Primary treatment of pelvic organ prolapse: pessary use versus prolapse surgery

According to the judgement of the gynaecologist, either a shelf (Falk) or a ring pessary (with or without central support) could be used. No other pessaries were placed during the trial. Preferably a ring pessary was placed. According to the judgement of the physician, a ring pessary with central support was placed in patients with apical descent, and a shelf (Falk) pessary was placed in patients with apical descent, extensive prolapse or lack of support of the ring pessary. Pessary fitting was performed at the outpatient clinic after fitting. A test pessary was used to determine the correct size of the pessary. The patient tested this pessary while walking around for 30 min. If the right size was found, the pessary was inserted and the fitting was considered successful. The type and size of pessary were recorded, as well as the number of different types and sizes. Side effects during pessary treatment including discharge, pain and blood loss were also recorded. After placement, all patients received instructions about pessary treatment. If the initial fit was unsuccessful another size or the other pessary type could be used, according to the judgement of the gynaecologist. Discontinuation and expulsion were recorded, and if pessary treatment was not successful, the reason for failure was also recorded.

A follow-up visit 6 weeks after placement was planned, and patients were instructed to return to the clinic if they had any complaints or if they lost the pessary. If the patients was satisfied with the pessary, follow-up visits every 3–4 months were planned for pessary cleaning and vaginal inspection. During the follow-up period, the type or size of the pessary could be changed. All participating physicians were skilled in pessary fitting and had performed at least 100 pessary fittings prior to the start of the trial.

The surgical intervention consisted of correction of all compartments that required surgery. The decisions as to which technique to use and which compartments to treat were left to the discretion of the gynaecologist, and depended on the results of the physical examination and the complaints. Cystocele repair involved conventional anterior colporrhaphy. For uterine descent, different techniques were allowed. These techniques could be vaginal hysterectomy with vault suspension or uterus-preserving techniques such as sacrospinous fixation, the Manchester-Fothergill procedure or a laparoscopic sacrohysteropexy. A coexisting rectocele was treated with conventional posterior colporrhaphy. When stress incontinence was diagnosed prior to surgery, the patient and surgeon decided whether to perform a concomitant incontinence procedure or whether first to perform prolapse surgery only with additional incontinence surgery later.

All procedures were performed under general anaesthesia or spinal anaesthesia. Prophylactic antibiotics were given peroperatively (metronidazole/cefalozine). As prophylaxis for thromboembolism, low molecular weight heparin was administered subcutaneously peroperatively and postoperatively. The data collected included the type of surgery, procedure time, estimated blood loss, length of hospital stay and perioperative complications. A urethral catheter was left in situ and was removed on the second postoperative day in women with anterior colporrhaphy or on the first day in other patients. If the procedure was complicated by a bladder lesion, the catheter was removed after 1 week. If urine retention occurred after removal of the catheter on the first day, the catheter was reinserted for another day.

Women were asked to complete a questionnaire before surgery, and at 3–6 months and 12 months after treatment. All patients were asked to undergo a pelvic examination before treatment, at 6 weeks after treatment and when clinically indicated. Pelvic examination after 1 year was performed in women who were still followed up by their physician. Randomized patients were invited for examination after 12 months. During pelvic examinations, POP-Q stage was determined [15] in the lithotomy position and during a Valsalva manoeuvre. Other findings of the pelvic examination including vaginal erosion, vaginal discharge, atrophy, bleeding and urinary incontinence were also recorded.

The primary outcome of the study was the functional outcome evaluated using the Urogenital Distress Inventory (UDI) [16] after 12 months. The UDI is a questionnaire that has been validated in Dutch for evaluating prolapse-related symptoms. The UDI is a disease-specific questionnaire comprising 17 questions for assessing the presence and experienced discomfort of pelvic floor problems. The UDI consists of five domains: discomfort/pain, urinary incontinence, overactive bladder, genital prolapse, and obstructive micturition. The UDI scores were calculated for al five domains [16]. The questionnaire also includes versions of the Defecatory Distress Inventory (DDI) [17] and the Incontinence Impact Questionnaire (IIQ) [16]. The questionnaires also include questions about sexual function. If participants did not respond they were sent the questionnaire a second time. If they ignored both questionnaires, they were contacted by telephone to find the reason.

Additional interventions could include physiotherapy and incontinence surgery in the pessary group, and physiotherapy, incontinence surgery or surgery for recurrent prolapse in the surgery group. Long-term complications and side effects were recorded in both groups. The remaining study parameters included age, body mass index, parity, menopausal status, presence of incontinence and use of oestrogens. Urinary stress incontinence was diagnosed with the patient in the lithotomy position with a full bladder. The patient was asked to cough several times. If urinary leakage was seen, urinary incontinence was diagnosed. In patients with a cystocele, the cystocele was redressed and the examination was repeated.

Disease-specific quality of life as evaluated using the UDI questionnaire was the primary endpoint. A difference between the two interventions of ten points in the prolapse domain of the UDI 12 months after treatment was considered clinically relevant [18]. Assuming a standard deviation of the score in this domain of 15 points, 72 patients were needed to show a statistically significant difference in the primary outcome (power of 80%, α error 0.05) [18]. Taking into account a 10% attrition rate, 80 patients (40 in each arm) needed to be included. After the RCT was halted, we focused on the cohort group which was part of the trial, since randomization turned out to be very difficult. We included patients until there were at least 40 patients in both groups.

The aim of the trial was to determine whether surgery was superior to pessary treatment in terms of the primary endpoint (prolapse domain of the UDI). For both nonrandomized and randomized groups, data were analysed according to the intention-to-treat principle. The domain scores were calculated for UDI, DDI and IIQ at baseline and after 12 months in both groups. The scores for these domains vary between zero and 100. A high score in a particular domain indicates more bothersome symptoms or worse quality of life. To examine differences between groups we used an unpaired t test or the Mann-Whitney test for continuous variables, depending on the distribution, and the chi-squared test was used for dichotomous variables. The Wilcoxon signed-ranks test was used to compare the domain scores before and after treatment in both groups separately. Two-sided significance tests were used, and p values <0.05 were considered to indicate statistical significance. For dichotomous outcomes, relative risks and 95% confidence intervals were calculated. All analyses were done with IBM SPSS statistics 22 (IBM, Armonk, NY).

In this study, we found that treatment preference limits patients’ willingness to undergo randomization. At 12 months the pessary group reported more symptoms in the prolapse domain of the UDI (the primary outcome) than the surgery group. Women in the pessary group were also more likely to need an additional intervention, and 28% needed surgical intervention in contrast to women in the surgery group of whom 5% needed reoperation. Younger patients, with a higher POP-Q stage and more severe urinary symptoms that affected their social life were more likely to choose POP surgery.

To our knowledge this is the first attempt to compare pessary and surgery as primary treatments for women with POP in a randomized trial. However, we were able to randomize only six women, since a large majority of the eligible women had a strong preference for one or other of the treatment options. Therefore, we decided to halt the trail and report the data as a prospective cohort study. Previous studies investigating women with POP have also shown strong patient preference for one or the other of two interventions [10]. Lamers et al. [10] found that the likelihood of preferring pessary treatment over surgery increases with increasing patient age. They concluded that surgery is preferred over pessary treatment as POP stage increases and if POP symptoms are more bothersome and affect general wellbeing, and that patients who are sexually active also tend to prefer surgery over conservative treatment [10]. In our study population, patient preference played an important role, to such a degree that randomization was very difficult.

Since most participating women were not randomized, selection bias was likely to have played a role. It is possible that doctors counselled women differently depending on patient characteristics such as age, comorbidity, sexual activity, POP stage and POP symptoms, and how they affect the woman’s daily activities. Because the groups were not allocated by randomization, the baseline characteristics were significantly different. The heterogeneity between the treatment groups is a reflection of normal daily practice. The pessary group in our study consisted of older women, as in previous studies [10]. The POP-Q stages of the pessary group were higher, in contrast to previous studies, but complaints were more bothersome in the surgery group, as has been found in previous studies [10].

The patients’ unwillingness to be randomized is an outcome in itself, as it reflects the strong differences between the two interventions with respect to invasiveness, risk profile and impact. The fact that the two treatment options are very different led to a strong preference for one or the other. However, in the OPUS trial [19], the randomization group and the patient-preference group did not show any differences in baseline characteristics or results. This might suggest that a RCT is not the only study design that can usefully add to the present literature, and the results of this prospective trial provide valuable conclusions. Following the advice of several epidemiologists, we did not perform separate analyses for our randomized and nonrandomized groups since the former was too small.

This prospective cohort study comparing pessary treatment and surgery as primary treatments for POP generated some important lessons. As a randomized clinical trial comparing the two strategies turned out not to be feasible, our prospective study results can be used to counsel women. A strategy of pessary treatment followed by surgery if needed, or a strategy with immediate surgery both seem to be effective options for the treatment of POP in women.

Both groups reported very low scores for almost all domains of the UDI, DDI and IIQ, which suggests that both treatments are effective. The pessary group reported significantly more symptoms in the prolapse domain of the UDI. However, the median scores were still very low (0, 10th to 90th percentile 0–33, in the pessary group; 0, 10th to 90th percentile 0–0, in the surgery group), which is probably clinically irrelevant. Previous studies have also shown that pessaries are effective in improving pelvic floor dysfunction [6‐9]. Most studies have shown improvements in both bulge and irritative bladder symptoms following pessary treatment [10], but two studies have also shown de novo stress urinary incontinence [9, 20]. Patient satisfaction rates with medium-term pessary use are high (70–92%) [20, 21]. Two prospective cohort studies comparing pessary treatment and surgery have shown similar improvements in urinary and bowel symptoms, sexual function, and quality of life at 12 months [11‐14].

The continuation rate of pessary treatment after 12 months was 60%, whereas 72% of the women in the pessary did not have an indication for surgical intervention. This result is very important since it illustrates that both treatments are very effective in treating urinary, defaecation and prolapse symptoms, since the domain scores for the UDI, DDI and IIQ after 12 months were very low, with a median score of 0.0 for most domains. Prospective trials have shown continuation rates between 50% and 80% after 1 year and between 14% and 48% after 5 years [10]. However, evidence concerning reinterventions is lacking. Although no cost analysis was performed, these results suggests that pessary treatment is more cost effective, since less surgery was performed and the domain scores for the UDI, DDI and IIQ at 12 months were low in both groups.

The patient population was only 113 patients despite the long study period. As the study was non-funded, we were not able to register all patients who declined to participate. Therefore, we cannot provide reliable data about a denominator. We included patients until there were 40 patients in each group. Since there were only six randomized patients, we decided, following advice from several epidemiologists, not to perform a separate analysis for the two groups, but to analyse both groups together and present the data as a prospective cohort study instead of a RCT with a prospective cohort alongside. Nevertheless, the results of the study provide evidence on continuation rates, additional interventions, quality of life and patient preference. These outcomes will help individualize counselling on treatment options to provide better guidance to patients with symptomatic POP in the treatment decision process. Younger patients with a lower POP-Q prolapse stage but more urinary symptoms that affect their social life are more likely to choose POP surgery. This agrees with the results of previous studies. A review on the topic shows that the probability of choosing pessary treatment over surgery increases with increasing patient age in accordance with the results of this study. Surgery was preferred over pessary treatmen if POP symptoms are more bothersome and affect the general wellbeing of the patient [10].

Women with POP stage II or higher treated with a pessary are bothered more by prolapse symptoms and more often undergo surgery in the first year of follow-up than patients who are treated surgically. However, pessary treatment allowed surgery to be avoided in 72% of women, although prolapse symptoms were less severe in those who had undergone surgery. These outcomes will help individualize counselling on treatment options and allow better guidance to patients with symptomatic POP in the treatment decision process.