Behavioral Monitoring and Neuromodulation of Feline Voiding Function

Abstract

Neuromodulation is a standard treatment for bladder dysfunction. Although preclinical studies continue to develop new approaches, these experiments are often performed under anesthesia that can affect normal bladder system activity. The goal of this study was to evaluate the use of typical neuromodulation parameters in an awake, behaving large-animal model. During an aseptic surgery, catheters were inserted into the bladder of two male felines and a cuff electrode was placed around the left pudendal nerve. Catheters and electrode leads were housed in an enclosure mounted on the lower back. After recovery, test sessions in an open enclosure were performed approximately weekly. In each session, body-temperature saline was infused while the bladder pressure was monitored until the animal voided one or more times. During some bladder fills, electrical stimulation of the pudendal nerve was applied (5 or 33 Hz; 1.5 or 2 × motor threshold - MT) to determine whether stimulation parameters affected the interval between voiding events or the voiding efficiency. Animals tolerated supra-threshold stimulation as long as the current was slowly increased to a target level. Five-Hz stimulation increased the voiding interval over 33 Hz and no stimulation trials while 33 Hz stimulation at 2 MT led to a greater voiding efficiency. These results follow trends observed in non-behavioral studies, and support the use of this animal model in further translational bladder neuromodulation research.

I. INTRODUCTION

Neuromodulation is a standard treatment for bladder dysfunction after conservative approaches have failed. Most preclinical bladder neuromodulation studies are performed using anesthetized animal models [1]–[4]. Anesthesia not only suppresses voluntary control but also alters the autonomic control of the bladder [5]–[7]. Many anesthetized studies use high stimulation amplitudes well above thresholds for activating motor fibers [1]–[4]. It is not well reported whether these amplitudes are tolerated by awake, behaving animals. For clinical implementation of preclinical developments, it is critical that necessary stimulation levels are fully tolerable.

Only a few studies have used non-anesthetized bladder neuromodulation models, such as an immobile spinal cord injury feline model [8], [9], a diabetic rat model [10], and a stationary sheep model [11], in part due to the challenges of instrumenting and performing experiments with behaving animals. To our knowledge, there are no studies reporting the effect of neuromodulation on bladder function in a fully mobile large animal. This is a valuable preclinical model that has utility for a variety of bladder dysfunction types [12]. In this pilot study with awake, behaving felines, we demonstrated that supra-threshold stimulation is tolerable and that bladder function can be modulated depending on the stimulation parameters.

II. Methods

A. Surgical Procedures

All experimental procedures were approved by the University of Michigan Institutional Animal Care and Use Committee. Aseptic techniques were used for all surgical procedures. Two male, domestic, short-hair felines (age: 1.0 years each; weight 6.4 and 5.1 kg; Liberty Research, Inc.) were anesthetized with a mixture of ketamine (6.6 mg/kg)-butorphanol (0.66 mg/kg)-dexmedetomidine (0.033 mg/kg) administered intramuscularly, intubated, and then maintained on isoflurane anesthesia (0.5–4%) during surgical procedures. Respiratory rate, heart rate, end-tidal CO₂, O₂ perfusion, and temperature were monitored continuously using a vitals monitor (Surgivet, Smiths Medical). Fluids (1:1 ratio of lactated Ringers solution and 5% dextrose) were infused intravenously via the cephalic vein at a rate of 5–10 ml/kg/hr.

After a laparotomy, two catheters (1 mm inner diameter, Silastic 508–005, Dow Corning) with injection ports mounted on the end were inserted into the bladder dome and secured in place with purse-string sutures. The catheter ends in the bladder were modified to have an arrow-like profile, with side wings that provided anchors inside the bladder. The catheters were tunneled subcutaneously to a dorsal midline incision just rostral to the tail. A custom-made bipolar electrode cuff (2 mm inner diameter) was placed around the left pudendal nerve via a postero-lateral incision. Catheter ports and electrode connections were housed within a 3-D printed photopolymer resin backpack (Formlabs) mounted rostral to the tail [13]. To mount the backpack, a stainless steel base plate was secured under the skin to the iliac crests, with stainless steel wire wrapped around bone screws embedded in the bone. The housing of the backpack was screwed into four transcutaneous posts of the baseplate after all incisions were closed. Fig. 1 shows a radiograph for one animal two weeks after surgery. After surgery completion, felines were weaned off anesthesia and ventilation and allowed to wake up on their own.

Figure 1.

Radiograph of Experiment 1 feline showing implanted materials.

Each feline was monitored for at least 5 days after surgery before testing sessions were started. A feline was considered recovered from surgery when it was walking, eating, and behaving normally. Analgesics were administered for 3 days during the post-surgery period. Animals were free-range housed with 2–4 fellow felines in a 413 ft² room with controlled temperature (19–21 °C) and relative humidity (35–60%), food and water available ad lib, and a 12 hr light/dark cycle. Enrichment was provided by toys and daily staff interaction.

B. Awake Testing Sessions

Test sessions were conducted approximately weekly in an enclosure consisting of two adjoined animal playpens (each with eight linked 0.61-m width by 1.22-m height panels) containing a litter box, non-absorbable litter, wet and dry food, and enrichments to encourage movement of the animal. Animals were either directly put in the cage or sedated first by an intramuscular injection of dexmedetomidine (0.019–0.039 mg/kg, Dexdomitor, MWI Veterinary Supply). Pressure lines were connected to the catheter ports prior to testing and were hung from the ceiling to reduce tangling and provide strain relief (Fig. 2). Bladder pressure was recorded by connection of fluid-filled tubing to the catheters, with one catheter also used as an infusion line. All pressure lines were flushed prior to experimentation to verify catheter patency. Pressure lines were connected to transducers (DPT-100, Utah Medical Products) and recorded with a PowerLab 8/35 (ADInstruments) or Grapevine Neural Interface Processor (Ripple) at 1 kHz.

Figure 2.

Experimental test setup. Bladder infusion and pressure monitoring performed through tethered, fluid-filled catheters.

An isolated Pulse Stimulator (Model 2100, A-M Systems) was connected to the wires of the pudendal cuff electrodes with alligator clips for delivering electrical stimulation. A push-button input to the recording system was used to track stimulus initiation and termination. The motor threshold (MT) was evaluated in each test session by identifying the minimum stimulation amplitude that elicited visible external anal sphincter contractions. If sedated, animals were injected intramuscularly with Antiseden (0.19–0.38 mg/kg, MWI Veterinary Supply) to reverse the effects of dexmedetomidine to begin the test session. Data recording began upon full awakening of the animal. Each test session consisted of one trial, where each trial consisted of multiple voiding events as a result of continuous infusion. In all test sessions, 41 °C sterile saline was infused at 0.5 ml/min into the bladder using a syringe pump (AS50 Infusion Pump, Baxter International). Bladder pressure was measured continuously.

Electrical stimulation was only delivered in a subset of bladder fills, applied continuously from the end of a previous void to the end of the current void. The stimulation amplitude was constant within a test session, set at 1.5 or 2.0 of the MT. The frequency during stimulation was either 5 Hz or 33 Hz with 200 μs pulse width. Stimulation frequency and amplitude combinations were randomized across trials. The stimulation amplitude was increased slowly, to limit pain or behavioral effects for the animal. In later sessions, non-absorbable litter (Kit4Cat) was used to allow for manual collection of urine and more accurate measurement of voided volumes.

Test sessions were recorded with video (C920 webcam, Logitech) for identifying voiding events and analyzing behavior. Animals were also continuously monitored with video in their housing room for overall health and behavior after implant procedures.

C. Data Analysis

Voiding intervals were calculated by determining the onset of each void based on reviewing the session video and bladder pressure data. Voiding efficiencies were calculated by dividing the voided volume by the infused volume plus an estimate of urine generated (1.1 ml/kg/hr [14]). Efficiencies were normalized within animals, to account for the assumed constant urine generation rate yielding efficiencies over 100% for a few voids. A two-way ANOVA with pair-wise comparisons was performed to evaluate the effect of frequency and amplitude on voiding interval and efficiency. Statistical significance was determined for α = 0.05.

III. Results and Discussion

In total, three animals received implants and underwent test sessions. The first animal (Experiment 0) had 11 bladder voids without stimulation and 8 bladder voids with sub-MT stimulation levels (70 μA, 200 μA) that were not normalized to MT. This animal was used to develop procedures and data from it is not reported here. For the two primary animals, ten sessions (9.2 ± 4.8 d interval) were recorded for the first animal (Experiment 1) and five (7.0 ± 1.6 d interval) for the second animal (Experiment 2), yielding 94 voids. Voiding efficiencies were obtained for 77 of these trials (82%). All data can be found at [15].

A. Awake Behavior and Motor Thresholds

Across test sessions, no signs of pain were observed as a result of stimulation. When no stimulation was applied, the animals walked around, rested, ate food, or played. At low stimulation amplitudes, there was no noticeable change in behavior. As the stimulation amplitude was slowly ramped to ~0.25 MT, the animal typically paused their current activity then resumed or walked around, suggesting perception. In some cases, the animal groomed their pelvic region. As the stimulation was increased to above MT, scrotum twitching became visually noticeable for 5 Hz stimulation. In a few trials, voluntary or involuntary left leg extension was observed while the animal was sitting or eating. However, no signs of pain were observed, such as vocalization, withdrawal, or shivering. Animals tolerated being in the enclosure for up to 2.3 hr.

The MT had a decreasing trend over time in both animals (Fig. 3). The average MT were 285.0 ± 56.4 μA and 494 ± 107.8 μA, for Experiments 1 and 2 respectively. These levels are comparable to those reported for anesthetized studies with a similar setup [2].

Figure 3.

Motor thresholds per testing date for each experiment.

The presence of the implant system itself modified overall animal behavior. During a multi-day observation period, the Experiment 1 feline visited a litter box every 2.1 hr on average, with 18.9% of voids occurring within 10 min of waking from sleep. The Experiment 2 feline had a shorter mean inter-void interval of 0.6 hr, with 18.2% of voids occurring within 10 min after waking. This contrasts with prior to surgery, as voiding intervals of 5.2 ± 3.7 hr and 3.3 ± 3.8 hr were observed for the two felines across a 24-hour period, with no voids within ten minutes of waking. We expect that a combination of the catheter within the bladders causing some irritation and general scarring around the implants led to a reduction in functional bladder volume.

B. Behavioral Bladder Pressure Monitoring

Awake bladder pressure traces were obtained for each session. Signal artifacts due to movement were not significant, particularly compared to pressure changes during a bladder contraction for voiding. Fig. 4 shows an example test session with multiple voids during periods of no stimulation and neuromodulation. The general characteristics of non-voiding contractions and voiding contractions were similar to those reported in anesthetized felines [16], [17]. However, we observed that the contraction amplitudes were generally higher (40–100 cmH₂O) than reported levels for anesthetized voids, and the onset of voiding contractions was often more rapid (data not reported). These observations support that anesthesia has a general suppressive effect on bladder activity.

Figure 4.

Example session with multiple voids across no-stimulation and stimulation periods. Shaded regions indicate periods of stimulation at each frequency (both at 1.5 MT).

Our setup did not allow the pressure transducer to move with animals. Thus, changes in posture could affect the relative pressure reading against the fixed-location transducers. During initial testing, we used a modified backpack that allowed for on-animal mounting of pressure transducers however, this was cumbersome for the animal and created stress.

C. Stimulation Effect on Voiding

We observed that the voiding interval can be modulated by stimulation frequency and amplitude. Stimulation at 5 Hz increased the median voiding interval for both 1.5 MT and 2 MT compared to no-stimulation and 33 Hz stimulation voids, with 2 MT yielding a stronger effect (Fig. 5). This observation aligns with previous anesthetized studies showing that increasing the stimulation amplitude leads to greater inhibition of bladder activity [4], [8], and that 5 Hz is effective at increasing the voiding interval or bladder capacity [1], [4].

Figure 5.

Effect of stimulation amplitude and frequency on interval between voids. Box plots indicate median values (middle line), 25^th and 75^th percentiles (lower and upper edges of boxes = interquartile range IQR), and minimum and maximum values (lower and upper error bars), with outliers that are 3 × IQR outside the 25^th or 75% percentile indicated by individual dots.

We also found that the voiding efficiency can be modulated, with the stimulation amplitude having a greater effect than the frequency (Fig. 6, Table I). At 1.5 MT, the median voiding efficiency showed a slight decrease for both 5 Hz and 33 Hz voids compared to no stimulation. At 2 MT, the voiding efficiency was significantly greater than at 1.5 MT (p = 0.017, Table II). Only 33 Hz at 2 MT increased the median voiding efficiency as compared to no stimulation periods, though the difference was not significant. The use of 33 Hz for increasing voiding efficiency aligns with prior studies with anesthetized animals [2], [3].

Figure 6.

Effect of stimulation amplitude and frequency on normalized voiding efficiency. Data within box plots as in Fig. 5.

TABLE I.

ANOVA Results (p-values) for effect of neuromodulation on voiding parameters

Factor	Voiding Interval	Voiding Efficiency
Frequency (no stim, 5 Hz, 33 Hz)	0.366	0.304
Amplitude (no stim, 1.5 MT, 2 MT)	0.208	0.034
Frequency × Amplitude	0.605	0.299

TABLE II.

Stimulus Amplitude Pair-Wise Comparison Results (p-values) for effect of amplitude alone on voiding efficiency

Factor Level Combination	Voiding Efficiency
No Stimulation – 1.5 MT	0.514
No Stimulation – 2 MT	0.107
1.5 MT – 2 MT	0.017

Regardless of stimulation amplitude, 5 Hz stimulation consistently resulted in longer voiding intervals and lower voiding efficiencies compared to 33 Hz. Regardless of stimulation frequency, 2 MT consistently resulted in both higher voiding intervals and higher voiding efficiencies. This suggests that higher-amplitude stimulation can potentially deliver better overall clinical results if given at a tolerable level. In contrast, there may not be a single stimulation frequency that most effectively modulates all urodynamic parameters, as different stimulation frequencies may be necessary to maximize clinical utility for both continence and micturition.

IV. Conclusion

This is the first study to investigate the effect of neuromodulation on bladder function in a healthy, behaving large-animal model. We found that in an awake animal, stimulating the pudendal nerve as high as 2 MT is well tolerated and can result in improved voiding efficiency and longer voiding intervals. While 5 Hz was preferential for continence-like effects such as increasing the voiding interval, 33 Hz was preferential for enhancing micturition, as has been reported previously. Further study, with a larger sample size will allow for increased quantification of urodynamic parameters, a wider exploration of the stimulation parameter space, and evaluating conditional rather than continuous stimulation.