Hemodynamically significant cardiac arrhythmias during general anesthesia for spine surgery: A case series and literature review.

BACKGROUND CONTEXT: Hemodynamically significant bradycardia and cardiac arrest (CA) are rare under general anesthesia (GA) for spine surgery. Although patient risks are well defined, emerging data implicate surgical, anesthetic and neurologic factors which should be considered in the immediate management and decision to continue or terminate surgery.
PURPOSE: To characterize causes and contributors to significant arrhythmias during spine surgery. We also provide an updated literature review to inform spine care teams and aid in the management of intraoperative bradycardia and CA. STUDY
DESIGN: Case series and literature review. PATIENT SAMPLE: Six patients who underwent spine surgery from 03/2016 to 01/2020 at a single institution and developed unexpected hemodynamically significant arrhythmia. OUTCOME MEASURES: Our primary outcome was to identify potential risk factors of interest for significant arrhythmia during spine surgery.
METHODS: Medical records of patients who underwent spine surgery from 03/2016 to 01/2020 at a single institution and developed unexpected hemodynamically significant arrhythmia during spine surgery were identified from a departmental Quality Assurance Database. We evaluated the presence/absence of patient, surgical, anesthetic and neurologic risk factors and estimated the most likely etiology of the event, immediate and subsequent management, whether surgery was postponed or continued and outcomes.
RESULTS: We found a temporal relationship of bradyarrhythmia and CA after somatosensory evoked potential (SSEP) stimulation in 4/6 cases and pharmacy/polypharmacy in 2/6. Surgery was completed in 4/6 patients, and terminated in 2/6 (subsequently completed in both). We found no adverse outcomes in any patients. Our literature review predominately identified case reports for guidance to support decision making. New literaure suggests peripheral nerve blocks and opioid-sparing anesthetic agents should also be considered.
CONCLUSIONS: Significant bradycardia and CA during spine surgery does not always require termination of the surgical procedure. Decision making should be undertaken in each case individually, with an updated awareness of potential causes. The study also suggests the need for large prospective studies to adequately assess incidence, risk factors and outcomes.

Background

Hemodynamically significant bradycardia and cardiac arrest are rare under general anesthesia for spine surgery [1], [2], [3]. Depending on the most likely etiology, surgery may be terminated to allow investigation, or continued after resolution of the event. Unexpected intraoperative arrhythmia may signal underlying coronary artery disease, valvulopathy, or conduction abnormality, and necessitate termination of surgery for diagnosis and intervention. Alternatively, arrhythmia may represent recognized side effects of sedative, hypnotic and analgesic agents used as part of a balanced anesthetic. These agents may be adjusted, and the surgery can proceed without further perioperative investigation.

Unfortunately, the etiology of intraoperative arrhythmia is likely to be unclear and multifactorial. Further, continuing or terminating surgery requires consensus of the intraoperative care team, in the absence of robust evidence to guide decision making. Indeed, we could identify no prospective studies reporting the incidence of significant intraoperative arrhythmias, immediate management strategies, and subsequent outcomes.

Given these gaps, we retrospectively reviewed a series of 6 patients to explore potential contributory factors to unexpected transient intraoperative bradycardia or asystole during spine surgery. We also reviewed the immediate management, subsequent decision making, and patient outcomes. This case series suggests the need for larger, prospective studies to adequately estimate the incidence, causes and consequences of significant arrhythmias during spine surgery. These data are needed to aid risk stratification and improve decision making for spine care teams.

Methods

This is a retrospective case series of 6 patients who underwent spine surgery by four surgeons at an orthopedic specialty hospital in New York City, NY. Cases were identified by searching the Department of Anesthesiology Quality Assurance database for reported instances of bradycardia or asystole of any duration during spine surgery between 03/2016 and 01/2020 at a single institution. Data were extracted by manual search of the electronic medical record. We defined the potential risk factors of interest for arrhythmia according to: patient factors [(history of cardiac, cardiovascular or neurological conditions, medications, age, ASA classification and body mass index (BMI)], surgical factors (diagnosis, type of surgery, time of day, time from incision to event, phase of surgery, any intraoperative neurophysiological monitoring (IONM) at the time of the event; either somatosensory evoked potentials, SSEPs, or motor evoked potentials, MEPs) and anesthetic factors (agents used, duration of nil per mouth (NPO) status and patient position). We also characterized details of the arrhythmia (bradycardia and/or asystole, and duration), subsequent management (any pharmacologic intervention, decision to proceed with or terminate surgery, any postoperative investigations) and the likely etiology. Results are presented descriptively.

Case reports

Summary details for the 6 patients are reported in Table 1. Patients ranged in age from 22 to 73 years. Four patients were male and 2 were female. Both female patients were overweight (Case 4: BMI 32.1 kg/m2; Case 5: BMI 31.3 kg/m2). Five patients were classified as ASA 1 or 2 and one patient as ASA 3 (Case 5). Five patients had no cardiovascular risk factors. One patient (Case 5) had a history of paroxysmal atrial fibrillation and hypertension, reported as well controlled on atenolol, amlodipine and flecanide. All 6 patients underwent routine preanesthetic medical evaluation and risk assessment. Four patients had normal preoperative electrocardiograph (ECG). One patient (Case 2) met voltage criteria for left ventricular hypertrophy (LVH). One patient (Case 3) had an ECG significant for sinus bradycardia, LVH and nonspecific T-wave abnormalities; a pre-operative echocardiogram showed no abnormalities. A standard anesthesia technique was followed for each case, as per our institutional protocol [4].

Table 1

Overview of patients’ demographics and contributory factors which lead to unexpected intraoperative bradycardia or asystole during spine surgery.

Variables	Case 1	Case 2	Case 3	Case 4	Case 5	Case 6
Age	26	22	57	72	73	61
Sex	Male	Male	Male	Female	Female	Male
Race	White	White	White	White	White	White
Hemoglobin (g/dL)	16	15.3	14.9	13.6	10.4 (chronic)	14.5
Cardiovascular risk factors	None	None	Sinus bradycardia noted on pre-operative ECG	None	Paroxysmal ArrhythmiaHypertension	None
Other relevant comorbidities	None	None	None	Chronic Hepatitis C	Bronchiectasis with mild pulmonary hypertension	Obstructive sleep apnea syndrome
Medication related to co-morbidities	None	None	AcetaminophenEnoxaparinHydromorphoneMethocarbamol	AcetaminophenAspirinCyclobenzaprineOxycodoneZolpidem	AtenololFlecanideAmlodipineTriamtereneClonazepamParoxetine	None
Hemodynamic relevant medication at the day of surgery	None	None	None	None	None	None
ASA	1	1	2	2	3	2
BMI	27.8	24.5	28.6	32.1	31.3	22
Indication	Lumbar disk herniation	Degenerative spondylolisthesis	Lumbar scoliosisSpinal stenosis	Degenerative disk diseaseSpinal stenosisLumbar scoliosis	Spinal stenosis	Degeneration of cervical intervertebral disk C4–6Myelopathy
Intervention	Microscopic discectomy right L5/S1	LLIF L4/5	OLIF L4-S1, LLIF L3–5	LLIF L4/5, PLIF L4/5	LLIF L2/3Extension Fusion L2-Pelvis (prior fusion L3-Pelvis)	ACDF C4–6
Time of day of surgery	Morning	Morning	Noon	Morning	Afternoon	Morning
Time to event from anesthetic Induction end (min)	8	44	102	87	18	36
Time to event from incision (min)	NA	NA	48	33	NA	NA
Phase of surgery	Drying of skin preparation	Time out	During exposure	During exposure	While positioning from Supine to lateral left	During exposure
Patient position	Prone	Supine	Lateral left	Prone	Lateral left decubitus	Supine
Use of IOMN (SSEP)	Yes	Yes	Yes	Yes	Yes	Yes
Time to event from IOMN stimulation (min)	1	1	11	1	0	15
IOMN Note	Tibial stimulation	Tibial and ulnar stimulation	Tibial and ulnar stimulation	Tibial stimulation	Tibial stimulation	Tibial and ulnar stimulation
Inhaled anesthetics	Isoflurane	Isoflurane	Isoflurane	Isoflurane	Sevoflurane	Isoflurane
Iv anesthetics	PropofolMidazolam	PropofolMidazolamKetamine	PropofolMidazolamKetamine	PropofolMidazolamKetamine	PropofolMidazolam	PropofolMidazolam
Opioids	FentanylHydromorphone	Fentanyl	FentanylMethadone	FentanylMethadone	Fentanyl	Fentanyl
Other anesthetic agents	CefazolinVecuroniumDexamethasone	CefazolinVecuroniumDexamethasone	CefazolinFamotidineVecuroniumDexamethasoneDexmedetomidine	CefazolinFamotidineVecuroniumDexamethasoneOndansetronLidocaineTranexamic AcidDexmedetomidine	VecuroniumLidocaine	VecuroniumSuccinylcholine
Use of peripheral nerve block	No	Yes	Yes	No	No	No
Time peripheral nerve block to event (min)	NA	36	92	NA	NA	NA
Duration of NPO status (h)	8.23	9.03	13.21	10.16	13.37	9.31
Iv fluid administered (ml)	500	1500	2000	1000	2500	1200
Estimated blood loss (ml)	20	50	600	100	200	No
Course of event	Continued	Continued	Continued	Aborted	Continued	Aborted
Reason	Recovered	Recovered	Recovered	Cardiac arrest	Recovered	Severe refractory hypotension
Duration event (sec)	<45	15	NA	45	<60	360
Cardiac rhythm	Asystole	Asystole	Bradycardia	Asystole	Asystole	Asystole followed by bradycardia and hypotension
Management	CPR. IONM technician alerted, and SSEPs stopped, followed by ROSC.	CPR. IONM technician alerted, and SSEPs stopped, followed by ROSC. Fluid	Iv fluids with multiple doses of ephedrine (10 mg x2) and glycopyrrolate (0.4 mg x1) with restoration of baseline heart rate and MAP. Dexmedetomidine infusion was stopped. Admitted to the ICU.Ephedrine	CPR, followed by ROSC, supine repositioning, PACU, cardiology consult	Supine repositioning; iv fluids and epinephrine. IONM technician alerted, and SSEPs stopped, followed by ROSC. Return to supine, returned	iv fluids, glycopyrrolate and ephedrine were administered. The patient's HR recovered to baseline, but the MAP was persistently low (52–60).
Postoperative cardiology investigation & evaluation	NA	Postoperative ECG: no LVH by voltage; heart murmur;outpatient cardiac MRI recommended	ECG without findings	ECG: Sinus rhythm, left axis deviation, RVCDTTE: moderate diastolic dysfunction	Bedside TTE: grossly normal RV and LV, dilated RA and LA, neutral volume status	Bedside TTE: grossly normal RV and LV; hyperdynamic appearance c/w hypovolemia; troponins: negative x3
Possible Cause	Vaso-vagal due to SSEP stimulation	Vaso-vagal due to SSEP stimulation	Dexmedetomidine infusion (underlying Sinus bradycardia)	Methadone/Dexmedetomidine induced heart block	Vaso-vagal due to SSEP stimulation	Vaso-vagal due to SSEP stimulation

LLIF Lateral Lumbar Interbody Fusion; OLIF Oblique Lateral Interbody Fusion; PLIF Posterior Lumbar interbody Fusion; ACDF Anterior Cervical Discectomy and Fusion; IOMN Intraoperative Neurophysiological Monitoring; SSEP Somatosensory Evoked Potential; NPO Nil Per Os; CPR Cardiopulmonary resuscitation; ROSC Return Of Spontaneous circulation; MAP Mean Arterial Pressure; ICU Intensive Care Unit; PACU Post Anesthesia Care Unit; HR Heart Rate; ECG Electrocardiogram; TTE Transthoracic Echocardiogram; RVCD right ventricular conduction delay.

Case 1: A 26-year old man underwent L5/S1 microdisectomy, indicated for herniated nucleus pulposus (HNP).

Event: asystole, lasting 45 s.

Conditions: Pre-incision, with the patient in the prone position. Anesthetics included inhaled isoflurane, propofol and ketamine infusions. IOMN technician performed tibial SSEPs immediately prior to the event.

Management: CPR. IONM technician alerted, and SSEPs stopped, followed by ROSC.

Decision making: The surgery was completed successfully and IONM was suspended for the duration of the procedure. No further arrhythmias noted.

Attribution: SSEP stimulation.

Case 2: A 22-year old male underwent lateral lumbar interbody fusion (LLIF) with posterior instrumentation at L4-L5 indicated for spondylolisthesis.

Event: transient bradycardia (to 10–15 bpm lasting 10 s) followed by asystole (lasting 10–15 s).

Conditions: Pre-incision, with the patient in the left lateral position, during surgical time out. Anesthetics included inhaled isoflurane, propofol, ketamine and dexmedetomidine infusions. IONM technician performed tibial and ulnar SSEPs immediately prior to the event.

Management: CPR. IONM technician alerted, and SSEPs stopped, followed by ROSC.

Decision making; The surgery was completed successfully and IONM was suspended for the duration of the procedures. No further arrhythmias noted. Given pre-operative findings on the ECG, a postoperative echocardiogram was performed and the patient was reviewed by cardiology. Finding were notable for LVH. The patient was referred for neurological review as an outpatient and diagnosed with postural orthostatic tachycardia syndrome (POTS). Attribution: SSEP stimulation, exacerbated by hypovolemia secondary to prolonged NPO status with underlying POTS.

Case 3: A 57-year old man underwent interbody fusion and posterior instrumentation at L4-S1 indicated for degenerative lumbar scoliosis with residual spinal stenosis.

Event: bradycardia to 11 bpm with hypotension (MAPs 40 s) lasting 25 s.

Conditions: During surgical exposure, with the patient in prone position. Anesthetics included inhaled isoflurane, propofol, ketamine and dexmedetomidine infusions, with opioid analgesics within the prior hour (hydromorphone, 2 mg, iv). IOMN technician performed tibial and ulnar SSEPs immediately prior to the event.

Management: Iv fluids were administered with multiple doses of ephedrine (10 mg x2) and glycopyrrolate (0.4 mg x1) with restoration of baseline heart rate and MAP. Dexmedetomidine infusion was stopped. The patient was admitted to the ICU for overnight telemetry.

Decision making: The surgery was completed without further episodes of bradycardia. A postoperative echocardiogram, ECG and monitoring were uneventful.

Attribution: Dexmedetomidine infusion, in a patient with underlying sinus bradycardia.

Case 4: A 72-year old female underwent a LLIF with posterior instrumentation at L4-L5 indicated for degenerative disk disease.

Event: Multiple transient episodes of second-degree heart block, Mobitz Type II followed by complete heart block and asystole, lasting 45 s.

Conditions: During surgical exposure, with the patient in prone position. Anesthetics included inhaled isoflurane, propofol, ketamine and dexmedetomidine infusions. Methadone (10 mg, iv) was administered within the previous hour. IOMN technician performed tibial SSEPs immediately prior to the event.

Management: CPR, followed by ROSC, supine repositioning.

Decision making: The procedure was terminated, and the patient was transferred to the ICU for monitoring and investigation. Evaluations for cardiac, anaphylactic, and thromboembolic etiologies were unremarkable. The patient underwent the planned surgery three days later without arrhythmia or complication. Methadone and dexmedetomidine were not administered.

Attribution: Methadone/dexmedetomidine-induced heart block.

Case 5: A 73-year old woman underwent a LLIF at L2-L3 with extension posterior fusion from L3-Pelvis, indicated for spinal stenosis.

Event: Asystole lasting 20 s. Preceded by 3 episodes of transient bradycardia, poorly responsive to ephedrine and glycopyrrolate.

Conditions: The bradycardic episodes occurred in the supine position while the IONM technician was titrating stimulus parameters to obtain baseline SSEPs. Anesthetic agents included inhaled isoflurane, propofol and ketamine infusions. Opioids were administered in the previous 30 min (fentanyl 100 µg iv). Asystole occurred prior to incision, during X-ray imaging with the patient in lateral position. The IONM technician performed tibial SSEPs immediately prior to the event.

Management: Supine repositioning; iv fluids and epinephrine were administered. The IONM technician was alerted and SSEP monitoring was terminated, followed by ROSC. A bedside transthoracic echocardiogram (TTE) was unremarkable with neutral volume status.

Decision making: The surgery was completed successfully and IONM was suspended for the duration of the procedures. No further arrhythmias noted.

Attribution: SSEP stimulation.

Case 6: A 61-year old male underwent anterior cervical decompression and fusion (ACDF) at C4-C6 for intervertebral disk degeneration.

Event: Asystole lasting 2 s, followed by bradycardia (15–20 bpm) lasting 80 s.

Conditions: During exposure with the patient in supine position. Anesthetics included inhaled isoflurane, propofol and ketamine infusions. The IONM technician performed tibial and ulnar SSEPs immediately prior to the event.

Management: iv fluids, glycopyrrolate and ephedrine were administered. The patient's HR recovered to baseline, but the MAP was persistently low (52–60). A bedside TTE was consistent with hypovolemia.

Decision making: The surgery was terminated, and the patient was transferred to the ICU for monitoring and further evaluation. Cardiology review and investigations were unremarkable. The planned surgery was completed the following day without arrhythmia or complications.

Attribution: SSEP stimulation, exacerbated by hypovolemia secondary to prolonged NPO status.

Discussion

In this retrospective case series, we found no adverse outcomes following hemodynamically significant arrhythmias in 6 patients undergoing spine surgery. Additionally, we attributed the most likely etiology of arrhythmia to SSEP monitoring (in 4/6 cases), or to pharmacy (in 2/6 cases). Each of these represent immediately reversible or modifiable causes of arrhythmia. Thus, it may not always be necessary to terminate spine surgery for investigation of these unexpected intraoperative events. These cases also highlight the importance of recognizing unique risks in spine surgery patients, and how emerging anesthetic, neurologic and surgical techniques may interact and contribute to the development of arrhythmias.

Risk factors for cardiac arrest during spine surgery have been well defined, including lumbar fusion, age over 65 years, obesity, cardiovascular disease, ethnicity and ASA status [1,2,8]. Bradycardia and asystole have been described under general anesthesia in combined surgical cohorts: Proposed mechanisms include unopposed parasympathetic activation, enhanced vasovagal response to decreased venous return and psychiatric stressors [5].

In contrast, few studies report significant bradycardia or transient asystole during spine surgery in otherwise healthy patients. Where described, the etiology of arrhythmia typically reflects venous thromboembolic events, preexisting cardiac abnormalities, anaphylactic shock or hypovolemia [6], [7], [8], [9]. The absence of structural heart disease or other defined risk factors associated with arrhythmias in our case series suggest that other mechanisms should be considered.

Previous case reports suggested bradycardia during spine surgery is caused via afferent parasympathetic stimulation during dural traction or electrocautery in the lumbosacral region [10], [11], [12], [13], [14], [15]. A lumbar-cardiac reflex has been proposed, in which parasympathetic stimulation leads to increased vagal tone, and consequent bradycardia and hypotension [16,17]. Typically, these reflex-mediated arrhythmias are terminated when direct or indirect manipulation of the spinal dura is discontinued [17]. In contrast, based on timing and phase of surgery, we did not find any arrhythmias attributable to a lumbar-cardiac reflex.

In our case series the only consistent contributor in all six patients was the SSEP stimulation immediately preceding the episode of bradycardia/asystole (Table 1). We therefore hypothesize SSEP stimulation may trigger a vasovagal reaction, similar to the proposed mechanism for dural traction-induced arrhythmia. Indeed, case reports describe bradycardia or/and asystole which normalize with cessation of MEPs or SSEPs [18,19], MEPs have also been implicated in the conversion from hemiblock to complete heart block in a case report [20], similar to our observation in Case 4. The mechanism by which SSEP stimulation causes arrhythmia may be via afferents from peripheral nerves which trigger a vagally mediated response which in turn depresses the sinoatrial node. It is unclear why a subset of patients experiences cardiac arrhythmia during IOMN stimulation, however an underlying predisposition or combination of factors is likely. This is highlighted by our analyses of Cases 2 and 6, in which hypovolemia was suspected as contributory. Alternatively, anatomic differences, such as cervical stenosis may render some patients more susceptible to the effects of IONM stimulation than others [21].

Patient positioning during spine surgery should also be considered in relation to unexpected arrhythmia. Prone positioning is associated with several physiologic changes, including decreased cardiac output, inferior vena cava compression, reduced venous return, and redistribution of pulmonary blood flow [6,22]. Compression of the lower extremity veins additionally raises the risk of intraoperative venous thromboembolism as a cause of arrhythmia [7]. In combination with intraoperative blood loss and hypovolemia, these position-related factors elevate the risk of cardiac arrhythmias with hemodynamic instability [5,23]. Although our analysis included just 6 cases, we did not find any consistent relationship between positioning, large blood loss and the development of arrhythmias.

The benefits of contemporary anesthetic and analgesic techniques must also be weighed against the risk for bradycardia and asystole. Dexmedetomidine is a highly selective α2 adreno-receptor agonist with sedative, anxiolytic, sympatholytic, and analgesic effects [24,25]. Given these advantages, dexmedetomidine is increasingly included in anesthetic regimens and enhanced recovery pathways for spine surgery [4]. However, a predictable side effect of dexmedetomidine is hemodynamically significant bradycardia, with potential to progress to asystole, as we observed in Case 3. This risk may be further elevated when dexmedetomidine is added to agents which prolong the QT interval. A major emphasis of pain management in spine surgery cohorts is to provide opioid-sparing, long lasting analgesia, and to prevent the conversion of acute to chronic pain. Methadone has recently been demonstrated to be of significant value in this regard [26]. However, methadone prolongs the QT interval, and has been associated with major cardiac arrhythmias and torsade de points when administered during spine surgery [27]. Consistent with these effects, we speculate that in Case 4, methadone and dexmedetomidine acted synergistically to exacerbate an underlying predisposition to bradycardia and increased vagal tone, culminating in asystole.

In two of our patients, a transversus abdominis plane (TAP) block was performed pre-operatively as an analgesic adjunct. Peripheral nerve blocks (PNBs) are increasingly applied to spine surgery as a method to provide opioid-sparing analgesia [28,29]. Although PNBs are relatively simple and safe to perform [30,31], local anesthetic toxicity syndrome classically manifests with sudden cardiac arrhythmias and hypotension, followed by cardiovascular collapse [32].

Our study suffers from the inherent limitations of a retrospective review. We chose a Quality Assurance Database to identify cases, which relies on voluntary reporting for inclusion. This likely underestimated the true incidence of arrhythmias in our spine surgery population. Data was extracted from the medical record, which assumes accurate entry, although intraoperative hemodynamic data is automatically imported into the records. Finally, our case series is small and derived from a specialty orthopedic surgery hospital, limiting generalizability.

Conclusion

Here we report 6 cases of significant bradycardia and/or asystole during spine surgery. Although multiple factors have been implicated, the cases highlight SSEP stimulation as a common etiology of arrhythmia. Prospective research is required to understand the temporal relationship and interactions between IONM modalities and arrhythmias. Allied to this, studies exploring risk mitigating strategies for IONM should be performed. For example, test stimulation before incision may help identify at-risk patients, as was suggested by Case 4 [19]. As anesthetic options evolve, and combinations of agents are used together to achieve analgesic goals, prospective trials will be required to understand the risks and benefits unique to spine surgery. Finally, our study suggests the true incidence of hemodynamically significant arrhythmias may be higher than has been previously reported. This question can only be answered by well-designed prospective study.

In the meantime, the decision to continue or terminate surgery should be decided based on patient condition and circumstances of each event. Surgeons and anesthesiologists should be aware of and prepared to treat significant cardiac arrhythmias during spine surgery even in otherwise healthy patients without known risk factors.

Declaration of Competing interest

The authors have no conflict of interest directly relevant to this work.