Three-Year Prospective Menstrual and Dysmenorrhea Outcomes after ThermachoiceIII ® in-Office Therapy

Introduction

The first global endometrial ablation device in clinical use was the Gynecare Thermachoice^® Uterine Balloon Therapy System introduced in 1997. Despite well-published historic data on the safety and efficacy of Thermal Balloon Ablation (TBA),^1,2 a MEDLINE^® and PubMed search a paucity of peer review published data on long term outcomes from use of the TBA method currently in clinical use (Gynecare ThermachoiceIII^® Uterine Balloon Therapy System Gynecare, Ethicon, Somerville, NJ).

Traditionally, with first generation TBA, therapy success was defined as the percentage of patients with normal or no bleeding at 3 years and without hysterectomy or repeat ablation throughout the 3 years of “long-term” follow-up. ³ Gervaise et al. reported a cumulative 3-year success rate, defined as eumenorrhea or less, of 83% for first-generation uterine balloon therapy and 76% for endometrial resection. ⁴ Two-year ⁵ and 3-year ³ follow-up data of a randomized controlled study comparing thermal balloon first generation with rollerball ablation ⁶ reported similar rates. At 3 years, the percentages of women who reported amenorrhea and eumenorrhea were 93 and 94%, respectively. ³ According to Loffer et. al., ³ 7% (8/114) of the thermal balloon first- generation group and 14% (14/99) of the rollerball group underwent hysterectomy at 3 years post-therapy, although the overall success rates were similar, at 86 and 82%, respectively. In a study by Bongers et al., ⁷ the cumulative repeat intervention rate for the thermal balloon group was 13% compared with 26% for the rollerball group. With this historic premise, it could be hypothesized that the percentage of eumenorrhea or less would increase, and the subsequent failure rate/hysterectomy risk would decrease with use of the third-generation thermal balloon technology, compared with historic data.

Previously, Chapa et al. have published the 3, 6, 12, and 24 month clinical outcomes form this original treatment cohort.^8–10 This manuscript is a continuation of that cohort's analysis extended to 36 months post-therapy. Only one other published, peer- reviewed manuscript has described the clinical outcomes after similar therapy, with a mean follow up of 30 months. ¹¹ No other published research has tracked these outcomes from local block, ¹² to short-term, to 36 months prospectively.

Materials and Methods

Patient recruitment was from November 2006 to November 2007, at Women's Specialty Center Dallas, a community-based obstetric and gynecologic medical clinic serving a largely lower socioeconomic population. Ethical board review approval was provided for the original cohort protocol. Patients presenting with a chief complaint of menorrhagia were screened for therapy and participation. Intervention included office- based diagnostic hysteroscopy, dilation and curettage, and Thermachoice III endometrial ablation. All procedures were performed in office using local, lower uterine block. Patient inclusion/exclusion criteria, patient demographics, pain management protocol, and pre-ablation work-up have been described in previous publications.^8–10,12

The primary endpoint was reduction in menstrual flow from baseline. The resulting bleeding patterns were labeled as: amenorrhea (complete lack of menstrual flow), hypomenorrhea (scant or minimum menstrual flow lasting <5 days requiring only the use of light pads) and failure (continued menstrual flow similar to pretreatment state, need for any further therapy, or any abnormal bleeding troublesome to the patient). Response to treatment was gathered either at the physician's office visit or by telephone contact. Hemoglobin values were taken at baseline and compared with 12, 24, and 36 month results as an objective measure of outcome. “Failure” patients underwent transvaginal ultrasound, saline-infusion-sonography, cervical cultures for gonorrhea/chlamydia, complete blood cell count (CBC), thyrotropin (TSH), and endometrial biopsy (EMB) assessment when able.

The secondary endpoint was dysmenorrhea reduction post-procedure. Patients were asked to grade their dysmenorrhea at baseline and again at 12, 24, and 36 months post-procedure using a 0 (none) to 10 (severe) visual analogue scale (VAS).

Results

The initial cohort included 148 patients. The baseline cohort's mean age was 41 (range 29–48), median 43. One hundred and twenty (81%) were available for evaluation at 36 months (Fig. 1). Table 1 shows the menstrual outcomes per year. No statistically significant differences were noted between the categories at 24 and 36 months. As stated in a previous publication, “the mean hemoglobin value at baseline was 9.4 (median 10.1, range 9.1–11.1, with 26% below the mean). ³ Hemoglobin values increased above baseline at both 24 and 36 months for all evaluable patients (Table 2), a trend originally described previously.^8–10

OPEN IN VIEWER

FIG. 1.

Group demographics.

Between 24 and 36 months, 3 additional patients (originally hypomenorrhic) were classed as “failure” placing a total of 8 patients (6.7%) total in that class in the 3 year period. Of the 8 “failure” patients, 6 underwent hysterectomy by study end (5% of final evaluable cohort). The 2 patients remaining in the failure group opted for a levonorgestrel-releasing intrauterine device (IUD) during year 2. The characteristics of the patient failure group are shown in Table 3 Hysterectomy data are presented in Table 4.

In the failure group of 8 patients, all underwent attempted EMB before hysterectomy was considered. Pipelle EMB was unsuccessful in 4/8 (50%) of these patients; 1 (12.5%) was nondiagnostic because of scant endometrial tissue, and 3 had “benign fragments of suspected proliferative endometrium.”

The secondary endpoint data (dysmenorrheal reduction) are presented in Table 5. Out of the original 148 patient cohort, 33 were initially excluded from analysis because of other pelvic pain syndromes, leaving 115 patients for evaluation at the beginning of this study.^10,11 Of this remaining group of 115, 79/115 (69%) complained of dysmenorrhea at baseline. This group of 79 became the study cohort for dysmenorrheal reduction: 79, 76, and 75 patients in that group were available for follow up at 12, 24, and 36 months, respectively. Dysmenorrhea reduction (graded by mean VAS score) remained significant (p value<0.05) at 36 months compared with baseline (Table 5).

Discussion

Previously, Chapa et al. published the clinical outcomes from this original cohort with endpoints of 6, 12, and, recently, 24 months.^8–10 The previous results demonstrated no statistical increase in amenorrhea from 3 to 6 months post-therapy, ⁸ but did show a statistically significant rise in amenorrhea at 1 year. ¹⁰ No significant change in categories was noted at 24 months, ¹⁰ and, now, at 36 months. Dysmenorrhea reduction was also noted at each study endpoint with statistically significant VAS scores being noted at 3 months post-therapy and being maintained for 36 months (Table 5).^8–10

These results are similar to those of other published reports on improvement in pelvic pain and cramping after endometrial destruction. Chandakas et al., in the United Kingdom, reported an 83% reduction in dysmenorrhea as well as in premenstrual symptoms at 1 year, with results persisting at 3 years. ¹³ Similarly, Meyer et al. reported 70 and 75% reduction in dysmenorrhea following TBA and rollerball, respectively, with improved quality of life. ⁶ Speroff et al. described the physiologic increase in prostaglandin F2 alpha levels in the endometrium in women with dysmenorrhea with a peak in levels during menstruation. ¹⁴ It is intuitive that as the endometrial basalis is scarred by ablative therapy, subsequent prostaglandin synthesis and propagation would be reduced.

This, coupled with decreased menstrual flow, would contribute to decrease menstrual pain and cramps. It is this author's personal opinion, similar to that of Speroff, that the reduced dysmenorrhea is the direct result of decreased prostaglandin F2 alpha from endometrial destruction, which provides symptomatic relief.

Menopausal status as an influencing factor on menstrual patterns must be addressed. At 36 months, 53% of our evaluable cohort was within the 31–40 age group. It is unlikely that menopausal status influenced our menstrual outcomes. Four patients complained of new onset vasomotor/climacteric symptoms by study end. At 12 month follow-up, 2 patients experienced climacteric symptoms, 1 patient reported the same by 24 months, with 1 additional reporting this between 24 and 36 months. These patients are best diagnosed as perimenopausal, because of the recent start/duration of symptoms. All patients were >45 years of age at 36 months. These 4 patients represent only 3.3% of our 120 patient evaluable cohort. All 4 patients were amenorrheic at study end.

A strength of this study design lies in its definition of “failure.” By including need for any other subsequent therapy rather than hysterectomy rate alone, it was possible to more accurately delineate the efficacy of TBA therapy. Three additional patients were classed as “failures” in the 24–36 month period because of patient reports of increased intermenstrual spotting/bleeding and cyclic pelvic pain.

Eight patients were considered failures by 36 months (6.7%) because of recurrent bleeding/cyclic pelvic pain. Interestingly, 3/6 hysterectomized patients possessed uterine fibroids, which were diagnosed at study start. On pathologic examination, these 3 also had adenomyosis (Table 4). The role uterine fibroids play in global endometrial ablation (GEA) failure is unclear from the literature, with some suggesting decreased efficacy of ablation procedures,^15,16 whereas others state no effect. ¹⁷ It must be noted that 87/148 of this original cohort had uterine fibroids, ⁸ as did 63 (52%) of the final evaluable cohort. It was decided to include patients for therapy with submucosal fibroids up to 4 cm in greatest dimension. That size criterion was chosen because prior data has shown efficacy in TBA equal to rollerball with submucosal myomas up to 3 cm. ¹⁸ Per thermal balloon instructions for use, submucosal fibroids are not listed under contraindications. Rather, it is stated that “the safety and effectiveness of the GYNECARE THERMACHOICE III UBT System has not been fully evaluated in patients with submucosal myomas.” ¹⁹ Adenomyosis has also been linked to higher subsequent failure post-ablation, ¹⁶ although its true mechanism of action is still unclear. Because of the small sample size of 8 patients in the failure category, it is not possible to provide clear prognostic factors at this time for subsequent post-ablation failure, nor was that the aim of this study.

The 6.7% failure rate (5% hysterectomy) at 36 months compares more favorably with published failure rates of prior generation thermal balloons. Longinotti et al. published a 10% probability of hysterectomy at 2–3 years by life table analysis among 3681 women treated from 1999 to 2004. ²⁰ By years of recruitment, prior generation thermal balloon devices were chosen for that cohort therapy.

As stated in the 24 month evaluation: “We must recognize one limitation of our study; we realize that the menstrual pattern outcome categories were not based on a validated scale. However, we chose that model to present a clinically adoptable tool all clinicians may use to track patient improvement in everyday practice. Additionally, the design of this study was not set for the identification of potential prognostic factors for post ablation outcomes.” ⁵ Evaluation was possible for 81% of patients from the original baseline cohort, and 19% were lost to follow-up. It is possible that results might have changed if the evaluable cohort had been larger than the actual.

As its strength, this study represents the first prospective 3 year outcome, peer- reviewed data with the third generation TBA device, and the procedure performed using local analgesia. This is second only to recently published data from Varma et al. In their prospective observational study (in a United Kingdom teaching hospital), 102 menorrhagic women underwent TBA between 2001 and 2005. Women underwent either Thermachoice first generation (n=51) or Thermachoice III (n=51) thermal balloon endometrial ablation (TBEA) performed in the outpatient setting using local anesthesia. Mean follow-up was 30 months. Multivariate analysis showed that using ThermachoiceIII rather than Thermachoice first generation increased the likelihood for amenorrhoea (p=0.001), and regardless of the type of Thermachoice device, higher mean intrauterine ablation pressure (p=0.014) and/or higher morphine rescue analgesia (p=0.030) correlated to better long-term patient satisfaction. ⁶ In their cohort, comparable to this one, “no need for further therapy” was noted in 92.2% of ThermachoiceIII-treated patients (article reference, Table 2). The opinion of those authors is shared by the authors of this article, that, “It is not unexpected that Thermachoice III has higher amenorrhoea rates, given its {impeller} design and better temperature distribution across the balloon. Our observation of increased intrauterine pressure and improved outcomes may be related to better surface contact of the balloon against the target endometrial surface and improved heat energy transfer.” ¹¹

Conclusions

This article should help guide physicians who provide global endometrial ablative therapy. Compared with historic data, a documented hysterectomy rate of 5% at 3 years contrasts with the 7% ³ to 10% ²⁰ rates with first generation balloon therapy. Overall, 94% of the patients did not require repeat intervention within 3 years compared with a historic rate of 88%. ³ The rise in hysterectomy rate over years 1, 2, and 3 (although improved over prior thermal balloon generations) must be prospectively followed to determine if a plateau or continual rise is noted. The goal of the authors of this study is to follow this original cohort for 5 years. For now, this is the second group to more adequately delineate clinical outcomes following this type of TBA system.