Aust. NZ J Ohstet Gynciecol 1998; 38: 2: 151

Heterotopic Pregnancy Complicating In Vitro Fertilization Neil Johnson: Penelope McComb? and Guy Gudex‘ National Women’s Hospital, Auckland, New Zeulund

Summary: A review was undertaken of the cases of heterotopic pregnancy resulting from in vitro fertilization/embryo transfer (IVF/ET) and frozen embryo replacement (FER) in a 6-year cohort of women at National Women’s Hospital in Auckland. The incidence of heterotopic pregnancy was 2.9% (5 cases) in 173 clinical pregnancies resulting from 901 embryo replacements. Of the 5 women with heterotopic pregnancy, 1 had unilateral tubal patency and 4 had bilateral tubal blockage; 3 had ‘high responder’ peak serum oestradiol levels (greater than 9,000 pmol/L) prior to oocyte pick-up (OPU); 3 had a serum human chorionic gonadotrophin beta subunit (beta-HCG) level greater than 600 IU/L on Day 14 following embryo transfer (ET) in the absence of a multiple intrauterine gestation on subsequent ultrasound scan. In the 4 women in whom unequivocal diagnosis of heterotopic pregnancy was not made on the initial ultrasound scan, there was delay in appropriate management, in 1 for more than 5 months. In conclusion, early IVF pregnancies require a transvaginal ultrasound scan performed by a sonographer experienced in the diagnosis of ectopic pregnancy and management of early pregnancy complications by clinicians in close consultation with the IVF centre itself. No single risk factor, laboratory test or combination of these is sensitive or specific enough to predict the occurrence of heterotopic pregnancy. The first-line surgical treatment of heterotopic pregnancy should be laparoscopic salpingectomy with excision of all except the intramural portion of the affected Fallopian tube.

Heterotopic pregnancy is the coexistence of an intrauterine pregnancy and an ectopic pregnancy. The condition usually requires surgical treatment of the ectopic gestation but frequently the intrauterine gestation results in a successful outcome with the birth of a baby.

Historically heterotopic pregnancy was rarely reported in the literature with an estimated incidence around 1 in 30,000 pregnancies (1). In the past 2 decades, with the widespread use of assisted reproductive techniques (ART), it has become clear that it is more common. In 1983 Reece found a frequency of 1 in 7,963 (2) and an incidence as high as 1 in 2,600 pregnancies resulting from ovulation induction was reported in Columbus, Ohio (3). In vitro fertilization (IVF) (4-6) and gamete intra- Fallopian transfer (GIFT) (5,6) are associated with a high incidence of heterotopic pregnancy.

This case series audit was designed to review cases of heterotopic pregnancy resulting from IVF in our

1 . Senior Registrar in Fertility. 2 . Embryologist. 3. Clinical DirectorferriliqPLUS and Consultant in Obstetrics and

Gynaecology. Address for correspondence: Dr Neil Johnson, 4 St. Aidan’s Crescent, Crossgate Moor, Durham DHI 4AP, England.

unit, to assess the factors useful in establishing an early diagnosis and to examine how the management of heterotopic pregnancy could be improved.

METHODS Computer records were studied of the outcome of

all pregnancies resulting from couples undergoing in vitro fertilizatiodembryo transfer (IVFET) cycles (total 684) and frozen embryo replacement (FER) cycles following embryo cryopreservation consequent upon previous IVF (total 217) during the years 1991 - 1996. A clinical pregnancy was defined as a positive pregnancy test in association with an ultrasonically detectable gestation sac (or histo- logical confirmation of trophoblastic tissue). The outcome of 173 clinical pregnancies from these cycles is summarized in table I .

The hospital records of those women recorded as having heterotopic pregnancies were also searched and 5 women were confirmed to have had heterotopic pregnancies. All 5 of these women had undergone IVFET using their husband’s or partner’s semen and all had embryo transfer (ET) 48 hours following oocyte retrieval (OPU - oocyte ‘pick-up’). Ovarian stimulation was carried out using either urofolli- trophin or human menopausal gonadotrophins following downregulation with a gonadorelin analogue. All 5 of the women were nulliparous and all had tubal factor implicated in their subfertility - 1 had

1 52

Table 1. Outcome of Clinical Pregnancies Resulting from IVF at National Women's Hospital, 1991-1996

Number Number resulting resulting

from from FER IVFET cycles cycles

(n=684) (n=217)

Singleton livebirth pregnancy 80 22 Multiple livebirth pregnancy 18 2

Ectopic pregnancy 6 1 Heterotopic pregnancy 5 0 Induced abortion 1 0 Spontaneous abortion 6 3 Missed abortion 10 I Empty sac 12 2

Fetal death 4 0

Total number

from IVFET + FER cycles

(n=901)

I02 20 4 7 5 1 9

11 14

Total clinical pregnancies 142 31 173 IVF/ET=in vitro fertilizatiodembryo transfer: FER=frozen embryo transfer

undergone left and right salpingectomies for previous left and right tubal ectopic pregnancies; 1 had ectopic gestations previously in both Fallopian tubes and had the tubes conserved at surgery although both were blocked; the other 3 had tubal damage from proven or suspected previous pelvic inflammatory disease. The characteristics of each of these women's cycles are recorded in table 2. In these cases, the diagnostic definition for the intrauterine pregnancy was based on ultrasound in addition to histology or the delivery of a baby; the diagnostic definition for ectopic pregnancy was based on surgical visualization and histological confirmation.

The management of some of these patients took place at centres other than our own since National Women's Hospital has been a referral centre for tertiary fertility services for the North Island and the northern part of the South Island of New Zealand.

RESULTS There were 173 clinical pregnancies resulting from

901 embryo replacements. There were 122 livebirths and 5 heterotopic pregnancies. The incidence of

heterotopic pregnancy in our series was 2.9 % . Heterotopic pregnancy occurred almost as frequently as tubal ectopic pregnancy (7 cases with an incidence 4.0%).

Table 2 details the characteristics of the IVF cycles of the 5 women whose pregnancies were complicated by heterotopic pregnancy. Table 3 details the course of the subsequent pregnancy in each case.

In Case 1 the woman was confirmed pregnant on Day 12 following ET by a serum human chorionic gonadotrophin beta subunit (beta HCG) estimation of 170 I U L and presented on Day 28 following ET with left iliac fossa aching. An ultrasound scan demonstrated 3 fetal hearts, 2 intrauterine with fetal poles of a size equivalent to a gestation of 6 weeks and 1 in association with a left adnexal mass. Diagnostic laparoscopy identified a left ampullary tubal ectopic with blood leaking from the fimbrial end of the tube. A left salpingectomy was performed at minilaparot- omy. The intrauterine twin gestation continued and an uncomplicated vaginal twin delivery occurred at a gestation of 34+ weeks following preterm prelabour rupture of the membranes.

In Case 2 there was delay in diagnosis of the ectopic pregnancy after confirmation of pregnancy by a Day 13 beta HCG of 565 IUL. The Day 18 beta HCG was the last estimation in the ensuing management as this took place at a hospital without easy access to a quantitative beta HCG assay at that time. A scan on Day 39 following ET revealed an intrauterine gestation of size equivalent to 8+ weeks and ovaries of 9cm (right) and 8cm (left) respectively, each with multiple follicles, consistent with mild ovarian hyperstimulation. The patient presented on Day 49 with abdominal pain and vaginal bleeding with clots. Ultrasound scan showed the fetal heart to be present with a gestation sac with a slightly irregular outline. Threatened miscarriage was diagnosed. Four hours later she passed bloodclot and fetal tissue vaginally (later confirmed histologically as products of conception). Thirty minutes after passing fetal tissue,

Table 2. Characteristics of IVF/ET Cycles of 5 Cases of Heterotopic Pregnancy Case No. Age Indication for IVF (and tubal status) Peak oestradiol (pmol/L) No. oocytes retrieved No. embryos transferred 1 26 Tubovarian adhesions (1 patient) 14,382 17 3 2 34 Two previous ectopic pregnancies 7,335 10 3

and salpingectomies - left complete, right partial (bilateral tubal blockage)

previous severe acute pelvic inflammatory disease (bilateral blockage)

with both tubes conserved (but bilateral tubal blockage)

blockage)

3 30 Bilateral hydrosalpinges from 10.445 8 3

4 35 Two previous ectopic pregnancies 1.626 1 1 2

5 3 1 Inoperable tubal disease (bilateral 19,188 12 3

IVFET=in vitro fertilization/embryo transfer

NEIL JOHNSON ET AL 1 5 3

Table 3. Outcome of Heterotopic Pregnancies at National Women’s Hospital 1991-1996 case No. Beta HCG level (IUL) Ultrasound scan result in

in relation to ET 1 Day 12 - 170

Day 16 - 958 Day 23 - 12,270

relation to ET Day 28 - 2 intrauterine fetal poles with fetal cardiac activity: a fetal cardiac pole with cardiac activity within a left adnexal mass

2

3

A

5

Day 13 - 565 Day 18 - 1,630

Day 39 - 8t week size fetal pole with cardiac activity Day 49 - fetal heart present, irregular-shaped gestation sac Day 50 - empty uterus, ovaries 7cm bilaterally with multiple follicles, fluid in pouch of Douglas and uterovesical pouch

Day 12 - 546 Day 17 - 718 Day 20 - 1,144 Day 28 - 12,500

Day 12 - 110 Day 23 - 24,600 Day 30 - 54,800 Day 33 - 56,600

Day 14 - 755 Day 21 - 13,762

Day 36 - empty intrauterine gestation sac Day 40 - empty sac (as above) and left adnexal mass

Day 24 - Single intrauterine gestation, free intraperitoneal fluid, a possible yolk sac inferomedial to right ovary with surrounding tissue ‘reaction’ - ?ectopic Day 33 - intrauterine fetal pole with fetal heart, a fetal pole with fetal heart inferomedial to right ovary, small amount free intraperitoneal fluid Day 19 - intrauterine gestation sac only Day 28 - transvaginal scan showed a gestation sac with yolk sac, fetal pole and cardiac activity; also an intrauterine gestation with fetal cardiac activity

Outcome of ectopic pregnancy Laparoscopy, mini laparotomy and left salpingectomy on Day 28; histology confirmed chorionic villi infiltrating the tube Laparotomy with right salpingectomy and excision of adnexal mass on Day 163

Outcome of intrauterine pregnancy Vaginal twin delivery at 3 4 t weeks following preterm prelabour rupture of the membranes.

Spontaneous miscarriage at gestation 9 weeks; histology confirmed products of conception

Laparoscopy, laparotomy and bilateral salpingectomy on Day 40; histology confirmed chorionic villi infiltrating the left Fallopian tube Laparotomy, partial right salpingectomy, bilateral Filschie clip application to proximal Fallopian tubes on Day 33; histology confirmed fetal parts and chorionic villi infiltrating the Fallopian tube

Day 28 - laparoscopy, laparotomy, right salpingectomy and Filscbie clip application to proximal left Fallopian tube

Evacuation of the uterus on Day 40; histology confirmed chorionic villi

Vaginal delivery at 41 weeks

Vaginal delivery at 40 weeks

ET=embryo transfer

she complained of abdominal and shoulder-tip pain, exhibited peritonism and required narcotic analgesia - peritonism secondary to ovarian cyst rupture was diagnosed. Ultrasound scan on Day 50 showed an empty uterus, ovaries of 7cm diameter bilaterally and some free fluid in the uterovesical pouch and the pouch of Douglas. Complete miscarriage was diagnosed. The haemoglobin value had dropped from 10.7 g/dL to 7.0 g/dL - she was transfused 3 units of blood and discharged 2 days later feeling much better with a haemoglobin value of 10.3 g/dL. The patient was readmitted on Day 58 following 24 hours of crdmpy abdominal pain with a severe exacerbation immediately prior to admission. There was peritonism and cervical excitation. Ultrasound scan revealed an empty uterus, right ovary l0cm diameter, left ovary 6cm diameter and no free fluid. The haemoglobin value was 9.9 g/dL. Ovarian hyperstimulation syndrome was diagnosed; the patient was hospitalized

for 14 days, given regular analgesics and discharged with oral iron therapy as the haemoglobin value was 9.2 g/dL. She was readmitted on Day 81 with persistent vaginal bleeding. Dilatation and curettage was performed (at which time a l0cm mass was noted in the pouch of Douglas) - a moderate amount of tissue was obtained and reported histologically as decidualized endometrium; the bleeding settled and she went home the next day; 163 days following ET the patient was admitted electively for a laparotomy for continuing pelvic pain in association with a persistent mass in the pouch of Douglas ultrasonically measuring 10 x 8 x 6cm and described as a complex mass with internal echoes. The findings at laparotomy were of a normal uterus and left ovary, the omentum was adherent to the pelvic peritoneum, a 7cm mass, originating from the remnant of the right Fallopian tube, was adherent to the posterior surface of the uterus, right broad ligament and ovary - it was

154 ACST. A x ) N.Z. JOVRNAI. OF OBSTETRICS AN) GYNAE(.OI.OC;Y

consistent with a surgical haemorrhagic cyst. The right adnexal mass was mobilized and the right tube and mass removed. Histology confirmed proximal Fallopian tube with intramural and interstitial fibrosis and non-specific inflammation. There was no convincing evidence of residual trophoblastic tissue within the mass but the overall appearances were suggestive of inflammation and reorganization following a tubal gestation.

In Case 3 the woman was confirmed pregnant by a Day 12 beta HCG of 546 IU/L and presented with left iliac fossa pain and vaginal bleeding on Day 40 following ET. An ultrasound scan 4 days previously had shown only an empty small intrauterine gestation sac. On Day 40 transvaginal ultrasound scan confirmed the empty intrauterine gestation sac with an irregular outline and now showed a left adnexal mass. Diagnostic laparoscopy confirmed a large left ampullary tubal ectopic pregnancy and a bilateral salpingectomy was performed (to prevent future risk of ectopic pregnancy at the patient’s request). An evacuation of the uterus was also performed at the patient’s request. Histology confirmed chorionic villi from both the uterine cavity and the Fallopian tube.

Following a confirmatory Day 12 beta HCG of 110 I U L the woman in Case 4 presented with abdominal pain and nausea on Day 24 following ET. Ultrasound scan confirmed an intrauterine gestation and raised the possibility of a coexisting right adnexal ectopic pregnancy. The differential diagnosis included heterotopic pregnancy and ovarian hyperstimulation syndrome. The patient was managed conservatively as an inpatient for 3 days, the symptoms settled and she was discharged. The pain recurred on Day 33 and rescan confirmed the likelihood of a heterotopic pregnancy. After laparoscopic confirmation, a laparotomy was performed with right salpingectomy and Filschie clip application at the comual end of the left Fallopian tube. Postoperative recovery was uncomplicated. The intrauterine pregnancy resulted in a vaginal delivery of a healthy baby at a gestation of 41 weeks.

In Case 5 the woman had a confirmatory Day 14 beta HCG of 755 IU/L and presented on Day 22 following ET with abdominal discomfort and bloating which was initially diagnosed as consistent with mild ovarian hyperstimulation in light of an ultrasound scan 3 days previously which had shown only an intrauterine gestation sac. The symptoms did not settle and rescan on Day 28 revealed an intrauterine gestation of a size equivalent to 6+ weeks and an ectopic gestation. Laparoscopy confirmed a right ampullary ectopic pregnancy; right salpingectomy and application of a Filschie clip to the comual end of the left Fallopian tube were performed by minilaparotomy. The intrauterine pregnancy resulted in a vaginal delivery of a healthy baby at 40 weeks.

DISCUSSION The 2.9% incidence of heterotopic pregnancy

complicating IVF pregnancy in our series is in keeping with the incidences reported elsewhere, ranging from 0.78% to 5% (7). Couples should be routinely counselled prior to an IVF/ET cycle of the risk of heterotopic pregnancy, in addition to the risk of tubal ectopic pregnancy, as heterotopic pregnancy was almost as common as tubal ectopic pregnancy following IVF/ET. General Practitioners and Obstetrician-Gynaecologists need to be aware of the possibility of heterotopic pregnancy in patients presenting with early pregnancy complications, especially if they conceived by assisted reproductive techniques.

Heterotopic pregnancy is an important condition resulting from IVF. Misdiagnosis or delayed diagnosis are still common in spite of the knowledge of its high incidence in the IVF population. There was delay in diagnosis in 4 out of the 5 cases in this series. One of the diagnostic criteria for ectopic pregnancy is the absence of an intrauterine gestation on ultrasound scan. In heterotopic pregnancy the sonographic presence of an intrauterine gestation does not exclude an ectopic gestation. Early IVF pregnancies require a transvaginal ultrasound scan performed by a sonographer experienced in the diagnosis of ectopic pregnancy.

Ovarian hyperstimulation syndrome was one of the initial differential diagnoses in 3 out of the 5 cases of heterotopic pregnancy. Whilst ovarian hyperstimulation is common following IVFET and often does present with abdominal pain, a high level of suspicion of heterotopic pregnancy should be main- tained in all cases of suspected ovarian hyperstimu- lation syndrome where pain is a significant feature, particularly if the peak oestradiol level prior to OPU has been too low to justify a diagnosis of hyperstimulation syndrome.

What are the consequences of misdiagnosis? In Case 2 the woman had a blood transfusion which may have been avoided. She also suffered ongoing pain related to a chronic ectopic pregnancy which only resolved after a laparotomy more than 5 months after the embryo replacement. This case also raises the possibility that missing the diagnosis of ectopic pregnancy could lead to the demise of the intrauterine pregnancy through the adverse effect of the intrauterine pregnancy brought about by intraperitoneal bleeding and inflammation caused by the coexisting tubal ectopic. Whilst there is no guarantee that diagnosis and surgical treatment of this patient on initial presentation at 8+ weeks would have resulted in a more successful outcome for the intrauterine gestation, it is interesting to note that after visualization of the fetal heart, the likelihood of spontaneous miscarriage in previously infertile women under 36 years old is very low at around 2% (8). Coexisting intrauterine pregnancies in heterotopic

NFIL JOHSSON t T A L 155

gestation resulting from IVF are ‘precious pregnancies’ in couples who usually have a long history of subfertility. The management of the woman in Case 2 underlines the necessity for early IVF pregnancy complications being managed in close consultation with clinicians in the IVF centre.

Unilateral tubal patency has been identified as a risk factor for heterotopic pregnancy following IVF resulting in a 3-fold increased incidence compared to bilateral patency or bilateral blockage (7). However, in common with Dimitry et a1 (4), more women in our series had previous evidence of bilateral tubal blockage (4) rather than unilateral tubal patency (1). Two of these women had previously had ectopic pregnancies affecting both Fallopian tubes. In Case 4 left and right salpingostomies had previously been performed. In Case 2 the woman had previously undergone left total salpingectomy and right partial salpingectomy - her ectopic pregnancy in this series occurred in the remnant of the right Fallopian tube. The authors believe there is no place for partial salpingectomy in the surgical treatment of tubal ectopic pregnancy (whether heterotopic or simply tubal ectopic) as this merely introduces the possibility of a future ectopic gestation in a redundant portion of the Fallopian tube. The place of salpingostomy for tubal ectopic pregnancy has now even been questioned - future pregnancy rates are only slightly reduced and recurrent ectopic pregnancy rates clearly reduced if laparoscopic salpingectomy is performed rather than laparoscopic salpingostomy (9). Laparoscopy should be considered as soon as heterotopic pregnancy is suspected. We advocate laparoscopic salpingectomy as the treatment of choice in heterotopic pregnancy where all except the intramural portion of the affected Fallopian tube is excised. Whilst no patient in this series had laparoscopic salpingectomy, the potential for reducing uterine manipulation and operating time by this approach (LO), thereby safeguarding the intrauterine gestation, and the documented cases of delivery of a healthy baby following laparoscopic surgery for heterotopic pregnancy (5,ll) support the first-line use of laparoscopic salpingectomy in heterotopic preg- nancy where the tubal pregnancy is unruptured.

A peak serum oestradiol level greater than 9,000 PmoK prior to OPU (a ‘higher responder’ level) has also been identified as a risk factor for heterotopic pregnancy (6). In our series 3 out of the 5 (Cases 1, 3, and 5 ) would fulfil this risk factor.

It has been suggested that elevated serum beta HCG levels on Day 10, 12 and 14 following ET in the absence of a subsequently confirmed intrauterine multiple gestation is another prognostic sign for a

heterotopic pregnancy (4). However the very wide normal range found in ongoing singleton intrauterine gestations makes interpretation of serum beta HCG levels in early pregnancy difficult. Guth et a1 (1995) (12) showed that if the beta HCG was greater than 600 I U L on Day 14 post-ET, all gestations were multiple. Extrapolating with a beta HCG doubling time of 1.6 days in early IVFRT pregnancies (13), only 3 of the 5 cases of heterotopic pregnancy in this series (Cases 2, 3, and 5 ) could be identified based on a beta HCG greater than 600 I U L on Day 14 post-ET in the absence of a multiple intrauterine gestation on subsequent early pregnancy ultrasound scan.

Acknowledgements We wish to thank the staff‘ of fertilityPLUS and

collaborating Consultants at other centres in New Zealand for their cooperation.

References 1. De Voe RW, Pratt JH. Simultaneous intra- and extrauterine

pregnancy. Am J Obstet Gynecol 1948; 15: 1191-1193. 2. Reece EA, Petrie RH, Sitmans MF, Finster M, Todd WD.

Combined intrauterine and extrauterine gestations: a review. Am J Obstet Gynecol 1983; 146: 323-330.

3. Richards SR, Stempel LE, Carlton BD. Heterotopic pregnancy: reappraisal of incidence. Am J Obstet Gynecol 1982: 142: 928- 930.

4. Dimitry ES, Margara R, Suback-Sharpe R, Winston R, Mills M. Nine cases of heterotopic pregnancy in 4 years or in vitro fertilization. Fertil Steril 1990; 53: 107-1 10.

5. Guirgis RR. Simultaneous intrauterine and ectopic pregnancies following in-vitro fertilization and gamete intra-fallopian transfer. Areview of nine cases. Hum Reprod 1990; 5: 484-486.

6.Molloy D, Hynes J, Deambrosis W, Hamson K, Keeping D, Hennessey J . Multiple-sited (heterotopic) pregnancy after in vitro fertilization and gamete intrafallopian transfer. Fertil Steril 1990;

7. Cohen J, Mayaux M, Guihard ML, Schwartz D. In vitro fertilization and embryo transfer: a collaborative study of I 1 63 pregnancies on the incidence and risk factors of ectopic pregnancies. Hum Reprod 1986; 1: 255-258.

8. Smith KE, Buyalos RP The profound impact of patient age on pregnancy outcome after early detection of fetal cardiac activity. Fertil Steril 1996; 65: 35-40.

9. Mol BWJ, Hajenius PJ, Engelsbel S et al. Is conservative surgery for tubal pregnancy preferable to salpingectomy ? An economic analysis. Br J Obstet Gynaecol 1997; 104: 834-839.

10. Johnson NP, Gillett WR. Assessment of the laparoscopic treatment of ectopic pregnancy. NZ Med J 1995; 108: 125-127.

11. Parker J, Watluns W, Robinson H, Byme D. Laparoscopic adnexal surgery during pregnancy: a case of heterotopic tubal pregnancy treated by laparoscopic salpingectomy. Aust NZ J Obstet Gynaecol 1995; 35: 208-210.

12. Guth B, Hudelson J , Higbie J et al. Predictive value of hCG level 14 days after embryo transfer. J Assist Reprod Genet 1995; 12: 13-14.

13. Zegers-Hochschild F, Altieri E, Fabres C, Fernandez E, Mackenna A, Orihuela P. Predictive value of human chorionic gonadotrophin in the outcome of early pregnancy after in-vitro fertilization and spontaneous conception. Hum Reprod 1994; 9: 1550-1555.

53: 1068-1071.