Improved sacral neuromodulation in the treatment of the hyperactive detrusor: signal modification in an animal model

Authors


Dr P.M. Braun, Department of Urology, University Hospital Kiel, D-24105 Kiel, Germany.
e-mail: pbraun@urology.uni-kiel.de

Abstract

OBJECTIVE

To investigate different stimulation signals for the peripheral nerve evaluation test (PNE, carried out before implanting a sacral neuromodulator for functional voiding dysfunction) in an animal model and to determine their efficacy, as up to 80% of patients do not respond to the PNE test.

MATERIALS AND METHODS

PNE foramen electrodes were placed in the S3 of 12 anaesthetized Göttingen minipigs. First, detrusor instabilities were induced by the intravesical instillation of formalin. A 10-min stimulation phase with both a quasi-trapezoidal (QT) signal and a rectangular signal followed. An interval of 30 min elapsed between the series of stimulations. The attained bladder pressure values were registered on a urodynamic unit and evaluated as contractions and amplitudes per minute. Six minipigs were treated in the same way but were not stimulated and served as a control group.

RESULTS

After formalin instillation, the mean (sd) number of involuntary detrusor contractions was 3.5 (0.8)/min and the sum of amplitudes 7.2 (1.1) cmH2O/min. Subsequent NaCl instillation and QT-stimulation reduced the contractions to 0.3 (0.3)/min and the sum of amplitudes to 0.8 (0.4) cmH2O/min. Stimulation with a rectangular signal, as used in the PNE test, followed after an interval of 10 min, giving 1.1 (0.1) contractions/min and a sum of amplitudes of 5.1 (2.4) cmH2O/min. Within the control group there was no significant reduction.

CONCLUSIONS

These results show that QT-stimulation suppresses uncontrollable detrusor contractions in the minipig more effectively than the conventional rectangular stimulation presently applied in sacral neuromodulation.

INTRODUCTION

Sacral neuromodulation for treating functional voiding dysfunction has become established in urology. Success rates of 20–80% were achieved with this method, particularly in refractory patients with detrusor hyper-reflexia, detrusor hypercontractility and chronic pelvic pain [1–5]. The final success can be assessed by a series of diagnostic tests carried out before implanting sacral neuromodulators. During the peripheral nerve evaluation (PNE) test, wire electrodes are percutaneously guided through a foramen cannulation of S3 and connected to the sacral roots of S3. A neuromodulator for chronic sacral neuromodulation is then implanted if there is at least a 50% improvement in symptoms [6].

The outcome of the PNE test is successful in only ≈ 20% of the patients tested, with the result that only a few patients can benefit from the implantation of a sacral neuromodulator. It has not been possible to improve this unsatisfactory outcome significantly, either by newly developed electrodes, which do not easily dislodge and have better nerve-electrode contact, or by using a bilateral PNE test [7–13].

Little is known about the exact mechanism of neuromodulation. Many stimulation parameters, e.g. of stimulation amplitude and frequency, have been investigated to improve its efficacy. As different nerve fibres react differently to different stimulation signal shapes and pulse widths [14–16], the aim of the present study was to investigate experimentally whether the signal type has the potential to improve sacral dorsal root neuromodulation to suppress bladder instability.

MATERIALS AND METHODS

After orotracheal intubation in 12 anaesthetized Göttingen minipigs (12–18 months old, body weight 25–33 kg), the electrocardiogram, pulse and breathing were monitored. A transurethral 8 F catheter was placed for bladder diversion. The skin was incised over the os sacrum and the dorsal openings of the foramen of S3 exposed (Fig. 1). A commonly available PNE wire electrode (no. 3057; cable width 0.65 mm, length of stimulation tip 7 mm, Medtronic Europe SA, Tolochenaz, Switzerland) was placed on both sides. The stimulation unit consisted of a 3245 A Universal Source generator (Hewlett-Packard, USA) connected to a personal computer using appropriate hardware for stimulation (Laboratory-Windows/CVI, programming language C, by the Fraunhofer Institute, St. Ingbert, Germany).

Figure 1.

Incision and placement of the electrodes into the sacral foramen

A rectal balloon was inserted into the rectum to measure pressure; the transurethral catheter and rectal catheter were connected to a urodynamic unit (Duet® MultiP, Dantec, Denmark) and before treatment an initial value was recorded for 10 min, having filled the bladder with 150 mL NaCl beforehand. Immediately afterwards the bladder was drained and 150 mL of 0.25% formalin solution instilled into the bladder for 10 min. The acute inflammation induced by this resulted in detrusor instability, which was recorded urodynamically [17,18].

In six animals, the bladder was then drained and, after a 10-min interval, refilled with 150 mL NaCl solution. The bladder pressures were then recorded again to evaluate whether formalin-induced instability occurred. Afterwards, the sacral nerves of S3 were stimulated with a quasi-trapezoidal (QT) ‘2/40’ signal (impulse width 1000 µs, 15 Hz, 2.0 V) for 10 min.

This procedure was then repeated but stimulating using a biphasic rectangular signal, normally used in neuromodulation (impulse width 210 µs, 15 Hz, 2 V; Fig. 2). This trial was conducted twice on each animal; in three animals the stimulation signal used first was QT, as described previously, but to eliminate systematic errors the rectangular signals were used first in the other three trials. All variables were measured twice to gain more data.

Figure 2.

Stimulation signals (biphasic rectangular signal and quasi-trapezoidal signal),both completely load compensated

The remaining six animals served as a control group. The protocol was the same but these animals were not stimulated, to gain reference data. All other conditions, e.g. measurement and recording times, were identical. The results were assessed statistically using the Wilcoxon signed-rank test at a significance level of P = 0.05 (Tamara Lenz, Institute of Medical Statistics, Biomathematics and Information Processing, University of Heidelberg).

The two variables evaluated for bladder instability with time were detrusor contractions per minute and the sum of amplitudes per minute. A detrusor contraction was defined as a bladder pressure increment from baseline. The sum of amplitudes per minute was defined as the sum of all bladder pressure amplitudes (in cmH2O) from baseline that occurred within 1 min. Both detrusor contractions and the sum of amplitudes were measured during 10-min recordings, from which could be inferred the value per minute.

RESULTS

There were no spontaneous detrusor contractions during the initial 10-min measurements under NaCl instillation (Fig. 3a and Table 1; the response to formalin and the values are shown in Table 1 and Fig. 3b). Under stimulation with the QT 2/40 signal the number of contraction and sum of amplitudes was reduced (Table 1 and Fig. 3c). At the end of this stimulation phase, there was an increase in the number of contractions and sum of amplitudes. In the control group the mean number of contractions and sum of amplitudes remained unchanged during the same investigation phase.

Figure 3.

Examples of urodynamic bladder pressure responses; a, NaCl (control); b, after instillation with formalin; c, during QT 2/40 stimulation; and d, during rectangular wave stimulation.

Table 1.  The number of contractions and sum of amplitudes per minute during the empty-state measurement (NaCl), after instillation with formalin (F), during stimulation with a rectangular (REC) and a QT 2/40 signal, during regeneration intervals, and in the control group
Sample*NaClFormalinQTAfter QTF1RECAfter RECF2
  • *

    Nos. 1–6, no. of animal; a or b, measurement series. Significant reduction (P < 0.05) in comparison with

  • *

    * F, F1 and F2, or

  • † F, F1 and F2 and REC.

Contractions/min
1a03.10.20.82.81.01.63.2
1b03.40.01.93.51.12.43.3
2a03.20.21.33.10.92.02.5
2b04.10.31.63.21.22.45.7
3a03.10.60.02.41.32.73.1
3b03.20.01.23.91.12.83.7
4a06.41.00.74.51.02.94.8
4b02.50.00.44.70.81.53.4
5a03.30.31.34.21.22.03.9
5b02.90.11.92.61.11.84.5
6a04.70.11.33.51.02.03.7
6b03.80.21.42.81.11.73.8
Mean (sd)0.03.6 (1.0)0.3 (0.3)1.2 (0.6)3.4 (0.8)1.1 (0.1)*2.2 (0.5)3.8 (0.9)
Sum of amplitudes/min
1a07.60.32.95.62.87.06.4
1b06.81.06.86.02.45.75.9
2a05.70.43.95.51.96.46.6
2b07.10.64.67.93.04.67.3
3a07.20.92.07.12.95.46.5
3b06.91.03.16.21.83.65.9
4a09.61.23.08.12.26.08.2
4b05.01.12.07.23.04.07.6
5a07.80.53.96.51.95.46.8
5b07.50.63.67.12.03.67.9
6a05.80.63.97.32.15.47.4
6b09.10.94.86.42.03.68.1
Mean (sd)0.07.2 (1.3)0.8 (0.3)3.7 (1.3)6.7 (0.8)2.3 (0.5)*5.1 (1.2)7.1 (0.8)

When the sacral nerves of S3 were stimulated with a biphasic rectangular signal the number of contractions and sum of amplitudes declined (Table 1); subsequent urodynamic evaluation with no stimulation showed an increase in both. Again, both were unchanged in the control group (Fig. 3d).

The examples of urodynamic bladder pressure recordings in Fig. 3 in one animal in the study group show that while no contractions occurred after NaCl instillation (Fig. 3a), there was clear and regular instability after instillation with formalin, which were significantly reduced by stimulation with the QT signal and somewhat less with the rectangular signal.

DISCUSSION

From clinical experience it is possible to treat non-paraplegic, anticholinergic-resistant detrusor hyperactivity by sacral neuromodulation [1–5,19]. Patients suitable for chronic neuromodulation treatment can be identified from the history, a voiding diary, quality-of-life questionnaire, pain questionnaire, video-urodynamics, and acute or subacute PNE tests. Despite these extensive diagnostics, in many patients (50–80%) the treatment fails for unknown reasons. A first attempt to reduce the number of those not responding to the PNE test was to introduce new wire electrodes which provided better nerve-electrode contact and were supposed to dislodge less easily. Although these new electrodes remained in place the success rate was no better [13]; the results of bilateral stimulation were also unsatisfactory [12].

One result of the present assessments in Göttingen minipigs was that detrusor instability can be induced by instilling formalin [17]. In addition, and similar to the hyperactivity observed in clinical routine practice, the instability was inhibited by low electrical stimulation. We chose a stimulation signal (QT 2/40) that had already been tested in our laboratory for the selective nerve stimulation of myelinated fibres [20]. In accordance with our results and those already published, the impulse shape and length have a decisive influence on the depolarization of fibres [14,15,21,22]. Neuromodulation has some effects on A-δ fibres, i.e. thin, myelinated nerve fibres [23], but short impulses (up to 250 µs), as used in current neuromodulation, primarily excite thicker nerve fibres [14–16].

Induced detrusor instability was inhibited by QT stimulation (Table 1), during which the number of contractions per minute was reduced by 92% and the sum of amplitudes per minute by 96% after instilling formalin and inducing detrusor instability. The results from the control group indicated that this inhibition was not caused by exhaustion of the detrusor but showed that there was a genuine suppression of the unstable detrusor.

The conventional rectangular stimulation signal, currently used in the PNE test also caused a decline of 42% in the number of contractions and of 15% in the sum of amplitudes. This decline was significantly less than that with QT stimulation. One reason for this might be that the relevant fibres for neuromodulation could have been stimulated by the QT signal more effectively.

As most stimulation parameters have been tested, and differences in hardware do not seem to affect the final outcome, it might be logical to use different stimulation signals [14–16]. One hypothesis as to why the QT signal may have given better results is that certain impulses can cause selective stimulation [21]; this is speculation and must be confirmed in additional trials by measuring the action-potential volleys of different nerve fibres, but the present results indicate that there are still options for a more optimal stimulation, which seem to activate the nerve fibres relevant for the PNE test.

In conclusion, results from Göttingen minipigs show that it is possible to inhibit formalin-induced detrusor instability by stimulating the posterior roots of S3. The QT signal gave significantly more successful inhibition of detrusor instability than achieved by the normal stimulation signal of the PNE test. That these trials were conducted in pigs (minipigs), whose neurophysiological similarities to the human lower urinary tract have been noted elsewhere [24], provides hope that these QT signals may also help to reduce bladder instability in humans and might be a first step towards reducing the number of those not responding to sacral neuromodulation.

Abbreviations
PNE

peripheral nerve test

QT

quasi-trapezoidal (signal).

Ancillary