The conservative management of interstitial pregnancy


Dr K. Jermy, Gynaecological Ultrasound and Minimal Access Surgery Unit, Department of Obstetrics and Gynaecology, St George's Hospital Medical School, Cranmer Terrace, London SW17 0RE, UK.


Objectives  To evaluate the effectiveness of systemic methotrexate in the treatment of interstitial pregnancy.

Design  Prospective observational study.

Setting  An Early Pregnancy Assessment Unit in a London Teaching Hospital.

Sample  Twenty consecutive women diagnosed with an interstitial pregnancy.

Methods  Women were diagnosed with an interstitial pregnancy based on transvaginal ultrasound findings. Single dose, intramuscular methotrexate was administered on day 0. A second dose of methotrexate was given if the β-hCG levels had not fallen by 15% between days four and seven. Weekly follow up continued until the serum β-hCG < 5 IU.

Main outcome measure  The resolution of serum β-hCG levels without the need for surgical intervention.

Results  Two hundred and ninety-three ectopic gestations were diagnosed over a 42-month period. Twenty of these were interstitial in nature, with a median initial serum β-hCG of 6452 IU. Of the 20 interstitial pregnancies, 17 cases received systemic methotrexate. Sixteen were treated successfully (94%), including all of the four cases with fetal heart activity present. A second methotrexate dose was given to six patients. Two cases were managed expectantly. Two cases underwent laparotomy and cornual resection: one elected for surgical management at the outset and one as a result of suspected ectopic rupture after two doses of methotrexate. There were no other complications.

Conclusions  Systemic methotrexate is a safe and highly effective treatment for interstitial pregnancy. Surgery can be avoided in the majority of women with this condition. Early recognition of the cornual pregnancy with transvaginal ultrasound is essential.


The incidence of ectopic pregnancy continues to increase and accounts for approximately 2% of reported pregnancies. Interstitial pregnancy complicates between 1% and 6% of all ectopic gestations.1,2 Historically, the location of such pregnancies within the wall of the uterus, surrounded by a thin rim of myometrium, has led to their late presentation, often with catastrophic complications such as uterine rupture and subsequent high morbidity and mortality.3 The establishment of early pregnancy assessment units, utilising high resolution transvaginal probes and with rapid access to quantitative β-hCG (human chorionic gonadotrophin) testing, has facilitated the early detection of the majority of ectopic gestations prior to tubal rupture. This, in turn, has meant that more conservative methods of treatment have been incorporated into the management options of those women diagnosed with an ectopic pregnancy without haemodynamic compromise. Appropriate case selection, often based on scoring systems,4 means that medical therapy—usually in the form of the antimetabolite chemotherapeutic agent methotrexate—and occasionally expectant management, can be safe and effective in managing an unruptured tubal ectopic pregnancy.4–6

However, there still remains a reluctance to use more conservative methods in the management of interstitial pregnancies, and their non-surgical management remains the subject of case reports. Conservative management options include the local injection of potassium chloride, the use of minimally invasive surgical techniques7,8 and the administration of methotrexate. This can be given either locally into the gestation9 or systemically, either as a single intramuscular dose or as multiple doses with folinic acid rescue.10

Conventional management typically results in cornual resection or hysterectomy at the time of laparotomy. This not only has an immediate impact on patient morbidity, but the consequences on subsequent term pregnancies are not clear and the risk of uterine rupture remains.

We have evaluated the management of 20 interstitial pregnancies in our unit over the last three and a half years, in an attempt to unify the management strategies in this uncommon form of ectopic pregnancy. We aim to conclude that systemic methotrexate provides a safe, effective method of treating interstitial pregnancy, and that the use of laparotomy with hysterectomy or cornual resection should be reserved for those patients presenting late with a ruptured interstitial pregnancy in whom there is haemodynamic compromise.


Between September 1998 and March 2002, 20 women were diagnosed with an interstitial pregnancy. All of the women presented to the early pregnancy assessment unit, with a positive urine β-hCG, either by self-referral, or from their general practitioner. All underwent a transvaginal ultrasound scan, using a 5-MHz transducer for B mode imaging (Aloka SSD 2000, Aloka, Japan). The diagnosis of interstitial pregnancy was made based on the following ultrasound criteria:

  • A regular endometrial echo, with no visible gestation present within it.
  • Products of conception located outside the endometrial echo, surrounded by a continuous rim of myometrium, within the interstitial area (Figs 1 and 2).
Figure 1.

Line drawing and transvaginal ultrasound image of a right cornual ectopic pregnancy at seven weeks amenorrhoea. The endometrial cavity is slightly distended with blood.

Figure 2.

Cornual ectopic pregnancy. Colour Doppler imaging demonstrating presumed intervillous flow.

The presence of free fluid within the pelvis and coexistent adnexal pathology was noted. All patients were haemodynamically stable, with normal liver and renal function and a normal full blood count, with no contraindications to methotrexate therapy. All but two of the patients were admitted to the ward and initially treated as inpatients.

The methotrexate was administered intramuscularly according to a single dose protocol, advocated by Stovall and Ling5 in the treatment of tubal ectopic pregnancy (Table 1).

Table 1.  Single dose intramuscular methotrexate protocol for treatment of unruptured ectopic pregnancy.
  1. β-hCG = human chorionic gonadotrophin; FBC = full blood count; G&S = blood group and serum save; U&E = urea and electrolytes.

0β-hCG, FBC, G&S, LFT, U&E
1β-hCG, intramuscular methotrexate: 50 mg/m2
7β-hCG, FBC, LFT
 2nd methotrexate injection, if β-hCG decrease is less than 15% from days 4 to 7

Serum was taken on days zero, four and seven (unless not feasible) in order to measure serum β-hCG, liver function tests and a full blood count. If the β-hCG levels had not fallen by 15% between days four and seven, and the patient remained systemically well, a second dose of intramuscular methotrexate was administered. No folinic acid rescue was given using this regime. Patients were advised to avoid a subsequent pregnancy within three months of methotrexate administration.

Patient follow up comprised of weekly serum β-hCG levels until levels were no longer detectable, and if a fetal heartbeat was present on the initial scan, daily ultrasound scans were performed until cardiac activity was no longer detected.

A successful outcome was defined as the resolution of serum β-hCG levels without the need for surgical intervention.


During the three and a half year study period, 293 ectopic gestations were diagnosed. Twenty of these were interstitial, with a prevalence of 6.8%. None were heterotopic in nature, and all of those women diagnosed with an interstitial pregnancy had conceived spontaneously.

Table 2 demonstrates the demography and sonographic findings of the group. The mean duration of amenorrhoea at presentation was 51.5 days and the median initial serum β-hCG was 6452 IU (range 32–31,381 IU)—excluding one patient who elected for primary surgical management. The diagnosis of interstitial pregnancy was made by ultrasound alone in 17 cases. The remaining three patients underwent a laparoscopy for suspected ectopic pregnancy, based on the finding of an empty uterus on ultrasonography, in conjunction with abnormally rising serum β-hCG values. Two of these patients (cases 2 and 15) also had a negative laparoscopy, and a subsequent transvaginal ultrasound confirmed an interstitial gestation. Four cases had demonstrable fetal cardiac activity on ultrasonography, and all of these were treated successfully with systemic methotrexate.

Table 2.  Demographics of the 20 patients.
CaseAgeAmenorrhoea (days)Presenting complaintsRisk factorsParitySize (mm)USS characteristicsInitial β-hCG (IU)Time until β-hCG undetectable (days)Treatment method
  1. USS = ultrasound; top = termination of pregnancy; PVB = vaginal bleeding; β-hCG = human chorionic gonadotrophin; mc = miscarriage; IU = international units; PID = pelvic inflammatory disease; MXT = methotrexate; IUCD = intrauterine contraceptive device; n/a = not available; ART = assisted reproductive techniques.

11852PVBNone0 + 1mc11 × 9 × 12Cardiac activity138837MXT-1 dose
21684PVBNone0n/aInhomogenous mass3225MXT-1 dose
32025PVBPID045 × 46 × 47Inhomogenous mass716855MXT-2 doses
43749Dating scanTubal disease1n/aEmpty uterus12,709n/aLaparotomy
53126PainIUCD in situ3 + 1mc27 × 26 × 28Cardiac activity703788MXT-2 doses
63449PVBNone226 × 26 × 27Cardiac activity327970MXT-1 dose
72963PVBPrior IUCD use1 + 3mc42 × 32 × 34Gestational sac31,38194MXT-1 dose
82656PVBIUCD in situ2 + 1top23 × 24 × 23Cardiac activity599556MXT-2 doses
93592PVB and painPrevious ART0 + 1top21 × 28 × 20Inhomogenous mass28556MXT-1 dose
103746PVBNone2 + 2mc33 × 33 × 33Inhomogenous mass88623MXT-1 dose
112670PVBSeptate uterus0 + 2mc21 × 25 × 24Gestational sac12515Expectant
123766Dating scanNone025 × 27 × 28Gestational sac215338MXT-1 dose
132533PVBNone016 × 14 × 18Embryonic pole17,398n/aLaparotomy
144159PVBNone1 + 1top36 × 38 × 37Embryonic pole6,52970MXT-2 doses
152839PVB and painNone016 × 15 × 13Embryonic pole892539MXT-2 doses
162651PVB and painNone1 + 2top29 × 30 × 30Empty uterus968461MXT-1 dose
173547PVBNone233 × 33 × 33Inhomogenous mass97427MXT-1 dose
182752PVB and painNone137 × 36 × 38Embryonic pole18,14534MXT-2 doses
192758PVBNone043 × 52 × 54Inhomogenous mass58080MXT-1 dose
203622PVBNone026 × 21 × 25Inhomogenous mass690940Expectant

Overall, 16 patients were treated successfully with systemic methotrexate; six of these patients requiring a second dose due to plateauing of β-hCG levels. Two patients were managed surgically with laparotomy and resection of the ectopic gestation; one patient elected for primary surgical management, whereas a second patient underwent an emergency procedure for suspected cornual rupture following two doses of systemic methotrexate (laparotomy findings diagnosed an intact cornual ectopic gestation). Two patients were managed expectantly without medical or surgical intervention.

All those with an initial β-hCG value under 5000 IU were treated successfully with one dose of methotrexate. All but two of the cases successfully treated with systemic methotrexate, with initial β-hCG values of more than 5000 IU, required two doses. The exceptions were cases 7 and 16, with initial β-hCG values of 31,381 and 9684 IU, respectively. These cases required only a single dose of methotrexate for resolution.

The median duration of hospital stay was seven days (range 0–40 days).

There were no complications arising from the use of methotrexate. All had normal serum biochemistry and haematology indices post-methotrexate administration except one patient, who had a transient rise in liver transaminases on day seven. One patient underwent cornuostomy at emergency laparotomy for a suspected ruptured ectopic pregnancy, without further complication.

Long term ultrasound follow up was obtained on one patient, and resolution of the uterine defect at one year following methotrexate administration was demonstrated (Fig. 3).

Figure 3.

Residual defect one year following methotrexate administration. This ultrasound is from the same case as in Fig. 1.


Ectopic pregnancy remains a significant cause of maternal mortality, accounting for 13 of the 17 deaths in early pregnancy, in the last triennial report in the United Kingdom.11 The establishment of early pregnancy assessment units with transvaginal ultrasound facilities, rapid, sensitive β-hCG assays and a heightened awareness has facilitated the earlier detection of all types of ectopic pregnancy.

The recognition that conventional surgical management of interstitial pregnancy has the potential for significant consequences on morbidity and subsequent fecundity has led to the introduction of more conservative methods of management; paralleling the management of tubal ectopic pregnancies.

Our results compare favourably with other studies using non-surgical methods of treating interstitial pregnancies (Table 3). Of the 17 cases that received systemic methotrexate, 16 were treated successfully (94%). Many studies have focussed on the use of local injection of methotrexate into the interstitial gestation, either transvaginally or laparoscopically, and both provide highly effective methods of administration. One of the arguments for local administration of methotrexate has been its favourable side effect profile. This, however, has been based on comparisons with multiple systemic dose regimes. The effectiveness of single dose systemic methotrexate, and more importantly its side effect profile, was demonstrated by Stovall and Ling in 1993 in the treatment of tubal gestations. Of 120 patients receiving single dose systemic methotrexate, side effects were virtually eliminated, while maintaining an efficacy equal to that of multiple systemic dose regimes. A systemic route of administration offers advantages over local injection of the ectopic gestation in that it is less invasive and is operator independent. Heterotopic pregnancy continues to remain an indication for local injection of potassium chloride.

Table 3.  Summary of methotrexate use in interstitial pregnancy: a review of the literature.
a. Systemic (intramuscular) methotrexate administration
Author (ref)Methotrexate regimeFolinic acid rescue?No. of casesInitial β-hCG (range in IU)Success rates*Notes
Fernandez et al.215 mg/day for 5 days or 1 mg/kg/day for 4 daysYes4n/a75% 
Karsdorp et al.121 mg/kg/day for 4 daysYes5**1500–16,800100%3 patients required 3 courses and 1 patient required 2 courses
Benifla et al.1315 mg/day for 5 days or 1 mg/kg/day for 4 daysNo6**364–43,00066% 
Hajenius et al.141 mg/kg/day for 4 daysYes8410–81,000100%3 patients required a 2nd course
Hafner et al.92 doses of 1 mg/kg 48 hours apartYes5793–41,15080% 
Present studySingle dose: 50 mg/m2No1732–31,38194%6 cases required a 2nd dose
Author (ref)Route of administrationFolinic acid rescue?No. of casesInitial β-hCG (range in IU)Success ratesNotes
  1. *Defined as the complete resolution of serum β-hCG levels, without the need for further surgical intervention.

  2. **One case received a local injection of methotrexate also.

Fernandez et al.2TransvaginalYes2n/a50%1 case received systematic methotrexate also
Benifla et al.13Laparoscopic and transvaginalNo6360–10,000100% 
Hafner et al.9TransvaginalNo5102–16,400100% 
Timor-Tritsch et al.15TransvaginalNo7n/a100%1 case needed a laparoscopy for haemostasis

Follow up tends to be prolonged, as initial β-hCG values tend to be higher than those encountered with tubal ectopic pregnancy. While all those women presenting with initial β-hCG values of less than 5000 IU were treated successfully with single dose methotrexate, almost all women with an initial β-hCG of greater than 5000 IU required two doses. This study therefore helps us to appropriately counsel patients regarding the likelihood of needing a second dose of methotrexate to achieve a successful outcome. Alternatively, this latter group may benefit from alternative dose strategies from the outset, or local administration of methotrexate.

Initially, all our patients were admitted for follow up as inpatients, until the serum β-hCG values had returned to normal. This has substantial implications on cost (hospital bedspace, time off work) and psychologically (time away from family, institutionalisation). These figures may be offset by managing women on an outpatient basis. As the study progressed, the need for hospitalisation was reviewed and several cases were managed exclusively as outpatients. This requires a dedicated early pregnancy assessment unit, allowing for effective follow up of serial β-hCG testing and as an access point for patient and doctor queries.

In the haemodynamically stable patient, the management options in those women diagnosed with an interstitial pregnancy will depend upon the skills and available services of the unit. The important lessons are

  • 1Early recognition of the interstitial pregnancy with transvaginal ultrasonography is essential, as only with early detection can conservative methods of management be maximised. The finding of an ‘empty uterus’ on scan should always prompt evaluation, not only of the adnexae, but also of the cornual region and the cervix.
  • 2Conservative methods, such as the use of methotrexate (systemic or local), and expectant management should be employed whenever possible.
  • 3When the initial serum β-hCG is greater than 5000 IU, a second systemic dose of methotrexate is likely to be required for a successful outcome.
  • 4There was no symptomatic or biochemical evidence of toxicity using this low dose methotrexate regime, and folinic acid rescue was not required.

Our work reinforces that of other studies, showing that laparotomy with hysterectomy or cornual resection should not be the first line management of interstitial pregnancy within the haemodynamically stable patient, and we propose that single dose systemic methotrexate is a safe and highly effective management option.

Accepted 10 August 2004