Background
It is unknown how initial cervix location and cervical support resistance to traction, which we term “apical support stiffness,” compare in women with different patterns of pelvic organ support. Defining a normal range of apical support stiffness is important to better understand the pathophysiology of apical support loss.
Objective
The aims of our study were to determine whether: (1) women with normal apical support on clinic Pelvic Organ Prolapse Quantification, but with vaginal wall prolapse (cystocele and/or rectocele), have the same intraoperative cervix location and apical support stiffness as women with normal pelvic support; and (2) all women with apical prolapse have abnormal intraoperative cervix location and apical support stiffness. A third objective was to identify clinical and biomechanical factors independently associated with clinic Pelvic Organ Prolapse Quantification point C.
Study Design
We conducted an observational study of women with a full spectrum of pelvic organ support scheduled to undergo gynecologic surgery. All women underwent a preoperative clinic examination, including Pelvic Organ Prolapse Quantification. Cervix starting location and the resistance (stiffness) of its supports to being moved steadily in the direction of a traction force that increased from 0-18 N was measured intraoperatively using a computer-controlled servoactuator device. Women were divided into 3 groups for analysis according to their pelvic support as classified using the clinic Pelvic Organ Prolapse Quantification: (1) “normal/normal” was women with normal apical (C < –5 cm) and vaginal (Ba and Bp < 0 cm) support; (2) normal/prolapse had normal apical support (C < –5 cm) but prolapse of the anterior or posterior vaginal walls (Ba and/or Bp ≥ 0 cm); and (3) prolapse/prolapse had both apical and vaginal wall prolapse (C > –5 cm and Ba and/or Bp ≥ 0 cm). Demographics, intraoperative cervix locations, and apical support stiffness values were then compared. Normal range of cervix location during clinic examination and operative testing was defined by the total range of values observed in the normal/normal group. The proportion of women in each group with cervix locations within and outside the normal range was determined. Linear regression was performed to identify variables independently associated with clinic Pelvic Organ Prolapse Quantification point C.
Results
In all, 52 women were included: 14 in the normal/normal group, 11 in the normal/prolapse group, and 27 in the prolapse/prolapse group. At 1 N of traction force in the operating room, 50% of women in the normal/prolapse group had cervix locations outside the normal range while 10% had apical support stiffness outside the normal range. Of women in the prolapse/prolapse group, 81% had cervix locations outside the normal range and 8% had apical support stiffness outside the normal range. Similar results for cervix locations were observed at 18 N of traction force; however the proportion of women with apical support stiffness outside the normal range increased to 50% in the normal/prolapse group and 59% in the prolapse/prolapse group. The prolapse/prolapse group had statistically lower apical support stiffness compared to the normal/normal group with increased traction from 1-18 N (0.47 ± 0.18 N/mm vs 0.63 ± 0.20 N/mm, P = .006), but all other comparisons were nonsignificant. After controlling for age, parity, body mass index, and apical support stiffness, cervix location at 1 N traction force remained an independent predictor of clinic Pelvic Organ Prolapse Quantification point C, but only in the prolapse/prolapse group.
Conclusion
Approximately 50% of women with cystocele and/or rectocele but normal apical support in the clinic had cervix locations outside the normal range under intraoperative traction, while 19% of women with uterine prolapse had normal apical support. Identifying women whose apical support falls outside a defined normal range may be a more accurate way to identify those who truly need a hysterectomy and/or an apical support procedure and to spare those who do not.
Introduction
Pelvic organ prolapse is a common indication for gynecologic surgery, with the annual number of women undergoing these procedures projected to reach >190,000 by 2020. While our understanding of the pathophysiology of pelvic organ prolapse has improved over the last decade–especially the importance of apical support –much remains unknown regarding the biomechanical properties of the affected tissues and pelvic structures. A clinical evaluation of apical support is used to inform surgeons’ decision-making as to whether a hysterectomy is needed as part of surgery for prolapse. However, there is a 50% disagreement rate among gynecologic surgeons about the level of apical support, assessed under traction in the operating room, that indicates the need for hysterectomy. Moreover, evidence about how to integrate Pelvic Organ Prolapse Quantification (POP-Q) and intraoperative findings in this assessment do not yet exist. At the heart of this issue is the fact that surgeons use their own assessment of whether the apical supports are normal under traction to make decisions about whether or not hysterectomy and/or apical suspension should be considered. An objective assessment of apical support stiffness along with outcome data could provide better information on which to base these important clinical decisions.
It is well established that the degree of uterine descent seen during maximal Valsalva in clinic is not as pronounced as that seen in the operating room under traction. This difference may be partially explained by different environmental and loading conditions (eg, Valsalva vs traction, leg position, anesthesia) to which the apical ligaments are subjected. To gain a better understanding of this, we developed a technique to measure the intraoperative mechanical properties of the apical ligaments. In pilot studies, we found that ligament stiffness only accounts for 19% of variation in POP-Q point C. What is missing now is a quantitative understanding of how various properties vary in women with different types of pelvic organ support. For example, do women with normal apical support, but with cystocele or rectocele on POP-Q examination, have the same mechanical properties for the cardinal/uterosacral complex as do those with normal support in all areas on clinic POP-Q? Conversely, do all women with apical prolapse on POP-Q examination have abnormal ligament properties? The answers to these questions could have direct implications for deciding whether or not the additional morbidity of a hysterectomy and/or apical suspension is justified at the time of an operation for prolapse.
The aims of our study were to: determine whether: (1) women with normal apical support on clinic POP-Q, but with vaginal wall prolapse (cystocele and/or rectocele), have the same intraoperative cervix location and apical support stiffness as women with normal pelvic support (null hypothesis is that there is no difference); and (2) all women with apical prolapse have abnormal intraoperative cervix location and apical support stiffness (null hypothesis is that all women with apical prolapse have abnormal intraoperative cervix location and apical support stiffness). A third objective was to identify clinical and biomechanical factors independently associated with clinic POP-Q point C.
Materials and Methods
Women with a full spectrum of pelvic organ support were recruited from the gynecology clinics at the University of Michigan from September 2012 through September 2013. Informed consent was obtained for all participants under an approved University of Michigan Institutional Review Board protocol (HUM00056743). Inclusion criteria included women ≥18 years of age who were planning to have gynecologic surgery and willing to undergo intraoperative testing. Exclusion criteria included: pregnancy (either currently or within the past year), prior hysterectomy or surgery for pelvic organ prolapse, uterine fibroids >12 weeks in size or known pelvic inflammatory disease, chronic steroid use, prior pelvic radiation, current treatment for cancer, history of organ transplant, history of vasovagal syncope, neurologic diseases or impairments, and mobility issues that would prohibit leg positioning in high lithotomy.
Preoperatively, women were examined and their pelvic organ support measured using the POP-Q at inclination of 45 degrees. Intraoperative testing was conducted on the day of their scheduled gynecologic surgery. The technique for making measurements of cervix location and apical ligament response to traction in the operating room using a computer-controlled servoactuator device has been previously described. To summarize, after induction of general anesthesia, the patient was positioned into high lithotomy and a short-blade Scherbak posterior-weighted vaginal speculum was placed. A single-tooth tenaculum was then placed across both the anterior and posterior cervical lips and the handle attached to the servoactuator device. Prior to activating the device, resting cervix location was determined by measuring the distance of the lateral cervical edge to the hymenal ring. The servoactuator then moved at a constant speed so as to apply an increasing tensile force (from 1-18 N) to the cervix while the position of the traction arm was simultaneously recorded, so that cervix locations at minimal (1 N) and maximal (18 N) force could then be determined. Because the pelvis does not move during testing, the location of the cervix was used as a proxy for ligament length. During early trials, a video recording of the hips from a lateral view was made to assure that the patient did not move on the table during traction. Change in cervix location was defined as the difference in cervix location (mm) from rest to either 1 N or 18 N, and the apical support stiffness was estimated by dividing the change in force (0-1 N and 1-18 N) by the measured change in cervical location.
Study participants were divided into 3 groups for analysis according to their pelvic support and cervix location as measured by a trained urogynecologist during the clinic POP-Q examination. We defined apical prolapse as C > –5 cm and vaginal wall prolapse as Ba and/or Bp ≥ 0 cm based on population norms. The normal/normal group included women with normal apical (C ≤ –5 cm) and vaginal (Ba and Bp < 0 cm) support; normal/prolapse included women with normal apical support (C ≤ –5 cm) but prolapse of the anterior or posterior vaginal walls (Ba and/or Bp ≥ 0 cm); and prolapse/prolapse included women with both apical and vaginal wall prolapse (C > –5 cm and Ba and/or Bp ≥ 0 cm).
The demographic and clinic POP-Q data were compared for the 3 pelvic support groups using simple linear regression models. Statistically significant differences between the groups were found for age and parity measures. A multivariable linear regression model that adjusted for age and parity allowed us to confirm that group differences in stiffness and displacement measures were not due to underlying differences in age and parity measures. After controlling for age and parity, cervix locations at the following traction forces were compared across and between groups: in clinic during maximal Valsalva and in the operating room at 1 N and 18 N of force. Normal range of cervix location during clinic examination and operative testing was defined by the range observed in the normal/normal group. The proportion of women in each group with cervix locations outside and within the normal range was then calculated for each of the 3 traction forces.
Pearson correlation coefficients were used to assess the association between POP-Q point C and the cervix location and stiffness measurements for all participants combined, as well as separately for each group. Multivariable linear regressions stratified by group were further used to assess the outcome of POP-Q point C as a function of cervix location at 1 N, stiffness, body mass index (BMI), age, and parity. The results of the regression models include r 2 and adjusted r 2 values that reflect the proportion of variance explained in POP-Q point C and are of primary interest. All analyses were conducted in software (Stata Statistical Software, Release 13; StataCorp LP, College Station, TX) Two-sided statistical significance was determined by an alpha value of 5%.
An a priori power calculation was not possible because the only prior data available did not have the relevant information regarding patient groupings. A post hoc power calculation based on group differences for ligament stiffness and cervix displacement demonstrated full (100%) statistical power when assuming the effect sizes presented in the “Results” section below.
Materials and Methods
Women with a full spectrum of pelvic organ support were recruited from the gynecology clinics at the University of Michigan from September 2012 through September 2013. Informed consent was obtained for all participants under an approved University of Michigan Institutional Review Board protocol (HUM00056743). Inclusion criteria included women ≥18 years of age who were planning to have gynecologic surgery and willing to undergo intraoperative testing. Exclusion criteria included: pregnancy (either currently or within the past year), prior hysterectomy or surgery for pelvic organ prolapse, uterine fibroids >12 weeks in size or known pelvic inflammatory disease, chronic steroid use, prior pelvic radiation, current treatment for cancer, history of organ transplant, history of vasovagal syncope, neurologic diseases or impairments, and mobility issues that would prohibit leg positioning in high lithotomy.
Preoperatively, women were examined and their pelvic organ support measured using the POP-Q at inclination of 45 degrees. Intraoperative testing was conducted on the day of their scheduled gynecologic surgery. The technique for making measurements of cervix location and apical ligament response to traction in the operating room using a computer-controlled servoactuator device has been previously described. To summarize, after induction of general anesthesia, the patient was positioned into high lithotomy and a short-blade Scherbak posterior-weighted vaginal speculum was placed. A single-tooth tenaculum was then placed across both the anterior and posterior cervical lips and the handle attached to the servoactuator device. Prior to activating the device, resting cervix location was determined by measuring the distance of the lateral cervical edge to the hymenal ring. The servoactuator then moved at a constant speed so as to apply an increasing tensile force (from 1-18 N) to the cervix while the position of the traction arm was simultaneously recorded, so that cervix locations at minimal (1 N) and maximal (18 N) force could then be determined. Because the pelvis does not move during testing, the location of the cervix was used as a proxy for ligament length. During early trials, a video recording of the hips from a lateral view was made to assure that the patient did not move on the table during traction. Change in cervix location was defined as the difference in cervix location (mm) from rest to either 1 N or 18 N, and the apical support stiffness was estimated by dividing the change in force (0-1 N and 1-18 N) by the measured change in cervical location.
Study participants were divided into 3 groups for analysis according to their pelvic support and cervix location as measured by a trained urogynecologist during the clinic POP-Q examination. We defined apical prolapse as C > –5 cm and vaginal wall prolapse as Ba and/or Bp ≥ 0 cm based on population norms. The normal/normal group included women with normal apical (C ≤ –5 cm) and vaginal (Ba and Bp < 0 cm) support; normal/prolapse included women with normal apical support (C ≤ –5 cm) but prolapse of the anterior or posterior vaginal walls (Ba and/or Bp ≥ 0 cm); and prolapse/prolapse included women with both apical and vaginal wall prolapse (C > –5 cm and Ba and/or Bp ≥ 0 cm).
The demographic and clinic POP-Q data were compared for the 3 pelvic support groups using simple linear regression models. Statistically significant differences between the groups were found for age and parity measures. A multivariable linear regression model that adjusted for age and parity allowed us to confirm that group differences in stiffness and displacement measures were not due to underlying differences in age and parity measures. After controlling for age and parity, cervix locations at the following traction forces were compared across and between groups: in clinic during maximal Valsalva and in the operating room at 1 N and 18 N of force. Normal range of cervix location during clinic examination and operative testing was defined by the range observed in the normal/normal group. The proportion of women in each group with cervix locations outside and within the normal range was then calculated for each of the 3 traction forces.
Pearson correlation coefficients were used to assess the association between POP-Q point C and the cervix location and stiffness measurements for all participants combined, as well as separately for each group. Multivariable linear regressions stratified by group were further used to assess the outcome of POP-Q point C as a function of cervix location at 1 N, stiffness, body mass index (BMI), age, and parity. The results of the regression models include r 2 and adjusted r 2 values that reflect the proportion of variance explained in POP-Q point C and are of primary interest. All analyses were conducted in software (Stata Statistical Software, Release 13; StataCorp LP, College Station, TX) Two-sided statistical significance was determined by an alpha value of 5%.
An a priori power calculation was not possible because the only prior data available did not have the relevant information regarding patient groupings. A post hoc power calculation based on group differences for ligament stiffness and cervix displacement demonstrated full (100%) statistical power when assuming the effect sizes presented in the “Results” section below.
Results
In all, 52 women were included in the study: 14 in the normal/normal group, 11 in the normal/prolapse group, and 27 in the prolapse/prolapse group. Data from 17 of these women were included in the description of the testing strategy. Of the 14 women in the normal/normal group, 5 had a midurethral sling, 7 had a total laparoscopic hysterectomy, 1 had a colposcopy, and 1 had a vaginal hysterectomy, none of which were done for an indication of pelvic organ prolapse. Procedures for the 11 women in the normal/prolapse group were as follows: vaginal hysterectomy ± anterior repair, posterior repair, uterosacral or sacrospinous ligament suspension, midurethral sling (N = 7); anterior and posterior repair and midurethral sling (N = 3); and midurethral sling and endometrial ablation (N = 1). Women in the prolapse/prolapse group underwent the following procedures: vaginal hysterectomy ± anterior repair, posterior repair, uterosacral or sacrospinous ligament suspension, midurethral sling (N = 17); laparoscopic supracervical hysterectomy, colpopexy ± anterior repair, posterior repair, midurethral sling (N = 5); total abdominal hysterectomy ± anterior repair, posterior repair, colpopexy (N = 3); midurethral sling (N = 1); and total laparoscopic hysterectomy and midurethral sling (N = 1). Demographics of the 3 groups are shown in Table 1 . Women in the normal/normal group were the youngest, had the lowest parity, and registered the highest BMI. By design, women in the normal/normal and normal/prolapse groups had similar cervix locations on POP-Q examination, while point C in the prolapse/prolapse group was closer to the hymen.
