Vancaillie, Thierry G. MD, FRANZCOG1; Anderson, Ted L. MD, PhD2; Johns, D Alan MD3
Female sterilization is the most common method of contraception in the United States, used in 17.8% of women aged 15–44 years,1 with approximately 684,000 procedures performed in the United States annually.2 Tubal ligation methods include resection of the tube or part thereof, the application of clips or rings onto the fallopian tube, and coagulation of the isthmic portion of the tube. The most common procedures to access the fallopian tubes are laparoscopy and minilaparotomy. Laparoscopic tubal ligation, the more common procedure, is typically performed under general anesthesia in an operating room. Minilaparotomy, often performed immediately after parturition, also involves anesthesia and one or more incisions, which are relatively small, as the name implies. Although these methods are generally safe and effective, they carry small, but noteworthy risks, related both to anesthesia and the procedure itself.
In November 2002, a transcervical sterilization procedure (Essure, Conceptus Inc., San Carlos, CA) was approved by the United States Food and Drug Administration.3,4 A nonincisional tubal occlusion method that accesses the fallopian tubes hysteroscopically offers several potential benefits over laparoscopic sterilization procedures. This less invasive approach does not require general anesthesia, which thereby may allow the procedure to be performed in an office setting. Further potential advantages include decreased postprocedural pain, decreased risk of adjacent organ/tissue damage, and more rapid recovery time.
The Essure procedure consists of inserting microcoils containing nickel titanium alloy, stainless steel, and polyethylene terephthalate fibers into the proximal section of each fallopian tube lumen.4,5 The insertion is believed to elicit a tissue in-growth response, thereby anchoring the coils and blocking the fallopian tubes.
The polymer matrix transcervical sterilization system (Adiana Permanent Contraception; Hologic, Inc., Bedford, MA) is a transcervical sterilization system based on the insertion of a polymer matrix totally integrated within the fallopian tube.6 The objective of this study was to evaluate the safety, placement efficacy, and reliability of this intratubal occlusion device for permanent contraception and to assess tolerability and overall satisfaction with the associated procedure.
METHODS AND MATERIALS
The Evaluation of the Adiana System for Transcervical Sterilization (EASE) clinical study was initiated in 2002. The study was designed as a prospective, multicenter, single-arm, international clinical study involving 18 investigators at 16 sites in the United States (14 sites), Mexico (one site), and Australia (one site). Institutional review board (IRB) approval of protocol and study oversight was obtained at each clinical site or from a central IRB (Independent Review Consulting, San Anselmo, CA). Study participants provided written informed consent after extensive counseling regarding study participation as well as sterilization per se. Subsequent surgical and radiographic procedures (hysterosalpingography, HSG) were conducted at university and private hospitals, ambulatory surgical centers, or private office settings.
The polymer matrix transcervical sterilization system has three components: a radio-frequency (RF) generator, a delivery catheter, and an implantable matrix. The RF generator is programmed to reach and maintain a temperature of 64°C at the catheter tip for a treatment duration of 60 seconds. The generator has an LCD display that guides the physician through the various steps of the procedure. The delivery catheter, designed to be introduced through the 5 French working channel of an operative hysteroscope, is used to access the fallopian tube and deliver the matrix into the intramural portion of the tube. The delivery catheter includes a bipolar electrode array and a position detection array to signal the physician that the catheter is in a tubular structure (ie, the tubal lumen). A black mark on the catheter provides visual evidence that the catheter is placed properly at the intramural portion of the tube. The matrix is a cylindrical piece of porous silicone, approximately 3.5 mm long and 1.5 mm in diameter.
The polymer matrix transcervical sterilization system procedure, designed to be performed in an outpatient or office setting, uses low-power RF energy and matrix implantation to achieve tubal occlusion, as illustrated in Figure 1. In the upper image, the position detection array is an electronic feature, which identifies that the catheter is within a tubular structure. The active element of the catheter is located within the intramural portion of the fallopian tube. There are four electrodes in a bipolar configuration on the outside of the catheter. The polymer matrix is held in the hollow of the catheter. In the middle image, at the completion of the 60-second RF application, the operator activates the release button on the handle, which initiates retraction of the outer sheath of the catheter. The push rod contained in the center of the catheter keeps the matrix in place. In the lower image, the matrix is positioned in apposition within the treated area of the fallopian tube.
After the physician has hysteroscopically accessed a fallopian tube, a pulse of RF energy (less than 3 W) is delivered to the intramural portion of the tube. The RF energy elicits a superficial thermal lesion approximately 0.5 mm deep. Next, the delivery catheter deposits a matrix in the same location as the lesion. The matrix is delivered by a push-button mechanism in the catheter handle. The delivery catheter is removed after matrix implantation. A new catheter is used to access the contralateral fallopian tube. The matrix provides a substrate for tissue ingrowth, leading to tubal occlusion and permanent sterilization.
Participants were recruited from general obstetric and gynecology practices or from IRB-approved print and radio advertisements. Women were eligible to participate if they were 18–45 years of age, generally healthy, had established fertility, and were seeking permanent contraception. Participants were required to acknowledge there was a potential risk of pregnancy when they relied on the device for contraception. In consultation with the Food and Drug Administration, recruitment aimed to obtain an age distribution of women similar to the U.S. Collaborative Review of Sterilization (CREST) study.7
Exclusion criteria included major medical conditions, use of cortisone or cytostatic medication, any underlying health condition that would adversely affect the ability to undergo surgical procedures (eg, cardiovascular conditions), pelvic inflammatory disease, uterine conditions that may affect tubal access (eg, synechiae), significant cervical or uterine pathology such as an abnormal Pap test or large leiomyomas, or any condition that might compromise compliance or long-term study follow-up.
The study was conducted under an Investigational Device Exemption from the Food and Drug Administration with a phased enrollment. During phase I, an initial cohort of 150 women was enrolled. Enrollment continued in phase II with total enrollment of 770 women. The primary efficacy endpoint for the study was the pregnancy prevention rate after 12 months of reliance on the polymer matrix transcervical sterilization system.
Before undergoing the procedure, complete medical histories were obtained for participants. In addition, each participant underwent a complete physical examination, with documentation of normal cervical cytology, negative testing for sexually transmitted diseases, and negative pregnancy test to confirm study eligibility.
Participants underwent hysteroscopy during the proliferative phase of their menstrual cycle, using glycine as a distention medium. Analgesia used during the procedure was left to the discretion of the investigator and ranged from topical or local only to intravenous sedation. Polymer matrix transcervical sterilization system delivery catheters were introduced sequentially into each fallopian tube under hysteroscopic visualization. Bipolar RF energy was delivered, and the matrix was placed within each tubal lumen, as described previously; matrix placement was confirmed at one week postprocedure using transvaginal ultrasonography.
Participants received counseling regarding the need for alternative contraception after placement of polymer matrix transcervical sterilization system implantable matrices, until bilateral occlusion could be confirmed at 3 months. If bilateral occlusion was not confirmed at 3 months, a second HSG was performed at 6 months postprocedure. Women who were patent in one or both tubes at 6 months were offered bilateral tubal sterilization and either withdrew from the study or were followed for safety evaluation only. Confirmation of bilateral occlusion by HSG initiated the reliance phase, during which women relied on the polymer matrix transcervical sterilization device for contraception. Women were evaluated by brief office visits at 3, 6, and 9 months into the reliance phase, and assessed with a complete examination at 1 year.
The primary efficacy endpoint for the study was the pregnancy prevention rate after 12 months of reliance on the polymer matrix transcervical sterilization system. Secondary endpoints included device placement rates as well as satisfaction and comfort with the placement procedure and during the reliance phase. Satisfaction and comfort outcomes were measured qualitatively and quantitatively, respectively. Comfort was assessed during the procedure and perioperative interval using a visual analog scale.8 Additionally, each woman was asked whether discomfort matched or exceeded her anticipation. Perioperative events, including spotting, bleeding, cramping, and nausea, were assessed and recorded. Short-term follow-up information regarding participant satisfaction was obtained at subsequent visits and telephone interviews.
The study was conducted under a common protocol with the intention of pooling the data for analysis. An analysis of variance or Cochran-Mantel-Haenszel test, as appropriate to the variable being analyzed, was conducted to evaluate the pooling of data across study sites through comparison of demographic characteristics and preoperative factors. A target accrual of 750 participants was planned to accommodate potential dropout. A total of 400 evaluable women was required to ensure that the study had 80% power to demonstrate a 1-year pregnancy rate of less than 5% using a one-tailed test of proportions. The calculation was performed with 5% type I error control and assumed a true pregnancy rate of 2.5%. The primary endpoint, 1-year pregnancy rate, was tabulated and a 95% one-tailed confidence interval was calculated using life-table methods. Continuous endpoints were summarized with sample size n, mean, median, standard deviation, and range. Categorical endpoints were summarized with sample size n, frequency counts, and percentages. All statistical analyses were conducted using SAS 8.2 (SAS Institute Inc., Cary, NC) software.
A total of 770 women (627 United States, 143 all other sites) were enrolled in the study. Of these, 125 either failed to meet inclusion criteria (50) or withdrew voluntarily. (75). Thus, 645 women were eligible to undergo the polymer matrix transcervical sterilization procedure and represent the intent-to-treat group. Age and ethnic diversity of the intent-to-treat group (Table 1) resembled historical controls of the CREST study.7 Participant accountability throughout the study is presented in Figure 2.
The polymer matrix transcervical sterilization system procedure was attempted in 645 women between November 2002 and May 2005. During the procedures, 33.2% (n=214) received topical or local anesthetic along with an nonsteroidal antiinflammatory drug or other nonnarcotic oral analgesic; 19.8% (n=128) received minimal sedation with an anxiolytic in combination with topical or local anesthetic, nonsteroidal antiinflammatory drug, or oral analgesia, and 47% (n=303) received conscious sedation with an intravenous agent.
The mean duration of the polymer matrix transcervical sterilization system procedure was 11:54 minutes (standard deviation 7:08), defined as interval between hysteroscope insertion to removal, with 90% of procedures completed in less than 20 minutes. The mean volume of glycine used during the procedure was 1,226 mL, with a mean fluid absorption of 182 mL.
In reviewing participant comfort levels after the procedure, 40.2% of women reported very little or no discomfort during the procedure; only 9.2% described it as very uncomfortable. Additionally, 80% (504 of 629) of women reported that the discomfort they experienced during the procedure and perioperative interval was the same or less than anticipated. At the time of discharge, women were asked to rate their pain using a visual analog scale with a range of 1–100, the mean score was 5.9 (standard deviation 11.03). In a 48-hour follow-up telephone call, when asked to describe their tolerance of the procedure, 98% (619 of 632) of women chose “well, very well, or excellent.” Nearly all women (98%, 626 of 636) returned to normal activities by 48 hours postprocedure; of these, 90% (560 of 626) resumed normal activities the day after the procedure.
Bilateral placement was achieved in 95% of women (611 of 645). Early in the study, women in whom bilateral placement was not achieved during the initial attempt (n=8) underwent a second placement attempt. Although bilateral placement was achieved in seven of these eight women, this practice was discontinued. Placement success was similar across sites and varied from 85.7% to 100%.
There were 34 women in whom bilateral placement was not successful (5%). Of these, unilateral placement was accomplished in 15 women. One woman ceased participation in the study before her 3-month waiting visit. The remaining 14 women were followed for safety endpoint and are included in the safety analysis.
Of the 611 women with successful bilateral placement who entered the 3-month waiting period, six were lost to follow-up or withdrew. One woman became pregnant during this interval while relying on alternative contraception. Thus, 604 women were evaluated for tubal occlusion by HSG at 3 months, of which bilateral occlusion was confirmed in 551. Of the 53 women with one or both tubes patent at 3 months, one woman was lost to follow-up before reevaluation, two women had alternative contraception failures, and five were not reevaluated at 6 months. The latter seven were followed for safety endpoints only. Reevaluation of the remaining 45 women by HSG at 6 months confirmed bilateral occlusion in 19 additional women. Bilateral occlusion was ultimately confirmed in 570 women.
Of the 645 women in the intent-to-treat group, 570 (88.4%) were ultimately able to rely on the polymer matrix transcervical sterilization system during the 1-year contraception reliance phase. Of the 570 women, 553 (97%) were compliant with all clinical follow-up and 17 were lost to follow-up. The efficacy calculation was based on these 553 women and more than 6,600 women-months of wearing.
Six pregnancies were reported in the first 12 months of reliance on the polymer matrix transcervical sterilization system, one of which was ectopic. Using life-table methods, this represents a cumulative failure rate of 1.07% (95% upper confidence bound of 2.1%). The one-year pregnancy prevention rate in the EASE study, derived with life-table methods, is 98.9% (95% confidence interval with a one-tailed, lower confidence bound of 98.2%).
All women in whom treatment was attempted (n=645) were followed for safety, and comfort if applicable. High satisfaction and comfort scores were attained during the reliance phase; 99% (524 of 531) of women were “very satisfied” or “satisfied” with the procedure, and 99% (525 of 532) of women reported their overall comfort as “excellent” or “very good.” With regard to comfort, fewer than 1% (2 of 532) rated the experience as “fair” and none rated it “intolerable.” There were no requests for matrix removal due to discomfort or pain.
Adverse events reported on the day of the treatment procedure are depicted in Table 2. On the day of procedure, cramping represented approximately one quarter of all reported adverse events. Vaginal spotting occurred at a similar frequency; other adverse events were reported at lesser frequencies. All these events were resolved within a mean duration of 3.0 days, except back pain, which resolved within a mean duration of 9.0 days. Hyponatremia (sodium 129 mEq/L) occurred in one case (0.15%), and was successfully treated with induced diuresis without sequelae.
No uterine or tubal perforations were reported. No injuries related to RF energy or matrix placement occurred. Most procedure-related adverse events were mild in duration and resolved without intervention.
Adverse events that occurred in the 12-month reliance phase are listed in Table 3. Dysmenorrhea represented one quarter of all reported adverse events during the reliance phase. Headache was the next most frequent adverse event, at approximately half the frequency of dysmenorrhea; other adverse events occurred at lower frequencies. No infections related to the polymer matrix transcervical sterilization system procedure were reported. No allergic or adverse reactions to the matrices occurred, and there were no known expulsions of the matrices. No polymer matrix transcervical sterilization system matrices required removal or were expelled.
Two device-related serious adverse events occurred during the first 12 months of reliance. An endometrial polyp, which was adjudicated by the Data Safety Monitoring Board as possibly related to device placement, was resected. Additionally, one isthmic ectopic pregnancy occurred and was treated successfully with methotrexate.
We present the results of a prospective, international, multicenter, single-arm clinical study of the polymer matrix transcervical sterilization system. A polymer matrix is deposited at the site of an RF-energy-created lesion within the intramural segment of the fallopian tube, leading to tissue ingrowth and permanent tubal occlusion. Well-tolerated by women without general anesthesia, the procedure is safe while providing greater than 98% pregnancy prevention.
With a mean time of just under 12 minutes, the polymer matrix transcervical sterilization system procedure is of short duration and is easily performed. Bilateral placement was achieved in 95% of attempts, compared with the 90% value reported in the pivotal study for the Essure transcervical sterilization device.9 The primary reasons for inability to cannulate tubes in the polymer matrix transcervical sterilization system study included extremely lateral tube location, poor visualization of ostia, and suspected tubal blockage.
Adverse events experienced during perioperative interval, such as cramping and some bleeding or spotting, are consistent with what might be expected after any operative hysteroscopic procedure. The incident of hyponatremia resolved with a single dose of diuretic.
The nonincisional, transcervical approach of the polymer matrix transcervical sterilization system offers functional and clinical advantages over more traditional laparoscopic sterilization techniques. No women undergoing this procedure required general anesthesia. Further, 53% of women did not require conscious sedation. In addition to the avoidance of incisional pain and risk for potential intraabdominal injury, women recovered rapidly, and postoperative discomfort was easily managed with oral analgesics. This compares favorably with a previous study of another transcervical device (Essure) and laparoscopic (Falope ring; Gyrus ACMI, Southborough, MA) sterilization, in which the transcervical group reported significantly lower pain scores immediately postoperatively as well as after 1 and 4 weeks.10 Importantly, most women undergoing the polymer matrix transcervical sterilization system procedure experienced a rapid return to normal activities within 1 to 2 days. This is in contrast to an average of 6 days for return to normal activities after laparoscopic tubal sterilization.11
The cumulative failure rate of the polymer matrix transcervical sterilization system using life-table methods is 1.07% at the primary endpoint of 1 year. To put this in a broader perspective, typical first-year failure rates of commonly used contraceptives include 16% for the diaphragm, 15% for condoms, 8% for oral contraceptive pills, and 3% for DepoProvera.12 When comparing the polymer matrix transcervical sterilization system with historical controls of the CREST study,7 the 1.07% failure rate falls within the range of other sterilization methods (0–2.4 %).
Hysteroscopic sterilization procedures should include subsequent evaluation by HSG. In a review of root causes for pregnancies after transcervical sterilization procedures, Levy et al13 reported that noncompliance with HSG or follow-up accounted for 46% of pregnancies, whereas 25% of pregnancies were related to misreading the X-ray or HSG. This issue gains relevance when considering that 50% of the pregnancies during year 1 in the EASE study were attributed to HSG misinterpretation. The polymer matrix transcervical sterilization system matrix is not radiopaque. Although matrices can be identified by transvaginal ultrasonography to confirm positioning, it is important to note that no clinical management decisions were made using transvaginal ultrasonography evaluation. We believe the clinical usefulness of transvaginal ultrasonography may be the secondary evaluation of women in whom HSG reveals tubal patency, to confirm matrix location. Taken together, these observations underscore the importance of using a standardized protocol when performing and interpreting the HSG to document tubal occlusion. Further, it is imperative that patients receive appropriate counseling not to rely on transcervical sterilization procedures for contraception before documentation of tubal occlusion.
An important consideration in counseling patients undergoing transcervical sterilization is the increased risk of ectopic pregnancy. However, this increased risk is not unique to hysteroscopic sterilization procedures. Peterson et al14 estimated the 10-year cumulative probability of ectopic pregnancy for all methods of tubal sterilization combined to be 7.3 per 1,000 procedures, although there seems to be considerable variability based on the sterilization method employed. One ectopic pregnancy was reported during the 1-year interval of this study.
In conclusion, the polymer matrix transcervical sterilization system is a safe and effective sterilization method that is well-tolerated, with high participant satisfaction. This system provides a pregnancy prevention rate of greater than 98%. When used appropriately in properly counseled patients, the polymer matrix transcervical sterilization system provides a low-risk, convenient option for permanent sterilization.