Complications of cerebrospinal fluid drainage in thoracoabdominal aortic procedures

Abstract

Background:

Cerebrospinal fluid (CSF) drainage is used to reduce spinal cord ischemia (SCI) in patients undergoing thoracoabdominal aortic procedures. Recent literature has found high rates of complication associated with CSF drainage, which has led to changes in practice. The aim of this study was to investigate rates of CSF drain–related complications in patients undergoing a thoracoabdominal aortic procedure with perioperative placement of a CSF drain.

Methods:

We conducted a single-centre retrospective cohort study. We defined major complications as intracranial hemorrhage, epidural hematoma or abscess, meningitis, and catheter retention requiring a reoperation. Minor complications assessed included drain-induced neurologic deficits, CSF leak, postdural puncture headache, asymptomatic blood in the CSF, drain failure, and catheter retention not requiring a reoperation. We recorded postoperative neurologic deficits as secondary outcomes.

Results:

There were 129 patients who met the inclusion criteria. We found 5 cases of permanent paraplegia in the overall cohort (3.9%), with only 2 occurring in the patients with prophylactic CSF drains (1.6%). There were no major CSF drain–related complications. The rate of minor complications was 17.8%. We found no association between complication rates and indication for procedure or type of operation.

Conclusion:

The lack of major complications in this series adds to existing variability in recent literature and provides support for continued use of this adjunct for SCI prevention. Further research is required to identify the etiology of significant differences in CSF drain complication rates seen at other centres.

Spinal cord ischemia (SCI) is a devastating complication of thoracoabdominal aortic procedures and increases morbidity and mortality in this population. ^1–4 The pathophysiology of SCI is linked to decreased spinal perfusion pressure secondary to decreased distal aortic pressure and flow to the anterior spinal arteries, and increased cerebrospinal fluid (CSF) pressure during aortic cross-clamp.^5–7 Previous studies have found the rate of permanent paraplegia after open thoracoabdominal aortic aneurysm (TAAA) repair to be as high as 16%.¹ Despite the lack of cross-clamping in thoracic endovascular aortic repair (TEVAR), the risk of SCI remains, with rates of postoperative neurologic dysfunction up to 31%^8,9 and permanent paraplegia up to 5%.^4,10–12

Given the substantial impact of SCI on patient morbidity and a mortality rate 10 times higher in patients with postoperative neurologic deficit,³ multiple adjuncts to increase intraoperative spinal perfusion have been developed, including left heart bypass,^13,14 hypothermic circulatory arrest,¹⁵ intraoperative somatosensory evoked potentials,¹⁶ and CSF drainage.³ In a randomized controlled trial of patients undergoing open repair of extent I and II TAAAs, Coselli and colleagues³ showed an 80% relative risk reduction of SCI with the use of prophylactic CSF drainage, with a postoperative rate of neurologic deficit of 2.6%, compared with 13.0% in the control group.³ This benefit was confirmed in a recent systematic review and meta-analysis showing CSF drainage as an effective strategy for prevention of SCI in TAAA repair.¹⁷

As with any procedure, CSF drainage is associated with potential complications, the most life-threatening being intracranial hemorrhage (ICH).^18–22 Whereas previous literature has found postoperative complications of CSF drainage to be as low as 1.5% after open TAAA repair, patients who experienced ICH have a 40% mortality rate.¹⁹ These findings were echoed in a recent review of a single centre’s TEVAR experience.²⁰ There has been a renewed interest in CSF drain complications owing to the higher complication rates reported in 3 recent studies.^23–25 In a retrospective review of patients undergoing fenestrated endovascular aortic repair (FEVAR) with greater than 4 cm of supraceliac coverage, a 45% CSF drain complication rate was observed, with the most common being a nonfunctioning drain requiring removal.²³

More worrisome were the results of 2 prospective cohort studies, which found high rates of severe CSF drain complications.^24,25 When looking at patients with extent I–III TAAAs, previous TEVAR, or previous thoracic aortic surgery undergoing FEVAR, Kitpanit and colleagues reported a 7.6% major complication rate. This included ICH, meningitis, neurologic deficits, and CSF drain fractures requiring reintervention.²⁴ The same study group experienced a 5.1% rate of SCI with 2 occurrences of paraplegia and 2 occurrences of paraparesis. This led the authors to conclude that the risks of CSF drainage may outweigh the benefits.²⁴ Similarly, in a prospective cohort of patients undergoing TEVAR or FEVAR with more than 5 cm of supraceliac coverage, Kärkkäinen and colleagues reported a 9% rate of moderate or severe CSF drain complications, including a 3% rate of ICH and 3% rate of spinal hematoma causing paraplegia in 1%.²⁵ The authors of this study have since changed their practice patterns and now reserve prophylactic CSF drains for patients with extent I and II TAAAs only.²⁵

The above research, showing high rates of CSF drain complications, prompted us to examine our centre’s rate of CSF drain–associated complications during repair of thoracic aortic pathology.

Methods

Study design

We performed a single-centre retrospective cohort study to investigate the outcomes of patients with placement of a CSF drain for spinal cord protection during the perioperative period of thoracic or thoracoabdominal aortic procedures. We screened the records of every patient who underwent 1 of the above surgeries between 2007 and 2021 for CSF drain usage, using operative dictations, intensive care unit (ICU) transfer notes, and discharge summaries.

We included patients in the study if they were taken to the operating room for a thoracic or thoracoabdominal procedure involving their aorta and had a CSF drain placed in the perioperative period. We excluded patients if they had preoperative paraplegia or substantial neurologic deficits, or if they underwent their procedure for aortic rupture. Additionally, patients who did not have sufficient data records for analysis of CSF drain complications were excluded from the study. Placement of a CSF drain was at the discretion of the attending surgeon. Our centre places a CSF drain in the following circumstances: greater than 15 cm of thoracic aorta coverage, coverage of the aorta between spinal levels T8 and T12, planned coverage of the left subclavian artery without concomitant carotid to subclavian bypass, substantial disease in spinal cord collateral vessels, or previous abdominal aortic surgery.

Spinal drain management

Preoperative notification to our centre’s ICU is completed 1–2 days before planned CSF drain insertion; both the ICU physician and nurse clinician are notified. This allows the unit to ensure availability of a postoperative bed as well as to schedule bedside staff with appropriate training to care for patients with a CSF drain.

The anesthesiology team preoperatively determines suitability of neuraxial intervention after review of coagulation parameters, presence of contraindications, and appropriate perioperative management of anticoagulant and antiplatelet cessation, in concordance with American Society of Regional Anesthesia guidelines.²⁶ These surgeries are all performed with anesthesiologists who have additional training in transesophageal echocardiography, which is a small and highly specialized group at our centre with an interest in caring for vascular surgery patients. Further information regarding CSF drain insertion technique can be found in Appendix 1, available at www.canjsurg.ca/lookup/doi/10.1503/cjs.003624/tab-related-content.

Intraoperatively, the drain is set at the level of the right atrium and with CSF drainage pressure of 10 mm Hg. Drainage is maintained to approximately 10 mL/h, although this may be modified based on intraoperative neuromonitoring findings to a maximal rate of 15 mL/h. Heparin anticoagulation may be instituted 1 hour after insertion, providing no frank blood was encountered during CSF drain placement.

Postoperatively, patients are transferred to a closed medical and surgical ICU for continued management. A notification of CSF drain is placed visibly at the bedside. Neurologic status is assessed in the immediate postoperative period after sedation and hourly by bedside nursing staff who have completed additional internal training for management of patients with CSF drains. Any changes to neurologic status prompt immediate notification of the on-call ICU physician, the in-house anesthesiologist, and the primary vascular surgeon of record. Additional calls are made to the on-call vascular surgery resident or fellow and the on-call vascular surgeon, if they are not the patient’s primary surgeon, to make them aware of the situation.

In patients who are awake and neurologically intact, the CSF drain is open at a pressure of 10 mm Hg with a maximum drainage of 10 mL/h. After this volume of drainage within an hour, the drain is clamped and reopened again the subsequent hour. Standard hemodynamic parameters are systolic blood pressure (SBP) of more than 130 mm Hg and mean arterial pressure (MAP) greater than 70. After a total of 24 hours, if no neurologic deficits are seen, the drain is clamped but remains inserted for a subsequent 24 hours. If at 48 hours, no neurologic deficits are seen, the drain is removed following cessation of anticoagulation in accordance with American Society of Regional Anesthesia guidelines. After drain removal, patients are kept supine for 6 hours of bed rest.

In patients who are neurologically compromised, the CSF drain is lowered to a pressure of 5 mm Hg with a maximum drainage of 15 mL/h. Hemodynamic targets are increased with an SBP goal of more than 140 and MAP greater than 80. The CSF drain is kept in situ, draining for 72 hours before clamping and removal following the protocol above; however, drain duration can vary on a case-by-case basis for patients who are neurocompromised.

Outcomes of interest

We collected basic demographics. We stratified indications for surgery into acute aortic syndrome (AAS), chronic dissection, and aneurysmal disease. Acute aortic syndrome included intramural hematoma, penetrating aortic ulcer, and acute aortic dissection. We used chronic dissection to describe patients with previous aortic dissection who had experienced aneurysmal dilatation, whereas aneurysmal disease encompassed all primary aortic aneurysms. When applicable, we recorded Crawford aneurysm extent and maximal axial aneurysm diameter. We categorized procedures as open, endovascular, and hybrid operations. Hybrid operations were defined as procedures with both an open and endovascular component; for example, a TEVAR with a carotid to subclavian bypass.

We grouped patients who received their spinal drain before the start of the operation into the preoperative group, and those who had drains inserted in the postoperative period in a postoperative group. A select few patients received a spinal drain both before and after the operation; we described them separately. We recorded the duration of drain insertion when it was present in the records. Outcomes of interest were CSF drain–related complications within 30 days of the procedure. Major complications we recorded included epidural hematoma, epidural abscess, ICH, meningitis, and catheter retention requiring a reoperation. Minor complications included drain-induced neurologic deficits, CSF leak, postdural puncture headache (PDPH), catheter retention not requiring a reoperation, asymptomatic blood in the CSF, and failure of the drain. Secondary outcomes recorded were postoperative neurologic deficits including permanent paraplegia and transient weakness or paresthesia.

Statistical analysis

We consulted a medical statistician for assistance in data analysis. Clinical and demographic information were expressed as mean ± standard deviation (SD) for continuous variables and percentages for discrete variables. We made comparisons between continuous variables with an unpaired t test, when necessary. We assessed whether surgical complications were independent of both indication for surgery and type of surgical procedure. We arranged these data into 2 contingency tables and tested them for independence with complications via Pearson χ² test, or Fisher exact test for small cell sizes. We plotted Pearson residuals to identify the cells most contributing to the total χ² score.

Ethics approval

This study was approved by the University of Calgary Research Ethics Board (REB21–1082).

Results

Demographics

Screening of all patients who underwent a thoracic or thoracoabdominal aortic procedure identified 129 patients who received a CSF drain for spinal cord protection. The mean (± SD) age of the cohort was 65.7 ± 12.3 years, and 56.6% were male (n = 73). The most common medical comorbidities in the patient population were hypertension (n = 101), dyslipidemia (n = 60), and coronary artery disease (n = 30). Most patients had a history of smoking (74.4%). A full breakdown of comorbidities can be found in Table 1.

Table 1.

Patient characteristics

Characteristic	No. (%)* of patients in full cohort n = 129
Patient characteristics
Mean age ± SD, yr	65.7 ± 12.3
Gender
Male	73 (56.6)
Female	56 (43.4)
Mean BMI ± SD	29.1 ± 6.4
Comorbidities
Hypertension	101 (78.3)
Dyslipidemia	60 (46.5)
Coronary artery disease	30 (23.3)
COPD	29 (22.4)
Chronic kidney disease	27 (20.9)
Diabetes mellitus	20 (15.5)
Smoking history
Smokes currently	44 (34.1)
Smoked previously	52 (40.3)
Previous aortic operation	64 (49.6)
Aortic characteristics
TAAA	57 (44.2)
Extent I	25 (43.8)
Extent II	5 (8.8)
Extent III	16 (28.1)
Extent IV	11 (19.3)
Mean size ± SD, cm	6.3 ± 1.2
Acute aortic syndrome	40 (31.0)
Acute dissection	18 (45.0)
PAU	13 (32.5)
IMH	9 (22.5)
Chronic dissection	31 (24.0)
Other indication	1

BMI = body mass index; COPD = chronic obstructive pulmonary disease; IMH = intramural hematoma; PAU = penetrating aortic ulcer; SD = standard deviation; TAAA = thoracoabdominal aortic aneurysm.

^*Unless otherwise specified.

Aortic characteristics

A majority of patients in the cohort had had a previous aortic operation (n = 64; 52.9%). The most common indication for surgery was a TAAA (n = 57; 44.2%), followed by AAS (n = 40; 31.0%) and chronic dissection (n = 31; 24.0%). One procedure was done for severe stenosis of the thoracic aorta. Mean size of aneurysms in this cohort was 6.3 ± 1.2 cm with a range of 4.4–9.9 cm. Of the TAAAs, Crawford extent I was the most prevalent (n = 24; 42.1%) (Table 1).

Most of the procedures were endovascular (n = 72, 55.8%), followed by hybrid (n = 40, 31.0%), and open (n = 17, 13.2%). Thirty of the hybrid procedures involved a left carotid to subclavian artery bypass.

Outcomes

Of all patients, 122 had CSF drains placed before the start of their aortic operation and 4 had drains placed postoperatively. Three patients had both preoperative and postoperative CSF drains. Further information on the 7 patients who did not receive our centre’s standard prophylactic CSF drain protocol is provided in Table 2.

Table 2.

Nonstandard cerebrospinal fluid drain placements

Patient age (yr), sex	Surgical indication	Procedure	Etiology of drain difference
Preoperative and postoperative CSFD placement
78, F	Acute dissection	TEVAR with left carotid to SCA bypass	Drain placed preoperatively, clotted off and was removed; second drain placed on POD1 owing to lower-extremity weakness
82, F	Penetrating aortic ulcer	TEVAR with left carotid to SCA bypass and CFA endarterectomy	Initial drain removed on POD1; second drain placed on POD2 to facilitate treatment of Type 1a endoleak
90, M	IMH	TEVAR	Preoperative drain removed owing to blood-tinged CSF; drain placed POD2 for left LE weakness
Postoperative CSFD placement
54, M	Extent IV aneurysm, 5.4 cm	TEVAR with left carotid to SCA bypass	Insertion on POD0 owing to transient LE neurologic deficit, likely secondary to MS given characteristic MRI findings
54, M	Extent IV aneurysm, 5.4 cm	Open surgical repair with visceral reimplantation	Insertion on POD0 owing to fluctuating neurologic deficits, same as above
60, M	Chronic dissection causing aortic occlusion	TEVAR	Paralysis on POD1 prompted CSFD insertion and attempted left carotid to SCA bypass; no improvement in symptoms
87, M	Penetrating aortic ulcer	TEVAR	Drain placed owing to flaccid paralysis on POD0; MRI confirmed anterior cord syndrome, permanent paralysis

CFA = common femoral artery; CSF = cerebrospinal fluid; CSFD = cerebrospinal fluid drain; F = female; IMH = intramural hematoma; IV = intravenous; LE = lower extremity; M = male; MRI = magnetic resonance imaging; MS = multiple sclerosis; POD = postoperative day; SCA = subclavian artery; TEVAR = thoracic endovascular aortic repair.

We found 5 instances of permanent postoperative paraplegia, for an overall rate of 3.9%; 2 of these occurred in patients who received a CSF drain in the preoperative period, for a rate of 1.6% in patients who had a prophylactic CSF drain (Table 3). One patient had an onset of paraplegia on postoperative day 0, with the remaining 4 showing deficits on postoperative day 1. There were 9 instances of transient neurologic deficits, for a rate of 7.0%. Five of these occurred on postoperative day 1, 2 on postoperative day 0, and 1 each on postoperative days 2 and 3. The mean (± SD) duration of spinal drainage in the 14 patients with postoperative neurologic deficits was 70.3 ± 37.0 hours compared with 35.7 ± 18.3 hours in the patients with no neurologic deficits, a difference that was found to be significant (p < 0.0001).

Table 3.

Postoperative neurologic complications

Patient age (yr), sex	Procedure details	Drain insertion	Neurologic status
Permanent paraplegia
76, M	Zone 3 TEVAR for IMH, three 150 mm overlapping stents, coverage of 300 mm total	Pre-op	Paralysis on POD1; currently has minimal lower extremity motor function and both neurogenic bowel and bladder
54, M	Open extent IV repair for chronic dissection with previous hemi-arch repair and TEVAR to celiac axis; IMA reimplanted	Pre-op	Paralysis on POD1 with MRI demonstrating ischemic changes to thoracic cord; requires wheelchair assistance and catheters
60, M	Zone 2 TEVAR for chronic dissection with acute aortic occlusion, 220 mm length	POD1	Presented with LE weakness secondary to occlusion; full paralysis on POD1 with MRI finding hyperintensity in keeping with ischemia; died on POD20
87, M	Zone 3 TEVAR for PAU at site of proximal previous TEVAR, deployment of 220 mm stent	POD0	Paralysis on POD0; MRI confirming ischemic changes to thoracic cord; wheelchair support for ambulation
90, M	Zone 2 TEVAR for PAU with IMH, 150 mm and 200 mm overlapping stents; total coverage of 300 mm	Pre-op and POD1	Left LE weakness on POD1; MRI does not identify any obvious ischemia; weakness still present
Transient neurologic deficits
78, F	Zone 2 TEVAR for acute dissection; coverage of 220 mm length	Pre-op and POD1	Postoperative drain placed for LE weakness with resolution of symptoms
77, M	Zone 3 TEVAR for TAAA with previous visceral debranching; distal landing zone proximal to renal arteries	Pre-op	Episodes of LE weakness on PODs 0, 1, and 2 that were reversed by unclamping of the CSF drain
71, F	Zone 3 TEVAR for TAAA; total coverage 280 mm	Pre-op	Nonspecific numbness bilaterally with drain clamping; resolved with unclamping
56, F	Zone 2 TEVAR with left carotid-subclavian bypass for IMH; 350 mm coverage	Pre-op	Transient numbness to right lateral thigh that resolved with drain unclamping
61, M	Zone 2 TEVAR with left carotid-subclavian for chronic dissection; 330 mm aortic coverage	Pre-op	Acute hypotensive episode on POD3 secondary to sepsis with transient LE paralysis; symptoms resolved with resuscitation
72, F	Zone 3 TEVAR for acute dissection; coverage down to celiac axis	Pre-op	Left leg weakness with MRI confirmed infarct at T5 level; weakness resolved by discharge with physiotherapy
72, F	Zone 3 TEVAR for PAU, complicated by infrarenal aortic rupture requiring open repair and cross-femoral bypass; 220 mm thoracic coverage	Pre-op	Developed bilateral weakness on POD2 with confirmed T4–T10 spinal infarct; resolved by transfer from ICU
71, F	Zone 3 TEVAR down to celiac axis for chronic dissection	Pre-op	Bilateral lower extremity weakness on POD1 that resolved with unclamping of the CSF drain
60, F	Zone 2 TEVAR down to celiac with left carotid-subclavian bypass for TAAA	Pre-op	Transient LE weakness; resolved with drain unclamping

CSF = cerebrospinal fluid; F = female; ICU = intensive care unit; IMA = inferior mesenteric artery; IMH = intramural hematoma; IV = intravenous; LE = lower extremity; M = male; MRI = magnetic resonance imaging; PAU = penetrating aortic ulcer; POD = postoperative day; SCA = subclavian artery; TEVAR = thoracic endovascular aortic repair.

We found no recorded instances of major CSF drain–related complications: epidural hematoma, epidural abscess, meningitis, ICH, or reoperation for catheter-related problems in the cohort. The rate of minor complications was 17.8% (n = 23) (Table 4). The most common minor complication was asymptomatic blood in the CSF (n = 8; 6.2), followed by PDPH (n = 6; 4.7%) and failure of the CSF drain (n = 5; 3.9%). We found 2 instances each of CSF leak and drain-induced paresthesia, for a rate of 1.6% each. With regard to the drain-induced paresthesia, both patients experienced lower extremity numbness with insertion, which resolved upon immediate removal of the drain. Both underwent a second attempt at drain insertion, which was successful, with no neurologic symptoms. One of these patients was noted to have difficult anatomy for insertion, with substantial obesity. One patient with a PDPH underwent a blood patch, with resolution of symptoms. One of the instances of drain failure was an inability to insert a CSF drain after multiple preoperative attempts, and ultimately the procedure proceeded without a drain in situ.

Table 4.

Cerebrospinal fluid drain complications

Type of complication	No. (%) of patients in full cohort n = 129*	No. (%) of patients in TAAA cohort n = 57	No. (%) of patients in AAS cohort n = 40	No. (%) of patients in chronic dissection cohort n = 31
Major complications	0 (0)	0 (0)	0 (0)	0 (0)
Epidural hematoma	0 (0)	0 (0)	0 (0)	0 (0)
Epidural abscess	0 (0)	0 (0)	0 (0)	0 (0)
ICH	0 (0)	0 (0)	0 (0)	0 (0)
Meningitis	0 (0)	0 (0)	0 (0)	0 (0)
Catheter reoperation	0 (0)	0 (0)	0 (0)	0 (0)
Minor complications	23 (17.8)	9 (15.8)	8 (20.0)	6 (19.4)
Asymptomatic bleed	8 (6.2)	1 (1.8)	3 (7.5)	4 (12.9)
PDPH	6 (4.7)	3 (5.3)	2 (5.0)	1 (3.2)
Drain failure	5 (3.9)	2 (3.5)	2 (5.0)	1 (3.2)
Paresthesia	2 (1.6)	2 (1.6)	0 (0)	0 (0)
CSF leak	2 (1.6)	1 (1.8)	1 (2.5)	0 (0)
Catheter nonoperative	0 (0)	0 (0)	0 (0)	0 (0)

AAS = acute aortic syndrome; CSF = cerebrospinal fluid; ICH = intracranial hemorrhage; PDPH = postdural puncture headache; TAAA = thoracoabdominal aortic aneurysm.

^*One surgery was done for severe stenosis of the thoracic aorta (i.e., coarctation), but this patient did not experience any of the CSF drain complications so was not included as a column in this table.

Eight patients died within 30 days of their operation (6.2%), at an average of 5.75 days postoperatively. None of the perioperative deaths were related to CSF drain placement. Three were related to ischemic colitis, and the remaining 5 were attributed (1 each) to multiorgan failure, cardiac arrest of unclear etiology, intra-abdominal hemorrhage, anoxic brain injury after cardiac arrest, and intraoperative retrograde dissection with cardiac tamponade.

We undertook further analysis to determine whether there was an association between CSF drain complications and indication for surgery or type of surgical procedure. We separated indication for surgery into TAAA (n = 57), AAS (n = 40), or chronic dissection (n = 31). We separated type of surgical procedure into endovascular-only (n = 72), hybrid (n = 40), and open-only procedures (n = 17). Neither indication for surgery (χ² = 4.80; p = 0.23) nor type of procedure (χ² = 2.69; p = 0.58) was associated with CSF drain complications.

Discussion

Spinal cord ischemia is a devastating complication of thoracoabdominal aortic procedures, which increases the morbidity and mortality of this patient population. Adjunctive use of CSF drainage has been shown to decrease the incidence of SCI. Despite the benefit of CSF drains, the procedure has inherent risks, which recent studies have reported at an unexpectedly high rate. The aim of this study was to determine our centre’s rate of CSF drain–related complications, to aid in a risk–benefit analysis for prophylactic CSF drainage in thoracic and thoracoabdominal aortic procedures. In our 15-year study period, we did not experience any major CSF drain complications, and our rate of minor complications was lower than that in the previous literature. Our rate of postoperative neurologic deficits was comparable to that in previous studies.

The lack of major CSF drain complications in our cohort is in contrast to recent single-centre reviews.^24,25 Kitpanit and colleagues reported a 7.6% rate of major CSF drain complications, including a rate of 2.6% for both subarachnoid hemorrhage and spinal hematoma, and a rate of 1.3% for both cerebellar hemorrhage and catheter fracture requiring reoperation.²⁴ In this cohort of patients undergoing FEVAR or branched endovascular aneurysm repair (BEVAR) for TAAA, CSF drainage was used in patients with extent I–III aneurysms and patients with previous endovascular or open thoracic aortic surgery.²⁴ Kärkkäinen and colleagues’ trial was also endovascular only, in a similar patient population, and reported a 9% rate of moderate or severe CSF drain–related complications.²⁵ This included patients with CSF leaks requiring blood patch, as well as frankly bloody insertions that caused postponement of the planned surgery, which were not included in our study’s defined major complications. We also reported a 3% event rate of both ICH and spinal hematoma, 1 of which resulted in permanent paraplegia in the affected patient.²⁵ As stated above, our centre’s minor CSF drain–related complication rate of 17.8% was also lower than that reported in other recent literature, which ranged from 25% to 44%.^23–25

When comparing our centre’s CSF drain management to other centres with higher rates of drain–related complications, we found that techniques for insertion were similar.^23–25 We noted differences when examining postoperative CSF drain management. Both Alqaim and colleagues and Kitpanit and colleagues set their CSF pressure to 12 mm Hg and MAP goals to more than 90 mm Hg — higher than our centre’s 10 mm Hg CSF pressure — and MAP greater than 70 for patients who were neurologically intact.^23,24 Kärkkäinen and colleagues used 10 mm Hg as their set point, but relied on intermittent opening of the CSF drain for 15 minutes out of every hour to achieve drainage goals, rather than continuous drainage.²⁵ Regarding volume drained, Kitpanit and colleagues and Kärkkäinen and colleagues both allow for a maximal drainage of 20 mL/h, which is greater than our centre’s 10 mL/h for patients who were neurologically intact, and 15 mL/h for patients with neurologic deficits.^24,25 In patients with neurologic deficits, Kitpanit and colleagues’ protocol used incremental drainage of 10–20 mL of CSF until symptoms resolved, rather than setting a new CSF pressure.²⁴ Patients at our centre are on bed rest until their drains are removed, typically at 48 hours after a 24-hour clamp trial, which differs from Alqaim and colleagues, who allowed patients to be up in a chair after a 4-hour clamp trial, subsequent to 24 hours of CSF drainage.²³ A recent systematic review was unable to find an association between CSF drain–related complications and drainage rate or amount of CSF drained.²¹

Our overall rates of permanent paraplegia and transient neurologic deficit of 3.9% and 7.0%, respectively, are consistent with the results of other contemporary studies that used adjunctive CSF drainage in thoracic or thoracoabdominal aortic procedures.^9,17,21 Of note, our rate includes patients who received preoperative prophylactic CSF drains as well as those who had postoperative drains placed. Taking into account only the patients in our cohort who received prophylactic CSF drains, we had 2 occurrences of permanent paraplegia, for a rate of 1.6%. The 3 other cases of paraplegia occurred in patients who had postoperative drains placed in an attempt to reverse spinal ischemia after detection of neurologic deficits. As our study was focused on CSF drain–related complications, we were unable to determine our centre’s overall rate of postoperative neurologic deficit, as we did not capture patients treated at our centre who did not receive a spinal drain. A recent publication using Vascular Quality Initiative data of patients with SCI after TEVAR, FEVAR, or BEVAR was in support of prophylactic CSF drain placement.²⁷ When comparing outcomes of patients with SCI who received prophylactic CSF drains versus those who received therapeutic CSF drains postoperatively, the authors found a significantly higher proportion of permanent paraplegia in patients who had therapeutic drainage (79% v. 54%; p = 0.04).²⁷

In our study, 30-day mortality after thoracic or thoracoabdominal aortic procedure with perioperative use of CSF drains was 6.2% (n = 8). None of these deaths was directly related to a CSF drain complication, which is in contrast to previous literature¹⁹ and can be attributed to our lack of major complications. A systematic review of 34 studies looking at CSF drain complications in the same patient population found a CSF drain–related mortality rate of 0.9% (confidence interval 0.6–1.4).²¹ All deaths occurred after a severe CSF drain–related complication.²¹

The lack of major complications associated with CSF drains in our cohort adds to existing variability in recent literature.^9,21 Differences in SCI prevention protocols between sites likely contribute to this heterogeneity and make it difficult to generalize existing evidence to single centres. Previous studies that found high rates of major CSF drain–related complications have adopted a more conservative approach to spinal drain use, reserving it for patients at higher risk for SCI.^24,25 Part of this hesitation toward continued CSF drain use may be a sequela of the litigious nature in the United States, where both of the above-mentioned studies were performed.

In contrast, the absence of major complications and CSF drain–related mortality in this present study lends support to the aggressive SCI prevention protocol at our site. Our spinal drain protocol is strictly adhered to and is managed by a small multidisciplinary group involving anesthesiologists, intensivists, and vascular surgeons. This likely contributes to our favourable results, as previous literature has suggested that a core group of clinicians and the use of preventive protocols is associated with a reduction in SCI and CSF drain–related complications.^8,9,23 Additionally, the use of a standardized SCI prevention protocol has been shown to allow for earlier detection and treatment, which is associated with better functional outcome.⁸ A recent clinical practice guideline published by members of our group emphasized the importance of multidisciplinary and regionalized management of patients with thoracic aortic pathology.²⁸ We believe the findings of our study support this, as our low rates of CSF drain complications and postoperative neurologic deficits can in part be attributed to our high-volume vascular surgical centre, where surgeons have additional training in thoracic pathology.

Limitations

The retrospective design may have introduced bias into the results. Patients with asymptomatic blood in the CSF did not routinely receive neuroimaging at our site, and the rate of ICH may therefore have been underestimated. Another limitation is the lack of significant association between indication for surgery or procedure type with CSF drain–related complications. Although this may support the decision-making strategy and aggressive use of prophylactic CSF drainage at our site, it results in an inability to comment on specific patient or procedure groups that may benefit from changes to the current CSF drain protocol. An advantage of this study design is reproducibility; other surgical groups can repeat our methods to inform site-specific guidance of prophylactic CSF drainage protocols, thus minimizing reliance on highly variable evidence that may not be locally applicable.

Conclusion

Our centre’s experience with prophylactic CSF drainage in thoracic and TAAAs with no major CSF drain–related complications and a low rate of paraplegia provides continued support for use of this adjunct. We will continue to be liberal with our use of CSF drains in our patient population, given our low complication rate and its proven benefit in reducing SCI. Further research is required to identify the etiology of substantial differences in CSF drain complication rates when comparing between tertiary vascular surgery centres.