To describe the initial placement of Elevate single-incision mesh kit device tips relative to the sacrospinous ligament, and to measure tip movement over a 6-month period from initial placement.
To describe the initial placement of Elevate single-incision mesh kit device tips relative to the sacrospinous ligament, and to measure tip movement over a 6-month period from initial placement.
Tertiary care urogynaecology centre in Calgary, Canada.
Women electing for surgical management of anterior vaginal wall prolapse.
Ten women underwent anterior prolapse repair using the Elevate single-incision mesh kit with a metallic fiducial marker attached to the tips of the surgical device. Women were imaged by magnetic resonance imaging (MRI) within 48 hours of surgery, and again 6 months later to investigate the position of the device tips and change ≥4 mm over the 6-month postoperative period.
Position of self-fixating tips within 48 hours of surgery, and at six months post-operative.
Anchor insertion was directly into the sacrospinous ligament in 10 of 20 insertion points (50%, 95% CI 27–73%). Movement was most often noted in the cranial-caudal direction: a change in location of ≥4 mm was observed for 8/20 anchors (40%, 95% CI 19–64%). Cranial-caudal movement was observed less frequently among sacrospinous anchors than among anchors inserted into other pelvic structures (1/10 versus 7/10, P = 0.020, difference in proportion −60%, 95% CI −94 to −26%). PFDI-20 scores improved statistically significantly by 6 months (P = 0.008, mean change −62.9%, 95% CI −105.1 to −20.7%), but PFIQ-7 scores did not change statistically significantly over the same time period (P = 0.523, mean change −12.4%, 95% CI −54.5 to 29.8%).
The novel self-fixating anchoring tips of this single-incision mesh kit do not reliably anchor into the sacrospinous ligament. The tips have been shown to move with time, although not all cases of anchor movement were associated with recurrent prolapse.
The use of permanent polypropylene mesh for pelvic organ prolapse repair has been shown to provide improved anatomical outcomes after surgery.[1, 2] Despite this, the use of mesh devices has come under increasing scrutiny by regulators in the USA, Canada, and Europe, because of concerns about complications (visceral and vascular injuries, mesh erosion, and vaginal or pelvic pain). In the literature these risks have been associated with trocar-based mesh kits.[4-7] To address concerns about risks associated with trocar placement, new single-incision mesh devices with novel anchoring techniques have been developed. These new kits can be licensed at present without new evidence of safety or effectiveness, instead relying on evidence based on predicate devices.
One such single-incision mesh kit was developed by American Medical Systems (AMS, Minnetonka, MN, USA). Marketed under the name Elevate®, the self-fixating tips of this device are designed and intended to push into the sacrospinous ligament and avoid the need for pulling an ‘arm’ of mesh through the ligament with trocars. This would have the theoretical advantage of avoiding pudendal nerve and vascular injury. Before licensing, the effectiveness of these tips was tested only for pull-out strength in cadavers. Since licensing, a single prospective cohort of 128 patients with 1-year follow-up found the Elevate kit to have an objective cure rate of 87.7%.
Our group was interested in introducing the Elevate kit into surgical practice for anterior repair of pelvic organ prolapse, but questioned whether the fixating tips could migrate from their initial placement over time, leading to recurrent prolapse or other complications. Before adopting the Elevate mesh kit into our practice, we set out to investigate where the anchors were anatomically located after surgery, and whether any movement of the anchor tips could be measured. Our study used magnetic resonance imaging (MRI) to trace fiducial markers attached adjacent to the Elevate anchors.
A prospective cohort study involving four urogynaecologists was carried out in a tertiary care hospital and an academic community hospital in Calgary, Alberta, Canada. Each participating surgeon had over 15 years of experience in prolapse surgery, including sacrospinous vault suspensions and use of permanent mesh in the form of other commercially available vaginal mesh kits, as well as free cut mesh anchored to the sacrospinous ligament with sutures. Prior to the trial, each surgeon had undertaken training in the use of Elevate devices. This was in the form of two independent cadaver labs sponsored by AMS. These labs were preceptored by two different surgeons that AMS considered to be expert implanters. During each anatomy lab, surgeons each had the opportunity to implant a minimum of two Elevate devices.
To take part in the study, patients had to have anterior vaginal wall prolapse with point Ba of 0 or greater on the pelvic organ prolapse quantification (POP-Q) score. The surgeon also had to have a specific reason for the use of permanent mesh. This included either prior failed native tissue repair, and/or very advanced prolapse. Patients were informed of the alternatives of native tissue repair or the traditional anterior permanent mesh procedure offered at our institution. Patients provided written consent to take part in the study and have a repair using the Elevate Anterior Prolapse Repair System. The Elevate system was not available outside the study protocol.
All Elevate procedures were undertaken according to AMS recommendations. Concomitant surgical procedures were permitted, and were conducted according to the usual practice of the surgeon (Table 1). All patients had postoperative vaginal packing for a minimum of 24 hours to limit early movement of the sacrospinous anchors during recovery. Patients stayed in hospital for 2–4 days postoperatively. Upon discharge, women were advised to avoid lifting, to use stool softeners, and to abstain from sexual intercourse for 6 weeks. A single study follow-up was undertaken at 6 months after surgery. All patients completed validated quality-of-life questionnaires before surgery and at their 6-month follow-up visit: the Pelvic Floor Distress Inventory (PFDI-20), Pelvic Floor Impact Questionnaire (PFIQ-7), and the Pelvic Organ Prolapse Sexual Function Questionnaire (PISQ-12).[10, 11]
|ID||Age (years)||Body mass index||Paritya||Prior hysterectomy||Concomitant procedures|
|7||54||24.4||0||No||Vaginal hysterectomy, TVT|
|Summary (median or %)||63, IQR 50–76||29, IQR 24–34||80% parous||90% yes||70% no concomitant procedures|
Early in the development of this study, we evaluated whether the polypropylene tips of the Elevate anchor could be visualised by MRI. This was done through test imaging performed with the Elevate mesh implanted into commercially available animal tissue. These test images proved the tips could not be visualised by the MRI technique. Therefore, to investigate the position of Elevate tips within 48 hours of surgery and at the 6-month follow-up, we needed to identify a fiducial marker for the Elevate tips. We chose to use Hemoclip® (Weck Surgical Instruments, Teleflex Medical, Durham, NC, USA) applied through the mesh adjacent to the fixation tip. Prior to the trial, we implanted four Elevate mesh kits (eight sacrospinous insertions) marked with Hemoclips into unembalmed cadavers. After each insertion, the mesh was pulled out: no movement of the markers occurred in these tests. The imaging technique selected for this study was MRI to maximise soft-tissue visualisation while limiting patient exposure to ionising radiation.
The details of the MRI protocol are available in Appendix S1. Our study had two primary outcomes: (1) evaluation of anchor placement, describing anchors as properly located in the sacrospinous ligament or placed into a different pelvic structure; and (2) measuring any change in position of the anchors over a 6-month period. A change of 4 mm in location between the first and second MRI was felt to be the threshold at which we could confidently say a true change in location had occurred. Movement of the fixations was compared between those originally placed directly into the sacrospinous ligament versus those that were placed into other pelvic structures. The proportions of anchors moving 4 mm or greater were compared.
Secondary outcome measures were change in mean point of maximum anterior vaginal wall descent (Ba) by POP-Q scoring before and after surgery, as well as change in standardised quality-of-life questionnaire scores (PFDI-20, PFIQ-7 and PISQ-12).
We attempted to control for bias: the measurements of the fiducial markers were performed blinded to patient outcomes; surgeons were unaware of MRI measurement results when performing the postoperative POP-Q scoring; participants did not know their MRI results during the study; radiologists were blinded to the order of MRI and clinical outcomes. It was not possible to control for sampling bias, because to become participants, women were identified as needing mesh-augmented surgery, so by definition were more difficult to treat and more likely to have symptomatic recurrent prolapse.
Before the start of our study, there was no evidence on which to base a sample size calculation. We chose to recruit ten women (20 insertions) to our study on the basis that this sample size would be adequate to identify whether movements had occurred. The technical characteristics of the MRI method were such that a 4 mm change in position would be identifiable (for details, see Appendix S1). Assuming that recurrent prolapses are related to anchor movement, we estimate a rate of movement of 15% based on rates from a prior publication describing the success rates of the Elevate device. Our sample of 20 insertions would be able to identify a 15% rate of movement ≥4 mm with a 95% confidence interval (95% CI) of 0–31%. After our study started, a paper was published that could help to justify our sample size decision. Cayrac reported that 50% of Pinnacle tips were properly placed in the sacrospinous ligament in a cadaveric study. If our study found a similar outcome, we could estimate a 50% incorrect placement rate with a 95% CI of 28–72%. The 6-month follow-up period was chosen because it was felt that any migration of the tip would most likely occur within the first months after surgery, before complete wound healing.
Data entry and management were carried out using excel 14.0.0, and analyses were carried out using sas 9.3 (SAS Institute Inc, Cary, NC, USA). Descriptive statistics (mean and standard deviation, median and interquartile range, proportions) were calculated for baseline and 6-month data, as appropriate. Proportions and exact binomial 95% CIs were reported for sacrospinous placement and ≥4 mm movement of anchors. The Fisher's exact test (FE) was used to compare the proportion of anchor insertions that were considered to have moved from sacrospinous placement versus other placement. Changes in Ba score and quality-of-life questionnaire scores at 6 months were compared with baseline using paired Student's t-tests. Our study followed the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) statement for reporting cohort research.
The study was approved by the Conjoint Health Ethics Research Board of the University of Calgary (ID# E-23814). Grant-in-aid funding was received for this project from AMS, but the company had no influence on the design, aim, analysis or interpretation of the study.
A total of ten women participated in the study. Details of patient characteristics and fiducial measurements for each woman appear in Table 1. No intraoperative or postoperative complications occurred.
The anatomical locations of the fiducial markers on the early postoperative MRIs were reviewed for all patients, and the data are presented in Table 2. Anchor insertion was directly into the sacrospinous ligament in ten of 20 insertion points (50%; 95% CI 27–73%). In the other ten insertion points, imaging revealed the anchors were unintentionally inserted into other pelvic structures. Six of the anchors were placed laterally and superiorly, ending up in the iliococcygeus muscle. In two insertion points, the anchor was placed into the ischiorectal fossa. All surgeons involved in the trial experienced misplacement of at least one anchor; no individual surgeon appeared to be more or less likely to place the anchor into the sacrospinous ligament. The proportion of misplaced anchors between the insertions on the right and the left was similar.
|ID||Right anchor||Left anchor|
|Placement||Scan #||Axial||CCa||Placement||Scan #||Axial||CCa|
|1||Sacrospinous ligament||1||25.9||0||Sacrospinous ligament||1||31.8||0|
|2||Ischiorectal fossa||1||17.2||0||Sacrospinous ligament||1||22.3||10|
|3||Ischiorectal fossa||1||23.6||4||Iliococcygeus muscle||1||31.3||4|
|4||Ischiorectal fossa||1||34.4||20||Ischiorectal fossa||1||34.4||20|
|5||Sacrospinous ligament||1||15.6||2||Sacrospinous ligament||1||17.5||12|
|6||Sacrospinous ligament||1||22.1||0||Ischiorectal fossa||1||22.1||−12|
|7||Ischiorectal fossa||1||19.3||8||Sacrospinous ligament||1||19.7||0|
|8||Iliococcygeus muscle||1||23.5||0||Sacrospinous ligament||1||22.3||0|
|9||Sacrospinous ligament||1||26.6||0||Sacrospinous ligament||1||27.8||0|
|10||Ischiorectal fossa||1||21.7||4||Ischiorectal fossa||1||36.4||−8|
On MRI axial slice measurements, compared with baseline, one of the 20 markers (5%, 95% CI 0.1–25%) had a change in location of ≥ 4 mm at the 6-month follow-up. This anchor had been located in the sacrospinous ligament on immediate postoperative imaging.
For measurements in the cranial-caudal direction (CC), eight of the 20 markers (40%, 95% CI 19–64%) had a change ≥ 4 mm. A measured change in location of ≥ 4 mm was noted for one of the ten sacrospinous anchors (10%), in comparison with the group of insertions into other pelvic structures (ischiorectal fossa or iliococcygeus muscle), in which seven of the ten anchors had a change above this threshold (70%). The difference in these proportions was statistically significant (−60%, 95% CI −94 to −26%, FE P = 0.020).
Table 3 presents POP-Q Ba score and quality-of-life scores for all women. Although recurrent prolapse was not a planned secondary outcome, two women presented to their surgeons during the study period with symptoms of recurrent prolapse. Both women had a Ba score of 0 on physical examination. In these two women, three out of four tips were found away from sacrospinous ligament on postoperative MRI. In the patient with one anchor placed in the sacrospinous ligament and the other placed into another pelvic structure (patient #8), the anchor placed into the sacrospinous ligament on the left-hand side did not have a significant change in location (1.5 mm on axial slice, 0 mm in CC), whereas the other anchor placed into the iliococcygeus muscle on the right-hand side moved by 22 mm in the cranial caudal direction. The other patient with recurrent prolapse had both anchors placed in other anatomical locations (patient #3), and both anchors were considered to have changed position on the 6-month MRI (right, 3.4 mm axial, 4 mm CC; left, 2.6 mm, 8 mm CC).
|Preoperative n = 10||6-month follow-up n = 10||Pa||Mean change (95% CI)|
|Mean postoperative Ba||1.6 ± 1.6||−0.8 ± 0.9||0.003||−2.4 (−3.7 to −1.1)|
|Mean PFDI-20||120.8 ± 68.3||57.9 ± 71.1||0.008||−62.9 (−105.1 to −20.7)|
|Mean PFIQ-7||103.8 ± 101.0||91.4 ± 96.5||0.523||−12.4 (−54.5 to 29.8)|
One woman's PFIQ-7 score worsened after she experienced a return of stress urinary incontinence following a tension-free vaginal tape (TVT) excision for persistent mesh erosion. She was not one of the two women who experienced anatomical anterior prolapse failures. All other women reported improved PFIQ-7 scores. PFDI-20 scores improved for all ten women. Median PISQ-12 scores improved for the four women who were sexually active (26 versus 13; IQR 23–41 versus IQR 12–14), although tests of statistical significance were not performed because of the small sample size.
In our study, MRI within 48 hours of the single-incision mesh kit prolapse repair revealed that only half (n = 10/20) of Elevate anchors were placed directly into the sacrospinous ligament. A change in anchor location ≥ 4 mm was often observed in the CC direction. Changes as large as 22 mm were observed, suggesting that the anchor can pull downwards out of its insertion point over time. Symptomatic recurrent prolapse occurred in two patients. Although our study had too few patients with symptomatic recurrence to allow us to investigate a possible relationship between anchor placement and symptomatic recurrence, our findings will be important to ensure that future studies are designed to include a sufficient number of patients to be able to investigate whether such a relationship exists.
The measured changes in position could be accounted for in a number of ways. Some may argue that the fiducial markers were not appropriate. We feel that this is unlikely, as we tested the markers in cadaver studies prior to this study and found that they did not fall off or move on the mesh, despite the anchors being pulled out completely. The poor surgical placement of half of the anchors could be blamed on poor surgical technique. This is unlikely for two reasons. First, each surgeon had more than 15 years of experience in pelvic floor reconstructive surgery. Second, each had attended two cadaver lab sessions taught by AMS mentorship surgeons, and were deemed to be successful in these training labs. At the time of the study procedures, the surgeons felt that all anchors had been properly inserted. We believe that poor placement of the anchors is inherent in the design of self-fixating devices. Perhaps the blind placement of the device and the design of the tips simply results in the device being difficult to place. Despite one or two training labs being a commonly accepted standard for the performance of new procedures, it may be insufficient with this device.
The movement detected by MRI in this study is likely to be a reflection of what happens in actual clinical practice: the anchors migrate small distances, particularly if they are not directly placed into the ligament. Reasons for imprecise placement may include difficulties with the technology itself, causing the anchor to miss its target, insufficient pressure on the ligament when firing the anchor device, insufficient dissection, or possibly anatomic variations in the thickness and strength of the sacrospinous ligament that cannot be anticipated or controlled by the surgeon. Given that low-density polypropylene is difficult to visualise on MRI, in cases of anchor insertion into the ischiorectal fossa we could not differentiate how this space was accessed. It is possible that the anchor was placed right through the levator ani muscle.
The imaging protocol used in this study was static MRI. As such, it does not evaluate the capacity of the anchors to move during increases in abdominal pressure, such as when a woman coughs or exercises. Going forward, evaluating the anchors in a dynamic fashion may provide further insight into how this device behaves in a woman's body, and whether a relationship exists between anchor movement and recurrent prolapse.
Our study was too small to add meaningful data to the literature about objective and subjective outcomes of this device. Our objective cure rates are within the range reported by larger studies.[5, 9, 14] We had two anatomical failures, and these two patients did not differ in baseline characteristics. One had incorrect placement of both anchors and the other had one improper placement.
The cadaveric pull-out studies performed by AMS prior to the Elevate kit being licensed in the USA implied no movement of the anchor would occur once it was placed. Given that anchor movement in the CC direction was observed by MRI, as well as high rates of movement after placement into other structures, this novel anchoring device is not performing as expected.
Our findings suggest that in the case of Elevate, pre-market testing did not predict how the device would function in practice. The high rate of insertion into unintended structures is concerning. As our data suggest that the anchors are capable of migration, which was not detected during cadaveric pull-out testing, this raises the question of whether pull-out strength testing in cadavers can actually be generalised to how anchors function in living, active patients.
Our study is the first to report on the placement of the anchoring device in anterior Elevate in vivo. The results were surprising considering the experience of the surgeons. This brings into question the benefit of using such anchoring devices if their placement is accurate only 50% of the time.
Surgeons who are considering adopting Elevate into their practice should be aware of the results of this study. It appears that two cadaveric training labs are not likely to be sufficient to reliably insert this device, even if one has experience in pelvic floor surgery. Surgeons should also be informed that the placement of anchors into the sacrospinous ligament is not reliably accurate in live patients, and that these anchors appear to have the capacity to move with time. With this information, surgeons can better decide if they feel it is an appropriate device to use in the treatment of prolapse.
Grant-in-aid funding was received for this project from AMS. Devices were purchased by Alberta Health Services as part of patient care. E.B. has received an unrestricted educational grant from Cook Medical for fellowship funding, as well as speaker's fees from Novo Nordisk and Astellas. M.R. and S.R. have received grant-in-aid funding from Johnson & Johnson and Boston Scientific for prior work. M.R. is on the Advisory Committee for Cook Myosite. M.M. and C.B. have received preceptorship compensation for teaching by Cook Medical. D.B. has no disclosures.
E.B., M.R., and S.R. were involved with the initial conception and design of the study. E.B. and D.B. were involved in data collection. E.B. and S.T. undertook analysis, and all authors (E.B., D.B., S.T., C.B., M.M., D.C., M.R., and S.R.) participated in data interpretation. E.B., S.R., and M.R. were involved in writing the intial draft of the article. All of the authors contributed to critically revising the article, and have approved the submitted version.
This study received approval from the Conjoint Health Research Ethics Board of the University of Calgary (approval ID# E-23814). All participants provided informed written consent for both participation in the study and for the surgical procedure.
Grant-in-aid funding was received for this project from AMS. The funder made no contribution to the study design, conduct, analysis, or writing.
The authors would like to acknowledge Hein Els for his contributions to the measurements and data collection from MRI images. That work was performed as part of his radiology fellowship. This work was presented at the International Urogynecology Association's 38th Annual Meeting on 31 May 2013 in Dublin, Ireland.