Perioperative Real-Time Near-Infrared Spectroscopy Monitoring of the Spinal Cord Collateral Network: Comparison of Two Clinically Available Systems

Objectives: Efficacy and feasibility of near-infrared spectroscopy monitoring of the paraspinal collateral network (cnNIRS) has been experimentally proven and clinically suggested to provide additional safety during and after thoracoabdominal aortic aneurysm repair. However, better understanding of available cnNIRS monitoring systems, concerning reaction time, sensitivity and specificity, would be advantageous. The aim of the study was to compare two different cnNIRS monitoring devices in three different clinical settings simulated in an established acute large animal model.

Methods: Two different NIRS systems (Foresight, Edwards and INVOS5100C, Medtronic) used at 4 paravertebral levels (mid and lower thoracic, upper and lower lumbar) were compared in an acute juvenile pig model (n = 19, 40–55 kg, all female). Three different experimental setups all leading to a reduction of spinal cord perfusion were used: segmental artery (SA) clipping (n = 5), SA coil-embolization (n = 5) and cross-clamping of thoracic aorta (n = 9; [Table 1]).

Table 1 Performed interventions in three experimental groups

SA clipping	SA coiling	Aortic X-clamp
Left upper + lower thoracotomy	Left upper thoracotomy	Left upper thoracotomy
Retroperitoneal exposure	7-F sheet in femoral artery
Exposure of SA	Identification of SA	Exposure of thoracic aorta
Sequential SA clipping (T4–L5)	Sequential SA coiling (T4–L5)	X-clamp on thoracic aorta

Results: At lower thoracic level statistically significant differences between monitoring systems were observed during SA clipping (p = 0.042) and aortic crossclamping (p = 0.006). No differences between devices were observed at midthoracic, upper, and lower lumbar levels. Considering longer duration of SA occlusion using coil-embolization technique compared to SA clipping (p < 0.001 on all levels) an additional comparison of cnNIRS measurements within each device (coiling vs. clipping) was performed: no difference was observed except in INVOS performance at upper lumbar level (p = 0.002).

Conclusion: Both devices provide adequate cnNIRS monitoring with similar performance at mid-thoracic, upper and lower lumbar levels. The occlusion method (100% instant with clipping vs. subtotal with coil embolization) may be equally reflected by both NIRS monitoring systems. INVOS seems to have a shorter reaction time and a more pronounced response to sequential occlusion of segmental arteries and aortic X-clamping at the lower thoracic level.