Elective spine surgery with continuation of clopidogrel anti-platelet therapy: Experiences from the community

Abstract

Purpose

This retrospective study aimed to assess the feasibility of continuing clopidogrel therapy during the perioperative period in elective cervical and thoracolumbar surgery.

Methods

After IRB approval, medical records of patients requiring one or two-level surgery over a two-year period (2015–2017) while receiving clopidogrel were reviewed for relevant outcomes. Over the same period, a control group of patients not receiving clopidogrel perioperatively was formed.

Result

In total, 136 patients were included: 37 clopidogrel and 99 control, with a mean age of 64.8 years. Between clopidogrel and control respectively, operative time was 86.7 min and 86.7 min (p = 0.620); blood loss was 127.0 cc and 117.5 cc (p = 0.480); drain output was 171.2 cc and 190.7 cc (p = 0.354); length of stay was 1.8 days and 1.5 days (p = 0.103). Two clopidogrel patients and 1 control patient had complications. Two clopidogrel patients and 1 control patient were readmitted within 30 days.

Conclusions

Remaining on clopidogrel therapy during elective spine surgery results in no difference in operative time, blood loss, drain output, length of stay, or readmission. Precaution should be taken in cervical procedures as the drain output in clopidogrel patients was increased and complications in this region can be severe.

1. Introduction

Clopidogrel (Plavix) is a widely used antiplatelet agent. It is considered a first line treatment for many atherothrombotic diseases such as acute ischemic stroke/transient ischemic attack (TIA), secondary prevention of recurrent stroke or TIA, unstable angina (UA) or non-ST elevation myocardial infarction (NSTEMI), ST-elevation myocardial infarction (STEMI), and post percutaneous coronary intervention (PCI) prevention of recurrent MI or stent thrombosis. According to the Medical Expenditure Panel Survey (MEPS) conducted by the Agency for Healthcare Research and Quality, clopidogrel is the most commonly purchased “cardiovascular agent” and 33rd most purchased drug with 21,349 purchases in 2016.¹

Because of its widespread use, perioperative management of clopidogrel is a topic of debate for surgeons. To reduce the risk of bleeding complications, anticoagulant medications such as clopidogrel are routinely discontinued prior to surgery. This common practice has been extrapolated from well-documented evidence in cardiac surgery literature that clopidogrel increases operative bleeding.^2,3 However, this potential risk must be weighed against the increased risk of thrombosis in the perioperative setting. Currently there are no standards of care related to the management of clopidogrel in the perioperative period and protocols for management vary. A 2007 survey among United Kingdom orthopaedic surgeons found that at least 59% would either discontinue clopidogrel and operate immediately or continue clopidogrel and operate immediately for a patient presenting with a hip fracture.⁴ A similar survey sent to U.S. orthopaedic surgery residency programs found that 73% would wait 3 days or less for urgent but not emergent cases.⁵

For elective orthopaedic cases, there seems to be a more conservative approach. A 2018 survey of 40 spine surgeons found that 80–87% (varied between anatomic regions) of surgeons always discontinue clopidogrel preoperatively.⁶ However, evidence in general surgery literature suggests that clopidogrel continuation may be safe even in elective settings.^{7, 8, 9, 10} Furthermore, a 2016 retrospective study of knee and hip arthroplasty patients found increased rates of blood transfusion in patients continuing clopidogrel but no difference in estimated blood loss, length of stay or adverse cardiac events.¹¹ Finally, a 2014 study evaluating clopidogrel use in elective spine surgery found no increase in postoperative bleeding for patients continuing clopidogrel.¹²

Given the conflicting literature on perioperative clopidogrel use and the scarcity of data in spine surgery, this retrospective study aims to assess the safety and feasibility of continuing clopidogrel therapy during the perioperative and postoperative period in elective cervical and thoracolumbar surgery.

2. Methods

To test the hypothesis that remaining on clopidogrel during elective spine surgery will not affect operative outcomes, all patients presenting to a single spine surgeon who required one or two-level operative intervention between 2015 and 2017 were identified using ICD Codes. After appropriate institutional review board approval was obtained, patient charts were retrospectively reviewed. Patients on clopidogrel therapy were included in the study group (“Group 1”). Patients who were not on clopidogrel comprised the control group (“Group 2”) and were collected consecutively over the same time period. Inclusion criteria included one or two-level surgery, prescribed clopidogrel therapy for the duration of the perioperative period, and at least 90 days of documented follow-up. Those requiring surgery of more than two levels, less than 90 days of follow-up or documented non-compliance with clopidogrel were excluded from the study. Medical records were reviewed for relevant outcomes. Outcomes included operative time, blood loss, drain output, complications, length of stay, re-admission, and mortality. Patients were stratified by anatomic region of surgical intervention. Wilcoxon rank-sum and Fisher’s exact tests were used to compare continuous and categorical variables between the groups, respectively. A P value of <0.05 was considered significant.

3. Results

In total, 136 patients were included: 37 patients receiving clopidogrel therapy (14 cervical, 23 lumbar), and 99 patients undergoing spine surgery without receiving clopidogrel (32 cervical, 67 lumbar). The mean age was 64.8 years. The patients were 54.8% male, 45.2% female. Mean ASA class was 3.1 in Group 1 and 2.7 in Group 2 (p-value <0.001).

3.1. Cervical spine operations

In the cervical operation subgroup, 64% of cases in Group 1 and 50% of cases in Group 2 were one-level operations. The specific cervical procedures in each group are shown in Table 1. The two groups had similar average operative times: 93.1 min for Group 1 and 86.5 min for Group 2. Average estimated blood loss was 85.7 cc in Group 1 and 82.8 cc in Group 2. Average drain output was higher in Group 1 (97.4 cc) versus Group 2 (43.1 cc) (p = 0.010). Mean length of stay was also higher in Group 1 (1.7 days) than in Group 2 (1.2 days), however, no other differences rose to the level of significance as detailed in Table 2.

Table 1.

Demographics of clopidogrel and control groups.

	Group 1 (clopidogrel)		Group 2 (no anticoagulation)		p-Value
N	37		99
Cervical Operations (%)	14 (37.8)		32 (32.3)		0.522
One Level	9 (24.3)		16 (16.2)
Anterior	7 (18.9)		15 (15.2)
Anterior and Posterior	1 (2.7)		0 (0.0)
Revision Anterior	1 (2.7)		1 (1.0)
Two Level	5 (13.5)		16 (16.2)
Anterior	5 (13.5)		13 (13.1)
Revision Anterior	0 (0.0)		3 (3.0)
Lumbar Operations (%)	23 (62.1)		67 (67.7)		0.877
One Level	14 (37.8)		42 (42.4)
Laminectomy	6 (16.2)		9 (9.1)
Laminectomy with discectomy	4 (10.8)		13 (13.1)
Re-exploration lami/disc	0 (0.0)		2 (2.0)
PSF with laminectomy	1 (2.7)		6 (6.1)
XLIF with laminectomy, PSF	1 (2.7)		0 (0.0)
ALIF	1 (2.7)		6 (6.1)
PLIF	1 (2.7)		0 (0.0)
XLIF	0 (0.0)		5 (5.1)
PSF	0 (0.0)		1 (1.0)
Two Level	9 (24.3)		25 (25.3)
Laminectomy	3 (8.1)		6 (6.1)
Laminectomy with Discectomy	1 (2.7)		1 (1.0)
Re-exploration lami/disc	0 (0.0)		2 (2.0)
PSF with laminectomy	4 (10.8)		10 (10.1)
ALIF	1 (2.7)		3 (3.0)
PLIF	0 (0.0)		1 (1.0)
XLIF	0 (0.0)		1 (1.0)
PSF	0 (0.0)		1 (1.0)
Mean Age (years) (SD, min-max)	63.2 (8.1, 44–85)		65.2 (7.9, 52–83)		0.218
Sex (%)	Male	Female	Male	Female
	20 (54.1)	17 (45.9)	54 (54.5)	45 (45.5)	1.000
Cervical	11 (78.6)	3 (21.4)	15 (46.9)	17 (53.1)	0.059
Lumbar	9 (39.1)	14 (60.9)	39 (58.2)	28 (41.8)	0.148
ASA Class (%)
2	3 (8.1)		30 (30.3)
3	28 (75.7)		53 (53.5)		<0.001
4	5 (13.5)		1 (1.0)

Table 2.

Comparison of outcomes between clopidogrel group and control group.

	Group 1 (clopidogrel)	Group 2 (no anticoagulation)	p-Value
N	37	99
Mean Operative Time (min) (SD, min-max)	86.7 (42.8, 39–215)	86.2 (33.2, 34–169)	0.620
Cervical Operations	93.1	86.5	0.722
Lumbar Operations	82.5	86.9	0.610
Estimated Blood Loss (cc) (SD, min-max)	127.0 (166.9, 50–800)	117.5 (119.2, 50–800)	0.480
Cervical Operations	85.7	82.8	0.176
Lumbar Operations	152.2	134.6	0.892
Total Drain Output (cc) (SD, min-max)	171.2 (260.1, 0–993)	190.7 (209.7, 5–1225)	0.354
Cervical Operations	97.4	43.1	0.010
Lumbar Operations	225.5	338.2	0.011
Length of Stay (days) (SD, min-max)	1.8 (1.0, 1–4)	1.5 (0.8, 1–5)	0.103
Cervical Operations	1.7	1.2	0.007
Lumbar Operations	1.9	1.7	0.685
Postoperative Complication within 90 days (%)	2 (5.4)	1 (1.0)	0.180
Readmission within 30 days (%)	2 (5.4)	1 (1.0)	0.180
90-day Mortality (%)	0 (0.0)	0 (0.0)	1.000

3.2. Lumbar spine operations

In lumbar cases, 61% of the cases in Group 1 and 63% of the cases in Group 2 were one-level operations. The specific lumbar procedures in each group are shown in Table 1. Operative time was similar between the two groups (82.5 min in Group 1 and 86.9 min in Group 2). Mean estimated blood loss was similar between Group 1 (152.2 cc) and Group 2 (134.6 cc). Interestingly, total drain output was significantly higher in Group 1 (338.2 cc) than in Group 2 (225.5 cc) (p = 0.011). Average length of stay was 1.9 days in Group 1 and 1.7 days in Group 2.

3.3. Postoperative complications, readmissions, mortality

Two patients in Group 1 and one patient in Group 2 experienced surgical complications (retropharyngeal edema, CSF leak, and CSF leak, respectively). Both CSF leaks were fixed intraoperatively. The patient receiving clopidogrel who experienced retropharyngeal edema was discharged on postoperative day 3 and readmitted two days later.

Two patients in Group 1 and one patient in Group 2 were re-admitted within 30 days (retro-tracheal swelling, UTI/pneumonia, and pneumonia, respectively). As mentioned above, the retropharyngeal edema patient was admitted on postoperative day 5 for symptoms of retropharyngeal edema. One patient in Group 1 was hospitalized on postoperative day 5 for concerns of an infection (urinary tract infection vs. pneumonia), and one patient in Group 2 was hospitalized on postoperative day 9 for symptoms of pneumonia. There were no deaths at 30 or 90 days. One of the cervical operations in Group 2 resulted in a CSF leak. The leak was repaired intraoperatively.

4. Discussion

Perioperative clopidogrel management is a topic with increasing relevance given the aging population and prevalence of cardiovascular disease. A study by Hawn et al. found that 23% of patients receiving PCI between 2000 and 2010 required noncardiac operations within 2 years of stent placement.¹³ Furthermore, musculoskeletal operations were the 4th most common noncardiac operation in their study. Facing this increasing population of high-risk patients, it is imperative that the effect of clopidogrel on orthopaedic surgery outcomes be thoroughly investigated.

With this aim in mind, studies in the orthopaedic surgery literature have been conflicting to date. At least two studies in the orthopaedic surgery literature have shown increased bleeding complications in patients on clopidogrel. Chechik et al. found more perioperative blood loss and longer lengths of stay in hip fracture patients taking clopidogrel.¹⁴ While, a study by Jacob et al. observed that hip fracture patients taking clopidogrel received more perioperative blood transfusions.¹¹ However, neither study found an increase in bleeding complications or mortality. Other studies investigating clopidogrel use in hip surgery have shown no difference in blood loss or transfusion rates after hip surgery.^{15, 16, 17}

Another consideration in patients receiving clopidogrel is the potential risk of perioperative thrombotic complications if antiplatelet therapy is discontinued. A study by Childers et al. showed an increase in perioperative major adverse cardiac events in patients discontinuing antiplatelet therapy for elective surgery.¹⁸ Furthermore, discontinuation of dual antiplatelet therapy was found to be a significant risk factor for “early” or “late” stent thrombosis after PCI.^{19, 20, 21} Another study by Collet et al. found that recent oral antiplatelet agent discontinuation doubled the risk of infarction and mortality within thirty days.²²

Given the evidence that premature discontinuation of clopidogrel brings risks of major cardiac complications, some surgeons opt to delay elective surgery until the duration of antiplatelet therapy is complete. However, one study evaluating the possible “rebound” phenomenon of clopidogrel discontinuation found that there was a significant increase in major cardiac events within 90 days of stopping clopidogrel regardless of how long the therapy lasted.²³ Furthermore, patients receiving clopidogrel for medical uses (not after PCI) may need antiplatelet therapy indefinitely. Thus, delaying elective spine surgery may not be feasible nor significantly safer solution in patients taking long-term clopidogrel.

The current study showed no statistical difference in operative time, blood loss, drain output, length of stay, postoperative complications, readmission or mortality between patients taking clopidogrel and a control group undergoing one-or-two level spine surgery. In fact, the only statistically significant finding between our overall groups was the average American Society of Anesthesiologists (ASA) physical status score with patients receiving clopidogrel having a significantly higher score (3.1) than those in the comparison group (2.7). This difference suggests that the clopidogrel population is “higher risk” than the comparison population based on their chronic disease history alone without taking antiplatelet use into consideration.

Other significant differences were elucidated when comparing subgroups based on cervical or lumbar region. When analyzing only cervical operations, patients on clopidogrel had higher total drain output when compared to patients not on clopidogrel (97.4 cc vs. 43.1 cc, p = 0.007). Although the two groups showed no difference overall, this increase in drain output during cervical operations is concerning. Complications in anterior cervical spine surgery tend to be more devastating than those in the lumbar spine. The risks include nerve, esophageal, and tracheal injury as well as airway compromise from hematoma formation or swelling. Furthermore, there was no statistical difference in complication rate between groups (p = 0.180), but the clinical significance of the one serious complication (retropharyngeal edema) arising from a two-level anterior cervical discectomy and fusion on a patient taking clopidogrel should not be understated. With the increase in total drain output and the serious nature of complications in the cervical region, the authors of this study suggest surgeons use caution when caring for patients on clopidogrel undergoing cervical operations. Precautions such as postoperative drains and inpatient perioperative management may still allow cervical operations on patients taking clopidogrel to be performed safely.

These findings are consistent with two other studies which showed no increase in bleeding or epidural hematoma formation in patients taking clopidogrel suggesting that discontinuation of clopidogrel or delay of surgery may not be necessary.^12,24 However, antiplatelet use during spine surgery is still a topic that has been scarcely investigated.

The retrospective nature of our study presents inherent weaknesses. For example, data accuracy is dependent upon valid chart documentation. Some variables studied (estimated blood loss) are subjective. Furthermore, although there were no statistically significant differences between groups overall, Group 1 did have higher drain output and length of stay within cervical operations. Given the small size of the study cohort and the lack of priori sample size calculation, a larger study would have greater statistical power to detect clinically meaningful differences as statistically significant.

In conclusion, remaining on Clopidogrel therapy during elective one-level or two-level cervical or lumbar surgery results in no significant difference in operative time, blood loss, drain output, length of stay, readmission rate, or 90-day mortality. Although there was no statistical difference in postoperative complication risk for patients receiving clopidogrel and undergoing cervical operations, special consideration in this population is warranted given the increased drain output and serious potential complications.

CRediT authorship contribution statement

John C. Prather: Writing - review & editing, Writing - original draft. Tyler P. Montgomery: Formal analysis. Doug Crowther: Formal analysis. Gerald McGwin: Formal analysis. Steven M. Theiss: Conceptualization.

Declaration of competing interest

None to declare.