a b s t r a c t

Objectives: A comparative examination of the clinical, laboratory, ultrasound findings, and operative characteris- tics of rare site located ectopic pregnancies.

Methods: Retrospective analysis of all department cases of rare site located ectopic pregnancies diagnosed and treated from December 2006 to December 2019.

Results: Thirty rare ectopic pregnancies were identified. Of these, 11 were ovarian, 10 were interstitial and 9 were tubal stump. The patients treated for ovarian pregnancy had significantly lower human chorionic gonadotropin (hCG) levels than patients treated for interstitial or stump pregnancies (2025 +- 1105 mIU/ml, 18,424 +- 2579 mIU/ml and 11,204 +- 9221 mIU/ml, respectively, p = 0.003). The main presenting symptom in patients with an ovarian pregnancy was abdominal pain (90.9%, 60.0% and 44.4%, respectively, p = 0.031). Signs of abdominal peritoneal irritation (i.e., rebound tenderness and guarding) were more frequent upon physical examination in patients with an ovarian pregnancy (72.2%, 30.0% and 22.2%, respectively, p = 0.044) who also exhibited the highest rates of syncope and hypovolemic shock upon admission compared to patients with an interstitial or stump pregnancy (54.5%, 10.0% and 11.1%, respectively, p = 0.031). Ovarian pregnancies were associated with the lowest Sonographic detection rates (9.1%, 80.0% and 100%, respectively, p = 0.0001) and with free fluid in the pouch of Douglas (72.7%, 20.0% and 22.2%, respectively, p = 0.02). Ovarian pregnancies experienced the highest rupture rate during surgery compared to interstitial or stump pregnancies (66.9%, 16.7% and 44.4%, re- spectively, p = 0.028), had a significantly higher estimated blood loss (1081 +- 647 ml, 760 +- 597 ml and 343 +- 318 ml, respectively, p = 0.003) and required blood transfusions in the perioperative period (63.6%, 20.0% and 11.1%, respectively, p = 0.025) significantly more often.

Conclusions: Ovarian pregnancy remains the most challenging diagnosis compared to interstitial and tubal stump

ectopic’s. Health care providers should recognize these rare site ectopic pregnancies and to handle these gyneco- logical emergencies promptly.

(C) 2021

1. Introduction

Over the last 20 years, there have been enormous changes in the di- agnosis and treatment of tubal pregnancies (the most common site of ectopic pregnancies) [1]. Today, early diagnosis has been made possible through the use of the sensitive human chorionic gonadotropin (hCG) test and the Transvaginal ultrasound. In many centers, most tubal preg- nancies can now be diagnosed before the rupture of the fallopian tube occurs, and only a few women require Urgent surgery and blood trans- fusions as a result of intra-abdominal bleeding and hemorrhagic shock [2]. Life-saving surgery has been replaced by procedures oriented to- wards the preservation of fertility and avoiding extensive surgery.

* Corresponding author at: Department of Obstetrics and Gynecology, Assaf Harofeh Medical Center, Zerifin 70300, Israel.

E-mail address: [email protected] (Y. Melcer).

Rare site ectopic pregnancies account for 0.4%-8.3% of all ectopic pregnancies [3-5] as compared to ovarian (0.5%-3.2%) pregnancies [3-5] and are defined as interstitial (2.4%) [3] and tubal stump ec- topic pregnancies (0.5%) [6]. Nevertheless, these events are still often misdiagnosed upon admission and sometimes are only correctly identi- fied after a rupture, which requires immediate surgery to avoid serious complications [4,5,7]. The failure to diagnose these cases may be due to a non-specific clinical presentation upon admission. The massive intra- abdominal bleeding may erroneously be attributed to the fragility and hypervascularity of the site of extra-tubal nidation [4,5]. Women with these pathologies primarily undergo surgery for the suspected rupture of a tubal Ectopic pregnancy or a hemorrhagic corpus luteum, so that the final diagnosis is only made intraoperatively [4,5].

There is scant research on the implantation sites of these ectopic pregnancies, at least in large population-based cohorts [3]. A few publi- cations that date back more than 30 years reported descriptive results

https://doi.org/10.1016/j.ajem.2021.11.025

0735-6757/(C) 2021

concerning the site of ectopic pregnancy in large samples either from specific geographic areas, [8-11] or from hospitals [12,13]. The remain- ing literature is composed of anecdotal case reports or small series, some of which include reviews of previous works. This points to the need for updated studies on the identification, diagnosis and treatment of rare site located ectopic pregnancies, including clinical, ultrasound and operative characteristics.

The current study aims were to report various aspects of rare site ec- topic pregnancies diagnosed over the course of 14 years, all in a single hospital setting.

1. Materials and methods

A computerized database search for all cases of ectopic pregnancies diagnosed and treated in our department was conducted from Decem- ber 2006 to December 2019. It was approved by the Institutional Ethics Committee (# 0344-19). Because of retrospective design of the current study informed consent was not required. Only cases of three rare ec- topic implantations were included; namely, ovarian, interstitial or stump pregnancies. Pregnancies in a cesarean section scar or ectopic tubal pregnancies were excluded.

The dataset included demographic characteristics, obstetric and gynecological history, clinical presentation, Sonographic findings and operative procedures. Blood loss was estimated as the existing hemoperitoneum at the time of the surgery in addition to the operative blood loss.

Three rare site ectopic pregnancy locations were recorded in the medical records: the ovary, the interstitial part of the tube, and the tubal stump after salpingectomy. The intra-operative findings for ovar- ian pregnancy and histopathological examination [4,5] met all four Spiegelberg criteria [14]: (1) an intact ipsilateral tube, clearly separate from the ovary; (2) a gestation occupying the normal position of the ovary; (3) a Gestational sac connected to the uterus by the utero- ovarian ligament; (4) ovarian tissue in the wall of the gestational sac. The following Sonographic criteria were used to define interstitial preg- nancy: (1) the uterine cavity was empty, (2) a gestational sac was lo- cated eccentrically near the superolateral aspect of the uterus, and

(3) there was a thin (<5 mm) myometrial layer surrounding the gesta- tional sac [15]. Data from the surgical history, the levels of hCG in the pa- tients’ blood and the transvaginal ultrasound [6] were all considered indications of a tubal stump pregnancy. All women underwent laparos- copy or laparotomy and the diagnoses were confirmed surgically.

SSPS software (SPSS Inc., version 24 Chicago, IL, USA) was imple-

mented for the statistical analyses. The descriptive parameters were expressed as the mean +- standard deviation or a percentage. Chi square tests or ANOVAs were used for statistical comparisons as appropriate.

A p value of <0.05 was considered statistically significant. This study was approved by the Institutional Review Board.

1. Results

During the study period (2006-2019), 117,422 deliveries and 1164 ectopic pregnancies (1.0%) were recorded in our institute. Thirty rare ectopic pregnancies were identified. Of these, as shown in the images in Fig. 1, 11 were ovarian, 10 were interstitial and 9 were tubal stump. Some of these cases have been previously reported by our group [4-6]. A comparison of the clinical characteristics of these rare ectopic pregnancies are presented in Table 1. There were no differences in the women’s age, gravidity, parity or pregnancies resulting from assisted re- productive technologies (ART) across the three rare tubal ectopic preg- nancy sites. Intrauterine device (IUD) use was more common among patients treated for ovarian pregnancy than among patients treated for interstitial or stump pregnancies (36.4%, 0% and 0%, respectively, p = 0.012). No significant differences were found for the mean hemo- globin levels on admission or for mean gestational age at the time of di- agnosis. However, the mean hCG level on admission was significantly lower among patients treated for ovarian pregnancy than among pa- tients treated for interstitial or stump pregnancies (2025 +- 1105 mIU/ ml, 18,424 +- 2579 mIU/ml and 11,204 +- 9221 mIU/ml, respectively,

p = 0.003).

As shown in Table 1, abdominal pain was the main presenting symp- tom in patients with an ovarian pregnancy as compared to patients with an interstitial or stump pregnancy (90.9%, 60.0% and 44.4%, respectively, p = 0.031). The physical exam revealed more frequent signs of perito- neal irritation (i.e., rebound tenderness and guarding) in patients with an ovarian pregnancy compared to patients with an interstitial or stump pregnancy (72.2%, 30.0% and 22.2%, respectively, p = 0.044). Patients with an ovarian pregnancy had the highest rates of syncope and hypovolemic shock upon admission compared to patients with an interstitial or stump pregnancy (54.5%, 10.0% and 11.1%, respectively,

p = 0.031).

In terms of ultrasound findings, ovarian pregnancies had the low- est sonographic detection rates compared to interstitial or stump pregnancies (9.1%, 80.0% and 100%, respectively, p = 0.0001). There was no difference in the rates of sonographic identification of embry- onic cardiac motion. However, free fluid in the pouch of Douglas was more common in patients with an ovarian pregnancy than in patients with interstitial or stump pregnancies (72.7%, 20.0% and 22.2%, re- spectively, p = 0.02). The pregnancy sac was detected significantly less often in patients with an ovarian pregnancy than in patients with an interstitial or stump pregnancy (9.1%, 70.0% and 55.6%, respectively, p = 0.013).

Fig. 1. Sonographic and laparoscopic images of ovarian and tubal stump pregnancy.

Table 1

Comparison of clinical characteristics, presentation, ultrasound and intraoperative findings for rare site ectopic pregnancies according to implantation site

	Ovarian	Interstitial	Stump	P value
	(n = 11)	(n = 10)	(n = 9)
Clinical characteristics Maternal age (years)	33.6 +- 4.0	33.6 +- 4.3	32.6 +- 2.9	0.785a
Pregnancy after Assisted reproductive technologies	2 (18.2)	2 (20.0)	5 (55.6)	0.135b
Gravidity	3.0 +- 1.4	3.7 +- 2.0	3.9 +- 1.4	0.445a
Parity	1.9 +- 1.4	2.0 +- 1.4	1.1 +- 1.1	0.282a
Current IUD	4 (36.4)	0	0	0.012b
Admission ?-HCG level (mIU/ml)	2025 +- 1105	18,424 +- 2579	11,204 +- 9221	0.003a
Admission hemoglobin (g/dl)	11.6 +- 1.9	11.8 +- 1.9	12.2 +- 0.9	0.745a
Gestational age at diagnosis (weeks)	6.4 +- 0.7	6.6 +- 1.2	7.1 +- 1.9	0.581a
Suspected sonographic diagnosis	1 (9.1)	8 (80.0)	10 (100.0)	0.0001b
Clinical presentation Abdominal pain	10 (90.9)	4 (60.0)	4 (44.4)	0.031b
Rebound tenderness	8 (72.7)	3 (30.0)	2 (22.2)	0.044b
circulatory collapse	6 (54.5)	1 (10.0)	1 (11.1)	0.031b
Ultrasound findings Peritoneal fluid	8 (72.7)	2 (20.0)	2 (22.2)	0.02b
Gestational sac	1 (9.1)	7 (70.0)	5 (55.6)	0.013b
Embryonic cardiac motion Intraoperative characteristicsc	0 (0%)	3 (30.0)	2 (22.2)	0.479b
Pregnancy side Right	7 (63.6)	4 (66.7)	5 (55.6)
Left	4 (36.4)	2 (33.3)	4 (44.4)	0.434b
Surgery Laparoscopy	(90.9)	(50.0)	(88.9)	0.38b
Laparotomy	(9.1)	(50.0)	(11.1)
Ruptured pregnancy	66.9	16.7	44.4	0.028a
Estimated blood loss (ml)	1081 +- 647	760 +- 597	343 +- 318	0.003a
Blood transfusion	7 (63.6)	2 (20.0)	1(11.1)	0.025b
Postoperative hospitalization days	1.8 +- 0.5	2.0 +- 0.5	1.6 +- 0.3	0.128a

Data are presented as percentages (%) or as the mean +- standard deviation.

^a ANOVA.

^b Chi-square test.

^c Intraoperative characteristics were assessed in 11 ovarian, 6 interstitial and 9 stump pregnancies.

Of these rare sided pregnancies, 11/11 (100%) of the ovarian preg- nancies, 6/10 (60.0%) of the interstitial, and 9/9 (100%) of the stump pregnancies required emergency surgery. The main surgical approach in all women was laparoscopy and was performed in 10/11 (90.9%), 3/6 (50.0%) and 8/9 (88.9%), respectively (Table 1). The indications for surgery were suspected rupture manifesting as acute abdominal pain, accompanied by Hemodynamic deterioration and/or free pelvic fluid observed on sonography. Four of the 10 (40%) women with an intersti- tial pregnancy underwent successful systemic methotrexate (MTX) treatment.

As shown in Table 1, there was no difference in distribution between the right and left sides. During surgery, rupture was detected signifi- cantly more frequently in ovarian pregnancies than in the interstitial or stump pregnancies (66.9%, 16.7% and 44.4%, respectively, p = 0.028). Ovarian pregnancies also involved a significantly higher esti- mated blood loss (1081 +- 647 ml, 760 +- 597 ml and 343 +- 318 ml, re- spectively, p = 0.003) and required a blood transfusion in the perioperative period (63.6%, 20.0% and 11.1%, respectively, p = 0.025) significantly more frequently. The duration of postoperative hospitaliza- tion was similar.

1. Discussion

Modern work-ups of ectopic pregnancies, including transvaginal ul- trasound and sensitive ?-hCG, enable an early accurate diagnosis of these cases, shifting the medical goal from lifesaving and reducing mor- bidity towards preserving fertility. Nevertheless, these same rare site ec- topic pregnancies still tend only to be diagnosed after rupture has occurred, thus necessitating emergency surgery. The diagnosis is based on the patients’ symptoms, the physical exam, as well as the lab- oratory and ultrasound findings. However, the findings here show that these diagnostic data tend to vary according to the implantation site.

Surprisingly, the mean hCG level on admission was significantly lower (p = 0.003) among patients treated for ovarian ectopic preg- nancy. During the physical exam, abdominal pain was the main present- ing symptom (p = 0.031) and peritoneal irritation signs were more common (p = 0.044) among women with an ovarian pregnancy. The pain was reported to be nonspecific. Women with an ovarian pregnancy had the highest rates of circulatory collapse (~55%, p = 0.031) which called for immediate surgical intervention. This points to the difficulties involved in achieving a correct Timely diagnosis to prevent acute deteri- oration in patient status.

Increasing awareness and progress in diagnostic modalities can help avoid the cascade of events reported here. In the current study, women with an ovarian pregnancy had the lowest rates (9%, p = 0.0001) of pre- operative sonographic diagnosis but these same women had signifi- cantly higher rates of rupture (p = 0.028), blood loss (p = 0.003) and required more blood transfusions (p = 0.025).

This may be because in the absence of a yolk sac or fetal heart motion in 2D ultrasound technology, an ovarian pregnancy may mimic a rup- tured corpus luteum. This underscores the need for better and more precise diagnostic tools. Comstock et al. [16] reported a case series in which they evaluated the ultrasonographic appearance of proven ovar- ian ectopic pregnancies. These included a wide echogenic ring with an internal echolucent area that clearly differed from the thin tubal ring ob- served in tubal pregnancies or a corpus luteum cyst. A yolk sac or fetal heart motion were rarely identified. By contrast, 3D ultrasound can suc- cessfully distinguish an ovarian pregnancy from a corpus luteum cyst, which may improve detection [17].

Ovarian pregnancies may fail to be diagnosed due to their non- specific clinical presentation upon admission. The massive intra- abdominal bleeding is often erroneously ascribed to the fragility and hypervascularity of the ovarian tissue [4,5,18-20]. Patients with primary ovarian pregnancy mostly undergo surgery for suspected rupture of a

tubal ectopic pregnancy or a hemorrhagic corpus luteum, so that the

final diagnosis is made intraoperatively [4,5,18-20].

The cause of ovarian pregnancy remains enigmatic. The hypothesis of a relationship between IUD and ovarian pregnancy is not new but is very controversial [18-20]. We found high rates of IUD use, ranging from 90% to 42.9%, but the rate of IUD use decreased significantly over time [4]. The significance of this finding is unclear since the rate of IUD use over time in our population is unknown. We may only assume that there was also a shift in the types of IUDs over time, which may be related to the decrease in the rate of the ovarian pregnancies (4).

The present study has a number of limitations deserving of note. Be- cause it utilized a retrospective design, there could be no control of ad- ditional associated factors. Records with incomplete data could not be completed. The small number of cases in our database was also a draw- back. This could have influenced the results through selection bias and/ or by impacting the statistical significance of the tests used. Neverthe- less, the cohort reported in this study is the largest to date over a lengthy period of time, which is an advantage given the rarity of this event.

It is important to mention that although the current study was lim- ited to the ovarian, interstitial and stump ectopic pregnancies, we should consider in the differential diagnosis other rare site located (cer- vical, abdominal and heterotopic) pregnancies in a pregnant woman based on the patients’ symptoms, the physical exam, as well as the lab- oratory and ultrasound findings.

Thus overall, despite growing awareness of rare site ectopic preg-

nancies, they still remain a diagnostic challenge and awaits enhance- ments in sonographic detection techniques. The diagnosis is often made at the time of surgery after the rupture has occurred. Health care providers should be better informed about a possible ovarian diag- nosis so that they can opt for quick surgical intervention in what may often be a gynecological emergency.

Conflicts of interest

None.

Funding and support

None.

References

1. Barnhart K, van Mello NM, Bourne T, Van Calster B, Bottomley C, et al. Pregnancy of unknown location: a consensus statement of nomenclature, definitions and out- come. Fertil Steril. 2011;95:857-66.
2. Lipscomb GH, Stovall TG, Ling FW. nonsurgical treatment of ectopic pregnancy. N Engl J Med. 2000;343:1325-9.
3. Bouyer J, Coste J, Fernandez H, Pouly JL, Job-Spira N. Sites of ectopic pregnancy: a 10- year population-based study of 1800 cases. Hum Reprod. 2002;17:3224-30.
4. Melcer Y, Smorgick N, Vaknin Z, Mendlovic S, Raziel A, Maymon R. Primary ovarian pregnancy: 43 years experience in a single institute and still a medical challenge. Isr Med Assoc J. 2015;17:687-90.
5. Melcer Y, Maymon R, Vaknin Z, Pansky M, Mendlovic S, Barel O, et al. Primary ovar- ian ectopic pregnancy: still a medical challenge. J Reprod Med. 2016;61:58-62.
6. Melcer Y, Naaman HZ, Hausman R, Vaknin Z, Levinsohn-Tavor O, Maymon R, et al. Tubal stump pregnancy after salpingectomy-does the time interval from surgical in- tervention to conception matter? J Obstet Gynaecol Res. 2021;47:2509-14.
7. Alalade AO, Smith FJE, Kendall CE, Odejinmi F. Evidence-based management of non- tubal ectopic pregnancies. J Obstet Gynaecol. 2017;37:982-91.
8. Hallatt JG. Primary ovarian pregnancy: a report of twenty-five cases. Am J Obstet Gynecol. 1982;143:55-60.
9. Hallatt JG, Grove JA. abdominal pregnancy: a study of twenty one consecutive cases. Am J Obstet Gynecol. 1985;152:444-9.
10. Herbertsson G, Magnusson S, Benediktsdottir K. Ovarian pregnancy and IUCD use in a defined complete population. Acta Obstet Gynecol Scand. 1987;66:607-10.
11. Atrash HK, Hughes JM, Hogue CJR. Ectopic pregnancy in the United States, 1970-1983, CDC surveillance summaries. Morbid Mortal Wk Rep. 1986;35:29-37.
12. Martin Jr JN, Sessums JK, Martin RW, Pryor JA, Morrison JC. Abdominal pregnancy: current concepts of management. Obstet Gynecol. 1988;71:549-57.
13. Al-Meshari AA, Chowdhury N, Adelusi B. Ovarian pregnancy. Int J Gynaecol Obstet. 1993;41:269-72.
14. Spiegelberg O. Zur casuistic der ovarial Schwangerschft. Archiv fur Gynaekologie. 1873;13:73-6.
15. Timor-Tritsch IE, Monteagudo A, Matera C, Veit CR. Sonographic evolution of cornual pregnancies treated without surgery. Obstet Gynecol. 1992;79:1044-9.
16. Comstock C, Huston K, Lee W. The ultrasonographic appearance of ovarian ectopic pregnancies. Obstet Gynecol. 2005;105:42-5.
17. Ghi T, Banfi A, Marconi R, Iaco PD, Pilu G, Aloysio DD, et al. Three-dimensional sono- graphic diagnosis of ovarian pregnancy. Ultrasound Obstet Gynecol. 2005;26:102-4.
18. Raziel A, Golan A, Pansky M, Ron-El R, Bukovsky I, Caspi E. Ovarian pregnancy: a re- port of twenty cases in one institution. Am J Obstet Gynecol. 1990;163:1182-5.
19. Raziel A, Mordechai E, Schachter M, Friedler S, Pansky M, Ron-El R. A comparison of the incidence, presentation, and management of ovarian pregnancies between two periods of time. J Am Assoc Gynecol Laparosc. 2004;11:191-4.
20. Raziel A, Schachter M, Mordechai E, Friedler S, Panski M, Ron-El R. Ovarian preg- nancy - a 12-year experience of 19 cases in one institution. Eur J Obstet Gynecol Reprod Biol. 2004;10:92-6.