Background & Aims

In humans, the onset of spinal cord stimulation (SCS)-induced sensation is linked to the neural recruitment of dorsal column fibres, which is represented by the evoked compound action potential (ECAP)1. Hence, the ECAP threshold (ECAPT) is utilised in the clinic to determine the optimal SCS dose2. However, the precise identification of the sensation threshold in animals is usually based on a motor response elicited in the animal at a certain stimulation intensity, referred to as the motor threshold (MT)3. The implementation of ECAPs in preclinical SCS models may allow for the substitution of MT with an objective method to identify animals’ sensation thresholds, leading to the improvement of translatability of the model. Thus, here we aimed to build on our previous work4,5 and present an automated method to improve the accuracy and reliability of ECAP analyses and determination of ECAPT in rats subjected to a model of neuropathic pain.

Methods

Adult male Sprague-Dawley rats (240-300 g; n=15) were subjected to the spared nerve injury (SNI) model. A custom-made six-contact lead was implanted epidurally, spanning T11-L3 as confirmed by X-ray. Stimulation and recordings were performed in anaesthetised and freely behaving rats using a custom designed multi-channel system4,5. The stimulation current was gradually increased in a stepwise manner from 0.0 mA (50 Hz 200 µs; 50 recordings/1 µA step size) until MT was detected. Manual and automated (MCS Studio; Saluda Medical) processing of recordings was conducted. This included the collection of input-output (IO) functions to identify the relationship between stimulation intensity and dorsal column activation4,5, followed by the estimation of the ECAPT. Manual generation of the IO functions included interpolation of the raw data points with a linear regression or an assumption-free spline curve (smoothing parameter=0.95) using Matlab (2023 release; Mathworks, Inc).

Results

Manual ECAPTs were determined by (1) offline visual inspection of the first observable ECAP (2) linear extrapolation of IO functions to the y-intercept5; (3) zero-crossing estimation of the fitted spline curve. Linear regression led to negative ECAPTs in the majority of the recordings under anaesthesia, potentially because the ECAP amplitude reached a plateau resulting in non-linear curves4. Although spline fitting is data-driven and assumption-free, it can give rise to highly non-linear curves leading to potentially unstable estimation of ECAPT, which was also observed in our data. As expected, the mean ECAPT estimated by visual inspection was higher compared to the mean ECAPTs obtained using the other two methods in both anaesthetised and freely behaving animals. Interestingly, mean ECAPTs were higher in the freely behaving condition compared to the anaesthetised condition for all three methods. Finally, manual ECAPT estimates were compared to ECAPTs estimated by the automated tool.

Conclusions

The reliability of the manually estimated ECAPT, using the different manual methods introduced in this study, is dependent on several factors, such as the signal-to-noise ratio of the recording as well as the linearity of the IO function data. The automated tool may potentially improve replicability and translatability of our findings, and overcome the challenges observed with the manual analysis of ECAP recordings. Moreover, the difference observed in the current required to generate an ECAP between anaesthetised and freely behaving rats demonstrates the importance of accounting for multiple factors including variations in dorsal column activation that occur with postural and physiological changes in preclinical models. This however requires further investigations.

References

1.Gmel GE, Santos Escapa R, Parker JL, Mugan D, Al-Kaisy A, Palmisani S. The Effect of Spinal Cord Stimulation Frequency on the Neural Response and Perceived Sensation in Patients With Chronic Pain. Front Neurosci. 2021;15:625835.
2.Parker J, Karantonis D, Single P. Hypothesis for the mechanism of action of ECAP-controlled closed-loop systems for spinal cord stimulation. Healthc Technol Lett. 2020 Jun;7(3):76-80.
3.Smits H, van Kleef M, Holsheimer J, Joosten EA. Experimental spinal cord stimulation and neuropathic pain: mechanism of action, technical aspects, and effectiveness. Pain Pract. 2013 Feb;13(2):154-68.
4.Dietz BE, Mugan D, Vuong QC, Obara I. Electrically Evoked Compound Action Potentials in Spinal Cord Stimulation: Implications for Preclinical Research Models. Neuromodulation. 2022 Jan;25(1):64-74.
5.Versantvoort EM, Dietz BE, Mugan D, Vuong QC, Luli S, Obara I. Evoked compound action potential (ECAP)-controlled closed-loop spinal cord stimulation in an experimental model of neuropathic pain in rats. Bioelectron Med. 2024 Jan 10;10(1):2.

Presenting Author

Ilona Obara

Poster Authors

Eline Versantvoort, MSc

Bsc, Msc

Newcastle University

Lead Author

Daniel Parker

MSc

Saluda Medical

Lead Author

Zubin Nanavati

MSc

Saluda Medical

Lead Author

Darayus Nanavati

MSc

Saluda Medical

Lead Author

Birte Dietz

PhD

Saluda Medical

Lead Author

Peter Single

MSc

Saluda Medical

Lead Author

Dave Mugan

MBA

Saluda Medical

Lead Author

Quoc Vuong

PhD

Newcastle University

Lead Author

Ilona Obara

PhD

Newcastle University

Lead Author

Topics

  • Models: Chronic Pain - Neuropathic