Transvaginal Mesh Surgery Is Sufficient to Repair Pelvic Organ Prolapse Compared to Laparoscopic Sacrocolpopexy: A Propensity Score Matching Analysis
Article information
Abstract
Purpose
Both laparoscopic sacrocolpopexy (LSC) and transvaginal mesh surgery (TVM) may provide favorable outcomes for patients with pelvic organ prolapse (POP). However, in our experience, LSC tends to be performed for severe POP and TVM for POP of a moderate or earlier stage. This study aimed to examine the postoperative results of the 2 groups including the preoperative parameters as closely as possible.
Methods
We retrospectively evaluated 238 patients with POP who underwent LSC or TVM at National Defense Medical College Hospital between September 2014 and January 2024. In our hospital, 94 patients were treated by LSC, whereas 144 were treated by TVM. After using nearest-neighbor propensity score matching, the surgical outcomes were compared between the matched 76 patients treated with LSC and 76 treated with TVM.
Results
To perform nearest-neighbor propensity score matching, the body mass index (BMI), POP stage, and number of laparotomies were used as explanatory variables. After propensity score matching, no significant differences in BMI, POP stage, and number of previous laparotomies were found between the matched LSC and TVM groups. Significant differences in age, blood loss, and operating time still existed even after propensity score matching. No significant differences in the occurrence of urinary incontinence or mesh exposure or prolapse recurrence were found between the matched LSC and TVM groups in all patients and patients with stage 4 POP. Moreover, no significant differences in the time to prolapse recurrence or urinary incontinence were found between the matched LSC and TVM groups (P=0.627 and P=0.235, respectively).
Conclusions
Based on the analysis using nearest-neighbor propensity score matching, both surgical methods were effective treatment approaches of POP. TVM would be sufficient to treat POP given its shorter operative time and low complication rates.
INTRODUCTION
Pelvic organ prolapse (POP) is characterized by the descent of one or more of the anterior or posterior vaginal walls, uterine cervix, or vaginal vault following hysterectomy [1]. Although POP can be painful and uncomfortable, it is not a life-threaten-ing condition. Cystocele, hysterocele, or both are frequent outcomes of vaginal prolapse, which can cause pelvic discomfort, a bulge or protrusion outside the vaginal opening, a feeling of fullness, and changes in bladder and bowel function among women. Mild vaginal prolapse might not be extremely uncomfortable or require prompt treatment. This form can be treated conservatively, such as through lifestyle changes, pelvic floor exercises, and vaginal pessaries; these treatments are sufficient to reduce symptoms [2,3]. Chronic symptoms including pelvic discomfort, a feeling of fullness or heaviness in the pelvis, and abnormal bladder and intestine function may indicate the need for surgical treatment [4].
Several surgical techniques can be selected for POP repair. Robot-assisted sacrocolpopexy has recently been recognized as the leading sacrocolpopexy approach, whereas in Japan, laparoscopic sacrocolpopexy (LSC) or transvaginal mesh surgery (TVM) have been performed for POP [5]. Both LSC and TVM may provide favorable outcomes [5,6]. In some studies, LSC outcomes tended to be more favorable than TVM outcomes in patients with POP stage 3 or higher [6,7]. However, in our previous studies, TVM outcomes are superior to or equal to those of LSC performed in patients with POP stage 3 or lower [8,9]. In other words, TVM may not be suitable for stage 4 POP. To compare the surgical outcomes of these 2 methods, parameters must be matched to exclude selection bias.
Up to this point, we have considered the body mass index (BMI), POP stage, and previous history of laparotomy in determining whether to perform LSC or TVM. For example, TVM was chosen for patients with BMI >30 kg/m2 and cystocele, whereas LSC was performed for patients with BMI <22 kg/m2 and stage 4 hysterocele. In other cases, TVM was performed for patients with a previous history of several laparotomies and cystocele, and LSC was indicated for patients with stage 3 hysterocele who have not experienced any laparotomies and BMI <20 kg/m2. Therefore, a propensity score matching analysis (nearest-neighbor propensity score matching) was required to reduce the potential for selection bias. In this study, patients with POP who underwent LSC or TVM were compared by eliminating significant differences in key factors as much as possible by propensity score matching. We conducted this study to confirm the expectation that both LSC and TVM would have favorable postoperative outcomes under comparable conditions.
MATERIALS AND METHODS
Patients
We retrospectively reviewed and analyzed clinicopathological data from the medical records of 238 patients with POP who underwent LSC or TVM at National Defense Medical College Hospital between September 2014 and January 2024. We examined patients who underwent either surgery within the abovementioned period, and followed them for at least 1 year postoperatively at our institution. Indications for surgery included vaginal prolapse symptoms, which meant the POP stage of ≥2 [10], or hydronephrosis and/or hydroureter caused by POP even in asymptomatic cases. All procedures performed in this study were in accordance with the tenets of the 2013 revision of the Declaration of Helsinki. The study protocol (Saitama, Japan; ID 4219) was accepted on August 21, 2020, by the institutional ethics committee of National Defense Medical College, and all patients who participated in this study provided written informed consent forms. This study included patients who underwent LSC or TVM for POP within the above-mentioned period, and excluded those who declined to participate.
The median follow-up period after surgery was 12.3 (interquartile range, 12.2–24.2) months. We recorded types of surgery, age, BMI, POP staging system, previous laparotomy, blood loss, operative time, and major complications intraoperatively.
Surgical Procedure
One urologist performed LSC in accordance with the method reported by Wattiez et al. [11]. Briefly, after port placement, the patient was positioned in a 15° Trendelenburg position to improve the visualization of the pelvic cavity by shifting the bowel cranially. The key steps were as follows:
(1) Anterior vaginal wall dissection: The space between the anterior vaginal wall and bladder was dissected, and the anterior wall mesh—either Polyform (Boston Scientific Japan, Japan) or ORIHIME (Kono Seisakusho, Japan)—was secured.
(2) Subtotal hysterectomy and adnexectomy: These procedures were performed if the uterus and adnexa were present.
(3) Posterior wall dissection: The posterior vaginal wall was separated from the rectum, and the posterior wall mesh was secured.
(4) Mesh retroperitonealization: The mesh was covered, and the pelvic peritoneum was partially closed.
(5) Peritoneal closure: The peritoneum was completely closed after securing the mesh to the anterior longitudinal ligament at the level of the promontory.
Three urologists performed TVM according to previously described methods [8,9]. Briefly, the procedure commenced with hydrodissection. Subsequently, the anterior vaginal wall was incised vertically, and the pubocervical fascia was fully dissected laterally using a blunt technique to expose the sacrospinous ligaments. Then, a skin incision was made 4 cm laterally and 3 cm inferior to the anal center. Targeting a position of one or 2 finger breadths medial to the ischial spine, a Shimada needle threaded with nylon monofilament sutures was advanced to further penetrate the sacrospinous ligaments from the incision. The arms of Polyform mesh or ORIHIME mesh were then removed with nylon monofilament loops, and an acceptable shape was created, spread, and secured under the bladder. The meshes were cut to fit the stencil paper with 2 arms in advance. In the Polyform group, the mesh arm width was approximately 3 cm. In the ORIHIME group, the initial arm width was set to 4.5−6 cm, which was later increased to approximately 7.5−8 cm [8,9]. Finally, traction was employed over the externalized arms to ensure proper alignment, and the vaginal wound was closed using 2-0 Vicryl (Johnson and Johnson, Tokyo) sutures.
Assessment Methods for Preoperative and Postoperative Parameters
Prolapse recurrence was defined as the most dependent portion being at POP stage ≥2, which means that the most distal prolapse portion is ≥−1 cm from the hymen plane, via physical examination according to Takazawa et al. [12]. The occurrence of postoperative urinary incontinence (UI) was defined by a daily life disturbance with incontinence regardless of the type of stress UI (SUI), urge UI (UUI), or mixed UI such as the use of several pads per day.
Statistical Analysis
The Kruskal-Wallis test or unpaired Student t-test was performed to compare the differences in variable data between the LSC and TVM groups before and after propensity score matching. Pearson chi-square test was used to determine the difference in the POP stage before and after propensity score matching, and the incidence of postoperative UI, mesh exposure, and prolapse recurrence after propensity score matching between the LSC and TVM groups. To perform nearest-neighbor propensity score matching, the BMI, POP stage, and number of laparotomies were used as explanatory variables, which are crucial in determining whether to proceed with LSC or TVM at our institution. The propensity score for each patient was determined by multiplying each variable (BMI, POP stage, and number of laparotomies) by a coefficient computed by logistic regression analysis. The caliper width was 0.2, and the matching ratio used for propensity score matching was 1:1. Furthermore, the difference in prolapse recurrence-free or incontinence occurrence-free survival rates were compared using the log-rank test between the groups after nearest-neighbor propensity score matching. Statistical analysis was performed using JMP PRO (ver. 17; SAS Institute, USA). A P-value <0.05 was considered statistically significant.
RESULTS
Patient Parameters
Before nearest-neighbor propensity score matching, the mean BMI value was 22.7 kg/m2 in the LSC and the median BMI value was 24.6 kg/m2 in the TVM group, which showed a significant difference (P<0.001). In this study, 20 and 12 patients from the LSC and TVM groups had stage 4 POP, respectively, showing a significant difference (P<0.001). Moreover, significant differences in the number of previous laparotomies, median blood loss, and median operating time were found between the LSC and TVM groups. Anterior vaginal wall injuries were observed in 8 patients in the LSC group, and peritoneal injuries were noted in 2 patients in the TVM group. Bladder injuries were seen in totally 5 patients (Table 1). However, no serious postoperative complications were observed. All of them were rated as Clavien-Dindo classification grade 1 [13].
Clinical characteristics of the enrolled patients before matching
Nearest-Neighbor Propensity Score Matching
The predicted probability as a propensity score was calculated using multiple logistic regression analysis. Using nearest-neighbor matching, 76 pairs were matched between the LSC and TVM groups. Following nearest-neighbor propensity score matching, no significant differences in BMI, POP stage, or number of previous laparotomies were found between the matched LSC and TVM groups. Significant differences in age, blood loss, and operating time still existed even after propensity score matching (Table 2). The details of the POP cases are shown in Table 3.
Clinical characteristics of the enrolled patients after matching
Details in pelvic organ prolapse
Moreover, no significant differences in the occurrence of UI or mesh exposure or prolapse recurrence were observed between the matched LSC and TVM groups (P=0.240, P=0.316, P=0.754, respectively) (Table 4). Hazard ratios of LSC to TVM were 1.176 for UI, 0.952 for mesh exposure, and 1.028 for POP recurrence, and 95% confidence intervals (TVM/LSC) were 0.850− 1.627 for UI, 0.689−1.316 for mesh exposure, and 0.744−1.420 for POP recurrence in all patients. In addition, no significant differences in the above-mentioned surgical outcomes were noted in patients with stage 4 POP (P=0.476, P=1.000, P=0.890, respectively) (Table 5). Hazard ratios (TVM/LSC) were 1.004 for UI, 0.861 for mesh exposure, and 0.550 for POP recurrence, and 95% confidence intervals (TVM/LSC) were 0.397−2.539 for UI, 0.346−2.143 for mesh exposure, and 0.223−1.355 for POP recurrence in patients with stage 4 POP. Moreover, no significant differences in the time to prolapse recurrence or UI were observed between the matched LSC and TVM groups (P=0.627 and P=0.235, respectively) (Fig. 1A and B).
Association between surgical methods and the presence of incontinence, mesh exposure, prolapse recurrence at 1 year postoperatively in all patients
Association between surgical methods and the presence of incontinence, mesh exposure, prolapse recurrence at 1 year postoperatively in patients with stage 4 POP
No significant difference in the time to prolapse recurrence (A) and urinary incontinence (B) was found between the matched groups of patients treated with LSC or TVM. LSC, laparoscopic sacrocolpopexy; TVM, tension-free vaginal mesh surgery.
DISCUSSION
In this study, significant differences in BMI, POP stage, and number of previous laparotomies were observed between the LSC and TVM groups before propensity score matching (Table 1). However, no significant differences in these variables were noted between these 2 groups after nearest-neighbor propensity score matching. Even after matching, the blood loss volume was significantly higher in the TVM group than in the LSC group, and the operating time was longer in the LSC group than in the TVM group. After propensity score matching, no significant differences in the incidence of UI, mesh exposure, and prolapse recurrence were observed between the 2 groups in all patients and even in those with stage 4 POP. Moreover, even after propensity score matching, no differences could be found in the time to prolapse recurrence or UI between these 2 groups. We have focused our analysis on the symptoms which would be most directly affected by surgery for POP: the presence or absence of incontinence, mesh exposure, and the prolapse itself.
The optimal surgical technique for POP is still being debated. Some studies have suggested that LSC offers several benefits, such as reduced intraoperative bleeding, faster recovery following surgery, extended postoperative effectiveness, and lower complication rates [14,15]. For young women who have active sexual lives, LSC is thought to be the best option. In contrast, a recent prospective multicenter clinical study indicated that TVM may have more favorable effects on patients’ quality of life, particularly their sexual quality of life [16].
In 2008 and 2011, the U.S. Food and Drug Administration (FDA) expressed concerns about TVM because of postoperative issues, such as discomfort, dyspareunia, mesh exposure, and mesh retraction [17]. Eventually, in 2019, the FDA ordered that all manufacturers of surgical meshes intended for transvaginal repair of anterior vaginal wall prolapse cease marketing and distribution of their products. This declaration sparked considerable controversy and ongoing debates about the appropriateness of TVM for POP. However, studies in Japan have found that TVM is associated with a low risk of postoperative prolapse recurrence, favorable results, and minimal mesh-related complications. In addition, the TVM group showed a notable increase in their quality of life [8,9,12,18]. Practically, only surgeons who are board certified in urology or gynecology and have undergone a training course in TVM are permitted to perform TVM in Japan. In fact, the mesh-related complications, such as dyspareunia, erosion, and retraction, have not been reported often in Japan as reported in the excellent results overseas [5,6,8,9].
One study did not find a significant difference in the postoperative recurrence rates of 5.1% for LSC and 8.5% for TVM [6]. In another study, the subjective anatomical success rates following LSC and TVM were 84.6% (33 of 39) and 86.7% (26 of 30), respectively [7]. Based on another study, the 5-year anatomical success rate following TVM was between 79% and 87%. The 1-year follow-up outcomes appeared to be representative of the 5-year outcomes [19]. In this study, all patients with TVMtreated POP had an anatomical success rate of 95.5% (stage 3) and 70.0% (stage 4) after nearest-neighbor matching, which was not substantially different from that of LSC-treated cases after propensity score matching (stage ≤3, 96.9%; stage 4, 72.7%).
In addition to the prolapse recurrence rate, no significant differences in the rates of mesh exposure or UI were found between the matched LSC or TVM groups, as reported in some studies [20,21]. One study revealed a very low rate of mesh erosion (3.1% for TVM and 2.6% for LSC), which was not much different from our results (0% for TVM and 1.3% for LSC after propensity score matching) [7]. UI developed in 35.8% of patients in the vaginal surgery group and 24.1% in the laparoscopic group before surgery in one study, and it persisted in 5.0% of the patients in the vaginal surgery group and 3.7% in the laparoscopic group postoperatively [22]. In this study, UI persisted in 10.5% of patients in the LSC group and 17.1% in the TVM group after propensity score matching, which were observed more frequently than reported above. Stress UI was the most frequent postoperative complication after either operation. Some data indicate the prevalence of this condition, ranging from 25% to 54.5% [23-26]. In general, the previous UI rate is a little high; however, it is comparable with our results.
In this study, we assessed postoperative outcomes among patients treated with LSC or TVM with comparable background variables using the propensity score matching technique. Although propensity score matching has somewhat reduced selection bias by balancing the known confounders, some undiscovered confounders may have affected the results. In addition, while potentially confounding variables such as age, presence of diabetes mellitus, and comorbidity could have been included, these factors did not influence decision-making regarding whether to proceed with LSC or TVM, as they did not directly affect the postoperative outcomes. Thus far, no evidence showed that these variables led to any postoperative complications. Currently, BMI, POP stage, and history of laparotomy are crucial factors in determining the surgical indication of LSC or TVM.
This study has several limitations. This retrospective study was conducted at a single center, and the ample size was relatively small, which may limit the generalizability of the findings. The median postoperative follow-up period was also relatively short. A postoperative follow-up period longer than 1 year could be an appropriate observation time for patients with benign diseases such as POP. Usually, a 1-year follow-up period is sufficient to detect recurrence or complications, but cases that require longer follow-up (e.g., patients in which the intraoperative findings suggest having fragile tissue) were followed for a longer period. However, future multicenter studies with larger sample sizes and longer follow-up periods could validate our findings and strengthen the evidence base. In addition, the statistical power in stage 4 subgroup is limited (only 11 patients in LSC vs. 10 in TVM). However, we do not believe that it is a problem because TVM showed non-inferior postoperative outcomes compared to LSC, even though the number of cases was small.
In conclusion, the results of our analysis using nearest-neighbor propensity score matching revealed that both surgical methods effectively ameliorated POP. Specifically, compared with LSC, TVM may be a viable alternative to LSC in selected patients, including some with advanced POP, based on shortterm outcomes. Accordingly, LSC could be an option for severe POP such as complete hysterocele.
Notes
Grant/Fund Support
This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Research Ethics
The study protocol (Saitama, Japan; ID 4219) was accepted on August 21, 2020, by the institutional ethics committee of National Defense Medical College, and all patients who participated in this study provided written informed consent forms.
Conflict of Interest
No potential conflict of interest relevant to this article was reported.
AUTHOR CONTRIBUTION STATEMENT
· Conceptualization: KKuroda, KH, KKawamura, AM, HK, KI
· Data curation: KKuroda, KH, KKawamura, AM, HK, KI
· Formal analysis: KKuroda, KH, KKawamura, AM, HK, KI
· Methodology: KKuroda, KH, KKawamura, AM, HK, KI
· Project administration: KKuroda, KH, KKawamura, AM, HK, KI
· Visualization: KKuroda, KH, KKawamura, AM, HK, KI
· Writing - original draft: KKuroda
· Writing - review & editing: KKuroda, KH, KKawamura, AM, HK, KI