Pelvic organ prolapse and uterine preservation: a cohort study (POP-UP study)

This is among the first studies comparing outcomes of the different variants of LSC with a particular focus on comparing these outcomes based on whether the uterus was spared or concomitantly removed. Of the total number of women who had an LSC procedure during the study period, 70% of women who presented with a significant apical POP requiring surgery had their uterus in situ. The majority of these women had a concomitant hysterectomy at the time of LSC. Our study demonstrated that LSC procedures with a concomitant total hysterectomy were associated with statistically significantly longer operating time and intra-operative blood loss. However, the median differences between groups were only 5 min and 50 ml respectively. In contrast, uterine sparing LSCs were associated with a significantly higher likelihood of a suboptimally placed mesh at 3 and 12 months postoperative and anterior compartment failures at 12 months. Nevertheless, other anatomical and patient reported outcomes were comparable in both groups. On head to head comparison of the different LSC variants there was no significant difference in anterior compartment failure rates. However, this observation should be interpreted with caution due to the small samples in some of the subgroups.

Other groups have reported higher incidence of anatomical failures in association with LSH [23, 24]. Saliba et al. compared outcomes of 64 LSCH + LSC versus 12 LSH procedures and the anatomical failure, defined as POP stage ≥ 2, was significantly higher in the LSH groups in both any and apical compartments (33.3% vs. 6.2% and 16.7% vs. 0% respectively), however, the study authors did not provide the actual length of follow-up [24]. Similarly, Gracia and colleagues reported significantly higher apical compartment failures, defined as C stage ≥ 2, when comparing 12 months outcomes after 15 LSH compared to 30 LSCH + LSC (53.2% vs. 10.0%). Anterior compartment recurrence (Ba stage ≥ 2) was also more common in their LSH cohort (72.4% vs. 33.3%) [23]. The reported incidence of anterior compartment failures concur with our findings of 21.1% vs. 8.8% in our LSH and LSCH + LSC subgroups respectively. Nevertheless, our low incidence of apical compartment recurrences both in the main and subgroup analyses are in stark contrast to the rates reported in these studies.

When comparing LSH and TLH + LSC, we did not have any apical compartment recurrences at 12 months compared to Pan et al. who reported 13.9% and 5.9% recurrence rates for the equivalent procedures in a cohort of 65 and 34 women who had LSH and TLH + LSC respectively, albeit, after an average follow-up of 34 months. Moreover, their anterior compartment failure rates were 13.9% versus 11.8% compared to 21.1% versus 5.2% in our study, while their posterior compartment recurrence incidences were 4.6% versus 5.9% and it was 0% and 15.8% in our LSH and TLH + LSC groups respectively [25]. The identified posterior compartment failure rate in our TLH + LSC was also higher than that reported by Illiano and associates (15.8% compared to 2.4%) [26]. Due to the nature of our study we were not able to explore the reasons behind the aforementioned differences in recurrence rates between our study and previous reports, which could be related to the operative technique, patient selection or duration of follow-up. Another reason for discrepancy in reported outcome rates between various studies is the POP-Q cut-off used to determine failure. Indeed, if we use the Ba > 0 cut-off for cystocele recurrence adopted in other studies [32, 37], our anterior compartment failure rates would have dropped to zero.

We identified a significantly higher likelihood of suboptimal mesh placement in our LSH group, which probably is an indicator of the relative technical difficulty of inserting the mesh in LSH compared to other variants of LSC. It would be of interest to assess if there is any association between mesh placement and clinical outcomes. However, we did not undertake such analysis because we believe our study is underpowered to test such hypothesis. The incidence of postoperative mesh-related complications in our study falls within the range of 1.0–2.6%. However, the incidence of mesh erosions were similar in our subgroup analyses unlike the differences reported by other authors [38‐40]. It is the technical challenge to achieve proper placement of the anterior mesh in LSH and be able to create a “de novo vaginal apex” that is considered to be a plausible reason for the higher anterior compartment failure in association with LSH and is the driver behind the suggestion of alternative modifications to the standard technique [41]. The process of refining the current LSH technique is crucially important since the number of women opting for uterine sparing surgery significantly falls if this technique is associated with inferior anatomical outcome [18, 19].

We appreciate that there are some limitations to our work. First, the retrospective nature of the study has an inherent risk of introducing selection and recall bias into our data. Due to the rigor in our hospital database and the high level of specialism required for the surgical procedures being assessed, it is extremely unlikely we would have missed any procedures or data that was collected. However, the issue of selection bias is more challenging to tackle except within a context of a randomized trial. Indeed, our 2 groups of interest had significant differences in their demographics and associated comorbidities. Second, although we report 12-month follow-up data, in POP surgery, this is considered relatively short. We recognize that the longer the follow-up the higher attrition rate, hence, the current study will form the basis for our LSC database that will enable us to increase our sample size and assess longer term outcomes. Although our sample size in the uterine sparing cohort was relatively small, a post hoc power calculation showed that the power of our study to identify the difference in anterior compartment failure rates between our main cohorts at a significance level of 0.05 was 70% (Additional file 4). Finally, it could be perceived that a report from a single center might limit the external validity of the study. However, the involvement of several independent trained surgeons, in a center accredited by the European Board & College of Obstetrics and Gynaecology (EBCOG) for training and the use of standardized operative technique and validated outcome measures make our findings generalizable. In contrast, the reporting on LSC variants based on whether the uterus was removed or spared using a comprehensive set of core outcomes and the novelty of the postoperative imaging information are major strengths to our study.