Sacral Neuromodulation (SNM) is a guideline recommended treatment for symptoms of overactive bladder (OAB). Therapy is set by selecting the stimulation program (electrode configuration) and adjusting the stimulation amplitude for each patient. To account for variation in the nerve-electrode interface arising from lead position and nerve anatomy, several stimulation programs are available that utilize different electrode configurations. Stimulation programs provide the user flexibility to create differently sized/shaped electrical fields by using various combinations of electrodes. Currently, patient feedback regarding perception of sensation is used when determining these parameters. One strategy for programming is to choose the program with the lowest sensory threshold (ST, amplitude at which sensation is first perceived), and set the amplitude of stimulation to the ST or slightly below ST. Previous studies have shown that sacral stimulation evoked electrical signals could be detected from implanted sacral leads [1,2]. Recent characterization of sacral evoked responses (SERs) showed that stimulation amplitudes eliciting a measurable SER (e.g., SER threshold) correlate with clinically used measures of stimulation response (e.g., ST) [3]. Extending our characterization of SERs, here we evaluate SER thresholds with changing electrode configurations and whether the SER threshold and STs correlate across multiple bipolar stimulation and sense electrode configurations.

The PEER 2 study is an ongoing multi-center, prospective, clinical feasibility study. Subjects with OAB that met all inclusion and no exclusion criteria were implanted with an InterStimTM tined lead. Following lead implant, SER recordings were collected from the lead using an investigational research system. SERs and STs were collected immediately post-operatively (PO) and at the end of therapy evaluation (post-trial) across six different electrode configurations. The electrode configurations are presented as the bipolar stimulation pair (e.g. configuration E3/E0 is stimulation across most proximal electrode 3 and most distal electrode 0), while sensing across the remaining two electrodes (E1/E2). Recordings were processed and analyzed to determine stimulation amplitudes associated with first detectable signal (SER threshold). Recordings were excluded from analysis if paired data (ST and SER threshold) were not available. Correlations were performed to evaluate the relationship between SER thresholds and ST with all electrode configurations grouped, and further correlation was determined for each individual electrode configuration.

To date, 39 subjects have completed data collection at implant and follow up. These subjects had a median (IQR) age of 66 (18) years, and at baseline urinary urge incontinence subjects (N=32) had a median (IQR) 6.3 (6.3) leaks/day and urgency frequency subjects (N=30) had 10.2 (4.6) voids/day. After lead placement, the average PO ST was 0.9±0.4 mA. The median (range) therapy exposure was 14 (7-14) days (N=39 subjects). There were no adverse events during the therapy evaluation when the investigational device was in use. There have been 3 adverse events reported following implant of the commercial neurostimulator which were related to the device, procedure, or therapy.

To evaluate whether SER threshold changes with electrode configuration within subjects, SER thresholds were plotted by electrode configuration (examples shown in Figure 1). The SER threshold varied by configuration within each subject. The median coefficient of variation (% CV,  standard deviation/mean SER threshold) of SER thresholds was quantified for each subject, the median CV across subjects was 24% at PO and 32% at post-trial, indicating that there is variation in SER threshold across configurations.  

With all configurations/programs tested grouped, SER thresholds were very strongly correlated with STs collected during PO (R=0.89, p<0.01, n=179 signals) and post-trial (R=0.58, p<0.01, n=211 signals), consistent with earlier reports from a smaller dataset. Further, for each tested electrode configuration, the SER threshold and ST were significantly correlated (R≥0.47, p≤0.01, as shown in Figure 2). The weakest and strongest correlations existed for a narrow sense configuration (stim E0/E1, sense E2/E3, R=0.47) and wide sense configuration (stim E1/E2, sense E0/E3, R=0.97), respectively.

The current state of SNM programming, where optimizing stimulation is based on a subjective patient-reported ST, presents an opportunity for improvement using a more objective measure. We found that within a subject, SER threshold changes with electrode configuration. This behavior is to be expected as each electrode configurations creates a slightly different stimulation field, and therefore may have a different threshold to activate the nerve. While the sensory or SER threshold may vary across configuration, we would expect the correlation between SER threshold and ST remain consistent while changing electrode configuration as each measure is a proxy for neural activation. Consistent with this, SER threshold and ST correlated for each configuration tested. The strength of correlation did vary slightly between configurations, potentially reflecting both the optimal stimulation location and sensing configuration.

These data support that lead-measured SERs may be a useful tool to optimize stimulation, serving as an objective measure of the body’s response to stimulation. SER and ST correlated for all electrode configurations, indicating that SERs could be a valuable tool to aid or replace subjective measures used for optimizing lead placement and programming, and further enable closed-loop SNM.

Goudelocke C, et al. Sensing Sacral Signals: Pelvic Electrically Evoked Recording (PEER) Feasibility Study Findings. Neurourol Urodyn. 2022;41 Suppl 1:S5-S254
Gmel GE, et al. Electrophysiological Responses in the Human S3 Nerve During Sacral Neuromodulation for Fecal Incontinence. Front Neurosci. 2021;15:1-10.
Goudelocke C, et al. Detecting and Decoding Signals – A Sacral Neuromodulation Lead Sensing Study. Presented at SUFU Annual Conference, March 10, 2023

Continence 7S1 (2023) 100935
DOI: 10.1016/j.cont.2023.100935