Between 2000 and 2009, the presence of a segmental reflex (R1 response), polysynaptic reflex (R2 response) and direct motor response (M-wave) on EMG of the external anal sphincter upon stimulation of the third sacral root were sequentially discovered [1]. These signals were linked to the presence of the motor response seen with the naked eye upon stimulation of the lead. It was postulated that the contraction of the pelvic floor musculature (PFM) represents an indirect motor response (R1 and R2 response) mediated by afferent input to the spinal cord.

This study assesses the relation between the visual observed contraction of the PFM and all different EMG signal components of the PFM upon stimulation of the third sacral root while comparing the use of intravaginal EMG to needle EMG.

This is a single tertiary center, prospective study conducted from 28/12/2017 to 15/01/2019 including patients with overactive bladder and non-obstructive urinary retention eligible for sacral neuromodulation (SNM). 

The lead was placed according to the standardized tined lead placement technique [2]. An intravaginal probe (MAPLe ®) [3] was placed and fixated to the perineum and monopolar needles (37mm x 27G, NATUS ®) were inserted through the external anal sphincter, at the left side at 2 and 4 o’clock and at the right side at 8 and 10 o’clock. The needles were placed approximately 3 cm deep, thereby recording EMG activity at the level of the puborectalis muscle. Each lead electrode (0-1-2-3) was stimulated with square wave pulses (pulse width: 210 µs, frequency: 5 Hz), with increasing stimulation amplitude (1-2-3-5-7-10 mA). 
During the procedure, the perineum was recorded with a camera and EMG signals were continuously recorded using a specialized device (NIM-Eclipse ®). The implanter was blinded for the EMG signals. Afterwards, the EMG signals were analyzed and the stimulation amplitude upon which a visual contraction and EMG motor response could be seen were assessed and noted, irrespectively of each other as the investigator was blinded for the patient ID. For further analyses, the intravaginal and needle EMG trace eliciting the largest EMG signals were taken. Of these traces, a distinction was made between the different EMG components (M wave, R1 response and R2 response) of the EMG signals recorded. The latency of the different EMG components at the EMG motor response threshold were noted. An example trace of intravaginal and needle EMG with the 3 EMG signal component can be found in figure 1. 

The association between the threshold noted by visual observation and the EMG motor response threshold of each signal component was assessed using the weighted kappa. Sensitivity of the PFM EMG in comparison to visual observation was assessed using the McNemar-Bower test. EMG response latency measured by intravaginal and needle EMG was compared using the paired t-test.

Fourteen  patients were included of which 11 were measured during the tined lead implantation, 5 of them were also measured during implantation of the definitive battery and 3 were measured during battery revision. In all patients the motor response threshold was determined for each lead electrode (tined lead has 4 electrodes), therefore 76 motor response thresholds are included. 

A visual motor response could be determined in 96.1% (73/76) of all cases of which 76.3% (58/76) were recorded upon current stimulation of 1 or 2 mA. 
An intravaginal EMG motor response for the 3 EMG signal components (M wave, R1 response and R2 response) was recorded in 97.4% (74/76), 42.1% (32/76) and 11.8% (9/76) of all cases respectively. Of these, respectively 80.2% (61/78), 3.9% (3/76) and none (0/76) were recorded upon current stimulation of 1 or 2 mA. 
A needle EMG motor response for the 3 EMG signal components (M wave, R1 response and R2 response) was recorded in 97.4% (74/76), 56.6% (43/76) and 15.8% (12/76) of all cases respectively. Of these, respectively 80.2% (61/78), 15.8% (12/76) and none (0/76) were recorded upon current stimulation of 1 or 2 mA.

Comparison between observation of a motor response by the naked eye and the current stimulation needed to elicit all different EMG signal components shows an almost perfect association between the visual response threshold and M wave motor response threshold for both the intravaginal and needle EMG (κ = 0.86 and κ = 0.85, respectively). However, no association between the visual motor response threshold and the R1 motor response threshold (κ = 0.004 and κ = 0.077, respectively) as well as the R2 motor response threshold (κ = -0.078 and κ = 0.015, respectively) could be withheld for intravaginal or needle EMG. The comparison between the visual motor response threshold and the EMG motor response threshold of the different EMG components for intravaginal and needle EMG can be found in figure 2.

The latency of the M wave differs significantly (p < 0.0001) between intravaginal EMG (3.35 +/- 0.75 ms) and needle EMG (3.57 +/- 0.72 ms). No significant difference (p > 0.05) is withheld between intravaginal and needle EMG for the R1 response (28,13 +/- 4.53 ms vs 28.17 +/- 4.71 ms, respectively) and R2 response (127.33 +/- 7.97 ms vs 123.22 +/- 7.17 ms, respectively).

Firstly, this study shows the motor response which is observed during stimulation of the third sacral root is a direct efferent motor response. These results refute the earlier observations, that the PFM motor response seen represents an indirect motor response, and therefore can completely change the idea that we have about the mechanism of action of SNM.

Secondly, this study shows the motor response can be reliably measured through the use of intravaginal and needle EMG.

Thirdly, additional information is gathered by the use of EMG (R1 and R2 response in addition to the M wave) and assessment of this addition information is possible.

Finally , the signals measured by intravaginal EMG through an intravaginal probe and needle EMG (inserted into the puborectalis muscle) seem to record similar muscle groups (as only minor differences in latency are seen).

This study shows that the PFM contractions as seen by the naked eye upon stimulation of the third sacral root is, in contrast to the current beliefs, a direct efferent motor response and can be reliably measured in patients under anesthesia through the use of intravaginal and needle EMG. In addition, indirect motor responses (R1 and R2 response), mediated by afferent input to the spinal cord, can be measured through intravaginal and needle EMG as well, although the added clinical value of this has to be further investigated.

Malaguti S. Interventional neurophysiology and an implantable system for neurostimulation of the sacral area. Functional neurology. 2009;24(4):207-219.
Matzel KE, Chartier-Kastler E, Knowles CH, et al. Sacral Neuromodulation: Standardized Electrode Placement Technique. Neuromodulation. 2017
Voorham-van der Zalm PJ, Voorham JC, van den Bos TW, et al. Reliability and differentiation of pelvic floor muscle electromyography measurements in healthy volunteers using a new device: the Multiple Array Probe Leiden (MAPLe). Neurourology and urodynamics. 2013;32(4):341-348.