The diaphragmatic pacemaker (DP) has proven its utility in replacing mechanical ventilation (MV) in patients with chronic spinal cord injury (SCI) and Amyotrophic Lateral Sclerosis (ALS), by improving the patients quality of life and reducing morbi-mortality and the associated health care costs. The anesthetic management of these patients and the particularities of the surgical procedure represent an anesthetic challenge. The objective of our study is to analyze the management and the intraoperative complications in the patients with DP in our institution.

Patients included are part of a program developed by the Spinal Cord Injury Unit of the Institut Guttmann of placement and strengthening of the diaphragm with NeuRx® Diaphragm Pacing Stimulation (DPS) System (Synapse Biomedical, Oberlin, OH, USA) device for patients suffering from neuromuscular disorders or upper spinal cord injuries dependent on MV. All of them were preselected by a multidisciplinary expert committee after the assessment of their clinical history and the evaluation of the phrenic nerve function using two complementary techniques like phrenic nerve stimulation and fluoroscopic evaluation of diaphragm movement. Pediatric patients were the exception, since they underwent surgery in order to obtain a definitive diagnosis on the phrenic nerve functionality. Patients were admitted in the center 24 hours prior to the intervention.

The surgical procedure of DP implantation consists in placing four intramuscular electrodes, two in each hemidiaphragm, using a conventional abdominal laparoscopy with carbon dioxide insufflation at 10 L/minute speed and pressure up to 15 mmHg. Four ports are inserted: one for the optical equipment, two for the mapping electrode and electrode insertion instruments and a smaller sized port as an exit site for the wires of the electrodes. Reverse Trendelenburg position is required for the procedure. It consists in locating the optimum point for electrode insertion. The process involves mapping between 30 and 50 different points in each hemidiaphragm by applying an electric stimulus of 2-24 mA at 100 µsec pulse widths11. It results in both qualitative and quantitative assessment of diaphragmatic movement: qualitative through laparoscopy and quantitative using the external assessment of the intraabdominal pressure during stimulation with temporal mapping electrode. Site of main electrode is identified as the location of each hemidiaphragm's change of maximum pressure and site of secondary electrode as replica of main site. Once the sites are identified in each hemidiaphragm, intramuscular electrode placement phase is initiated. Response to desired stimulation is checked subsequently. Finally the electrodes are tunneled out to the corresponding percutaneous exit site and an electrocardiogram strip is recorded with all electrodes active in order to confirm there is no capture of cardiac rhythm. In the case of pediatric age patients equipment and incisions were adapted to patient's weight and size. However the surgical procedure was the same.

Upon arrival to the operating room, heart rate (HR), non-invasive arterial pressure (NIAP), pulse oximetry oxygen saturation (SaO2) and capnography (Carescape Monitor B850, GE Healthcare, Finland) were standardly monitored. In patients with previous history of ischemic cardiopathology and/or difficult to manage autonomic dysreflexia crisis, invasive arterial pressure was monitored through radial artery catheterization using the Seldinger technique (Leathercath Arterial; Vygon Ecoven, France). In most cases no premedication was provided. In pediatric patients prophylactic atropine was required for induction, it was given in doses of 0.01 mg/kg. Induction was carried out either intravenously with propofol in doses raging 1.5-2 mg/kg in adult patients and 3 mg/kg in pediatric patients or by inhalation with sevoflurane in pediatric patients. After a prior priming of the circuit, a series of three forced inspirations were carried out using reservoir bag at concentrations of 6% of the anesthetic until reaching a minimal alveolar concentration (MAC) of 2.0-3.5%. Anesthetic maintenance was carried out following anesthetist's criterion, with sevoflurane of 2-2.5% MAC or continuous intravenous infusion with propofol for Total Intravenous Anesthesia (TIVA) maintaining the infusion at 10 mg/kg/h for 30 minutes, followed by 8 mg/kg/h for another 30 minutes and 6 mg/kg/h until the end of the intervention, maintaining bispectral index (BIS) value of 40-60. Patients were disconnected from their usual ventilation system and transferred to the GE Datex-Ohmeda Aespire 3000 (GE Healthcare, Finland) system on volume control mode, with tidal volume of 6 mL/kg, respiratory rate between 10 and 15 breaths per minute and PEEP +7 cmH₂O with a mixture of O2/air at 50% in order to reach oxygen saturation superior to 95% before initiating the surgical procedure. Given the type of procedure, it was attempted to minimize the use of muscle relaxants, since these are not required in patients with tracheostomy. Rocuronium at 0.4 mg/kg at the anesthetist's criterion was used for managing the airway in patients who needed orotracheal intubation (OTI). In these cases, neuromuscular relaxation was monitored (NMT Neuromuscular Transmission MechanoSensor GE Healthcare, Finland) and a Train of Four (TOF) measurement of 100% was obtained prior to administering muscle relaxants. OTI was performed when TOF answer was 0 and diaphragmatic mapping phase initiated solely if TOF answer was 4 and with a percentage superior to 90%. When these levels of response were not achieved, sugammadex 2 mg/kg was used to reverse the blocking in case of induction with rocuronium or neostigmine 0.04-0.07 mg/kg.

Intraoperative analgesia was carried out using continuous infusion of remifentanil (0.5-1 µg/kg/min) until the end of the intervention. Surgeons infiltrated laparoscopic ports with bupivacaine 0.5% with vasoconstrictor before incision. Patients requiring OTI received a fentanyl bolus (3 µg/kg) in the induction and in all cases at the end of the surgery a fentanyl bolus (1.5 µg/kg) was administered in order to avoid postoperative hyperalgesia associated to remifentanil.

Isolated boluses of propofol (0.5-1 mg/kg) or fentanyl (1-1.5 µg/kg) were administered when patients required anesthetic deepening for a correct tolerance of the laparoscopic technique without using muscle relaxants. Patients received bolus of 5 mg ephedrine repeatedly until normal values were reached if intraoperative arterial hypotension appeared, defined as a 20% decrease of baseline or mean arterial pressure (MAP) inferior to 60 mmHg.

Patients with dysreflexive events manifesting as hypertensive crisis with pressure 20% higher than baseline associated to bradycardia were treated using anesthetic deepening and/or hypotensive drugs (bolus of urapidil 25 mg every 10 minutes) until reaching a correct pressure control.

At the end of the intervention all patients were connected to their usual MV, either invasive or not, with supplemental oxygen. They also received intravenous analgesic treatment in the ward with paracetamol, nonsteroidal anti-inflammatory drugs (NSAIDs) dexketoprofen or metamizole magnesium and tramadol as rescue medication during 24-48 hours.

In the immediate postoperative period, following protocols a control X-ray was carried out in order to rule out pneumothorax. If present, it was drained percutaneously. Patients remained in post-anesthesia care unit (PACU) for at least two hours. They were transferred to ward while maintaining continuous monitoring of ECG, NIAP and SaO₂ during 24 h once they were aware and oriented in the three spheres and the following criteria were met: analgesia is optimized visual analogue scale (VAS) inferior to 2), dysreflexive events and postoperative spasms ruled out, arterial pressure is normalized, heart rate to baseline values and oxygen saturation higher than 95%. Stimulation of diaphragm was initiated in the ward during the first 3-4 days following surgery in SCI and after one week in ALS depending on the medical team in charge of the stimulation process. Correct conditioning of the diaphragm and weaning from MV required repeated sessions. Surgical team carried out a 15 days post-intervention follow-up and given that it was a neuroprosthesis, they also did follow-ups one month and one year after the surgery. Once patients were released, the team in charge of neurostimulation carried out postoperative follow-ups at three, six and twelve months. Subsequently they conducted annual controls.

The information compiled included demographic data, ASA classification, diagnostic that led to device placement, chronically used MV type and tracheostomy use.

Type of anesthesia administered was also indicated, including specific accessory monitoring, drugs used, airway management, surgery duration and placement or not of permanent electrodes.

All incidents or complications occurred during the procedure were also recorded, whether they were related to anesthetic or surgical technique. During intraoperative period there were recorded any arrhythmias related to electric stimulation during mapping phase, disadaptation to MV or poor tolerance to the pneumoperitoneum (peak pressure > 35 cmH2O and plateau pressure > 30 cmH2O, desaturation < 90%) and presence of autonomic dysreflexia crisis related to painful stimulus (blood pressure > 20% of baseline or systolic blood pressure > 155 mmHg and diastolic blood pressure DBP > 100 mmHg). As for surgical complications in the immediate postoperative period, pneumothorax appearance was given special attention.

Inclusion Criteria:

• diaphragmatic pacer placement surgery

Exclusion Criteria:

• no access to patients history

all patients conducted to diaphragmatic pacer placement from December 2007 to July 2015 at our institution