The relationship between uterine leiomyomata and pelvic floor symptoms.

A total of 211 eligible women were approached for the study. Five declined participation, 7 were unable to participate owing to language barriers. Fifty-four women did not complete the PFDI questionnaire; they were considered non-responders and excluded from analysis. 145 women completed the PFDI questionnaire and form our study group. Seventy-two women (49.6 %) returned the voiding diary. Comparisons were made between women with a uterus ≤12-week size (group I) and women with a uterus >12 week size (group II). Demographics of each group are shown in Table 1 ; there were no significant differences between groups. Self-reported pelvic floor symptoms are given in Table 2 . The most common bothersome symptoms were pressure in the low abdomen (74 %), urinary frequency (64 %), heaviness in the low abdomen (64 %), pain in the low abdomen (62 %), loss of urine with cough, sneeze, or other activity (59 %), loss of urine related to urgency (48 %), and strong sense of urgency to have a bowel movement (41 %). There was no difference between groups in reported bother from urinary frequency (63.1 % vs 64.3 %, p =0.89). Participants in group I reported more straining to have a bowel movement ( p =0.042). Participants in group II reported increased feeling of incomplete bladder emptying ( p =0.007) and increased sensation of difficulty emptying their bladder ( p =0.008). There was no significant difference in other pelvic floor symptoms. Scale scores for the components of the PFDI-20 and the PFIQ-7 are shown in Table 3 . There was no difference between groups with respect to the subscores for urinary, pelvic organ prolapse or colorectal/anal symptoms. In the subgroup of women who returned the voiding diary, the average number of voids in 24 h for women in group I was 8.1 (±9.8) ( n =30), while the number of voids for women in group II was 6.3 (±3.3) ( n =44; p =0.34). The number of voids reported by the voiding diary was significantly associated with the positive response to question 15 of the PFDI (Do you usually experience frequent urination?) ( p =0.36; p <0.001). There was no difference between groups in the number of urgency episodes in 24 h; group I had an average of 2.4 (±4.1) urgency episodes, compared with 1.2 (±1.6) for group II ( p =0.13). Since there are few data to suggest whether a size cut-off of 12 weeks is truly clinically relevant, different sized cut-offs were examined and receiver-operator characteristic (ROC) curves were performed. ROC curves were performed for each question and for the composite scores of the POPDI, UDI, CRADI and PFIQ. There was no clear cut-off point of leiomyoma size that predicted pelvic floor complaints (AUC for all curves <0.8). Ultrasound images were available for review for 139 of the 145 women who completed the questionnaires. Leiomyomata were noted in the anterior location in 32 cases (average diameter of leiomyoma 5.7 ± 3.9 cm), posterior in 52 cases (6.2 ±3.2 cm), submucosal in 6 (4.1 ± 0.9 cm), fundal in 35 (8.9 ± 4.3 cm), and pedunculated in 18 (7.8 ± 3.9 cm). Leiomyomata were noted to be predominantly anterior in 18 cases, predominantly posterior in 47 cases, and predominantly in another location in 58 cases. The remaining 16 ultrasounds did not show a single predominant location, and were excluded from this portion of the analysis. Responses to the PFDI-20 were analyzed by multiple regression. There was no significant difference in degree of bother for any individual question when grouped by leiomyoma location (anterior, posterior or other). Table 4 shows the scale scores for the components of the PFDI-20 and the PFIQ-7 and the average number of voids per day, where no significant difference was found when grouped by leiomyoma location. Planned 1 year follow up was performed. Subjects were contacted by telephone to complete a subset of the original PFDI-20 questionnaire. Of the original 145 subjects, 69 (47.6 %) responded to at least one question at 1 year follow up. Three declined to participate, and the remainder were disconnected, moved, wrong number or did not answer. Of those that responded, 40 had undergone interval surgery (at our institution: 19 total abdominal hysterectomy, 5 supracervical hysterectomy, 3 myomectomy, 3 laparoscopic hysterectomy, 3 vaginal hysterectomy, 4 hysteroscopy; at another institution: 1 hysterectomy of unclear type and 2 subjects reported surgery but were unsure of the type). Seven subjects had undergone anti-incontinence procedures (six slings and one Burch urethropexy) and were excluded from analysis for a total of 30 subjects. Table 5 shows the mean response for each question at enrollment and at 1 year follow up grouped by interval surgery. Significant improvement in most symptoms was seen in subjects who had undergone interval surgery, while those who had not had surgery had no significant change in symptoms on any question.

We performed a prospective, case–control survey of women seen in the Los Angeles County + University of Southern California Gynecology clinic between July 2010 and June 2011. The inclusion criteria were: women aged 18 or older presenting for gynecological care with uterine leiomyomata diagnosed by physical examination or ultrasound. The exclusion criteria were: inability to give informed consent; prior hysterectomy, anti-incontinence or prolapse surgery; adnexal mass >5 cm; referral for prolapse; endometriosis; interstitial cystitis; pelvic malignancy; and/or pelvic radiation. The institutional review board at the University of Southern California approved this protocol. Demographic information, including age, gravity, vaginal parity, medical comorbidities, menopausal status, prior surgery, and medications were recorded. Physical examination findings, including height, weight, estimated uterine size on bimanual examination, and grade of vaginal prolapse were recorded. Ultrasound reports or images (when available) were reviewed to confirm the presence of uterine leiomyomata. Ultrasound images were reviewed to assess the ultrasound estimated uterine volume and the location of uterine leiomyomata. Only leiomyomata of 3 cm or greater were considered clinically significant and were included for analysis. The size of 3 cm was chosen to be consistent with prior published studies [ 15 ] and because the average myometrial thickness at the fundus is approximately 2.5 cm [ 16 ]; therefore, a leiomyoma of greater than 2.5 cm is likely to distort the uterine contour. Location of the leiomyomata was assigned to anterior predominant, posterior predominant, or other location predominant by a reviewing radiologist. The anterior location was defined as superior to the endometrium, cranial to the cervix, and in the caudal two thirds of the uterus. Likewise, posterior location was defined as inferior to the endometrium, cranial to the cervix, and in the caudal two thirds of the uterus. Participants completed the validated short forms of the Pelvic Floor Distress Inventory (PFDI-20) and the Pelvic Floor Impact Questionnaire (PFIQ-7). These questionnaires have been validated in English and in Spanish [ 17 , 18 ]. A response of 0 (no symptoms) or 1 (no bother) were grouped for data analysis into the category of “no bother.” Responses of 2 through 4 (somewhat bothered, moderately bothered, and quite a bit bothered) were grouped for data analysis into the category of “bother.” PFDI scale scores were calculated by multiplication of the mean value of all answered items by 25 (range 0–100). The PFDI-20 summary score was the sum of the three subscales (range 0–300); if a subscale was not available for analysis, the summary score was excluded. The PFIQ-7 was scaled similarly, where the mean value for all answered items on the corresponding subscale were multiplied by (100/3) to obtain a single subscale score (range 0–100) and the PFIQ-7 summary score was the sum of the three subscales (range 0–300) [ 17 , 18 ]. Subjects were given the option to complete the questionnaire during their clinic visit, or to mail it to the investigators. Subjects were also asked to complete and mail in a 3-day voiding diary in a self-addressed, stamped envelope. At 1 year after enrollment, subjects were contacted via telephone interview to review whether they had had any interval surgery and to answer a shortened version (9 questions) of the PFDI-20. The subset of questions was chosen to include the question on pelvic pressure (Q1), the two questions found to be significantly different between groups on initial analysis (Q5 and Q7), and the complete UDI-6 (Q15–20). Participants with a uterus ≤12-week size (group I) were compared with those with a uterine size >12 week size (group II). There are few data to suggest what uterine size is clinically relevant; thus, the study was powered to detect differences using different cut-offs and additional comparisons planned. The period of 12 weeks was chosen because clinically, it appears that this is the size during pregnancy when the uterus enlarges out of the pelvis and theoretically would have a bulk effect on the proximal organs. The comparison group of patients with small uterine leiomyomata was chosen to address whether it is the bulk effect of the enlarged uterus on the pelvic floor rather than the presence of leiomyomatous changes that is associated with symptoms, and to attempt to minimize bias, since a comparison group of patients with no leiomyomata would likely include women who presented to our tertiary care clinic for evaluation of pain, prolapse, defecatory disorders, and/or incontinence. Differences between groups were compared using an unpaired t test, Chi-squared test or a Mann–Whitney U test, multiple regression or one-way ANOVA, as indicated. Fisher’s exact test was used for categories with fewer than five responses. Receiver operator characteristic curves were performed for each question of the PFDI and for the composite scales of each questionnaire, each outcome was plotted against uterine size in weeks, and an area under the curve (AUC) of <0.8 was considered significant. The primary outcome was subjective urinary frequency, defined as the affirmative response to question #15 of the PFDI “Do you usually experience frequent urination?” Assuming a baseline subjective urinary frequency prevalence of 25 % [ 19 ], to detect a difference of 25 %, with α=0.05 and β=0.2, at a 1:2 ratio we required 42 cases (group II) and 84 controls (group I). Assuming a 70 % completion rate, the target enrollment was 60 cases and 120 controls. Secondary outcomes were planned to evaluate the reported urinary frequency in a voiding diary. Assuming the median number of voids per day is 8 (range 4–18) [ 20 ], with α=0.05 and β=0.2 (adjusted for non-parametric nature of comparison) a total of 25 subjects per group is required to detect a difference of 3 voids per day.

Pelvic floor complaints are common among women presenting for gynecological care. We found the prevalence of urinary frequency to be nearly 65 %; stress incontinence 60 %; and urgency incontinence 50 % in women with uterine leiomyomata. This is significantly higher than the 25 % of women with urinary urgency and 15–33 % with urinary incontinence reported in the general population [ 19 , 21 , 22 ], but similar to the 60 % of women with urinary urgency and 45–54 % with urinary incontinence reported in the prospective cross-sectional study of women with large leiomyomata planning to undergo hysterectomy or myomectomy [ 22 ]. One recent prospective cross-sectional study looked at 78 women who answered the Bristol Female Lower Urinary Tract Symptom Scored Form (BLUTS-SF) questionnaire. Similar to the current study, they found that the most prevalent symptoms were urinary urgency (59 %) and urinary incontinence (45–54 %). When subdivided by individual symptoms, they found no difference in the mean uterine size for the symptoms of nocturia, frequency, urgency incontinence, stress incontinence, or voiding dysfunction. They did find that subjects complaining of urinary urgency had larger uterine volumes (516 cm 3 vs 894 cm 3 , p =0.017) than those who denied urgency [ 15 ]. Similarly, our study did not find a difference in the symptoms of frequency, urgency incontinence or stress incontinence by uterine size. The PFDI-20 does not have a question specifically related to urinary urgency; however, our study did evaluate urgency as self-reported by a voiding diary. We did not find a difference between groups in the number of urge-related voiding episodes over 3 days, but only 49.6 % of our sample completed the voiding diary. The diary data were powered to detect a difference in number of voids per day and was potentially underpowered to detect a true difference in number of urgency episodes per day. When compared with women with small leiomyomatous uteri, those with large leiomyomatous uteri are more likely to complain of difficulty with bladder emptying. It is unclear whether these symptoms are due to the pressure of the uterus on the bladder causing a sensation of incomplete bladder emptying or an actual physical obstruction of the bladder outlet leading to urinary retention. The absence of other obstructive symptoms, such as urinary frequency and urgency incontinence, suggests the former. Post-void residuals were not measured in this study, which may help differentiate between pressure on the bladder and physical bladder outlet obstruction. Surprisingly, women with smaller leiomyomatous uteri were more likely to endorse straining to defecate compared with women with larger leiomyomatous uteri. We would expect any effect of leiomyomata on defecatory symptoms to be worse with larger leiomyomata, be it due to mass effect, location of the leiomyoma or to some unknown secreted factor. This finding could be explained by the fact that uteri with small leiomyomata remain in the pelvis and may exert more pressure effect on the rectosigmoid. However, it may also represent statistical variation owing to the large number of comparisons or selection bias of the women presenting to the clinic. This study recruited from a tertiary referral center, where most patients are referred for some gynecological complaint. Patients with small leiomyomatous uteri are more likely to have had their leiomyomata detected while evaluating another complaint, such as pain, constipation or bleeding. Community-dwelling women with leiomyomata being seen for routine health maintenance would have made a better control group, but we did not have access to this type of population. The current study failed to show an association between leiomyoma location and urinary, prolapse, or colorectal/anal symptoms. However, a recent prospective cross-sectional study did show an association between anterior leiomyomata and voiding dysfunction [ 15 ]. The effect of leiomyoma location is likely highly variable, owing to: the rotational effect of lateral leiomyomata, the anterior displacement of the fundus due to posterior leiomyomata, the tendency of the bladder to scar the anterior uterus after cesarean section, and the high anatomical variability of the uterine position within the pelvis. At 1-year follow-up, we found that women who had undergone surgery were significantly more likely to report improvement in symptoms. This is consistent with other studies that have shown improvement in urinary symptoms after hysterectomy or myomectomy [ 14 , 23 – 27 ]. It is postulated that surgical intervention alone may have an effect on pelvic floor symptoms [ 24 ], an idea that is supported by our findings that uterine size is not necessarily associated with pelvic floor symptoms. Our findings suggest that bulky uterine leiomyomata are associated with a sensation of incomplete bladder emptying. Other pelvic floor symptoms do not appear to be affected by uterine size. Despite these findings, it is certainly possible that a certain subset of leiomyomata may cause additional pelvic floor symptoms. It may be that the presently used cut-off of 3 cm is too inclusive, and only larger anterior myomas affect the bladder, a notion supported by Parker-Autry et al., who found that urinary symptoms were associated with a larger anterior leiomyoma volume, compared with those without urinary symptoms [ 15 ]. Solitary leiomyomata may have a stronger effect based on location than those with multiple leiomyomata in various locations. Furthermore, lateral or broad ligament leiomyomata (which are particularly difficult to assess on ultrasound) may rotate the uterus, which could kink the ureter, or apply irregular pressure to the bladder, potentially causing urinary symptoms. Finally, certain low anterior or posterior leiomyomata may push the cervix anteriorly, obstructing the urethra, as has been described in extreme cases where the enlarged uterus is incarcerated under the sacral promontory. Such incarceration of the fundus can pivot the uterus along its short access such that the cervix is pressed anteriorly against the pubic symphysis, obstructing the urethra and resulting in acute urinary retention. The strengths of this study include its prospective nature, large sample size, the use of a validated questionnaire, objective measures (voiding diary) and the comparison with a control group of women with small leiomyomata. Weaknesses include the lack of generalizability because of a patient population that consisted primarily of Hispanic, low-income women recruited from a tertiary care facility, and the subjective nature of questionnaire studies. Also, although 199 women agreed to participate, only 74 % completed the PFDI questionnaire, which may misrepresent the true prevalence of symptoms owing to response bias. Even with our relatively large sample size, the effect of outliers may affect our results, particularly in the questions with few affirmative answers. While leiomyomata are a significant health concern and the most common indication for hysterectomy [ 28 ], little is known about the effects of leiomyomata on the lower urinary tract and pelvic floor. Our study indicates that women with large leiomyomatous uteri are more likely to report symptoms of incomplete bladder emptying than women with small uteri. Furthermore, women with leiomyomata report high rates of pelvic floor symptoms, regardless of overall uterine size, and these symptoms may improve after surgical intervention. The evaluation of uterine leiomyomata should include a complete assessment of pelvic floor symptoms, and women who have large leiomyomata may expect symptoms of incomplete bladder emptying to improve after surgical intervention.

Uterine leiomyomata are benign, monoclonal tumors of the smooth muscle of the myometrium [ 1 ]. The exact prevalence is unknown, since not all leiomyomata are symptomatic; however, estimates range from 40 to 70 % depending on age, race, and other risk factors [ 2 , 3 ]. Many women with leiomyomata are asymptomatic [ 3 , 4 ]. However, of women with symptoms, up to 62 % have more than one [ 5 ]. The most commonly cited symptoms of leiomyomata are: abnormal bleeding, pelvic pain, urinary symptoms, “bulk symptoms,” colorectal symptoms, and infertility [ 1 , 6 – 9 ]. Despite being commonly described symptoms of leiomyomata, there are few experimental data to support a relationship between leiomyomata and urinary or colorectal symptoms. Case reports have described uterine leiomyomata causing urinary retention and/or hydronephrosis [ 10 – 12 ]. It is proposed that large leiomyomata might cause urinary frequency, urinary urgency, constipation or tenesmus via mass effect on the bladder and rectosigmoid [ 8 , 13 , 14 ] and anterior leiomyomata appear to be more associated with lower urinary tract symptoms than leiomyomata in other locations [ 15 ]. The purpose of our study is to examine the prevalence of pelvic floor symptoms in women with leiomyomatous uteri of more than 12 weeks compared with women with leiomyomatous uteri less than or equal to 12 weeks and to evaluate the resolution of these symptoms after surgical intervention.