ROSETTA: Refractory Overactive Bladder—Sacral Neuromodulation Versus Botulinum Toxin Assessment
ROSETTA was a nine-center, open-label randomized superiority trial in women with refractory urgency urinary incontinence comparing one cystoscopic 200 U onabotulinumtoxinA injection with staged InterStim sacral neuromodulation; the injection produced a small additional reduction in daily episodes over six months, with more urinary tract infection and intermittent catheterization.[1]
- Design and population
- Multicenter, open-label randomized superiority trial. 386 randomized women at nine US centers with at least six urgency-incontinence episodes on a three-day diary despite supervised behavioral or physical therapy and at least two anticholinergics or inability to use them.
- Arms
- Cystoscopic intradetrusor onabotulinumtoxinA 200 U versus staged sacral neuromodulation (InterStim)
- Report and follow-up
- 2016 · Mean monthly diary change through six months
| Endpoint | Results | Effect | Finding |
|---|---|---|---|
| Mean daily episode reduction averaged over six months; between-group difference in reduction, onabotulinumtoxinA minus SNM | 3.9 fewer/day (n=190) vs 3.3 fewer/day (n=174) | difference 0.63; CI 0.13–1.14; P 0.01 | Met |
Harms: Treated urinary tract infection through six months: 66/191 (35%) vs 20/178 (11%); Intermittent catheterization after toxin: 38/191 (20%) at any visit; 16/191 (8%) at one month; 4/191 (2%) at six months vs not applicable; SNM device revision or removal through six months: not applicable vs 6/178 (3%)
Design
The trial randomized 386 women after at least one supervised behavioral or physical therapy intervention and at least two anticholinergic medications, unless those drugs were intolerable or contraindicated. Entry required at least six urgency-incontinence episodes on a three-day diary. Relevant neurologic disease, previous treatment with either study intervention, and postvoid residual above 150 mL were exclusions. The assigned-treatment groups in the primary report contained 381 eligible women; 364 contributed at least one postbaseline diary to the modified intention-to-treat analysis.[1]
Participating implanters and injectors had each performed at least 10 procedures of their assigned type and reviewed standardized technique videos. The primary endpoint was the change from baseline in mean daily urgency-incontinence episodes across monthly three-day diaries over six months. The prespecified superiority test used two-sided alpha 0.05; the study was powered around a two-episode daily difference, while secondary comparisons were exploratory without multiplicity adjustment.[1]
Results
In the 364-woman primary analysis, mean daily episodes fell by 3.9 with onabotulinumtoxinA and 3.3 with sacral neuromodulation across six months: a mean difference in reduction of 0.63 episode per day (95% CI 0.13–1.14; P = 0.01), meeting the prespecified superiority test. Among women with at least four monthly diaries, complete resolution occurred in 35/178 (20%) versus 6/166 (4%); this was an exploratory outcome. Treatment-satisfaction scores were 67.7 versus 59.8 (mean difference 7.8, 95% CI 1.6–14.1).[1]
The planned two-year analysis followed initial clinical responders, with outcome data for 260/298 (87%). Mean daily episode reductions across 24 months were 3.88 with toxin and 3.50 with neuromodulation (mean difference 0.38, 95% CI −0.14 to 0.89; P = 0.15). Complete resolution at month 24 was estimated at 5% in each group. Treatment-satisfaction scores across follow-up remained higher after toxin (68.1 versus 58.9; reported difference for neuromodulation minus toxin −9.1, 95% CI −14.4 to −3.9).[2]
In the trial-based health-sector economic analysis, mean two-year costs in 2018 US dollars were $7,460 for toxin and $35,680 for neuromodulation, a $28,220 difference. Among participants with QALY data (150 toxin, 134 neuromodulation), adjusted quality-adjusted life-years were 1.41 and 1.39, respectively (difference for neuromodulation minus toxin −0.02, 95% CI −0.09 to 0.05).[3]
Surgery and treatment delivery
For neuromodulation, first-stage lead placement took place in the operating suite under local and monitored anesthesia care. Teams tested sensory and motor responses at each electrode, then permitted program adjustments during a 7–14-day test. A reduction of at least 50% on a three-day diary qualified the patient for permanent generator implantation; nonresponders had the lead removed. Of 169 eligible assigned women who underwent first-stage placement, 142 (84%) responded and 139 received the generator. The toxin arm received a 200 U cystoscopic intradetrusor injection in clinic; 189/192 randomized women received the full dose and 159/192 (83%) met the one-month response definition.[1]
Postinjection intermittent catheterization began for residual above 300 mL, or above 200 mL with incomplete-voiding symptoms.[1] Through two years, 115/159 (72%) initial toxin responders requested a second injection after a median 350 days; 101 received another 200 U and 14 were reduced to 100 U after prolonged catheterization.[2] Among implanted initial neuromodulation responders, 81/139 (58%) needed reprogramming.[1][2]
Adverse events
Through six months, treated urinary tract infection occurred in 66/191 (35%) toxin recipients and 20/178 (11%) neuromodulation recipients. Intermittent catheterization after toxin was required at some visit by 38/191 (20%), including 16/191 (8%) at one month and 4/191 (2%) at six months; median catheterization duration for any reason was 37 days. Neuromodulation revision or removal occurred in 6/178 (3%).[1]
Over two years, recurrent UTI among women without baseline recurrent UTI occurred in 40/167 (24%) after toxin and 16/154 (10.4%) after neuromodulation. Catheterization at any point after toxin occurred in 45/189 (24%). Among 139 implanted initial neuromodulation responders, 4 (3%) underwent revision and 12 (8.6%) had device removal; the reported removal reasons were infection, reduced efficacy, patient preference, and pain.[2]
Limitations
Treatments and assessments were open-label. The six-month primary analysis missed roughly 10% of monthly diaries, with more missing in the neuromodulation arm.[1] The two-year efficacy analysis was restricted to initial responders and permitted repeat toxin injections and neuromodulator reprogramming. The economic analysis used trial-era 2018 US costs; its five-year estimates extrapolated two-year outcomes.[1][2][3]
In conversation
- Builds on ABC (2012). The earlier 100 U toxin-versus-anticholinergic trial informed ROSETTA's dose and refractory-population rationale.
Same setting Nitti OAB Trial (2012), InSite (2014), ARTISAN-SNM (2019)
The ABC trial compared a lower toxin dose with anticholinergic medication, whereas ROSETTA enrolled women with refractory symptoms and compared a higher toxin dose directly with staged neuromodulation.[1]
Guidelines and approvals
- AUA/SUFU (2024). For OAB with inadequate response or intolerable adverse effects from pharmacotherapy or behavioral therapy, offer sacral neuromodulation, tibial nerve stimulation, and/or intradetrusor botulinum toxin .[4]
- EAU female LUTS guideline (2026). Offer bladder-wall onabotulinumtoxinA 100 U for OAB or urgency urinary incontinence refractory to conservative or drug treatment ; warn about limited response duration, UTI, and possible prolonged intermittent catheterization .[5]
- EAU female LUTS guideline (2026). Offer sacral nerve stimulation for OAB or urgency urinary incontinence refractory to anticholinergic therapy .[5]
Report history
| Published | Report | Cutoff | Headline results | Citation |
|---|---|---|---|---|
| 2016-10-04 | Six-month randomized comparison · primary | — | Mean daily episode reduction averaged over six months; between-group difference in reduction, onabotulinumtoxinA minus SNM: 3.9 fewer/day (n=190) vs 3.3 fewer/day (n=174); difference 0.63; CI 0.13–1.14; P 0.01 Met Complete urgency-incontinence resolution with at least four monthly diaries through six months; risk difference reported as SNM minus toxin: 35/178 (20%) vs 6/166 (4%); difference -16; CI -26–-5; P <0.001 Six-month Overactive Bladder Satisfaction of Treatment score; mean difference toxin minus SNM: 67.7 vs 59.8; difference 7.8; CI 1.6–14.1; P 0.01 | Amundsen CL, Richter HE, Menefee SA, et al. "OnabotulinumtoxinA vs Sacral Neuromodulation on Refractory Urgency Urinary Incontinence in Women: A Randomized Clinical Trial." *JAMA.* 2016;316(13):1366-1374. doi:10.1001/jama.2016.14617 DOI |
| 2018-02-24 | Two-year clinical-responder follow-up · updated | — | Mean daily episode reduction across 24 months in initial responders; mean difference in reduction, toxin minus SNM: 3.88 fewer/day vs 3.50 fewer/day; difference 0.38; CI -0.14–0.89; P 0.15 Mean Overactive Bladder Satisfaction score across 6–24 months; difference SNM minus toxin: 68.1 vs 58.9; difference -9.1; CI -14.4–-3.9; P <0.001 Estimated proportion with no urgency-incontinence episodes at month 24: 5% vs 5%; — | Amundsen CL, Komesu YM, Chermansky C, et al. "Two-Year Outcomes of Sacral Neuromodulation Versus OnabotulinumtoxinA for Refractory Urgency Urinary Incontinence: A Randomized Trial." *Eur Urol.* 2018;74(1):66-73. doi:10.1016/j.eururo.2018.02.011 DOI |
| 2019-11-18 | Trial-based economic analysis · secondary | — | Mean per-person health-sector cost in 2018 US dollars; difference SNM minus toxin: $7,460 vs $35,680; difference 28220; P <0.001 Adjusted quality-adjusted life-years; difference SNM minus toxin: 1.41 vs 1.39; difference -0.02; CI -0.09–0.05; P 0.60 | Harvie HS, Amundsen CL, Neuwahl SJ, et al. "Cost-Effectiveness of Sacral Neuromodulation versus OnabotulinumtoxinA for Refractory Urgency Urinary Incontinence: Results of the ROSETTA Randomized Trial." *J Urol.* 2020;203(5):969-977. doi:10.1097/JU.0000000000000656 DOI |
References
1. Amundsen CL, Richter HE, Menefee SA, et al. "OnabotulinumtoxinA vs Sacral Neuromodulation on Refractory Urgency Urinary Incontinence in Women: A Randomized Clinical Trial." JAMA. 2016;316(13):1366-1374. doi:10.1001/jama.2016.14617
2. Amundsen CL, Komesu YM, Chermansky C, et al. "Two-Year Outcomes of Sacral Neuromodulation Versus OnabotulinumtoxinA for Refractory Urgency Urinary Incontinence: A Randomized Trial." Eur Urol. 2018;74(1):66-73. doi:10.1016/j.eururo.2018.02.011
3. Harvie HS, Amundsen CL, Neuwahl SJ, et al. "Cost-Effectiveness of Sacral Neuromodulation versus OnabotulinumtoxinA for Refractory Urgency Urinary Incontinence: Results of the ROSETTA Randomized Trial." J Urol. 2020;203(5):969-977. doi:10.1097/JU.0000000000000656
4. American Urological Association and Society of Urodynamics, Female Pelvic Medicine & Urogenital Reconstruction. The AUA/SUFU Guideline on the Diagnosis and Treatment of Idiopathic Overactive Bladder. 2024. Statement 25.
5. European Association of Urology. EAU Guidelines on Management of Non-Neurogenic Female Lower Urinary Tract Symptoms. Limited update March 2026. Sections 4.1.4.c.1 and 4.1.4.c.2.