Fluoroscopically guided lumbar spine interlaminar and transforaminal epidural injections: inadvertent intravascular injection

Abstract

Background

Inadvertent intravascular injection is a rare but catastrophic complication of lumbar epidural injections.

Purpose

To determine risk factors for inadvertent intravascular injection in fluoroscopically guided lumbar spine epidural injections.

Material and Methods

A total of 212 patients who presented for lumbar interlaminar or transforaminal injection were prospectively enrolled. Patient demographics, history of surgery, injection side, site and approach, and volume of contrast injected were recorded.

Results

There were 89 (42%) interlaminar and 123 (58%) transforaminal injections. For 36 (17%) patients, there had been surgery at the injected or adjacent lumbar level. There were 25 (12%) inadvertent intravascular injections, with an incidence of 2/93 (2%) for interlaminar and 23/119 (19%) for transforaminal injections. The patients with inadvertent intravascular injection were older (P = 0.017) and had prior surgery at or adjacent to the level of injection (P < 0.0001). Transforaminal approach had a higher intravasation rate than interlaminar injections, both when comparing the entire cohort (P = 0.0001) and only patients without prior surgery (P = 0.01). In multivariable logistic regression analysis, transforaminal injections (odds ratio [OR] 9.77, 95% confidence interval [CI] 2.14–44.6, P = 0.003) and prior surgery at or adjacent to the level of injection (OR 5.71, 95% CI 2.15–15.15, P < 0.001) were independently associated with increased risk of inadvertent intravascular injections.

Conclusion

Inadvertent intravascular injection occurred in 12% of our lumbar injection cohort and is more common with transforaminal injections, in older patients, and with prior lumbar surgery at or adjacent to the level of injection.

Keywords

Fluoroscopy epidural injection intravasation spine complication

Introduction

Low back pain and radiculopathy are common problems which contribute to approximately 40% of all absences from work (1 –4). Lumbar spine injection procedures are a routine treatment for back pain which have both diagnostic and therapeutic value (5 –7). Complications are rare, the most serious including bacterial and aseptic meningitis, arachnoiditis, and paralysis (8 –22). Intravascular injection, which occurs in 6.4%–9.2% of spine injection procedures, can potentially result in seizure, cardiac arrest, or anesthetic toxicity (23) Therefore, it is important to evaluate for inadvertent intravascular injection.

The purpose of this study is to determine the risk factors for inadvertent intravascular injection during interlaminar and transforaminal epidural injections.

Material and Methods

Participant selection

Institutional review board (IRB) approval was obtained for this study, which was compliant with the Health Insurance Portability and Accountability Act (HIPAA). A waiver of informed consent was obtained from our IRB before performing this prospective study because it was observational and did not require any deviation from routine standard of care.

Injection technique

All injections were performed with fluoroscopic guidance using a 22-gauge spinal needle by two musculoskeletal radiologists with eight and two years of experience using either paramedian interlaminar or transforaminal approach described in Palmer 2016 (24). Decision for injection approach was determined by a combination of: ordering physician request; magnetic resonance imaging findings; and patient clinical findings (midline versus lateralized pain, back pain with or without radicular symptoms). In a 3-mL syringe, 3 mL of iodinated contrast (Omnipaque, 300 mgI/mL, GE Healthcare, Piscataway, NJ, USA) was drawn up into the syringe. A minimum of 1 mL of iodinated contrast was injected under 15 frames/s live fluoroscopy through connective tubing to confirm needle position, confirm flow of contrast into the epidural or epiradicular space, and to evaluate for intravascular flow (Figs. 1 and 2). Injection of contrast was stopped upon immediate detection of intravascular flow and the volume of contrast injected to that point was recorded. Subsequently, all patients were injected with betamethasone 2 mL (6 mg/mL, American Regent Inc., Shirley, NY, USA), and 5 mL of a 2 : 8, 1 : 9, or 0 : 10 ratio mixture of lidocaine 1% (AuroMedics Pharma LLC, E. Windsor, NJ, USA), and injectable normal saline (Fresenius Kabl, Lake Zurich, IL, USA).

Fig. 1.

A 64-year-old woman presenting for L5-S1 transforaminal injection. Fluoroscopic image demonstrates inadvertent vascular injection into both central and peripheral blood vessels (dashed arrows). There has been laminectomy at the level above (solid arrows).

Fig. 2.

A 72-year-old man presenting for L2-L3 interlaminar injection. Fluoroscopic image demonstrates inadvertent injection into both anterior and posterior blood vessels (arrowheads). There has been laminectomy from L3-S1 (arrow). A total of 1 mL of contrast was injected before these vessels were visualized, compared with an average of 0.3 mL for the other inadvertent intravascular injections. The needle was repositioned multiple times, but the blood vessels could not be avoided, and the procedure was aborted.

Patient demographics, body mass index (BMI), injection side, site and approach, and volume of contrast injected when blood vessels became visible were recorded. Patient medical records and imaging studies were also reviewed for a history of surgery at the injected level or an adjacent lumbar level. An adjacent level was defined as a contiguous level with the level of injection, for example, if the injection was at L3-L4, surgeries at L2-L3, L3-L4, and L4-L5 were documented. Type of surgery performed was also recorded.

Statistical analysis

Categorical data were presented as frequencies and percentages and continuous data were presented using means and SD. Univariate analyses to determine differences between patients with versus without inadvertent intravascular injection were done using Fisher’s exact test for categorical variables and Student’s t-test for continuous variables. Multivariable logistic regression analysis was performed to determine factors independently associated with inadvertent intravascular injection. Results from logistic regression analysis were presented as adjusted odds ratios (OR) with corresponding 95% confidence intervals (CI) and Wald P values. A two-tailed alpha level of P < 0.05 was considered statistically significant. All statistical analyses were performed using JMP Pro version 14.0.0 (SAS Institute Inc., Cary, NC, USA) and Stata version 16.0 (StataCorp LLC., College Station, TX, USA).

Power analysis

For detecting a difference of 15% in the rate of inadvertent intravascular injections between patients with interlaminar injections and patients with transforaminal lumbar injections, a sample size of 82 patients per group will provide 80% power to detect this difference based on Fisher’s exact test and assuming a two-tailed alpha level of 0.05. This is assuming an outcome rate for of inadvertent intravascular injections of 5% in the interlaminar injection group and 20% in the transforaminal injection group. The achieved sample size in this prospective study of 93 patients with interlaminar injections and 119 patients with transforaminal lumbar injections provides adequate statistical power to detect this effect size. Power and sample size calculations were performed using nQuery Advisor version 8 (Statistical Solutions Ltd., Cork, Ireland).

Results

Participants

Between December 2017 and July 2019, 212 patients who presented for interlaminar and transforaminal lumbar injection were prospectively enrolled. There were 100/212 (47%) men and 112/212 (53%) women (mean age = 64 ± 16 years, age range = 16–90 years). The BMI was 28.8 ± 6.6 kg/m² (range = 15.5–46.0 kg/m²). There were 99/212 (47%) left-sided injections and 113/212 (53%) right-sided injections. Distribution of injection levels is detailed in Table 1. There were 89/212 (42%) interlaminar and 123/212 (58%) transforaminal injections. There were 36/212 (17%) patients with prior surgery at or adjacent to the level of the injection (21 posterior decompression, 14 posterior spinal fusion with or without decompression, and one anterior discectomy and fusion).

Table 1.

Number of injections at each of the injection levels.

Level	Number of injections (n = 212)
L1-L2	6 (3) 4 interlaminar, 2 transforaminal
L2-L3	22 (10) 12 interlaminar, 10 transforaminal
L3-L4	44 (21) 28 interlaminar, 16 transforaminal
L4-L5	69 (33) 41 interlaminar, 28 transforaminal
L5-S1	56 (26) 4 interlaminar, 52 transforaminal
S1-S2	15 (7) 0 interlaminar, 15 transforaminal

Values are given as n (%).

Inadvertent injections

There were 25/212 (12%) inadvertent intravascular injections (Table 2). The mean age of the participants with inadvertent intravascular injection was higher, and this difference was statistically significant (P =0.017). There was no significant difference between the gender distributions of the participants and the side of the injection (P ≥ 0.05). Inadvertent intravascular injections were significantly more common in injections performed by the transforaminal approach (P = 0.0001), occurring with an incidence of 2/89 (2%) for interlaminar injections and 23/123 (19%) for transforaminal injections. Intravascular injection was significantly more common among patients when the injection was performed at or adjacent to a level with prior surgery (P < 0.0001). The overall incidence of inadvertent intravascular injection among patients with surgery was 13/36 (36%) and, of these patients, 12/13 (92%) had an injection by a transforaminal approach. In multivariable logistic regression analysis (Table 3), transforaminal injections were independently associated with increased risk of inadvertent intravascular injections (OR = 9.77, 95% CI = 2.14–44.6, P = 0.003). Prior surgery at level or adjacent level was also an independent risk factor for inadvertent intravascular injections (OR = 5.71, 95% CI = 2.15–15.15, P < 0.001).

Table 2.

Univariate analysis of clinical and injection characteristics for inadvertent intravascular injection.

	Inadvertent intravascular injection (n = 25)	No intravascular injection (n = 187)	P
Gender	Male: 14 (56) Female: 11 (44)	Male: 98 (52) Female: 89 (48)	0.83
Age (years)	70 ± 14	63 ± 16	0.017
Side	Left: 12 (48) Right: 13 (52)	Left: 87 (47) Right: 100 (54)	0.99
Injection type	Interlaminar: 2 (8) Transforaminal: 23 (92)	Interlaminar: 87 (47) Transforaminal: 100 (53)	0.0001
Surgery at injection level or adjacent level	No: 12 (48) Yes: 13 (52)	No: 164 (88) Yes: 23 (12)	<0.0001

Values are given as n (%).The bold values are statistically significant.

Table 3.

Multivariable logistic regression analysis of clinical and injection characteristics for inadvertent intravascular injection.

Covariate	Adjusted OR	95% CI	P
Gender
Male	1.07	(0.42–2.75)	0.89
Female	Reference	–	–
Age (years)	1.03	(0.99–1.06)	0.115
Side
Left	Reference	–	–
Right	1.16	(0.46–2.98)	0.752
Injection type
Interlaminar	Reference	–	–
Transforaminal	9.77	(2.14–44.6)	0.003
Surgery at injection level or adjacent level
No	Reference	–	–
Yes	5.71	(2.15–15.15)	<0.001

CI, confidence interval; OR, odds ratio.The bold values are statistically significant (p < 0.05).

The patients who had previous spine surgery were removed, and repeat comparison between interlaminar and transforaminal injections was performed. There were 176 injections in patients who had never had surgery, 79/176 (44%) were interlaminar and 97/176 (55%) were transforaminal. Inadvertent intravascular injections remained significantly more common in injections performed by the transforaminal approach (P = 0.01), occurring with an incidence of 1/79 (1%) for interlaminar injections and 11/97 (11%) for transforaminal injections.

For the two interlaminar spinal injections with intravascular injection, the blood vessels were detected at approximately 1 mL of injected. For the transforaminal interlaminar injections, the blood vessels were detected at 0.29 ± 0.23 mL (range = 0.1–1 mL) of injected contrast.

Complications

All injections were completed, including adequate injection of contrast and complete injection of steroid and anesthetic doses. There were no injection-related complications for any of the patients included in the study.

Discussion

In the present study, there was an overall 12% incidence of inadvertent intravascular injection, with higher rates in transforaminal injections than in interlaminar injections. The highest rates were detected in older patients and in injections at or adjacent to a lumber vertebral level with prior surgery.

The overall incidence of intravascular injection is similar to that described by previous reports. Yu et al. (25) reported an incidence of 9% for transforaminal injections, although the trial dose of contrast was only 0.3 mL and the injections were performed under CT guidance, which is less sensitive than fluoroscopic guidance for the detection of intravascular flow (24). Furman et al. (26) reported an overall incidence of 11.2% intravascular injection, with a significantly higher rate at S1 (21.3%) compared with the other lumbar levels (8.1%). Test injection was performed using 1–2 mL of non-ionic contrast. Sullivan et al. (23) found an incidence of 10.8% for transforaminal injections and 1.9% for interlaminar injections, but the volume of contrast injected was not specified. Smuck et al. (27) demonstrated an incidence of 13.1% for transforaminal injections using a test injection of up to 0.5 mL of contrast. Hong et al. (28) found the highest incidence of intravascular flow for transforaminal injections, 15.5%, and also injected up to 3 mL, the largest test volume of contrast. It was found that intravascular flow was demonstrated in an average of 0.3 mL of contrast injection, and it is possible that some of the studies with lower injected contrast volumes did not see the intravascular flow, which would have become apparent at higher contrast volumes.

An extensive venous and arterial plexus present within the epineural and epidural spaces has been described (24,29). In particular, the high density of vessels at the level of the neural foramen is likely responsible for reports of higher rates of intravascular injection in this location (25). Disc herniations have also been shown to result in neovascularity surrounding the disc fragments, some of which may help with regression of disc material (30 –32). Furthermore, healing after surgery results in angiogenesis, which does not occur in a predictable pattern (33,34). Our study supports prior spinal surgery as a risk factor for inadvertent intravascular injection. However, even with the surgical cases removed, transforaminal approach demonstrated a higher inadvertent intravascular injection rate. As a transforaminal approach is often chosen in situations where surgery such as posterior decompression precludes interlaminar approach, extra caution should be taken in these procedures to detect inadvertent intravascular injection.

Paralysis and spinal cord injury are rare but catastrophic complications of epidural injections (11 –22). The risk of spinal cord injury is higher when a particulate steroid such as betamethasone is used, presumably because of the embolism of these particles, which can aggregate to total particle sizes of 100 μm, a size which is sufficient for occlusion of small end-vessels (35). The rationale for the use of particulate steroids for epidural injections is primarily based on experience with intramuscular injections, where the local depot of particulate steroids is hypothesized to release a constant steroid dose for a longer period of time (36). Three studies comparing particulate and non-particulate steroid injections have not found a difference in amount of pain relief, but have suggested a shorter duration of pain relief for non-particulate dexamethasone, which may also require more injections for similar efficacy (14,37,38). In addition, there are other proposed mechanisms of catastrophic vascular events aside from intravascular injection, including vasospasm, compression by epidural hemorrhage, and mechanical disruption of arteries (35,39–42). In our study, inadvertent intravascular injection was demonstrated on average with a 0.3-mL injection, although as much as 1 mL of contrast was required to visualize the vessels. This suggests that at least 1 mL of contrast should be injected to detect inadvertent intravascular injection.

The advantages of this study include one institution with standardized protocols, and prospective enrollment of participants to ensure careful assessment for intravascular flow at the time of injection. However, accurately determining whether the injected blood vessels are arterial or venous is difficult under live fluoroscopy. Therefore, it remains difficult to determine which intravascular injections would be expected to result in increased risk for vascular complication (19). In addition, a limitation of this study is the small number of intravascular injections with interlaminar procedures, which makes it difficult to draw definitive conclusions about interlaminar epidural injections.

In conclusion, our findings suggest inadvertent intravascular injection occurs with higher frequency in older patients, in injections at or adjacent to a lumbar vertebral level with prior surgery, and with transforaminal injections.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Connie Y Chang

References

Hashemi

Webster

Clancy

EA.

Trends in disability duration and cost of workers’ compensation low back pain claims (1988–1996). J Occup Environ Med 1998; 40:1110–1119.

Guo

Tanaka

Halperin

, et al. Back pain prevalence in US industry and estimates of lost workdays. Am J Public Health 1999; 89:1029–1035.

Eley

Magnetic resonance imaging for low back injuries: appropriate use in managing workers’ compensation claims. AAOHN J 2006; 54:429–433.

Turner

Franklin

Fulton-Kehoe

, et al. Worker recovery expectations and fear-avoidance predict work disability in a population-based workers’ compensation back pain sample. Spine 2006; 31:682–689.

Derby

Kine

Saal

, et al. Response to steroid and duration of radicular pain as predictors of surgical outcome. Spine 1992; 17:S176–183.

Dooley

McBroom

Taguchi

, et al. Nerve root infiltration in the diagnosis of radicular pain. Spine 1988; 13:79–83.

Slipman

Issac

The role of diagnostic selective nerve root blocks in the management of spinal pain.

Pain Physician 2001; 4:214–226.

Abram

O’Connor

TC.

Complications associated with epidural steroid injections. Reg Anesth 1996; 21:149–162.

McLain

Fry

Hecht

ST.

Transient paralysis associated with epidural steroid injection. J Spinal Disord 1997; 10:441–444.

10.

Houten

Errico

TJ.

Paraplegia after lumbosacral nerve root block: report of three cases. Spine J 2002; 2:70–75.

11.

Huntoon

Martin

DP.

Paralysis after transforaminal epidural injection and previous spinal surgery. Reg Anesth Pain Med 2004; 29:494–495.

12.

Somayaji

Saifuddin

Casey

ATH

, et al. Spinal cord infarction following therapeutic computed tomography-guided left L2 nerve root injection. Spine 2005; 30:E106–108.

13.

Neal

Kopp

Pasternak

, et al. Anatomy and pathophysiology of spinal cord injury associated with regional anesthesia and pain medicine: 2015 update. Reg Anesth Pain Med 2015; 40:506–525.

14.

Kennedy

Dreyfuss

Aprill

, et al. Paraplegia following image-guided transforaminal lumbar spine epidural steroid injection: two case reports. Pain Med 2009; 10:1389–1394.

15.

Glaser

Shah

RV.

Root cause analysis of paraplegia following transforaminal epidural steroid injections: the “unsafe” triangle. Pain Physician 2010; 13:237–244.

16.

Helm

Glaser

Falco

, et al. A medical-legal review regarding the standard of care for epidural injections, with particular reference to a closed case. Pain Physician 2010; 13:145–150.

17.

Thefenne

Dubecq

Zing

, et al. A rare case of paraplegia complicating a lumbar epidural infiltration. Ann Phys Rehabil Med 2010; 53:575–583.

18.

Chang Chien

Candido

Knezevic

NN.

Digital subtraction angiography does not reliably prevent paraplegia associated with lumbar transforaminal epidural steroid injection.

Pain Physician 2012; 15:515–523.

19.

Nagpal

Chang-Chien

Benfield

, et al. Digital subtraction angiography use during epidural steroid injections does not reliably distinguish artery from vein. Pain Physician 2016; 19:255–266.

20.

Laemmel

Segal

Mirshahi

, et al. Deleterious effects of intra-arterial administration of particulate steroids on microvascular perfusion in a mouse model. Radiology 2016; 279:731–740.

21.

Rathmell

Benzon

Dreyfuss

, et al. Safeguards to prevent neurologic complications after epidural steroid injections: consensus opinions from a multidisciplinary working group and national organizations. Anesthesiology 2015; 122:974–984.

22.

Atluri

Glaser

Shah

, et al. Needle position analysis in cases of paralysis from transforaminal epidurals: consider alternative approaches to traditional technique. Pain Physician 2013; 16:321–334.

23.

Sullivan

Willick

Chira-Adisai

, et al. Incidence of intravascular uptake in lumbar spinal injection procedures. Spine 2000; 25:481–486.

24.

Palmer

WE.

Spinal injections for pain management. Radiology 2016; 281:669–688.

25.

Lagemann

Ghodadra

, et al. Extraforaminal needle tip position reduces risk of intravascular injection in CT-fluoroscopic lumbar transforaminal epidural steroid injections. J Spine Surg 2016; 2:246–255.

26.

Furman

O’Brien

Zgleszewski

TM.

Incidence of intravascular penetration in transforaminal lumbosacral epidural steroid injections. Spine 2000; 25:2628–2632.

27.

Smuck

Fuller

Yoder

, et al. Incidence of simultaneous epidural and vascular injection during lumbosacral transforaminal epidural injections. Spine J 2007; 7:79–82.

28.

Hong

Kim

Huh

, et al. Analysis of inadvertent intradiscal and intravascular injection during lumbar transforaminal epidural steroid injections: a prospective study. Reg Anesth Pain Med 2013; 38:520–525.

29.

Mandell

Czuczman

Gaviola

, et al. The lumbar neural foramen and transforaminal epidural steroid injections: an anatomic review with key safety considerations in planning the percutaneous approach. AJR Am J Roentgenol 2017; 209:W26–W35.

30.

David

Ciurea

Iencean

, et al. Angiogenesis in the degeneration of the lumbar intervertebral disc. J Med Life 2010; 3:154–161.

31.

Rätsep

Minajeva

Asser

Relationship between neovascularization and degenerative changes in herniated lumbar intervertebral discs. Eur Spine J 2013; 22:2474–2480.

32.

Pytel

Wollmann

Fessler

, et al. Degenerative spine disease: pathologic findings in 985 surgical specimens. Am J Clin Pathol 2006; 125:193–202.

33.

Trabucchi

Radaelli

Marazzi

, et al. The role of mast cells in wound healing. Int J Tissue React 1988; 10:367–372.

34.

Arnold

West

Kumar

Wound healing: the effect of macrophage and tumour derived angiogenesis factors on skin graft vascularization.

Br J Exp Pathol 1987; 68:569–574.

35.

Derby

Lee

S-H

Date

, et al. Size and aggregation of corticosteroids used for epidural injections. Pain Med 2008; 9:227–234.

36.

Abraham

Demiraj

Ungemach

FR.

Comparison of the hypothalamic-pituitary-adrenal axis susceptibility upon single-dose i.m. depot versus long-acting i.v. triamcinolone acetonide therapy: a direct pharmacokinetic correlation. J Endocrinol 2006; 191:491–496.

37.

Kim

Brown

Efficacy and safety of lumbar epidural dexamethasone versus methylprednisolone in the treatment of lumbar radiculopathy: a comparison of soluble versus particulate steroids. Clin J Pain 2011; 27:518–522.

38.

El-Yahchouchi

Geske

Carter

, et al. The noninferiority of the nonparticulate steroid dexamethasone vs the particulate steroids betamethasone and triamcinolone in lumbar transforaminal epidural steroid injections. Pain Med 2013; 14:1650–1657.

39.

Dietrich

Suter

Pfirrmann

CWA

, et al. Supraacetabular fossa (pseudodefect of acetabular cartilage): frequency at MR arthrography and comparison of findings at MR arthrography and arthroscopy. Radiology 2012; 263:484–491.

40.

Ziai

Ardelt

Llinas

RH.

Brainstem stroke following uncomplicated cervical epidural steroid injection. Arch Neurol 2006; 63:1643–1646.

41.

Rozin

Koehler

, et al. Death during transforaminal epidural steroid nerve root block (C7) due to perforation of the left vertebral artery. Am J Forensic Med Pathol 2003; 24:351–355.

42.

Wallace

Fukui

Williams

, et al. Complications of cervical selective nerve root blocks performed with fluoroscopic guidance. AJR Am J Roentgenol 2007; 188:1218–1221.