IMPROVE: Perioperative Vaginal Estrogen with Native-Tissue Apical Prolapse Repair
IMPROVE was a three-center, double-masked superiority trial in 199 postmenopausal women planning native-tissue vaginal apical repair that compared perioperative vaginal conjugated estrogen cream with identical placebo and found no improvement in its prespecified 12-month composite failure endpoint (model-estimated failure 19% versus 9%; adjusted HR 1.97, 95% CI 0.92–4.22; P = 0.08).[1]
- Design and population
- Three-center, double-masked randomized superiority trial. Postmenopausal women aged at least 48 years with bothersome apical and/or anterior prolapse at least to the hymen planning native-tissue transvaginal apical repair.
- Arms
- Perioperative vaginal conjugated estrogen cream versus identical placebo cream
- Report and follow-up
- 2023 · 12 months after surgery; 186/199 randomized participants underwent surgery and had postoperative assessment
| Endpoint | Results | Effect | Finding |
|---|---|---|---|
| Model-estimated 12-month failure incidence; adjusted hazard ratio for estrogen versus placebo | 19% (20 events) vs 9% (10 events) | HR 1.97; CI 0.92–4.22; P 0.08 | Not met |
Harms: Bladder perforation: 3/93 (3.2%) vs 6/93 (6.5%); Ureteral injury: 5/93 (5.4%) vs 7/93 (7.5%); At least one postoperative urinary tract infection: 18/93 (19.4%) vs 23/93 (24.7%)
Design
Eligible patients had symptomatic anterior and/or apical descent at least to the hymen and were planning transvaginal native-tissue repair. Vaginal atrophy symptoms were not required. Current smokers, patients using hormonal treatments affecting the vagina, and those with prior apical repair or prolapse mesh were excluded. Participants, surgeons and outcome assessors were masked to cream assignment; the apical suspension route was chosen by the surgeon.[1]
Failure meant any anterior or posterior wall descent beyond the hymen, apical descent beyond one-third of total vaginal length, a bothersome bulge, or further prolapse treatment by surgery or requested pessary. The comparison used time to first failure in the 93 operated and assessed participants per arm. The trial tested superiority at a two-sided alpha of 0.05. Its 12-month percentages are model estimates from interval-censored time-to-event analysis, rather than event counts divided by the operated group sizes.[1]
Results
There were 20 composite failure events with estrogen and 10 with placebo. Anatomical descent was the most frequent failure component. In a secondary patient-reported measure, 88/89 (98.9%) estrogen recipients and 84/86 (97.7%) placebo recipients rated themselves much or very much better at 12 months (P = 0.62).[1]
At operation, masked surgeons rated apical tissue quality 3.6 versus 3.3 on a 1-to-5 scale (mean difference 0.3 points, 95% CI 0.06–0.56; P = 0.02). Among participants with bothersome atrophy symptoms at baseline, the adjusted score for the most bothersome symptom at 12 months was 1.06 after estrogen and 1.62 after placebo; the between-group difference in change was −0.87 points (95% CI −1.67 to −0.07; P = 0.003).[1]
At 36 months, 164/186 operated participants supplied follow-up data. Model-estimated composite failure was 32.6% with estrogen and 26.8% with placebo (adjusted HR 1.55, 95% CI 0.90–2.66; P = 0.11). Of respondents, 75/80 (93.8%) and 72/76 (94.7%) reported being much or very much better, including 51 of 55 participants classified as surgical failures.[2]
Surgery and treatment delivery
Patients inserted 1 g of conjugated estrogen cream (0.625 mg/g) or placebo nightly for 2 weeks, then twice weekly to complete at least 5 total weeks before surgery. Cream resumed at discharge and continued twice weekly through 12 postoperative months. By diary, 77% of estrogen recipients and 74% of placebo recipients used at least half of scheduled cream through that period.[1]
The standardized apical repair used uterosacral or sacrospinous ligament suspension with permanent and delayed absorbable monofilament sutures. Vaginal hysterectomy was performed when the uterus remained. Of the 93 operated patients per group, uterosacral suspension was performed in 62 versus 75 and sacrospinous suspension in 29 versus 17. Concomitant anterior repair occurred in 82 versus 73, posterior repair in 61 versus 59, and midurethral sling in 53 versus 53.[1]
After the assigned cream ended, participants could use vaginal estrogen for symptoms. At 36 months, 40/82 (48.8%) originally assigned estrogen and 47/78 (60.3%) originally assigned placebo reported vaginal estrogen use.[2]
Adverse events
Intraoperative bladder perforation was reported in 3/93 (3.2%) estrogen recipients and 6/93 (6.5%) placebo recipients; ureteral injury in 5/93 (5.4%) and 7/93 (7.5%). At least one postoperative urinary tract infection occurred in 18/93 (19.4%) and 23/93 (24.7%) through 12 months (P = 0.38).[1]
After postoperative month 6, granulation tissue was found in 16/90 (17.8%) versus 7/87 (8.0%), and suture exposure in 14/90 (15.6%) versus 15/87 (17.2%). Neither finding led to repeat operation in the report.[1]
Limitations
The modified intention-to-treat analysis excluded randomized participants who did not undergo surgery, and the failure composite included apical descent before the apex reached the hymen. Atrophy outcomes used a study-designed, nonvalidated questionnaire. The primary failure estimate had a wide confidence interval. After the assigned cream stopped at 12 months, vaginal estrogen use crossed between original groups, complicating interpretation of the later comparison.[1][2]
In conversation
- Builds on OPTIMAL (2014). Used OPTIMAL's uterosacral and sacrospinous suspension framework and a similar composite failure definition.
OPTIMAL informed the two native-tissue apical suspension routes and failure definition used here. IMPROVE tested the estrogen adjunct while surgeons selected the suspension route.[1]
Report history
| Published | Report | Cutoff | Headline results | Citation |
|---|---|---|---|---|
| 2023-08-15 | Twelve-month randomized report · primary | — | Model-estimated 12-month failure incidence; adjusted hazard ratio for estrogen versus placebo: 19% (20 events) vs 9% (10 events); HR 1.97; CI 0.92–4.22; P 0.08 Not met Much or very much better on PGI-I at 12 months: 88/89 (98.9%) vs 84/86 (97.7%); P 0.62 Mean rating at surgery on a 1-to-5 scale; mean difference in points: 3.6 (SD 0.8) vs 3.3 (SD 0.9); difference 0.3; CI 0.06–0.56; P 0.02 Adjusted mean symptom score at 12 months among those bothered at baseline; mean difference in points: 1.06 (baseline 3.20) vs 1.62 (baseline 2.89); difference -0.87; CI -1.67–-0.07; P 0.003 | Rahn DD, Richter HE, Sung VW, et al. Perioperative Vaginal Estrogen as Adjunct to Native Tissue Vaginal Apical Prolapse Repair: A Randomized Clinical Trial. *JAMA.* 2023;330(7):615-625. doi:10.1001/jama.2023.12317 DOI |
| 2024-05-04 | Three-year follow-up · updated | — | Model-estimated failure through 36 months; adjusted hazard ratio for estrogen versus placebo: 32.6% vs 26.8%; HR 1.55; CI 0.9–2.66; P 0.11 Much or very much better on PGI-I at 36 months: 75/80 (93.8%) vs 72/76 (94.7%); P >0.99 | Rahn DD, Richter HE, Sung VW, et al. Three-year outcomes of a randomized clinical trial of perioperative vaginal estrogen as adjunct to native tissue vaginal apical prolapse repair. *Am J Obstet Gynecol.* 2024;231(2):263.e1-263.e10. doi:10.1016/j.ajog.2024.04.042 DOI |
References
1. Rahn DD, Richter HE, Sung VW, et al. "Perioperative Vaginal Estrogen as Adjunct to Native Tissue Vaginal Apical Prolapse Repair: A Randomized Clinical Trial." JAMA. 2023;330(7):615-625. doi:10.1001/jama.2023.12317
2. Rahn DD, Richter HE, Sung VW, et al. "Three-year outcomes of a randomized clinical trial of perioperative vaginal estrogen as adjunct to native tissue vaginal apical prolapse repair." Am J Obstet Gynecol. 2024;231(2):263.e1-263.e10. doi:10.1016/j.ajog.2024.04.042