Reoperation rates for recurrence of fibroids after abdominal myomectomy in women with large uterus.

Uterine fibroids, or leiomyomas, are the most common benign tumors in premenopausal women. They are usually discovered incidentally by imaging in asymptomatic women, but 20%-50% of women develop symptoms such as abnormal uterine bleeding, pelvic pressure, pain or urinary or bowel complaints [ 1 ]. Some risk factors for the formation of these benign tumors include: Black race, age, family history, premenopausal state and hypertension, while use of hormonal contraception, smoking in low BMI women and low parity are protective [ 2 ]. Black women have the highest lifetime risk of having fibroids and suffer from more severe symptoms, which interfere with daily functioning [ 3 ]. Non-surgical treatment of uterine fibroids include expectant management [ 4 ] and medical therapy, which include estrogen-progestin contraceptives, progestins [ 5 , 6 ], levonorgesterel releasing intrauterine devices [ 7 ] and implants, progesterone receptor modulators [ 8 ] such as ulipristal acetate [ 9 ], mifepristone [ 10 ], and GnRH agonists [ 11 ], aromatase inhibitors [ 12 ] and selective estrogen receptor modulators such as raloxifene [ 13 ]. These approaches may provide symptom relief, especially in situations where bleeding is the main complaint. Although about three quarters of women report short-term improvement over the first year of treatment, long-term failure rates are high [ 14 ], and about 50% will have surgery within 24 months [ 15 ]. The surgical approach is the principal treatment of fibroids, especially in cases where bulk-symptoms are the dominant issue or if infertility is attributable to the presence of myomas. Hysterectomy as well as modern surgical approaches such as uterine artery embolization, radio frequency ablation and magnetic resonance-guided focused ultrasound can provide a cure for symptoms. Myomectomy, which may be performed laparoscopically or hysteroscopically, is recommended for women, who wish to maintain fertility. With any uterine conserving approach, however, the risk of fibroid recurrence remains. Around 15–33% of fibroids recur after myomectomy, and around 10%-21% of women undergo a hysterectomy within five to ten years [ 16 , 17 ]. Published rates and time to recurrence vary widely and include 12–15%, 31–43%, 51–62%, and 84% at 1, 3, 5, and 8 years respectively [ 18 – 22 ]. In this study, we focus on a group of 64 women, who were admitted for uterine surgery after having previously undergone an open myomectomy for a large symptomatic fibroid uterus with a mean size equivalent to 20.9 ±0.5 weeks gestation (range: 12–30 week-size). While the recurrence rates published in the literature address recurrence of fibroids in general [ 23 ] or the rates of reoperation following myomectomy [ 17 ], our study specifically addresses reoperation rates in patients, who have undergone myomectomy using an open approach for an initial diagnosis of large symptomatic fibroid uterus. The current literature also does not typically address the reasons underlying the need for repeat surgery. We seek to characterize the timing, risk factors as well as the presenting symptoms which led patients to undergo repeat surgery in this patient population.

During the study period, 592 primary open myomectomies for large symptomatic fibroid uterus were performed and these patients were followed. Out of this group, 72 uterine surgeries were performed in patients who had previously undergone myomectomy. Mean age at reoperation was 44.1 ± 0.6 years (range 33–53) and racial composition was as follows: Non-Hispanic White: 29.0%, African American: 52.2%, Hispanic: 10.1%, Other: 8.7%. The mean time from initial surgery to second surgery was 7.9 ±0.3 years (range: 5–15 years). Five hundred and ninety-two patients were followed and seventy-two patients (12%) from this group chose to undergo uterine surgery for their symptoms at a later date. Data was available for n = 69 patients ( Fig 1 ). Our cohort consisted of 72 patients who underwent reoperation for a recurrence of symptoms attributable to fibroid uterus after having undergone a primary abdominal myomectomy for large myomas. Data was available on n = 69 patients. Indications for uterine surgery after previous myomectomy in our sample consisted of: abnormal uterine bleeding secondary to fibroids (37.7%), dyspareunia/dysmenorrhea refractory to medical management (24.6%), pelvic pressure (13.0%), infertility (17.4%), as well as other symptoms, such as urinary issues, attributed to fibroid uterus (7.8%). Although patients were extensively counselled on alternatives, including hysterectomy, to repeat open myomectomy, patients cited the following reasons for uterine conserving surgery: desire to maintain fertility (40%), belief that having a uterus/cervix enhances sexual pleasure (50%), belief that having a cervix prevents prolapse (10%). The time between the first surgery and the patient presenting for her second uterine surgery was, in this study, not correlated with race, the number of fibroids or the aggregate weight of fibroids removed during the initial surgery nor with the operative time or EBL at initial procedure, unlike reported in other studies [ 25 ]. Time to reoperation was, however, significantly correlated with both age at initial surgery (correlation = -0.40, R2 = 0.16, p<0.001) and uterine size at reoperation (correlation = 0.42, R2 = 0.18, p<0.001) ( Fig 2A and 2B ). Published non-modifiable risk factors for recurrence include uterine size, premenopausal status and age [ 26 ], which our results confirmed. Reoperation a decade or more after the initial surgery was significantly correlated with younger age at the initial surgery (p = 0.031) and a larger uterus (p<0.008) ( Fig 2C and 2D ). Time between first surgery and patient presenting for repeat uterine surgery was correlated with ( A) age at initial surgery (correlation = -0.46, R2 = 0.21, p<0.001) and ( B) uterine size at reoperation (correlation = 0.41, R2 = 0.16, p<0.001). Reoperation a decade or more after initial surgery was significantly correlated with ( C) younger age at the initial surgery(* p<0.031) and ( D) larger uterus (* p<0.008). When presenting for repeat surgery, mean uterine size was 12.2 ±0.2 weeks (range: 10–20 weeks) consistent with ultrasound. This was a significantly (p<0.001) smaller than when they presented for their initial surgery, which had a mean uterine size of 20.8 ±0.5 weeks (range: 12–30 weeks). Presentation at recurrence of fibroids was highly variable and included bleeding, pelvic pressure and pain with a considerable number of patients complaining of a combination of these symptoms. Indication for pain was associated with older ages, while infertility was associated with younger ages (mean age of 46.7 vs 38.9 respectively, p<0.001). Uterine size at reoperation was found to be independent of race. The Triamcinolone group (n = 348) was compared to the “other method group” (n = 244) and no significant differences were found regarding age, number of fibroids removed, aggregate weight of fibroids, blood loss at surgery and initial uterine size. Data was analyzed to determine if the use of intraperitoneal triamcinolone with dextran (n = 36) or adhesion barriers such as Seprafilm (n = 2) or Interceed (n = 10) at the time of the initial myomectomy influenced timing of fibroid recurrence over untreated patients (n = 21). Neither steroids nor adhesion barriers significantly delayed timing for reoperation. Data was analyzed to determine if the use of triamcinolone would alter the rate of patients with symptoms necessitating repeat uterine surgery. Data from a previous study on postoperative adhesions consisted of a group of 348 patients in the arm receiving triamcinolone. These patients were followed longitudinally and 36 patients (10%) underwent subsequent uterine surgery during the study period. For comparison, 33 women, out of the 244 primary myomectomies that were performed using either no adhesion barriers or commercially available products as described previously, underwent repeat surgery with all but two choosing repeat myomectomy. This difference (10% vs 13%) was statistically significant (p<0.001).

This study followed a group of women who chose to undergo repeat surgery for recurrence of symptomatic fibroid uterus after an initial open myomectomy for large size myomatous uterus. Time to reoperation was correlated to age and to fibroid size but not to race or number of fibroids removed, which may be an important part of patient counselling, particularly in younger women. Further research is necessary to assess how gynecologists can prevent the recurrence of fibroids in patients who undergo uterus sparing procedures. Novel interventions, such as intraperitoneal triamcinolone at the time of initial surgery, may potentially play a role in reducing reoperation rates.

Admissions for uterine surgery after previous myomectomy from March 1, 1998 to June 2010 were identified by the St. Vincent’s Catholic Medical Center Manhattan medical records office and through office charts under IRB approval (#0104191MX). A subset of this deidentified dataset was previously queried for a separate unrelated study evaluating postoperative adhesions. Tabulated fields included, data from the initial surgery (age, uterine size, number and aggregate weight of fibroids removed, surgical time and estimated blood loss and initial surgeon name). Uterus size was determined clinically via bimanual exam performed by the attending surgeon and described in menstrual weeks as with the gravid uterus as well as by abdominal and transvaginal ultrasound. All women were continuously enrolled and followed during the study period (range: 1 to 12 years). Fields from the office visit when the surgeon evaluated the patient with new complaints included: symptoms, alternate treatments received thus far as well as uterine size. Patients were evaluated preoperatively with pelvic exam, transvaginal ultrasound, endometrial biopsy, liquid-based cervical smear, complete blood count, type and screen, coagulation studies, thyroid function tests, follicular stimulating hormone and serum human chorionic gonadotropin (hCG). Chart review documented that patients were appropriately counselled on alternative approaches including watchful waiting, medical management and minimally invasive surgical approaches including repeat myomectomy and hysterectomy. Data collected from the electronic hospital chart at reoperation included age, uterine size, surgical complications and name of surgeon performing the procedure. Total number of myomectomies performed as well as total number which involved the instillation of the intraperitoneal steroid triamcinolone, during a previous study at the institution, were obtained from medical records. We recorded, from the operative note, if any anti-adhesive methods were employed at all at the initial surgery or if the patient received no such treatment. This decision is left to the surgeon at our institution. Such methods potentially included either the use of adhesion barriers: principally at our institution Interceed (Johnson&Johnson, New Brunswick, NJ) or Seprafilm (Baxter, Deerfield, IL); or the instillation of intraperitoneal steroids. This latter method consisted of intraperitoneal placement of 200 mg triamcinolone acetonide suspension (Bristol-Myers-Squibb Pharmaceutical, Princeton NJ) in 500 mL dextran in the peritoneal cavity at the time of closure as previously published by our group [ 24 ]. Operative technique consisted of a Pfannenstiel incision, pitressin (20U in 60 mL normal saline) was slowly injected into the center of each myoma using an 18-guage spinal needle. Uterine incisions were made using electrocautery in a transverse direction to avoid the arcuate vessels and were carried through the serosa, myometrium and pseudocapsule using needlepoint electrode. The myoma was grasped with a tenaculum and the overlying myometrium and pseudocapsule were bluntly dissected off the myoma. Care was taken not to enter the uterine cavity unless the fibroid was classified as International Federation of Gynecology and Obstetrics (FIGO) type 2, 3 or 2–5, in which case removing the entire fibroid was prioritized. The defect was closed in multiple layers using running 0-polyglactin (Vicryl, Johnson&Johnson) suture. Data previously collected in an Excel Spreadsheet was analyzed using Prism (Graphpad, San Diego, CA) and Wizard Pro for Mac. Descriptive statistics were used to characterize the dataset. Pearson correlation was performed to test for correlation, Student t-test was used to compare groups assuming unequal variance (F-test) and two tails. Pearson Chi-squared test was performed on categorical data. Significance was set with p<0.05. This study was approved by the institutional review board (IRB) at St. Vincent’s Catholic Medical Centers Manhattan, (IRB# 010419M1X), and involved a retrospective chart review. All data was de-identified and informed consent was waved by the IRB for this retrospective chart review. All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and national research committee and with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards.

(XLSX) Click here for additional data file.