Sacral neuromodulation is clinically used for refractory urinary incontinence and generally uses continuous stimulation. Recent studies have indicated conditional SNM, which applies stimulation intermittently, can produce equivalent functional outcome to continuous SNM [1]. Conditional SNM could reduce energy use and is a potential research tool for optimizing SNM use. However, conditional SNM requires feedback on bladder state (pressure or volume) to enable stimulation decisions. 
This proof-of-concept study was to demonstrate feasibility of automatically triggering SNM using real-time wireless bladder pressure data in urologically normal sheep. Sheep were implanted with Summit™ RC+S, an implantable research stimulator allowing for sub-second control of stimulation through wireless communication. A minimally invasive bladder pressure (MIBP) sensor and a computer algorithm detected the onset of bladder contractions and automatically controlled the Summit device (Figure 1). Continuous and four conditional SNM paradigms were tested over 5 days in four conscious sheep.

The MIBP device used low-power flexible electronics housed in a medical silicone housing which curled into a pigtail shape after insertion to remain in the bladder. The MIBP transmitted vesical pressure data at 10 Hz wirelessly. MIBP data were processed by a Context-Aware Thresholding (CAT) algorithm in a custom Labview (National Instruments) program to detect bladder contractions in real time. The Labview program communicated with the Summit system software to remotely control an implanted neurostimulator (INS) (Figure 1). 
SNM studies were performed on four Polypay sheep (age: 15.9 ± 1.1 months; weight: 57.3 ± 2.4 kg). Surgical and urological monitoring methods were detailed previously [2]. Briefly, animals were anesthetized for bilateral implantation of two Model 3889 InterStim® quadripolar leads in S3 (n=2) or S4 (n=2) based on intraoperative perianal response to stimulation. Leads were connected to the INS (Summit RC+S, B35300R). All animals were recovered for at least 30 days prior to the study. 
Sheep stood in a sling frame during repeat, single fill cystometry to assess bladder capacity at baseline (no stimulation) or in response to SNM protocols. Bladder pressure was recorded (Mikro-CathTM, Millar, Houston, TX) in 3 sheep through a 12 Fr filling catheter (Lubri-Sil®, Bard Medical, Covington, GA) and 1 sheep with an 8 Fr catheter due to urethral sensitivity. Warmed saline was infused at 30 ml per minute (Flo-Gard R pump, Baxter, Deerfield, IL) until voiding or a pressure of 30-mmHg was reached. Capacity was defined as the saline amount infused. 
Sheep remained in the sling frame during insertion of the MIBP without analgesia or antibiotics. MIBPs were inserted through the urethra similarly to catheterization. After insertion, a wireless antenna was placed near the animal’s hip to receive MIBP pressure data. The MIBP was removed using conscious cystoscopy. 
Animals received three baseline cystometric fills followed by a random progression of conditional or continuous SNM trials. Baseline capacity per experiment day was calculated as the average of 3-6 fills without stimulation. Testing consisted of two conditional physiologic paradigms, two pressure-triggered paradigms, and continuous SNM (Figure 2). For continuous SNM, stimulation was activated prior to filling start. For physiologically-timed paradigms, SNM was either delivered during the first half of the fill cycle or the second half of the fill cycle. Fill cycle duration was determined per sheep, per day, by averaging initial baseline fills. Physiologic conditional SNM was triggered manually based on infused bladder volume. 
The two pressure-triggered stimulation paradigms were oscillating and latched-stimulation. In the oscillating paradigm, the pressure signal was used to trigger stimulation on and off based on the CAT algorithm. The latched paradigm used CAT to detect the start of bladder contractions after an initial 30-s lockout period, then started and maintained SNM until the void. The lockout period was used to prevent CAT false positives due to pressurization artifacts when bladder capacity was below 15 mL.

The MIBP was easy to insert and extract and did not cause any noticeable discomfort or urinary tract complications. Throughout cystometry, the MIBP showed repeatable artifacts on initial pressurization (below 15 mL) and after voiding. We conclude that the MIBP was measuring contact force from the detrusor in addition to bladder pressure which limited the ultimate pressure accuracy compared to catheter values. MIBP data was time-correlated with catheter-measured contractions.
132 cystometric trials were performed with the MIBP in place (55 baseline, 19 continuous SNM, 20 oscillating SNM, 17 latched SNM, 4 first-half SNM, 4 second-half SNM). 13 trials were discarded from analysis due to early voids below 50% baseline capacity, likely caused by catheter-related spasticity. Pooled analysis showed capacity changes relative to baseline of: 157% (continuous SNM), 147% (oscillating SNM), 112% (latched SNM), 84% (first-half SNM), and 268% (second-half SNM). 
In general, the CAT algorithm was sufficiently accurate to detect detrusor contractions, triggering SNM 2.8 times (median, excluding discarded trials) per bladder fill cycle. SNM was occasionally falsely triggered due to MIBP data reception loss or pressurization artifacts. Including all stimulation paradigms, a total of 1,224 commands were sent to the Summit system. A median latency of 72 ms occurred between data received into the Labview system and the implanted RC+S pulse generator changing SNM state.

To our knowledge, this study is the first to demonstrate a wireless system using real-time bladder pressure data to automatically trigger SNM. While our approach was limited to conscious animals with short-range communication to PC software, an integrated system could conceivably be developed in future work. The median time to send bladder data, process it by CAT, and communicate to the Summit system was within the range of human reaction time, indicating that rapid decisions and activation of SNM is feasible. Occasionally SNM was errantly triggered due to data interruptions from the MIBP radio link. Our results indicate basic feasibility of automatic, conditional SNM; reliable reception of the feedback signal was a critical link of this approach. 
This study’s limitations prevent statistical conclusions on the efficacy or mechanisms of conditional SNM paradigms. It was limited to five days of observation, tested on only four animals, and stimulation paradigms were tested 1-6 times per animal. This study also used a urologically normal animal model which has been shown to increase bladder capacity during continuous SNM [2]. Conditional, physiologically-timed SNM has also been demonstrated to affect bladder capacity in this continent model [3]. However, conditional neuromodulation could perform differently in the case of an incontinent model, and is an area of future research

This study showed that both algorithm-controlled and physiologically-timed conditional SNM paradigms were technically feasible with this system, as conditional SNM was reliably triggered in the sheep. Outcomes from oscillating and latched paradigms suggested that some animals responded more strongly than others. This could have been due to a number of variable but is also representative of clinical use as the effectiveness of SNM varies between individuals. Future research with greater statistical power could quantify the efficacy of conditional SNM for improving bladder capacity.

S. Siegel, K. Kreder, E. Takacs, R. McNamara, and F. Kan, “Prospective randomized feasibility study assessing the effect of cyclic snm on UUI in women,” Female Pelvic Med. Reconstr. Surg., 2018
T. S. Brink, P. L. Zimmerman, M. A. Mattson, X. Su, and D. E. Nelson, “A chronic, conscious large animal platform to quantify therapeutic effects of sacral neuromodulation on bladder function,” J. Urol., vol. 194, no. 1, pp. 252–258, 2015
B. A. Potts et al., “Timing of sacral neurostimulation is important for increasing bladder capacity in the anesthetized rat,” Am. J. Physiol. Renal Physiol., vol. 317, no. 5, pp. F1183–F1188, 2019