Introduction: Intraoperative Predictors of Sacral Neuromodulation Implantation & Treatment Response - Results from the ROSETTA Trial
Methods: As part of the prospective, randomized, multi-center ROSETTA trial, intra-op details of SNM stage 1 were collected: responsive electrodes, amplitudes, and strength of motor and sensory response (Likert scale). Surgeons were instructed to achieve intra-op motor or sensory response at amplitudes < 5V on = 2 electrodes. Participants with stage 1 success (= 50% reduction in UUIE on 3-day diary) received a stage 2 implant. Demographics, baseline OAB symptoms, and therapy response were recorded in the primary trial. An intra-op amplitude-response score for each electrode was calculated, ranging from 0 (no response at high amplitude) to 90 (maximum response at 1 V). Statistical methods included Fisher’s exact test, as well as logistic and linear regression.
Results: Of 161 women with intra-op data, 141 (88%) had a positive motor or sensory response at < 5V on = 2 electrodes; 139 (86%) had stage 1 success, while 22 (14%) did not. Demographics, baseline symptoms, and surgery duration did not differ by stage 1 result.
Stage 1 success did not differ by number of electrodes with intra-op motor and/or sensory response, average amplitude at responsive electrodes, or minimum amplitude producing responses. Relative to other electrodes, a best amplitude-response score for pelvic bellows at electrode 3 was associated with stage 1 failure (11/22 (50%) vs 36/138 (26%); p=0.0409). This was the strongest predictor of Stage 1 response in stepwise and LASSO logistic modeling. Similarly, in a linear model of Stage 1 percent improvement in daily UUIE, a best amplitude-response score for pelvic bellows at electrode 3 was associated with lower percent improvement (mean ± standard deviation = 66% ± 31, N=47 vs 79% ± 25, N=113, p=0.008).
After stage 2, mean daily UUIE reduction at 24 months was less for patients with an intra-op sensation response at electrode 3 (mean ± SD = -2.5 ± 2.2, N=115) than without (-5.0 ± 3.2, N=22, p=0.005). No intra-op variables predicted change in UUIE at 6 months.
Conclusions: Intraoperative data during stage 1 sacral neuromodulation show a limited ability to predict trial stimulation outcome, but were constrained by small sample size. Future research into optimal stimulus parameters is needed. Source of