Study design: Laboratory investigation using serial slow-®ll cystometrograms. Objectives: To examine the acute e ects of di erent modes of dorsal penile nerve stimulation on detrusor hyperre¯exia, bladder capacity and bladder compliance in spinal cord injury (SCI). Setting: Spinal Injuries Unit, Royal National Orthopaedic Hospital, Stanmore, Middlesex, UK. Methods: Fourteen SCI patients were examined. Microtip transducer catheters enabled continuous measurement of anal sphincter, urethral sphincter and intravesical pressures. Control cystometrograms were followed by stimulation of the dorsal penile nerve at 15 Hz, 200 ms pulse width and amplitude equal to twice that which produced a pudendo-anal re¯ex. Stimulation was either continuous or in bursts of one minute triggered by a rise in detrusor pressure of 10 cm water (conditional). Further control cystometrograms were then performed to examine the residual e ects of stimulation. Results: Bladder capacity increased signi®cantly during three initial control ®lls. Continuous stimulation (n=6) signi®cantly increased bladder capacity by a mean of 110% (+Standard Deviation 85%) or 173 ml (+146 ml), and bladder compliance by a mean of 53% (+31%). Conditional stimulation in a di erent group of patients (n=6) signi®cantly increased bladder capacity, by 144% (+127%) or 230 ml (+143 ml). In the conditional neuromodulation experiments, the gap between suppressed contractions fell reliably as bladder volume increased, and the time from start of stimulation to peak of intravesical pressure and 50% decline in intravesical pressure rise was 2.8 s (+0.9 s) and 7.6 s (+1.0s) respectively. The two methods of stimulation were compared in six patients; in four out of six conditional neuromodulation resulted in a higher mean bladder capacity than continuous, but the di erence was not signi®cant. Conclusions: Both conditional and continuous stimulation signi®cantly increase bladder capacity. The conditional mode is probably at least as e ective as the continuous, suggesting that it could be used in an implanted device for bladder suppression. Spinal Cord (2001) 39, 420 ± 428

Study design: Investigation of ®ve patients receiving an implant, using laboratory cystometry and self-catheterisation at home. Objectives: To use the established Finetech-Brindley sacral root stimulator to increase bladder capacity by neuromodulation, eliminating the need for posterior rhizotomy, as well as achieving bladder emptying by neurostimulation. Setting: Spinal Injuries Unit, Royal National Orthopaedic Hospital, Stanmore, Middlesex, UK. Methods: Five patients underwent implantation of a Finetech-Brindley stimulator without rhizotomy of the posterior roots. This was either a two channel extradural device (four cases) or a three channel intrathecal device (one case). In each patient, the implant was con®gured as a Sacral Posterior and Anterior Root Stimulator (SPARS). Postoperatively, repeated provocations using rapid instillation of 60 ml saline were used to determine the relative thresholds for neuromodulation using each channel. The eect of continuous neuromodulation was examined in the laboratory using slow ®ll cystometrograms, and conditional stimulation was also studied (neuromodulation for 1 min to suppress hyperre¯exic contractions as they occurred). In one patient, neuromodulation was applied continuously at home, and volumes at self catheterisation recorded in a diary. Results: Re¯ex erections were preserved in each patient. In three patients, detrusor hyperre¯exia persisted postoperatively and neuromodulation via the implant was studied. In these three patients, the con®guration was: S2 mixed roots bilaterally (channel B), and S34 bilaterally (channel A). Both channels could be used to suppress provoked hyperre¯exic contractions, with the S2 channel eective at a shorter pulse width than S34 in a majority of cases. Continuous stimulation more than doubled bladder capacity in two out of three patients during slow ®ll cystometry. Conditional stimulation was highly eective. In the one patient who used continuous stimulation at home, bladder capacity was more than doubled and the eect was comparable with anticholinergic medication. Bladder pressures 470 cm water could be achieved with intense stimulation in three patients, but detrusor-external urethral sphincter dyssynergia (DSD) prevented complete emptying. Conclusions: Neuromodulation via a SPARS was eective and may replace the need for posterior rhizotomy. However, persisting DSD may prevent complete bladder emptying and warrants further investigation.