Uterine leiomyomas (myomas or fibroids) are a common cause of hysterectomy and an important clinical problem in gynecology. For women who wish to retain the potential for childbearing, abdominal myomectomy has been the prevailing option since first described in the 1890s.1 Recently, a variety of endoscopic alternatives to hysterectomy have also become widely used for removing leiomyomas, including laparoscopic and hysteroscopic myomectomy.
As more women choose alternatives to hysterectomy, the study of outcomes after these procedures is necessary to appropriately counsel patients regarding the choice of procedures. Good outcomes research has not been conducted for myomectomy, as it has for hysterectomy.2 However, clinical experience and pooled results of numerous small studies suggest that there is excellent resolution of symptoms of both menorrhagia and pelvic pressure after abdominal myomectomy.3
With all conservative procedures, however, there is a risk not seen with hysterectomy: what has been termed the recurrence of leiomyomas or, more appropriately, the risk of formation of new leiomyomas. The detection of new myomas using transvaginal ultrasound suggests that up to half of all women have detectable myomas 5 years after abdominal myomectomy.4–6 The need for an additional myomectomy or hysterectomy after abdominal myomectomy is also high and appears to be between 15% and 20% after approximately 10 years.3,7,8
Few studies have examined predictors of a second surgery after initial abdominal myomectomy. Childbirth after a myomectomy appears to decrease the recurrence risk.4,6 The number of myomas present at the time of initial surgery appears to affect recurrence, with an 11% risk of subsequent surgery among women with a single myoma and a recurrence risk of 26% among women with multiple myomas.7 There have been conflicting reports over whether the preoperative use of gonadotropin‐releasing hormone (GnRH)‐agonist increases recurrence risk.5,9
We sought to determine whether demographic and anthropometric characteristics present at the time of initial surgery could predict leiomyoma recurrence after abdominal myomectomy. Understanding these risk factors may allow us to appropriately counsel patients and develop strategies to prevent recurrent disease.
MATERIALS AND METHODS
We assessed the risk of a second surgery in a cohort of 73 women undergoing abdominal myomectomy 8 or more years before analysis. We term this initial surgery the index surgery. All of the index surgeries were performed by the same experienced attending surgeon, thus eliminating variability in surgical technique. All women were premenopausal, white, and none received GnRH‐agonist treatment. A retrospective analysis of hospital and office records recorded variables at the time of the index surgery including patient variables such as age, weight, gravidity, parity, and uterine size, as well as intraoperative and pathologic information. The medical records review also ascertained subsequent treatment for leiomyomas at our institution. These patients are part of a cohort described elsewhere.10
A survey was sent to all patients to gain further information regarding reproductive risk factors, which appear to influence leiomyoma formation. We also asked patients to report any procedures for leiomyomas performed at other institutions. Operative notes were reviewed for all self‐reported surgeries. The Institutional Review Board of Brigham and Women's Hospital approved the protocol for this study. All study procedures are in accordance with ethical standards set forth in the revised Declaration of Helsinki.
We classified recurrence in two different ways. We first used the traditional definition, a second abdominal myomectomy or hysterectomy during the period of observation. This is termed “major surgery” in our analysis. However, given the increasing array of minimally invasive surgeries, we also examined any surgery related to uterine leiomyomas as a second endpoint. This included laparoscopic and hysteroscopic myomectomy as well as dilatation and curettage (D&C) related to bleeding, but not other etiologies such as miscarriage. Operative notes were reviewed to determine inclusion. In our analysis, this is termed “any surgery.”
Information from both the medical records review and the survey were coded and entered into an Access database (Microsoft, Redmond, Washington, DC). Statistical analysis was carried out using SAS for Windows statistical software, 8.0 (SAS Institute, Cary, NC). In our description of the cohort, we reported means and standard deviations for continuous variables as well as frequency distributions for categoric variables. To assess differences between groups of women, Mantel‐Haenszel χ2 tests were performed for categoric variables, Student t tests were performed for normally distributed continuous variables, and Wilcoxon rank sum tests were performed for non‐normally distributed continuous variables. We used Cox proportional hazards regression to derive hazard ratios (HRs) and 95% confidence intervals (CIs) for time to fibroid recurrence (any surgery versus no surgery) stratified by selected covariates.11 We created two separate multivariate models in which we adjusted for: 1) age at index surgery only, and 2) age at index surgery as well as other predictors of recurrence in our dataset. The assumption of proportional hazards was verified using log(−log) survival plots for all univariate models.
Of the 73 women in our original cohort, we excluded three patients with documented myomectomies before the index surgery and two patients who did not have pathologically confirmed leiomyomas at the index surgery. Of the remaining 68 women, we further excluded three women who could not be followed beyond the date of surgery. Thus, 65 patients comprised our study sample. The mean length of follow‐up for the study sample was 83.6 ± 35.0 months. We were able to obtain completed surveys for 47 (72%) of these 65 subjects. Comparison of responders to nonresponders showed this subset to be a representative sample of the cohort (data not shown). In addition, the proportion of women who had a second surgery after initial myomectomy was similar for both the restricted dataset of 47 women who completed the survey (34% recurrence) and the full dataset of 65 women (35% recurrence).
Selected sociodemographic and anthropometric factors for the cohort, present at the time of index myomectomy are summarized in Table 1. Our sample is fairly typical of women with leiomyomas with most women in at least their fourth decade of life with a large proportion reporting menorrhagia (41%) or pelvic pressure (47%) (data not shown). At the time of the index surgery, a mean of approximately 10 myomas were removed per case with 8 cm as the mean size of the largest leiomyoma removed (Table 1).
Over the period of follow‐up, ten of the 65 women (15%) had a subsequent major surgery. If we include minimally invasive techniques, over one‐third of all women (35%) required second surgery. The majority of these surgeries were endoscopic myomectomies. Only two patients underwent a D&C related to abnormal bleeding, and one of these also underwent an endoscopic procedure. The mean interval to major surgery was 71.6 ± 36.5 months with 60.4 ± 38.6 months for any surgery. There was little difference between those who did or did not require a second surgery regarding age, body mass index, gravidity, or parity (Table 2). In addition, clinical symptoms preceding the initial myomectomy such as infertility, menorrhagia, dysmenorrhea, pelvic pressure, or pain did not distinguish women who did and did not go on to have subsequent surgeries.
There was, however, a significant inverse relationship between the clinically measured uterine size and risk of recurrence (Table 2). Women with no further surgery had significantly larger uteri than women having subsequent major surgery (14.5 ± 4.6 versus 10.9 ± 3.1 weeks, P = .032) or women having any surgery (14.5 ± 4.6 versus 11.2 ± 3.3 weeks, P = .006).
The same inverse relationship between uterine size and recurrence was also seen with the pathologically measured size of the largest myoma removed. Women with no recurrence had a significantly larger dominant fibroid than women undergoing major surgery (9.0 ± 4.3 versus 5.8 ± 3.0 cm, P = .030) or women having any subsequent surgery (9.0 ± 4.3 versus 6.5 ± 2.9 cm, P = .016). A similar trend was seen in size measured by preoperative ultrasound (Table 2).
Other than pathologic size of the largest fibroid removed, no other operative finding at the time of initial myomectomy distinguished those who did or did not go on to have recurrent surgeries. We did observe, however, that a greater proportion of women who had recurrent surgeries had hospital stays of 4 days or longer.
Uterine size as a predictor of any recurrent surgery also demonstrates significantly elevated relative risk when analyzed as age‐adjusted or multivariate HRs (Table 3). Recurrence was significantly decreased in women with myomatous uteri greater than 12 gestational weeks (HR 0.1, 95% CI 0.01, 0.4) after adjustment for age, weight gain, age at menarche, and history of endometriosis.
Although body mass index at the time of the index surgery did not differ between groups (24.9 ± 4.1 versus 24.7 ± 5.2 kg/m2), self‐reported weight gain at the time of the follow‐up survey appeared to be strongly predictive of recurrence. Women gaining at least 30 lb since age 18 had a nearly five‐fold increase in relative risk of recurrence when compared with women with less than a 10‐lb weight gain (HR 4.8, 95% CI 1.2, 18.5) after similar multivariate adjustments.
Decreasing education appeared to increase the risk of subsequent surgery for myomas just as it does for hysterectomy.12 Risk of recurrence was also increased in women who reported a history of endometriosis (multivariate HR 5.2, 95% CI 1.3, 20.2). Although it may be biologically plausible that two estrogen‐dependent diseases could interact to produce more significant disease, the fact that endometriosis itself might have led to the decision to proceed with another surgery cannot be ruled out. Finally, both later age at menarche and longer cycle lengths appeared to be correlated with increased risk of additional surgery (Table 3).
Neither oral contraceptive use nor parity protected against recurrent surgery, as they appear to do for initial fibroid formation. In fact, after multivariate modeling, parity caused a significant increase in risk of recurrent surgery (HR 5.0, 95% CI 1.1, 22.5).
A number of interesting issues are raised by our study. First, by several different measures, women with smaller uteri and/or myomas at the time of index surgery appear to have an increased risk of subsequent surgery. Although previous authors have focused on the number of myomas, the size may be equally important.7
Smaller leiomyomas may be more likely to be missed at the time of surgery. Alternatively, this may be a clinical marker for a different pathogenic pathway. Among myomas with cytogenetic abnormalities, translocations between chromosomes 12 and 14 [t(12;14)] have been associated with larger myomas, and deletions of the long arm of chromosome 7 [del(7)(q22q32)] have been associated with smaller leiomyomas.13,14 Some of these karyotypic subtypes may be more “virulent” and result in higher rates of recurrence. Studies correlating cytogenetic studies with clinical outcomes will be useful in elucidating the relationship between uterine and leiomyoma size and the risk of a second surgery.
These findings may also influence recommendations for prophylactic myomectomy: performing surgery while the uterus is relatively small to prevent increased surgical morbidity with a greater uterine size. Although the ACOG concluded there is no evidence supporting this practice, this study suggests there may indeed be harm.15 In the hands of a surgeon proficient in myomectomy, there may be an advantage in delaying surgery.
This study provides the impetus for examining recurrent surgery in women undergoing treatment with GnRH‐agonist before myomectomy because this medication has been used precisely to decrease both leiomyoma and uterine size.16–18 Studies with short‐term follow‐up have differed regarding whether there is an increased chance of recurrence after myomectomy with GnRH‐agonist.5,9
Weight gain after puberty may be important in increasing new myoma formation or growth of existing myomas after myomectomy. Studies are inconsistent regarding whether increased body mass index is predictive of increased risk of leiomyomas.19–22 However, data from the Nurses' Health Study support a role for weight gain after age 18 in increasing the relative risk of symptomatic leiomyomas.22 Certainly, if prevention of obesity led to a decrease in the risk of recurrent myoma surgery, this would be a benign and beneficial intervention.
It is notable that the relationship between weight gain and fibroid recurrence is seen in our relatively lean cohort. Although previous studies have suggested that increased weight gain may place women at increased risk of diverse diseases including stroke and asthma, most associations have been seen in women who are much heavier than those in our cohort.23,24 Although this may be due to other related factors such as exercise, the molecular genetics of leiomyomas may also be involved in this relationship. High mobility group A2 (HMGA2), initially termed high mobility group I‐C (HMGI‐C), is a transcriptional factor, which is upregulated in leiomyomas expressing the 12;14 chromosomal translocation.25 This gene is also involved in the regulation of body weight and obesity.26
The process that influences myoma recurrence may date back to at least the time of puberty, given the association of recurrence with age of menarche seen in this study. The direction of this association is also opposite to that seen for initial formation of fibroids where early menarche appears to increase risk.27
The fact that late age at menarche, increasing oral contraceptive use, and multiparity lead to a decreased risk of fibroid formation in epidemiologic surveys and increased recurrence in our study suggests that there are independent effects on recurrence after initial myomectomy from formation of clinically detectable fibroids. The fact that fibroid formation and growth appear to involve both transformation from a normal myocyte into a leiomyoma as well as a period of growth acceleration provides several different points where differential regulation may occur.
A major limitation to this study is that black women are not included in our cohort. Black women have an increased incidence and prevalence of disease and have also been found to have more severe disease at the time of hysterectomy.27–29 Thus, studying levels of recurrence and predictors of recurrence in this high‐risk population is especially important.
1. Bonney V. The technique and result of myomectomy. Lancet 1931;1:171–7.
2. Carlson KJ, Miller BA, Fowler FJ Jr. The Maine women's health study: I. Outcomes of hysterectomy. Obstet Gynecol 1994;83:556–65.
3. Buttram VC Jr, Reiter RC. Uterine leiomyomata: Etiology, symptomatology, and management. Fertil Steril 1981;36:433–45.
4. Candiani GB, Fedele L, Parazzini F, Villa L. Risk of recurrence after myomectomy. Br J Obstet Gynaecol 1991;98:385–9.
5. Fedele L, Vercellini P, Bianchi S, Brioschi D, Dorta M. Treatment with GnRH agonists before myomectomy and the risk of short-term myoma recurrence. Br J Obstet Gynaecol 1990;97:393–6.
6. Fedele L, Parazzini F, Luchini L, Mezzopane R, Tozzi L, Villa L. Recurrence of fibroids after myomectomy: A transvaginal ultrasonographic study. Hum Reprod 1995; 10:1795–6.
7. Malone LJ. Myomectomy: Recurrence after removal of solitary and multiple myomas. Obstet Gynecol 1969;34:200–3.
8. Acien P, Quereda F. Abdominal myomectomy: Results of a simple operative technique. Fertil Steril 1996;65:41–51.
9. Friedman AJ, Daly M, Juneau-Norcross M, Fine C, Rein MS. Recurrence of myomas after myomectomy in women pretreated with leuprolide acetate depot or placebo. Fertil Steril 1992;58:205–8.
10. Reilly R, Nour N. Abdominal myomectomy is associated with few surgical complications. J Gynecol Techniques 1998;4:107–12.
11. Allison P. Survival analysis using the SAS system: A practical guide. Cary, NC: SAS Institute, 1995.
12. Harlow BL, Barbieri RL. Influence of education on risk of hysterectomy before age 45 years. Am J Epidemiol 1999; 150:843–7.
13. Rein MS, Friedman AJ, Barbieri RL, Pavelka K, Fletcher JA, Morton CC. Cytogenetic abnormalities in uterine leiomyomata. Obstet Gynecol 1991;77:923–6.
14. Ligon AH, Morton CC. Genetics of uterine leiomyomata. Genes Chromosomes Cancer 2000;28:235–45.
15. American College of Obstetricians and Gynecologists. Surgical alternatives to hysterectomy in the management of leiomyomas. Washington, DC: American College of Obstetricians and Gynecologists; 2000.
16. Gerris J, Degueldre M, Peters AA, Romao F, Stjernquist M, al-Taher H. The place of Zoladex in deferred surgery for uterine fibroids. Zoladex Myoma Study Group. Hormone Res 1996;45:279–84.
17. Stovall TG, Muneyyirci-Delale O, Summitt RL Jr, Scialli AR. GnRH agonist and iron versus placebo and iron in the anemic patient before surgery for leiomyomas: A randomized controlled trial. Leuprolide Acetate Study Group. Obstet Gynecol 1995;86:65–71.
18. Zullo F, Pellicano M, De Stefano R, Zupi E, Mastrantonio P. A prospective randomized study to evaluate leuprolide acetate treatment before laparoscopic myomectomy: Efficacy and ultrasonographic predictors. Am J Obstet Gynecol 1998;178:108–12.
19. Ross RK, Pike MC, Vessey MP, Bull D, Yeates D, Casagrande JT. Risk factors for uterine fibroids: Reduced risk associated with oral contraceptives [published erratum appears in Br Med J (Clin Res Ed) 1986;293:1027]. Br Med J (Clin Res Ed) 1986;293:359–62.
20. Parazzini F, La Vecchia C, Negri E, Cecchetti G, Fedele L. Epidemiologic characteristics of women with uterine fibroids: A case-control study. Obstet Gynecol 1988;72:853–7.
21. Samadi AR, Lee NC, Flanders WD, Boring JR 3rd, Parris EB. Risk factors for self-reported uterine fibroids: A case-control study. Am J Public Health 1996;86:858–62.
22. Marshall LM, Spiegelman D, Manson JE, Goldman MB, Barbieri RL, Stampfer MJ, et al. Risk of uterine leiomyomata among premenopausal women in relation to body size and cigarette smoking. Epidemiology 1998;9:511–7.
23. Manson JE, Willett WC, Stampfer MJ, Colditz GA, Hunter DJ, Hankinson SE, et al. Body weight and mortality among women. N Engl J Med 1995;333:677–85.
24. Rexrode KM, Hennekens CH, Willett WC, Colditz GA, Stampfer MJ, Rich-Edwards JW, et al. A prospective study of body mass index, weight change, and risk of stroke in women. JAMA 1997;277:1539–45.
25. Schoenberg Fejzo M, Ashar HR, Krauter KS, Powell WL, Rein MS, Weremowicz S, et al. Translocation breakpoints upstream of the HMGIC gene in uterine leiomyomata suggest dysregulation of this gene by a mechanism different from that in lipomas. Genes Chromosomes Cancer 1996;17:1–6.
26. Anand A, Chada K. In vivo modulation of Hmgic reduces obesity. Nat Genet 2000;24:377–80.
27. Faerstein E, Szklo M, Rosenshein N. Risk factors for uterine leiomyoma: A practice-based case-control study. I. African-American heritage, reproductive history, body size, and smoking. Am J Epidemiol 2001;153:1–10.
28. Kjerulff KH, Erickson BA, Langenberg PW. Chronic gynecological conditions reported by US women: Findings from the National Health Interview Survey, 1984 to 1992. Am J Public Health 1996;86:195–9.
29. Marshall LM, Spiegelman D, Barbieri RL, Goldman MB, Manson JE, Colditz GA, et al. Variation in the incidence of uterine leiomyoma among premenopausal women by age and race. Obstet Gynecol 1997;90:967–73.