PROSPECT: Mesh, Graft, or Standard Repair for Primary Transvaginal Prolapse Surgery

Bottom line Negative

Adding a polypropylene mesh inlay or a biological graft to primary transvaginal anterior or posterior prolapse repair did not improve 1-year Pelvic Organ Prolapse Symptom Scores versus native-tissue repair alone; 51 of 434 women who received synthetic mesh had a mesh complication within 2 years.

Neither augmentation improved POP-SS at 1 year; 12% of mesh recipients had a mesh complication by 2 years

PROSPECT was a pair of pragmatic, multicenter, parallel-group randomized trials in 1,352 women at 35 UK hospitals having a primary transvaginal anterior or posterior compartment prolapse repair. Standard native-tissue repair alone was compared with the same repair plus a polypropylene mesh inlay (the mesh trial) or a biological graft (the graft trial). Neither augmentation improved the 1-year Pelvic Organ Prolapse Symptom Score (POP-SS; 5.4 versus 5.5 with mesh and 5.5 versus 5.6 with graft), and 51 of 434 women who received synthetic mesh had a mesh complication within 2 years.^[1]

Design and population

Two pragmatic multicenter parallel-group randomized trials, surgeons unmasked. Women at 35 UK hospitals undergoing primary transvaginal anterior or posterior compartment prolapse repair.

Arms

Standard repair with a polypropylene mesh inlay versus standard repair with a biological graft inlay versus standard native-tissue repair alone

Report and follow-up

2016 · 1-year primary outcomes and clinic review; questionnaires to 2 years (79% to 82% of each arm returned the 24-month questionnaire)

Endpoint	Results	Effect	Finding
1-year Pelvic Organ Prolapse Symptom Score (0 to 28; lower is better), mesh trial; mean difference (mesh minus standard)	5.4 (SD 5.5; n=395) vs 5.5 (SD 5.1; n=389)	difference 0; CI -0.7–0.71; P 0.99	Not met
1-year POP-SS, graft trial; mean difference (graft minus standard)	5.5 (SD 5.6; n=342) vs 5.6 (SD 5.6; n=337)	difference -0.15; CI -0.93–0.63; P 0.71	Not met
1-year prolapse-related quality-of-life visual analogue score (0 to 10; lower is better), mesh trial; mean difference	2.0 (SD 2.7; n=389) vs 2.2 (SD 2.7; n=380)	difference 0.13; CI -0.25–0.51; P 0.50	Not met
1-year prolapse-related quality-of-life score, graft trial; mean difference	2.2 (SD 2.8; n=335) vs 2.4 (SD 2.9; n=330)	difference 0.13; CI -0.3–0.56; P 0.54	Not met
1-year POP-Q leading edge beyond the hymen (stage 2b, 3 or 4), mesh trial; risk ratio (mesh versus standard)	47/338 (14%) vs 54/336 (16%)	RR 1.12; CI 0.79–1.6; P 0.52	Not met
1-year POP-Q leading edge beyond the hymen (stage 2b, 3 or 4), graft trial; risk ratio (graft versus standard)	47/303 (16%) vs 54/298 (18%)	RR 1.14; CI 0.8–1.62; P 0.47	Not met

Harms: Serious adverse events in year 1, excluding mesh complications: 31/430 (7%) mesh trial; 23/367 (6%) graft trial vs 34/435 (8%) vs 36/368 (10%); Mesh complications, year 1: 2/430 (<1%) mesh trial; 2/367 (<1%) graft trial vs 32/435 (7%) vs 2/368 (<1%); Mesh complications, year 2 (new in that year): 1/430 (<1%) mesh trial; 1/367 (<1%) graft trial vs 25/435 (6%) vs 1/368 (<1%)

ISRCTN60695184 · Report DOI

Design

Women under the care of a participating surgeon were potentially eligible when a decision had been made for primary surgery for anterior or posterior vaginal wall prolapse; women unable or unwilling to give informed consent or unable to complete questionnaires were ineligible. Concomitant uterine, vault or continence surgery was allowed, and women having a repeat repair in the same compartment were reported separately. Of 2,478 eligible women, 1,352 (55%) were randomized and 4 were excluded after randomization; the other 1,126 were not randomized because of clinical decisions (379), participant decisions (613), or other reasons (134). Sixty-five gynecologic surgeons, all proficient in transvaginal anterior and posterior repair, operated.^[1]

Because not every surgeon offered all three options, women were randomized within three strata: 1:1:1 to standard repair, mesh or graft (stratum A, 252 women with standard repair counted in both trials), or 1:1 to standard repair versus mesh (stratum B) or versus graft (stratum C). The mesh trial analyzed strata A and B and the graft trial strata A and C, so the standard-repair group is shared in part. Minimization covered age, planned repair type, concomitant continence or upper-compartment surgery, and surgeon.^[1]

The primary clinical outcome was the POP-SS at 1 year (seven items scored 0 to 4, total 0 to 28), with a prespecified minimally important difference of 2 points; the second primary outcome was prolapse-related quality of life on a 0 to 10 visual analogue scale. Objective failure was a POP-Q leading edge beyond the hymen (above 0 cm) at a 1-year clinic review. The trial planned 400 women per arm to detect a 0.25 SD difference in POP-SS with 90% power and α of 0.025, analyzed by modified intention to treat with complete cases, and reported confidence intervals not adjusted for multiple comparisons.^[1]

Results

At 1 year the mean POP-SS was 5.4 (SD 5.5) after standard repair and 5.5 (SD 5.1) after mesh augmentation (mean difference 0.00, 95% CI −0.70 to 0.71; P = 0.99), and 5.5 (SD 5.6) versus 5.6 (SD 5.6) in the graft trial (mean difference −0.15, 95% CI −0.93 to 0.63; P = 0.71). Both intervals excluded the 2-point minimally important difference, so the primary endpoint was not met. Prolapse-related quality-of-life scores were 2.0 versus 2.2 in the mesh trial (mean difference 0.13, 95% CI −0.25 to 0.51; P = 0.50) and 2.2 versus 2.4 in the graft trial (mean difference 0.13, 95% CI −0.30 to 0.56; P = 0.54).^[1]

At 2 years the POP-SS was 4.9 versus 5.3 with mesh (mean difference 0.32, 95% CI −0.39 to 1.03) and 4.9 versus 5.5 with graft (mean difference 0.32, 95% CI −0.48 to 1.12). At least one prolapse symptom (POP-SS above 0) was present in 328/395 (83%) versus 329/389 (85%) in the mesh trial and 283/342 (83%) versus 276/337 (82%) in the graft trial at 1 year. A leading edge beyond the hymen was found in 47/338 (14%) after standard repair versus 54/336 (16%) after mesh (risk ratio 1.12, 95% CI 0.79 to 1.60) and in 47/303 (16%) versus 54/298 (18%) in the graft trial (risk ratio 1.14, 95% CI 0.80 to 1.62). A per-protocol analysis that excluded women who did not receive the allocated operation did not change the primary result.^[1]

New prolapse operations in the first year were 6/395 (2%) versus 12/389 (3%) in the mesh trial and 7/342 (2%) versus 10/337 (3%) in the graft trial; in the second year they were 16/348 (5%) versus 15/343 (4%) and 15/299 (5%) versus 15/300 (5%). Overall, 6% of women had a new prolapse operation in the same or another compartment within 2 years. Differences in other secondary outcomes were sparse and unadjusted for multiplicity: the vaginal symptoms score at 1 year was higher after graft (9.0 versus 7.1; mean difference 1.31, 95% CI 0.04 to 2.59), as was the proportion reporting something coming down at 2 years (120/299 [40%] versus 91/298 [31%]; risk ratio 1.26, 95% CI 1.01 to 1.58).^[1]

The 4- and 6-year questionnaire follow-up retained 72% and 65% of randomized women and collected no examination-based anatomy. The 6-year POP-SS was 5.4 versus 6.3 in the mesh trial (mean difference 0.85, 95% CI −0.05 to 1.75) and 5.9 versus 6.2 in the graft trial (mean difference 0.27, 95% CI −0.64 to 1.18); at 4 years the mesh difference was 1.01 (95% CI 0.20 to 1.83), below the 2-point threshold. Functional cure at 6 years, defined as no sense of something coming down, no prolapse surgery and no pessary, was 144/268 (53.7%) after native-tissue repair versus 150/274 (54.7%) after mesh, and 120/233 (51.5%) versus 122/247 (49.4%) in the graft trial. Cumulative new prolapse operations by 6 years were 32/266 (12.0%) versus 28/271 (10.3%) in the mesh trial and 27/231 (11.7%) versus 29/245 (11.8%) in the graft trial. Complete satisfaction with surgery at 6 years was reported by 115/268 (42.9%) after mesh versus 144/260 (55.4%) after native-tissue repair (P < 0.001), and by 108/234 (46.2%) after graft versus 125/228 (54.8%; P = 0.255).^[2]

Surgery and treatment delivery

Surgeons used their usual techniques, with midline or fascial plication permitted, and any available product. Every mesh was type 1 monofilament macroporous polypropylene (19 to 44 g/m²; hybrid coated mesh allowed), and the grafts were porcine acellular collagen, porcine small intestinal submucosa or bovine dermis. Mesh or graft was placed below the fascial layer when possible and secured with peripheral sutures. Consultants or doctors who had completed specialty training did 78% and 81% of procedures in the standard and mesh arms of the mesh trial and 69% and 75% in the standard and graft arms of the graft trial.^[1]

Adherence to allocation differed by arm. Standard repair was received by 403/425 (95%) in the mesh trial and 342/359 (95%) in the graft trial, mesh by 341/425 (80%) and graft by 294/363 (81%) of women who had surgery; 60 women allocated to mesh and 57 allocated to graft had standard repair instead. Twenty-three women had no surgery and contributed baseline data only. Primary outcome data at 1 year were available for 784/865 (91%) in the mesh trial and 679/735 (92%) in the graft trial, and examiners were masked to allocation for 581/657 (88%) and 539/573 (94%) of the 1-year examinations.^[1]

Adverse events

Serious adverse events in year 1, excluding mesh complications, occurred in 31/430 (7%) versus 34/435 (8%) in the mesh trial (risk ratio 1.08, 95% CI 0.68 to 1.72) and 23/367 (6%) versus 36/368 (10%) in the graft trial (risk ratio 1.57, 95% CI 0.95 to 2.59). Severe dyspareunia at 1 year was reported by 8/186 (4%) versus 9/173 (5%) in the mesh trial and 9/149 (6%) versus 8/165 (5%) in the graft trial.^[1]

In the mesh trial, mesh complications occurred in 32/435 (7%) of the mesh arm in year 1 and 25/435 (6%) in year 2, with partial surgical removal or oversewing in 23/435 (5%) and 17/435 (4%). Of the 25 women who had partial removal in year 1 (2 in the standard arm), 18 (72%) were asymptomatic and 16 (64%) had exposures under 1 cm²; one woman had total removal within 2 weeks for severe infection. Of the 17 in year 2, 13 (76%) were asymptomatic and 10 (59%) had exposures under 1 cm². Among women who received synthetic mesh by any route, including concomitant vault, uterine or continence procedures, 51/434 (12%) had a mesh complication within 2 years and 37 had surgical removal; counting only mesh used as the anterior or posterior repair, the figure was 41/284 (14%). In the graft trial, 2/368 (<1%) of the graft arm and 2/367 (<1%) of the standard arm had a mesh complication in year 1, all with concomitant synthetic mesh, and 1/368 (<1%) and 1/367 (<1%) in year 2.^[1]

By 6 years, mesh had been removed in 22/263 (8.4%) of mesh-arm respondents versus 4/255 (1.6%) after native-tissue repair (risk ratio 5.24, 95% CI 1.84 to 14.96); 2/231 (0.9%) of graft-arm respondents had mesh removed. Severe dyspareunia at 6 years was reported by 6/105 (5.7%) after mesh versus 4/93 (4.3%) after native tissue, and severe urinary incontinence by 29/267 (10.9%) versus 18/264 (6.8%; risk ratio 1.58, 95% CI 0.97 to 2.59).^[2]

A secondary analysis of all 2,632 women operated in PROSPECT, randomized or in the comprehensive cohort and analyzed by repair received (1,712 native tissue, 482 mesh inlay, 360 xenograft, 78 mesh kit), described complications without inferential comparisons. Mesh exposure occurred in 12.0% of women who received mesh, 64.2% asymptomatic, and surgical removal in 7.9% (95% CI 5.5 to 11.1%). Bladder injury occurred in 0.4% and bowel injury in 0.2%; prolonged catheterization beyond 10 days in 3.9%; de novo urinary incontinence at 24 months in 6.2%; vaginal adhesions in 1.7% (3.3% after xenograft); and readmission for non-mesh complications over 2 years in 4.3%. De novo severe dyspareunia at 24 months occurred in 3.5% after native tissue, 3.5% after xenograft, 1.4% after mesh inlay and 4.8% after mesh kit.^[3]

Limitations

Surgeons could not be masked, and the pragmatic design allowed any mesh or graft and each surgeon's technique. About one in five women allocated to mesh or graft did not receive it, and the analysis was complete-case, with 2-year questionnaires returned by 79% to 82% of each arm.^[1]

Only two of the three strata could be used in any one comparison, so each analysis excluded the women randomized in the other two-arm stratum. Fifteen women were classed as POP-Q stage 0 or 1 before surgery, which the authors read as inconsistent surgical indications. Examination-based anatomy was collected only at 1 year; the 4- and 6-year reports used questionnaires, retained 65% of randomized women at 6 years, and cannot detect asymptomatic mesh exposure.^[1][2]

The trials tested mesh inlays; mesh kits were not used for primary repairs within the randomized trials.^[3]

In conversation

• Built on by VUE (2020). The same group's earlier trial; its methods, outcome measures and symptom-score standard deviation informed VUE's design and sample size.

Same setting Altman (Nordic TVM) (2011)

The 2024 Cochrane update (51 randomized trials, 7,846 women) found less prolapse awareness after permanent transvaginal mesh than after native-tissue repair (risk ratio 0.83, 95% CI 0.73 to 0.95; 17 studies) and less recurrence on examination (risk ratio 0.42, 95% CI 0.32 to 0.55; 25 studies), but more repeat surgery for the composite of prolapse, stress incontinence or mesh exposure (risk ratio 1.56, 95% CI 1.07 to 2.26; 27 studies).^[4] For biological graft versus native tissue it found no difference in awareness (risk ratio 1.06, 95% CI 0.73 to 1.56; 8 studies).^[4] The PROSPECT authors noted that the pooled trials included uterine and vault prolapse, mixed first and repeat repairs and mesh kits, while PROSPECT randomized first repairs in the anterior or posterior compartment using inlays.^[1]

Report history

PROSPECT report history

Published	Report	Cutoff	Headline results	Citation
2016-12-20	2-year randomized report · primary	—	1-year Pelvic Organ Prolapse Symptom Score (0 to 28; lower is better), mesh trial; mean difference (mesh minus standard): 5.4 (SD 5.5; n=395) vs 5.5 (SD 5.1; n=389); difference 0; CI -0.7–0.71; P 0.99 Not met 1-year POP-SS, graft trial; mean difference (graft minus standard): 5.5 (SD 5.6; n=342) vs 5.6 (SD 5.6; n=337); difference -0.15; CI -0.93–0.63; P 0.71 Not met 1-year prolapse-related quality-of-life visual analogue score (0 to 10; lower is better), mesh trial; mean difference: 2.0 (SD 2.7; n=389) vs 2.2 (SD 2.7; n=380); difference 0.13; CI -0.25–0.51; P 0.50 Not met 1-year prolapse-related quality-of-life score, graft trial; mean difference: 2.2 (SD 2.8; n=335) vs 2.4 (SD 2.9; n=330); difference 0.13; CI -0.3–0.56; P 0.54 Not met 2-year POP-SS, mesh trial; mean difference (mesh minus standard): 4.9 (SD 5.1; n=347) vs 5.3 (SD 5.1; n=342); difference 0.32; CI -0.39–1.03; P 0.37 2-year POP-SS, graft trial; mean difference (graft minus standard): 4.9 (SD 5.1; n=298) vs 5.5 (SD 5.7; n=299); difference 0.32; CI -0.48–1.12; P 0.43 1-year POP-Q leading edge beyond the hymen (stage 2b, 3 or 4), mesh trial; risk ratio (mesh versus standard): 47/338 (14%) vs 54/336 (16%); RR 1.12; CI 0.79–1.6; P 0.52 Not met 1-year POP-Q leading edge beyond the hymen (stage 2b, 3 or 4), graft trial; risk ratio (graft versus standard): 47/303 (16%) vs 54/298 (18%); RR 1.14; CI 0.8–1.62; P 0.47 Not met 1-year at least one prolapse symptom (POP-SS above 0), mesh trial; risk ratio: 328/395 (83%) vs 329/389 (85%); RR 1.01; CI 0.95–1.08; P 0.64 1-year at least one prolapse symptom (POP-SS above 0), graft trial; risk ratio: 283/342 (83%) vs 276/337 (82%); RR 0.99; CI 0.93–1.06; P 0.85 New prolapse operation in the first year after surgery, mesh trial; risk ratio: 6/395 (2%) vs 12/389 (3%); RR 1.99; CI 0.76–5.24; P 0.16 New prolapse operation in the first year, graft trial; risk ratio: 7/342 (2%) vs 10/337 (3%); RR 1.44; CI 0.56–3.73; P 0.45 New prolapse operation in the second year, mesh trial; risk ratio: 16/348 (5%) vs 15/343 (4%); RR 0.94; CI 0.47–1.88; P 0.87 New prolapse operation in the second year, graft trial; risk ratio: 15/299 (5%) vs 15/300 (5%); RR 0.99; CI 0.49–1.98; P 0.98 1-year International Consultation on Incontinence vaginal symptoms score (higher is worse), graft trial; mean difference: 7.1 (SD 6.9; n=294) vs 9.0 (SD 9.1; n=294); difference 1.31; CI 0.04–2.59; P 0.04 2-year any report of 'something coming down', graft trial; risk ratio: 91/298 (31%) vs 120/299 (40%); RR 1.26; CI 1.01–1.58; P 0.04	Glazener CMA, Breeman S, Elders A, et al. Mesh, graft, or standard repair for women having primary transvaginal anterior or posterior compartment prolapse surgery: two parallel-group, multicentre, randomised, controlled trials (PROSPECT). *Lancet.* 2017;389(10067):381-392. DOI
2021-09-27	Complications, all 2,632 operated women · secondary	—	De novo severe dyspareunia at 24 months by repair received: 3.5% vs 1.4% vs 3.5%; — Descriptive De novo vaginal tightness at 24 months by repair received: 1.6% vs 0.3% vs 1.2%; — Descriptive	Reid FM, Elders A, Breeman S, et al. How common are complications following polypropylene mesh, biological xenograft and native tissue surgery for pelvic organ prolapse? A secondary analysis from the PROSPECT trial. *BJOG.* 2021;128(13):2180-2189. DOI
2022-08-26	4- and 6-year follow-up · updated	—	6-year POP-SS, mesh trial; mean difference (mesh minus native tissue): 5.4 (SD 5.4; n=269) vs 6.3 (SD 6.2; n=272); difference 0.85; CI -0.05–1.75; P 0.063 Not met 6-year POP-SS, graft trial; mean difference (graft minus native tissue): 5.9 (SD 5.4; n=237) vs 6.2 (SD 5.8; n=248); difference 0.27; CI -0.64–1.18; P 0.565 Not met 6-year functional cure (no sense of something coming down, no prolapse surgery, no pessary), mesh trial; risk ratio: 144/268 (53.7%) vs 150/274 (54.7%); RR 1.02; CI 0.88–1.19; P 0.797 Not met 6-year functional cure, graft trial; risk ratio: 120/233 (51.5%) vs 122/247 (49.4%); RR 0.99; CI 0.83–1.18; P 0.892 Not met 4-year POP-SS, mesh trial; mean difference (mesh minus native tissue): 5.1 (SD 5.0; n=297) vs 6.1 (SD 6.0; n=311); difference 1.01; CI 0.2–1.83; P 0.015 4-year POP-SS, graft trial; mean difference (graft minus native tissue): 5.3 (SD 5.2; n=266) vs 5.8 (SD 5.5; n=262); difference 0.54; CI -0.31–1.39; P 0.212 Cumulative new prolapse operation within 6 years, mesh trial; risk ratio: 32/266 (12.0%) vs 28/271 (10.3%); RR 0.87; CI 0.54–1.4; P 0.572 Cumulative new prolapse operation within 6 years, graft trial; risk ratio: 27/231 (11.7%) vs 29/245 (11.8%); RR 1; CI 0.61–1.63; P 0.995 6-year completely satisfied with surgery, mesh trial: 144/260 (55.4%) vs 115/268 (42.9%); P <0.001 6-year completely satisfied with surgery, graft trial: 125/228 (54.8%) vs 108/234 (46.2%); P 0.255	Reid FM, Aucott L, Glazener CMA, et al. PROSPECT: 4- and 6-year follow-up of a randomised trial of surgery for vaginal prolapse. *Int Urogynecol J.* 2023;34(1):67-78. DOI

---

References

1. Glazener CMA, Breeman S, Elders A, et al. "Mesh, Graft, or Standard Repair for Women Having Primary Transvaginal Anterior or Posterior Compartment Prolapse Surgery: Two Parallel-Group, Multicentre, Randomised, Controlled Trials (PROSPECT)." Lancet. 2017;389(10067):381-392. doi:10.1016/S0140-6736(16)31596-3

2. Reid FM, Aucott L, Glazener CMA, et al. "PROSPECT: 4- and 6-Year Follow-Up of a Randomised Trial of Surgery for Vaginal Prolapse." Int Urogynecol J. 2023;34(1):67-78. doi:10.1007/s00192-022-05308-0

3. Reid FM, Elders A, Breeman S, et al. "How Common Are Complications Following Polypropylene Mesh, Biological Xenograft and Native Tissue Surgery for Pelvic Organ Prolapse? A Secondary Analysis from the PROSPECT Trial." BJOG. 2021;128(13):2180-2189. doi:10.1111/1471-0528.16897

4. Yeung E, Baessler K, Christmann-Schmid C, et al. "Transvaginal Mesh or Grafts or Native Tissue Repair for Vaginal Prolapse." Cochrane Database Syst Rev. 2024;3:CD012079. doi:10.1002/14651858.CD012079.pub2