 4 and 7. Thus, the two-dose protocol gives two doses in the first week and has provisions to give up to two more. In this way, the number of visits and surveillance laboratories are the same as the single dose, but more methotrexate is given sooner in hopes of maximizing success rate (without increasing complexity or the number of visits). This regimen was demonstrated to be safe in a three-center trial; the overall success rate will be determined as it is used more widely.

TABLE 4 Treatment with Multiple-dose Methotrexate

Indications for systemic methotrexate for uncomplicated ectopic pregnancy:

  • No rupture (hemodynamic stability)
  • UTZ size >4 cm
  • Β-hCG >10,000 mIU/mL
  • Positive fetal heartbeat: proceed with caution
  • Willingness of patient to comply with subsequent treatment monitoring

UTZ, ultrasound; hCG, human chorionic gonadotropin.

Safeguards and Counseling

Prior to instituting methotrexate therapy, physicians should evaluate baseline laboratory values. The patient should be screened with a complete blood count, liver function tests, and serum creatinine. A chest x-ray should be considered in women reporting a history of prior pulmonary disease due to their risk of developing methotrexate-related interstitial pneumonitis.

During methotrexate therapy, a woman should be examined by a single examiner only once, to diminish the risk of causing mechanical trauma and tubal rupture. The physician and the patient must recognize that transient pain (“separating” or “tearing pain”) is common. Transient pelvic pain from tubal bleeding or hematoma formation at the ectopic site frequently occurs 3 to 7 days after the start of therapy, lasts 4 to 12 hours, and is presumably due to tubal abortion. Perhaps the most difficult aspect of methotrexate therapy is learning to distinguish the transient abdominal pain of successful therapy from that of a rupturing ectopic pregnancy. Physicians must therefore carefully observe for clinical indications that an operation is necessary (Tables 4, 5). Thus, surgical intervention is required when pain is worsening and persistent beyond 12 hours. Orthostatic hypotension or a falling hematocrit should lead to immediate surgery. Sometimes, it is necessary to hospitalize the patient with pain for observation (usually about 24 hours) to insure a correct diagnosis. In addition, colicky abdominal pain is common during the first 2 or 3 days of methotrexate therapy, and the woman should avoid gas-producing foods such as leeks and cabbage. Women receiving methotrexate should discontinue prenatal vitamins, as they contain folic acid, and should especially avoid any additional folic acid supplementation. Finally, the patient should avoid exposure to the sun because photosensitivity can be a complication of methotrexate.

TABLE 5 Dealing with Methotrexate Failure

Operate when high suspicion of rupture:

  • Pain is severe and persistent, regardless of β-hCG levels
  • Falling hematocrit
  • Orthostatic hypotension

Consider operating when signs of treatment failure:

  • Levels of β-hCG do not decline by at least 15% between days 4 and 7 of treatment
  • Levels of β-hCG increase or plateau after first week of treatment

Methotrexate by Direct Injection

In 1987, Feichtinger and Kemeter instilled 10 mg (1 mL) of methotrexate into an ectopic gestational sac under transvaginal ultrasonography, and resolution occurred within 2 weeks. Direct injection delivers concentrations of methotrexate to the site of implantation at higher concentrations than those achieved with systemic administration. Less systemic distribution of the drug should decrease the overall toxicity. However, this approach has the substantial disadvantage of requiring laparoscopic or ultrasound needle guidance.

Outcomes in 21 studies involving direct injection of methotrexate with either laparoscopic or transvaginal ultrasound guidance are presented in Table 3. Between 1989 and 1997, 75.1% of 668 cases of ectopic pregnancy were treated successfully with methotrexate by direct injection, and some patients required more than one injection. Tubal patency and subsequent pregnancy rates were comparable to conservative laparoscopic surgery and systemic methotrexate: 80.2% of the women tested had patent oviducts, and of the women desiring pregnancy, 57.2% had a subsequent intrauterine pregnancy and 5.9% developed a recurrent ectopic pregnancy.

Randomized, controlled trials have demonstrated successful treatment with methotrexate by direct injection in 86.2% of the patients. Again, successful therapy included some patients who received more than one injection. Tubal patency was present in 85.1% of the women evaluated, and intrauterine pregnancy occurred in 73.1% of the women desiring subsequent fertility. One of the earlier randomized, controlled trials was discontinued because three of seven patients assigned to laparoscopic injection of methotrexate required additional laparoscopic surgery. Even with the higher success rate in the randomized trials, this technique is more cumbersome than systemic methotrexate. Given the overall tolerability and high success rate of systemic methotrexate, it continues to be the most accepted nonsurgical treatment modality.

Side Effects

High doses of methotrexate can cause bone marrow suppression, hepatotoxicity, stomatitis, pulmonary fibrosis, alopecia, and photosensitivity. These side effects are infrequent in the short treatment schedules used in ectopic pregnancy and can be attenuated by the administration of leucovorin (citrovorum factor). The side effects of methotrexate resolve within 3 to 4 days after the therapy is discontinued. Impaired liver function is the most common side effect. Other side effects include stomatitis, gastritis and enteritis, and bone marrow suppression. Local therapy by direct injection of methotrexate into the ectopic gestation resulted in fewer side effects, likely because of less systemic absorption. Even with local injection, impaired liver function tests, gastritis and enteritis, and bone marrow suppression can occur. Additional case reports exist in the literature. Cases of life-threatening neutropenia and febrile morbidity can occur after single or multidose intramuscular methotrexate, requiring hospitalization. Cases of transient interstitial pneumonitis from methotrexate therapy for ectopic pregnancy have been observed. Reversible alopecia (a loss of 33% to 50% of the scalp hair) on two separate occasions following single-dose therapy for an ectopic pregnancy has also been reported. Rarely, hematosalpinx and pelvic hematoceles have been noted as late sequelae of methotrexate following the normalization of β-hCG levels. These patients have pelvic pain, abnormal bleeding, and a pelvic mass, requiring surgical intervention, 3 to 5 months after therapy.

Fortunately, the side effects reported with methotrexate used to treat ectopic pregnancy have mostly been minor. Out of 100 patients treated in one study, only two patients developed stomatitis and three had transient elevation of transaminases, all resolving spontaneously. Another study that used the single-dose regimen had only one patient reporting nausea and vomiting following methotrexate treatment. Thus, with its overall good tolerability, methotrexate remains the first choice before surgical therapy.

Direct Injection of Cytotoxic Agents

Prostaglandins, hyperosmolar glucose, potassium chloride, and saline by direct injection have been tried as therapeutic alternatives to methotrexate. The limited experience with these agents, poor success rates, and the need for laparoscopic or transvaginal aspiration makes these unattractive treatment alternatives.

Evidence-Based Recommendation

Multiple-dose systemic methotrexate is the first-line medical treatment for ectopic pregnancy. Nearly half of patients under a multidose protocol will require fewer doses for ectopic pregnancy resolution. (Strength of recommendation: A.)

Surgical Treatment

Since the first successful salpingectomy (resection of involved fallopian tube segment with implanted trophoblastic tissue) performed by Tait in 1884, ectopic pregnancies traditionally have been treated by salpingectomy, usually by laparotomy. Historically, ectopic pregnancies were diagnosed at the time of emergency surgery, when concern for the patient's life superseded any concerns for her future fertility. It was not until 1953, when Stromme performed the first conservative procedure (salpingostomy, or removal of only the ectopic pregnancy with conservation of the tube) for ectopic pregnancy, that subsequent successful pregnancy outcomes were reported, confirming the potential for fertility preservation after salpingostomy. These surgical techniques have been modified for endoscopy. The laparoscopic approach is associated with less blood loss, less analgesia requirement, and a shorter duration of hospital stay. In addition, cost analysis has demonstrated significant savings in randomized trials. When evaluating subsequent fertility, intrauterine pregnancy rates are comparable for laparoscopy and laparotomy, as are rates of recurrent ectopic pregnancy.

Ruptured Ectopic Pregnancy

Early diagnosis and treatment of ectopic pregnancy avoids rupture in most cases. In the 1970s, 13.5% to 17.8% of patients with ectopic pregnancies arrived for treatment in hypovolemic shock, whereas in the early 1980s, only 4.4% of patients arrived in this condition. Today, either laparotomy or laparoscopy with salpingectomy is the first choice for rupture.

Once contraindicated over concern of decreased venous return from intraperitoneal insufflation, laparoscopic salpingectomy can be successful in patients in hypovolemic shock. Still, in critical instances when expeditious entry into the peritoneal cavity and tamponade of bleeding is necessary, rapid laparotomy to stem the bleeding is the preferred method. Nearly all patients in hypovolemic shock require blood transfusions; those with large red blood cell requirements also need fresh frozen plasma. In the hands of a highly skilled laparoscopist, with adequate cardiac monitoring and anesthesia, laparoscopic salpingectomy is an acceptable alternative to laparotomy even when there has been extensive intraperitoneal bleeding. At present, it is the surgeon's choice of laparoscopy or laparotomy for ruptured ectopic pregnancy.

Stable Ectopic Pregnancy

If methotrexate is contraindicated, laparoscopic salpingostomy is the first surgical choice. Alternatively, salpingectomy can be performed either during laparotomy or laparoscopy by using cautery or sutures (laparoscopic or endoloops). A review of data from nine studies showed that subsequent to salpingostomy, 53% of patients have intrauterine pregnancies compared with 49.3% after salpingectomy. Recurrent ectopic pregnancy rates were slightly higher after conservative surgery, 14.8% compared with 9.9%. Other studies have suggested a higher intrauterine pregnancy rate in women after salpingostomy, but at the cost of a possible higher risk of recurrent ectopic after 3 years of follow-up. Laparoscopic salpingectomy is preferred over salpingostomy in cases of uncontrollable bleeding not resolving with conservative measures when extensive tubal damage is present, if the ectopic pregnancy has recurred in the same tube, if it is a large pregnancy (>5 cm), and if sterilization is desired.

The recommended conservative surgical procedure for an ampullary ectopic pregnancy is linear salpingostomy, because the ectopic nidation typically is located between the endosalpinx and serosa rather than in the tubal lumen. A linear salpingostomy is created through a longitudinal incision by electrocautery, scissors, or laser over the bulging antimesenteric border of the fallopian tube. The products of conception are removed with forceps or gentle flushing or suction. After maintaining hemostasis, the incision is left to heal by secondary intention or closed primarily. There appears to be no additional benefit to suturing the tubal defect closed, as studies have shown no difference in subsequent tubal patency rates, postoperative adhesion rates, or cumulative pregnancy rates.

Historically, isthmic segment pregnancies were routinely treated with segmental excision followed by intraoperative or delayed microsurgical anastomosis. The tubal lumen is narrower and the muscularis is thicker in the isthmus than in the ampulla, predisposing the isthmus to greater damage after salpingostomy and greater rates of proximal tubal obstruction. With today's high success rates of in vitro fertilization (IVF), tubal anastomosis is rarely performed and the resected tubal segment is bypassed altogether by use of ART. Manual fimbrial expression, also known as milking, should not be used unless the trophoblastic tissue is already aborting spontaneously through the fimbriae.

Laparoscopic salpingostomy and fimbrial expression have been evaluated in 32 studies and are presented in Table 3. Of the 1,626 patients treated between 1980 and 1997, treatment was successful in 93.4% (defined as requiring no additional therapy). Of the patients evaluated for tubal patency by using either hysterosalpingography or laparoscopy, 76% had patent tubes. Of the women desiring subsequent fertility, 56.6% had an intrauterine pregnancy and 13.4% developed another ectopic pregnancy.

Persistent Ectopic Pregnancy Following Salpingostomy

Persistent ectopic pregnancy is diagnosed by a plateauing or rising β-hCG concentration following conservative surgical therapy. The β-hCG level should be checked on postoperative day 1, keeping in mind that a drop of <50%>

The increased rate of persistent ectopic pregnancies has been a criticism of conservative laparoscopic therapy when compared with laparotomy. A decision analysis that compared prophylactic methotrexate with linear salpingostomy against no methotrexate in a group of 1,000 women concluded that prophylactic methotrexate at the time of surgery was preferable if certain clinical conditions are met as follows: (a) the incidence of persistent ectopic pregnancy is greater than 9% with observation alone after salpingostomy, (b) the incidence of persistence is less than 5% when prophylactic methotrexate is given, (c) the probability of ectopic pregnancy rupture is greater than 7.3% with a persistent ectopic pregnancy, and (d) the complication rate associated with prophylactic methotrexate is less than 18%. Because the great majority of clinical circumstances meet these recommendations, prophylactic methotrexate administration is recommended.

Evidence-Based Recommendation

Due to lower morbidity and equal efficacy, laparoscopic surgery is preferable to laparotomy in the treatment of bleeding or complicated ectopic pregnancy. Salpingectomy by laparotomy is reserved for ectopic ruptures with a hemodynamically unstable patient. (Strength of recommendation: A.)

Ectopic Pregnancy and Assisted Reproductive Technology


The risk of ectopic pregnancy is increased in patients undergoing an ART procedure. This increased risk has been attributed to the cause of infertility for which most patients seek treatment, that is, tubal factor infertility. Information on ectopic pregnancies resulting from ART comes from data obtained from institutions in the United States and Canada reporting to the Society for Assisted Reproductive Technology. The rate of pregnancies that resulted in ectopic pregnancies after IVF in 1999 was 3%, with newer figures from 2004 reporting the ectopic rate to be about 2%. The latter included outcome of ART cycles using fresh, nondonor eggs or embryos in approximately 76,000 embryo transfers. This lower percentage likely reflected the trend toward performing salpingectomies when hydrosalpinges are present to improve the success of ART.


As in naturally occurring ectopic pregnancies, the fallopian tube is the most common site for ectopic pregnancies following IVF. Data obtained from three case-control studies reveal that 82.2% of ectopic pregnancies were tubal. When tubal location was specified, 92.7% were ampullary and 7.3% interstitial. Extratubal ectopic nidations were as follows: 4.6% ovarian or abdominal, 1.5% cervical, and 11.7% heterotopic pregnancies (Fig. 1).

Tubal Pathology

Tubal pathology is the most important predisposing factor for ectopic pregnancy in patients undergoing IVF. Ectopic pregnancies are four times higher in patients with tubal factor infertility compared with patients with normal tubes. Hydrosalpinges are associated more commonly with ectopic pregnancy than other types of tubal pathology. Prior tubal reconstructive surgery (salpingostomy) increases the risk of ectopic pregnancy by 10% above that in patients with tubal factor infertility without prior surgery.

Thus, it is not surprising that patients with previous pelvic inflammatory disease have a sixfold increase in ectopic pregnancy after IVF. However, a history of prior ectopic pregnancy does not seem as important a risk factor in IVF cycles as in natural cycles.

Salpingectomy, particularly with hydrosalpinx, has been shown to decrease risks of ectopic pregnancy while increasing pregnancy rates after IVF. Meta-analysis has demonstrated that the presence of hydrosalpinges decreases the chance for viable pregnancy by approximately 50% when compared with patients with tubal disease but without hydrosalpinges. The implantation rate was also noted to be 50% lower with a higher chance of miscarriage and ectopic gestation. The ultimate conclusion is that when a hydrosalpinx is present, there is a decreased pregnancy rate with resultant decreased delivery rate following IVF. In addition, patients who undergo salpingectomy or proximal tubal occlusion prior to oocyte retrieval and transfer are at decreased risk for pelvic infection as well as future ectopic pregnancy.

Ovulation Induction

Hormone alterations during ovulation induction theoretically alter tubal function. In animal models, estrogen administration results in functional tubal blockage and embryo arrest in the fallopian tube. In humans, steroid hormones alter tubal function and contractility, thus affecting tubal peristalsis. There remains controversy as to whether ovulation-inducing agents, including clomiphene citrate, increase ectopic pregnancy rates, but it will be difficult to separate out the impact of the therapeutic agent from occult tubal disease.

Embryo Transfer

Knutzen and associates injected 40 µL of radiopaque fluid in mock embryo transfers and found that the material entered the tubes either partially or totally in 44% of subjects, suggesting that misplacement of embryos into the fallopian tubes leads to ectopic pregnancy. Embryo catheter placement also was implicated in the increased risk of ectopic pregnancies, which occurred more frequently in patients who underwent deep fundal transfer versus midcavity placement. Although transfer techniques may increase the chances of embryos reaching the fallopian tubes, it is the tubal pathology preventing the embryos from moving back into the uterus and resulting in an ectopic pregnancy.

Heterotopic Pregnancy

Heterotopic pregnancies occur in 1% to 3% of pregnancies following ART procedures and are usually diagnosed incidentally on routine follow-up ultrasonographic studies. This increased prevalence of heterotopic pregnancies following ART may be related to ovarian hyperstimulation and multiple ovum development. Of 111 reported heterotopic pregnancies following ART, 88.3% were tubal, 6.3% cornual, 2.7% abdominal, 1.8% cervical, and 0.9% ovarian.

Evidence-Based Recommendation

Heterotopic and extratubal ectopic pregnancies are more frequent following ART than with natural cycles. Salpingectomy or proximal tubal occlusion of a preexisting hydrosalpinx prior to IVF helps to prevent tubal ectopic pregnancies while increasing pregnancy rates following ART. (Strength of recommendation: B.)

Expectant Management

Ectopic pregnancies may resolve spontaneously. In a cavalier experiment in 1955, Lund hospitalized 119 women with ectopic pregnancy for observation. All were at least 6 weeks gestation. Some required multiple blood transfusions, and many were hemodynamically unstable. However, 68 resolved without surgery being required. Twelve additional studies reported in the literature since Lund's study found similar results (Table 3). Of the ectopic pregnancies, 67.2% resolved without surgery. Thus, both conservative medical and surgical therapy overtreats at least 50% of women with ectopic pregnancy. Falling β-hCG levels under 1,000 mIU/mL have been followed with conservative expectant management. Although patients with an equivocal diagnosis of ectopic pregnancy may be treated in this fashion, there are no data to support expectant management in clinical practice. In addition, despite close follow-up and even in the context of declining β-hCG levels, tubal rupture may still occur.

Evidence-Based Recommendation

Expectant management of ectopic pregnancy may be considered an appropriate conservative therapy for some patients with low initial (1,000 mIU/mL) and falling β-hCG levels. Both clinicians and patients need to be aware of the potential risks of choosing expectant management over proven therapies.

Cost Analysis

The last estimated U.S. costs for ectopic pregnancy are over 15 years old. In 1990, total costs for ectopic pregnancies were estimated to be $1.1 billion. Direct costs, expenditures for health care, accounted for 77% of the total costs, and the remainder were incurred as a result of lost wages or household responsibilities not performed due to illness (indirect costs). Direct costs from hospital charges were estimated at $6,079 per case, with hospital accommodations (mean length of stay, 3.47 days) and operating room charges accounting for the majority of the hospital expense, 36% and 40%, respectively. An additional $3,254 for professional fees increased inpatient charges to $9,333, and $149 for postoperative follow-up visits increased the total direct cost to $9,482 per case. Indirect costs for a 28-day disability were estimated at $250.5 million, 67% as a result of lost wages and the remainder from lost household duties. These costs are likely substantially higher today.

European studies by Mol and colleagues have attempted to evaluate costs, but it is important to remember that they figured in longer hospital stays and more sick days than are customary in the United States and that the costs they estimated are within a socialized medical system. A study undertaken to compare the costs of systemic methotrexate with surgery concluded that there would be a reduction in overall costs if patients were treated without confirmatory laparoscopy when β-hCG levels were below 3,000 mIU/mL; otherwise, there was not a substantial cost saving over surgery. Because a confirmatory laparoscopy is no longer required for diagnosis, the lower cost for medical therapy is more realistic. Compared with the cost of a laparoscopic salpingostomy, methotrexate results in an estimated 20% decrease in the cost of treatment.

A decision analysis by Morlock and associates created a model to estimate the costs incurred by treating ectopic pregnancy by methotrexate or by laparoscopic salpingostomy. They felt that such an analysis was important because although previous studies found cost advantages with methotrexate, they had not adequately considered failure of ectopic resolution after only one dose of methotrexate or the potential side effects and complications of methotrexate use. They also felt that several European studies had calculated the costs of inpatient laparoscopy, not currently standard care in the United States where treatment is routinely done in the outpatient setting. Incorporating all of the assumptions about failed methotrexate treatment and costs of surgery and hospitalization for medical failures, the authors found a $3,011 cost saving with methotrexate treatment compared with laparoscopy. Even when they altered the model by assuming the lowest resolution rate for methotrexate-treated ectopic pregnancy of 57% and highest complication rates, the model still supported the use of methotrexate with a saving of $760.

Finally, it should be noted that Ailawadi and colleagues performed a decision analysis comparing the complicating rate and cost of diagnosis ectopic pregnancy with evaluation of the uterus prior to medical management versus presumptively treating women with a presumed ectopic pregnancy with methotrexate without confirming the diagnosis with a dilation and curettage (D&C). Surprisingly, the outcomes were quite similar. Thus, there is no advantage to taking the "shortcut"of treating women presumed to but not confirmed to have an ectopic pregnancy in terms of cost and/or complications. Data supporting the definitive diagnosis was that there were fewer visits required by patients after performance of an evacuation of the cavity, as fewer women needed medical management and evaluation of serial β-hCG concentration. Moreover, a more accurate prognosis can be given to a woman regarding recurrence of miscarriage, ectopic pregnancy, or overall fecundity if a miscarriage is accurately distinguished from an ectopic pregnancy.

Evidence-Based Recommendation

Systemic methotrexate for unruptured ectopic pregnancy is less expensive than surgery, and direct costs are decreased substantially with methotrexate therapy. In addition to its cost effectiveness, systemic methotrexate does not subject patients to the risks of surgery. This cost benefit, however, diminishes with higher β-hCG titers and even disappears with levels greater than 3,000 mIU/mL because of single-dose methotrexate treatment failures and increased complications. (Strength of recommendation: B.)

Rare Types of Ectopic Pregnancy

Abdominal Pregnancy

The incidence of abdominal pregnancy is estimated at 1 in 8,000 births and represents 1.4% of all ectopic pregnancies. The prognosis is poor, with an estimated maternal mortality rate of 5.1 per 1,000 cases. The risk of dying from an abdominal pregnancy is 7.7 times higher than from other forms of ectopic pregnancy. The high rate of morbidity and mortality from abdominal pregnancy often results from a delay in diagnosis.

Abdominal pregnancies can be categorized as primary or secondary. These ectopic pregnancies may become apparent anywhere throughout gestation, from the first trimester to fetal viability. Symptoms may vary from those considered normal for pregnancy to severe abdominal pain, intra-abdominal hemorrhage, and hemodynamic instability. Primary abdominal pregnancies are rare and are thought to occur as a result of primary peritoneal implantation. They usually abort early in the first trimester due to hemorrhagic disruption of the implantation site and hemoperitoneum. Secondary abdominal pregnancies occur with reimplantation after a partial tubal abortion or intraligamentary extension following tubal rupture. Historical criteria to distinguish between primary and secondary abdominal pregnancies are moot, because treatment is guided by the clinical picture.

Ultrasonography is the diagnostic tool of choice and usually can identify the empty uterus along with the extrauterine products of conception. If the fetus is near viability, hospitalization is recommended. If time permits, bowel preparation, administration of prophylactic antibiotics, and adequate blood replacement should be made available prior to an operative delivery. Unless the placenta is implanted on major vessels or vital structures, it should be removed. Although complications may occur, including sepsis, abscess formation, secondary hemorrhage, intestinal obstruction, wound dehiscence, amniotic fluid cyst formation, hypofibrinogenemia, and preeclampsia, the placenta can be left in place to prevent further hemorrhage at the time of surgery. In contrast to the typical tubal ectopic pregnancy, methotrexate is unlikely to accelerate retained placental absorption because the trophoblastic cells are no longer actively dividing.

Ovarian Pregnancy

Ovarian pregnancy, the most common form of abdominal pregnancy, is rare, accounting for less than 3% of all ectopic gestations. Clinical findings are similar to those of tubal ectopic gestations: abdominal pain, amenorrhea, and abnormal vaginal bleeding. In addition, hemodynamic instability as a result of rupture occurs in 30% of patients. Women with ovarian pregnancies are usually young and multiparous, but the factors leading to ovarian pregnancies are not clear.

The diagnosis usually is made by the pathologist because many ovarian pregnancies are mistaken for a ruptured corpus luteum or other ovarian tumors. Only 28% of cases were diagnosed correctly at time of laparotomy. The recommended treatment is cystectomy, wedge resection, or oophorectomy during laparotomy, although laparoscopic removal has been successful.

Cornual Pregnancy

Cornual or interstitial pregnancy accounts for 4.7% of ectopic gestations and carries a 2.2% maternal mortality. Clinically, a pregnancy implanted at this site where the fallopian tube is traversing the muscular wall of the uterus is seen as a swelling lateral to the round ligament. Almost all cases are diagnosed after the patient is symptomatic. The most frequent symptoms are menstrual aberration, abdominal pain, abnormal vaginal bleeding, and shock, resulting from the brisk hemorrhage associated with uterine rupture. Due to myometrial distensibility, rupture is usually delayed, occurring at 9 to 12 weeks gestation.

A unique risk factor for interstitial pregnancy is previous salpingectomy, present in about 25% of patients. Only a high index of suspicion and repeated ultrasonographic examination with Doppler flow studies allows early diagnosis. With a timely early diagnosis, alternatives to the traditional cornual resection during laparotomy have been performed successfully. These include laparoscopic cornual resection, systemic methotrexate administration, local injection of methotrexate, potassium chloride injection, and removal by hysteroscopy. Regardless of the initial treatment attempted, if uncontrolled hemorrhage occurs, immediate laparotomy with uterine repair or hysterectomy is warranted to stop the blood loss.

Cervical Pregnancy

The incidence of cervical pregnancy ranges from 1 in 2,500 to 1 in 12,422 pregnancies. The most common predisposing factor is a prior D&C, present in 68.6% of patients. Interestingly, 31% of these were performed for termination of pregnancy. Other predisposing factors implicated in cervical pregnancies are previous cesarean delivery and IVF.

The most common initial symptom of cervical pregnancy is painless vaginal bleeding. These extrauterine pregnancies are usually diagnosed incidentally during routine ultrasonography or at the time of surgery for a suspected abortion in progress. In reported cases, 91% of patients sought treatment for vaginal bleeding, and 29.2% had massive bleeding. Not surprisingly, abdominal pain occurred with vaginal bleeding in only 25.8% of cases. The cervix is usually enlarged, globular, or distended. On occasion, it appears cyanotic, hyperemic, and soft in consistency. Sonography and magnetic resonance imaging have improved diagnosis of cervical pregnancy. Up to 81.8% of patients have been diagnosed correctly with ultrasonographic identification of the gestational sac in the cervix below a closed internal cervical os, with trophoblastic invasion into the endocervical tissue.

When the patient is hemodynamically stable, conservative therapy commonly is employed. There are no large studies-only several case series for clinical guidance. These have shown that use of methotrexate, local prostaglanding or hyperosmolar glucose injection, curettage, or a combination of these methods have been successful. Prior to curettage, uterine artery embolization minimizes the substantial risk of postevacuation hemorrhage. Systemic and local treatment with various agents carries an overall success rate of 81.3%. Unfortunately, massive hemorrhage may occur despite conservative measures, and hysterectomy may be the only lifesaving option.

Heterotopic Pregnancy

Heterotopic pregnancy is the coexistence of an intrauterine and ectopic gestation. In 1948, the spontaneous heterotopic pregnancy rate was calculated as 1 in 30,000 pregnancies, based on an ectopic pregnancy incidence of 0.37% and dizygous twinning rate of 0.8%. In the 1980s, the calculation rose to 1 in 10,000 due to an increased ectopic pregnancy rate. Today, heterotopic pregnancies occur in 1 in 3,889 to 1 in 6,778 pregnancies in the general population. Out of almost 133,000 pregnancies reported in the U.S. ART Registry between 1999 and 2002, 207 heterotopic pregnancies were reported, an incidence of about 1:640. In a review of 66 heterotopic pregnancies by Reece and associates, 93.9% were tubal and 6.1% ovarian.

Simultaneous existence of intra- and extrauterine pregnancies poses several diagnostic pitfalls. Heterotopic pregnancies are diagnosed in most cases after clinical signs and symptoms develop, and 50% of patients are admitted for emergency surgery following rupture. The delay in diagnosis is secondary to the finding of an intrauterine pregnancy, which provides false reassurance of absence of pathology, with the assumption that any symptoms will be self-limited.

Similar to tubal ectopic pregnancies, the most common complaint is lower abdominal pain. Routine ultrasonography detects only about 50% of tubal heterotopic pregnancies, and the remainder are diagnosed during laparoscopy or laparotomy when patients become symptomatic. Serial levels of the β-hCG are not helpful due to the effect of the intrauterine pregnancy.

If patients are hemodynamically unstable, exploratory laparotomy is warranted. If the diagnosis is suspected or the patient is symptomatic but hemodynamically stable, laparoscopy can be performed. Expectant management is not recommended, because β-hCG levels cannot be monitored adequately and the course of the ectopic cannot be followed. Systemic methotrexate is contraindicated if a viable intrauterine pregnancy is present and desired. Local injection of methotrexate with potassium chloride has been noted as successful in a small case series.

Summary Points

  • In most circumstances, ectopic pregnancy can be diagnosed before symptoms develop and treated definitively with few complications.
  • Quantitative β-hCG testing, ultrasonography, and curettage allow early diagnosis of ectopic pregnancy and use of medical therapy as the initial therapy option.
  • Conservative surgical therapy and medical therapy for ectopic pregnancy are comparable in terms of success rates and subsequent fertility. Medical therapy is the preferred choice because of the freedom from surgical complications and lower cost.
  • Surgery is the treatment of choice for hemorrhage, medical failures, neglected cases, and when medical therapy is contraindicated.
  • Multiple-dose methotrexate is preferable to single-dose methotrexate, direct injection, or tubal cannulation and is the first choice for unruptured, uncomplicated ectopic pregnancy.
  • Laparoscopic salpingostomy or salpingectomy is favored for cases of intra-abdominal hemorrhage, medical failure, neglected cases, and complex cases when medical therapy is contraindicated.
  • Prophylactic postoperative systemic methotrexate (a single dose) can prevent virtually all cases of persistent ectopic pregnancy following salpingostomy.
  • Salpingectomy prior to IVF decreases ectopic pregnancy incidence while increasing pregnancy rates in select patients with preexisting tubal disease.