In the 1980s, studies reported on the relative safety and ease of ultrasound-guided (US-guided) oocyte retrieval as compared with the classical laparoscopic approach for in-vitro fertilization (IVF) [1,2]. Furthermore, ultrasound-guided retrieval was shown to have higher numbers of mature oocytes and oocyte fertilization rates when compared with the laparoscopic approach [2]. This could be related to the fact that US-guided follicular aspiration allowed better assessment of the complete emptying of the follicles as one study indicated that the cumulus-oocyte complex often is likely to be aspirated when the follicle is nearly completely collapsed [3]. Additional disadvantages to laparoscopic approach include the need for general anesthesia and all potential complications associated with laparoscopic surgery.

As a result, also in the 1980s, US-guided oocyte retrieval was modified and was conducted via combination of transabdominaltransvesical [4,5] or transurethral-transvesical approach [6]. These techniques however, were associated with frequent complications, mostly involving the urinary tract, including infection, urinary retention, and hematuria [7]. Because of such complications, subsequent studies reported that physicians and patients preferred a transvaginal US-guided over abdominal US for both follicular monitoring and aspiration [8,9]. With accumulation of further evidence on the feasibility, safety and efficacy of the transvaginal US-guided oocyte retrieval, today this approach has been universally adopted and it has become the standard of care worldwide. The transvaginal approach however, poses a bothersome problem when the ovaries are not accessible transvaginally due to variations in pelvic organ anatomy from biological variability or pelvis/abdominal disease. Historically, in such cases, clinicians had to revert back to laparoscopic follicular aspiration, or these women were denied IVF.

An alternative method developed for these cases is a transabdominal US-guided retrieval, which avoids the increased time, cost, and potential complications of laparoscopy. One retrospective series and 5 case reports have been published using this latter approach [10-15]. In most, if not all, of these cases however, a needle-guide was used to facilitate oocyte retrieval and, invariably, more than one operator was required to complete the process; one to use the ultrasound and another to aspirate the follicles. Furthermore, the majority of these cases required a general or regional anesthesia.

In the present study, we describe our experience using a modified transabdominal US-guided follicular aspiration technique, which does not require a needle guide and both scanning and aspiration are performed by a single operator, when one or both ovaries are inaccessible transvaginally. All cases were done under conscious sedation.

Institutional review board approval was obtained. We retrospectively reviewed 816 IVF cycles with or without Intra Cytoplasmic Sperm Injection (ICSI) over 3 years done in our relatively new IVF center, and we identified 13 cycles (1.6%) in 13 patients involving transabdominal US-guided follicular aspiration. In three cases, both ovaries were aspirated transabdominally. In the remaining nine cases, follicles were aspirated transabdominally from one and transvaginally from the contralateral ovary. The indication for IVF/ ICSI was a male factor (decreased sperm count, motility, normal morphology, or a combination of them) in 6 of those cases, and a female factor in the remaining 7 patients (5 with polycystic ovarian syndrome, and 2 with decreased ovarian reserve).

Patients were prepared for oocyte retrieval using a standard ovarian long-stimulation protocol. Down regulation was achieved using Gonadotropin Releasing Hormone (GnRH) agonist (Decapeptyl, Ferring, Germany), then ovarian stimulation was initiated using an exogenous recombinant gonadotropin (Puregon, MSD, Netherlands), with a starting dose of 150 to 225 International Unit (IU). The dose was then adjusted in tandem with ovarian follicular development as monitored by serial serum estradiol and transvaginal ultrasound. Oocyte retrieval was performed 36 hours after the administration of 10,000 iu of human Chorionic Gonadotropin (hCG-Choriomon, IBSA, Switzerland), which was administrated when at least three follicles reached 17 millimetre (mm) in diameter.

The need for transabdominal aspiration was recognized during follicular monitoring when one or both ovaries were not clearly visualized transvaginally and were much more apparent transabdominally. At the time of follicular aspiration, we attempted in all cases to access the ovaries transvaginally using the usual technique of applying abdominal pressure to push the ovaries into the pelvis. If transvaginal aspiration was not possible, we converted to transabdominal retrieval. Under the same conscious sedation (Fentanyl and Midazolam) used for transvaginal aspiration, the patient was repositioned from dorsal lithotomy to supine. The urinary bladder was emptied by patient voiding or in-and-out catheterization just prior to attempted transvaginal approach. When the transvaginal approach failed to access one or both ovaries, the same operator reverted to scanning the patient’s abdomen using regular ultrasound gel and an abdominal probe (3.5 MHz Probe; Philips iU22, Philips Medical Systems, Bothell, WA, USA) to identify an area in the abdominal wall for the most feasible and safest access to the ovary(s).

The gel was wiped off and, using sterile techniques, the predetermined abdominal wall area, including skin and subcutaneous tissues, was injected with local anesthetic (1% Lidocaine Hydrochloride). Under ultrasound guidance, a standard 17-gauge retrieval needle (Cook Medical, Brisbane, Australia) was inserted through the abdominal wall by the same operator without the use of a needle guide (Figure 1A and 1B). All but one ovary required a single ovarian puncture to retrieve all oocytes. Only one ovary required two punctures, as some of the follicles could not be aspirated with the first attempt, however both ovarian punctures were done through the same transabdominal needle insertion. There were no failed attempts.

Figure 1: (A) The single operator is holding the US probe with one hand and inserting the oocyte-retrieval needle with his other hand-No sterile cover is used for the US cord as it never touches the needle or abdominal wall puncture site. (B) Transabdominal ultrasound image showing the retrieval needle tip within one follicle-No “biopsy line or path” is seen, as there was no needle guide used-Arrows are delineating the needle path.

Age, Body Mass Index (BMI), peak estradiol level reached, and the total number of follicles, number of “good size” follicles (>17 mm), as well as the total number of oocytes retrieved transabdominally and transvaginally were recorded (Table 1). Additionally, number of total, damaged, and mature oocytes retrieved, as well as the number of normally fertilized oocytes, fertilization rates, number of “good quality” embryos (8-cell embryo with <25% fragmentation rate on luteal day-3) and their percentage, were also recorded (Table 2). We also calculated the implantation, pregnancy, clinical pregnancy, and live birth rates (Table 2). The mean and standard deviation were used for description. The number of oocytes retrieved transabdominally and transvaginally were compared using paired t-test and statistical significance was considered at P<.05

Table 1: Patients’ characteristics and number of oocytes retrieved abdominally and vaginally.

Table 2: In vitro fertilization and pregnancy outcomes.

The mean and standard deviation (± SD) for age and BMI were 30.9 (± 3.3), and 31.5 (± 3.8) kg/m², respectively (Table 1).

The mean (± SD) number of oocytes retrieved transabdominally was less than that retrieved transvaginally: 8.4 (± 4.8) vs. 10.5 (± 6.8), however that was not statistically significant (P=0.93). Additionally, the mean number of “good size” follicles seen abdominally was only 7.9, which is almost equal to the mean number of oocytes retrieved transabdominally. Furthermore, the mean number of “good size” follicles seen abdominally and vaginally was 14.8, which is almost equal the mean number of mature oocytes retrieved in total (Tables 1 and 2).

The mean number of damaged oocytes was 0.5, while the mean and (SD) fertilization rate was 78.1% (± 16.2). The mean and (SD) percentage of “good quality” embryos was 51.9% (± 19.8) (Table 2).

The mean Implantation Rate (IR) was 36.2%, and out of the 13 patients, only 12 had Embryo Transfer (ET), as one patient developed Ovarian Hyperstimulation Syndrome (OHSS) and her embryos were cryopreserved for future transfer. Of the 12 patients, one had an ectopic pregnancy, treated with methotrexate and six had at least one intrauterine gestational sac seen by ultrasound, giving us an overall pregnancy rate (PR) of 58.3% and a Clinical Pregnancy Rate (CPR) of 50%. Three of these 6 patients had both ovaries retrieved transabdominally. Two of the 6 patients conceived with a twin. One of the twin cases aborted at 19 weeks due to an incompetent cervix and another singleton miscarried at 8 weeks of gestation. The remaining 4 patients delivered at term and all babies are alive and well, giving us a Live Birth Rate (LBR) of 33.3%. The patient who developed OHSS never came back for ET. Of the remaining 5 patients who did not conceive on their fresh cycle, one conceived from a Frozen Embryo Transfer (FET), and she delivered at term a singleton healthy baby. It is also important to note that one of the patients who did not conceive from her fresh cycle, had her oocytes fertilized by frozen sperms from testicular biopsy.

Three out of the 12 patients had both their ovaries retrieved abdominally. Their mean IR was 55.7%, while their overall PR, CPR, and LBR were 100%, 100%, and 66.6%, respectively.

We described our experience with a modified transabdominal follicular aspiration method in women with one or both inaccessible ovaries via the traditional transvaginal technique. A major advantage of our technique is the feasibility of transabdominal oocyte retrieval by a single operator. This is not just advantageous from a logistic point of view-as the elimination of the need for another experienced person -but it also allows the operator better control on the nominal and insignificant movements; one needs to apply, in order to move from one follicle to another. The procedure is done without the need of a needle guide (free-hand technique). Those two measures give higher chance to accomplish the procedure with a single ovarian puncture. Our technique has other advantages including obviating the need for general or regional anesthesia. Although it appears easier and safer than the traditional transabdominal approach, our comparative groups are not sufficient to allow such a claim.

Due to the relatively decreased elasticity of the abdominal wall, transabdominal follicular aspiration usually, if not always, requires multiple ovarian punctures, done via multiple abdominal wall entries, which increases patient discomfort and requirements for analgesia during and after the procedure, and can potentially increase the risk of abdominal wall residual scar. Additionally, multiple transabdominal punctures increase the risk of injury to underlying viscera, including bowel and blood vessels resulting in unwelcome sequellae. By eliminating the use of a needle guide, the operator will have a wider range of movement allowing retrieval of all oocytes through a single ovarian puncture. In the largest published study to date, including 69 women requiring transabdominal oocyte retrieval, all procedures were done under general or spinal anesthesia and multiple ovarian punctures were required to retrieve the follicles [15]. In contradistinction, all our procedures were done under the same conscious sedation used for transvaginal retrieval of the contralateral ovary, and in all but one case, a single ovarian puncture was adequate to retrieve all oocytes. Although one ovary required two punctures, both were done through the same transabdominal puncture. No “extra” pain killers were dispensed for these patients to use at home. Our technique, including elimination of a rigid needle guide, provides more flexibility and a wider range of hand movement to compensate for the relatively reduced mobility of the abdominal wall.

Although multiple transabdominal wall punctures potentially increase the risk of bowel and vessel injury, no such injuries were documented in the study by Barton et al., however one patient developed high-grade fever of undermined origin that later responded to intravenous empirical antibiotic therapy over four days [15]. None of our 13 patients had any complication or developed fever. It is important to note that we routinely prophylax all patients with a single dose of intravenous antibiotic (Cefazolin 1 g) administered within 30 minutes of oocyte retrieval. We do not use any antiseptic solution to clean the abdomen and we do not cover the US cord with a special sterile cover. Instead, we use sterile 0.9% normal saline to wash the abdomen, just as we do for the “less clean” vagina during transvaginal retrieval. Barton et al. described using Betadine antiseptic solution to wash the abdomen, but did not mention if prophylactic antibiotics were used.

Our results indicate that we had a lower number of oocytes retrieved transabdominally than transvaginally. Although this was not statistically significant, but it can also be attributable to the fact that there were fewer follicles present in the transabdominally retrieved ovaries rather than due to a shortcoming of the technique, as our data also showed that the mean number of “good size” follicles seen abdominally correlated well with the mean number of oocytes retrieved transabdominally, and the mean number of “good size” follicles seen abdominally and vaginally correlated with the mean number of mature oocytes retrieved in total (Tables 1 and 2). Furthermore, if we examine the 3 cases who had both ovaries retrieved transabdominally, we note that the number of oocytes retrieved were 15, 14, and 16, respectively which is comparable to the mean number of oocytes retrieved transvaginally in the remaining 9 patients. The number of damaged oocytes was negligible while the fertilization rate was comparable to accepted standards. All but 2 patients had enough embryos to cryopreserve (data not showing), and the pregnancy, clinical pregnancy, and live birth rates were within accepted norms.

Our study is not without limitations. It is a simple descriptive study of a small cohort of women undergoing IVF who had inaccessible ovaries via the traditional transvaginal route. Our approach has clinical implications. This small proportion (1.6%) of women with inaccessible ovaries should not be denied IVF or undergo other potentially risky retrievals. At this stage we cannot determine if pregnancy rates are equivalent from oocytes retrieved transabdominally and transvaginally. We do know however, that out of 6 clinical pregnancies we had, at least 3 occurred from oocytes retrieved transabdominally, because they never had any oocytes retrieved transvaginally. We aim to put this into consideration with future patients, and segregate oocytes retrieved and embryos produced via the transabdominal or transvaginal route.

Transvaginal inaccessibility of the ovaries is uncommon. The study of Barton et al. including over 17,000 patients reported that only 0.4% of them required transabdominal aspiration [15]. Although this number is small, however with enhanced treatment options for couples who otherwise would have been unable to conceive due to anatomical reasons, this scenario is likely to be encountered by IVF physicians. Our results showed a prevalence of 1.6%.

Transvaginal oocyte retrieval remains the preferred and easier route for patients and physicians. However in situations where the transvaginal approach is difficult or impossible to perform, transabdominal aspiration of one of both ovaries should be considered. Discussing pros and cons and potential complications with the couple is a must. The use of a needle-guide should always be considered when the experience to operate without it, is lacking. The “free-hand” transabdominal approach that does not rely on a needle-guide is not necessarily new, as it is usually utilized by maternal-fetal-medicine doctors when performing amniocentesis or chorionic villus sampling, and it is may be used by few IVF doctors when performing oocyte retrieval, however no one have described their experience and outcomes with this rarely needed approach. In the present study we have shown that our approach is feasible, relatively safe and efficacious in the small number of women applied.

We acknowledge the nurses of the Assisted Reproductive Technology Unit at the International Medical Center (Mrs. Ginalyn Frias, Mrs. Rimmy Varunny, Mrs. Theresa Pauline David, Ms. Venus Velasco, Ms. Ivory Israel) for their exceptional skills and support to make this possible. We also acknowledge Mrs. Adela Mohammed for her valuable patients’ support and education.