Periradicular infiltration of the lumbar spine: testing the robustness of an interventional ultra-low-dose protocol at different body mass index levels

Abstract

Background

Computed tomography (CT)-guided periradicular infiltration remains a frequent interventional procedure for treatment of low back pain.

Purpose

To present an interventional ultra-low-dose protocol for CT-guided periradicular infiltration therapy and assess its application at different body mass index (BMI) levels.

Material and Methods

Over a period of 14 months, 79 patients underwent 183 CT-guided interventions for single-site lumbar periradicular therapy using an ultra-low-dose protocol with a basic setup of 100 kV and 5 mAs. Procedures were performed via intermittent fluoroscopy. A retrospective review was performed to analyze the parameters tube current and tube voltage, dose-length product, and BMI.

Results

The interventional ultra-low-dose protocol allowed safe treatment of 91.1% of the patients without a need for adapting the protocol. In seven patients with a higher BMI (range, 31–38 kg/m²; mean, 34 kg/m²), the tube current had to be increased to retain sufficient image quality. Only patients with a BMI of 30 and higher showed a significant correlation between BMI and dose-length product (P value = 0.02), resulting in a slightly increased dose (P value = 0.002).

Conclusion

The protocol presented for the interventional part of CT-guided periradicular infiltration allows to safely treat patients with a median calculated effective dose of 0.045 mSv (converted from a dose-length-product of 2.26 mGy*cm). Patients with a BMI of 30 and higher required a higher calculated effective dose with just one patient slightly exceeding 0.1 mSv.

Keywords

Interventional radiology computed tomography (CT)spine lumbar injection radiation dosage

Introduction

Low back pain is a widespread syndrome (1,2), with a high economic burden due to long-term medical management and absence from work (3). Therefore, there is a need for effective treatment strategies. One of many options, especially in patients with lumbar nerve root irritation (4), is periradicular infiltration, which has been shown to improve patient outcome (5,6). Two strategies have emerged into routine: while at the epidural space injection (ESI) the cortisone is applied into the epidural space; and the periradicular infiltration aims at the spinal nerve emerging through the intervertebral foramen.

Periradicular infiltration is predominantly performed using an interlaminar or transforaminal access routes (7). Infiltration can be performed without imaging or under guidance of fluoroscopy, computed tomography (CT) or magnetic resonance imaging (MRI) (8,9). There is an ongoing debate about the most eligible procedure in terms of individual weighing of safety, radiation exposure, and costs (10,11). CT-guided injection has proven to offer both high precision and high patient safety for some reasons: CT provides much higher spatial resolution than simple fluoroscopic guidance and allows clear visualization of the target area, placement of the needle tip, and anatomic abnormalities such as osseous stenosis (12). Hence, this procedure is routinely performed under CT guidance. However, a generally accepted examination protocol striking the right balance between safety and dose reduction is still lacking. Therefore, radiation exposure varies and is often a cause of concern (13) for both interventionalists (14) and patients. The former is at risk of radiation-induced effects such as cataractogenesis (15), while the latter might be exposed to stochastic radiation effects (16,17); there is a general risk of DNA double-strand breaks reported at higher doses (18). Reports on the calculated effective overall dose for periradicular therapy are in the range of 0.21–3.35 mSv per intervention (19,20).

Since patients with low back pain tend to undergo multiple treatment sessions, dose reduction has become an important issue (10,21). While the radiation of the planning scan affects a larger volume, the dose applied during the intervention is confined to the slice from which images are obtained, where it accumulates and increases with the number of images acquired. Recently, an iterative reconstruction technique has been implemented for fluoroscopy and interventional CT allowing for a decrease of the dose to the small area exposed during the intervention. This is why we investigated an ultra-low-dose interventional protocol. Furthermore, as the number of obese patients increases (22), a focus was on investigating its robustness in relation to different body mass index (BMI) levels.

Material and Methods

Patients

This institutional review board approved study included 79 patients undergoing a single site periradicular infiltration therapy of the lumbar spine for pain relief over a period of 14 months.

Inclusion criteria were: patients aged over 18 years with acute or chronic lumbar nerve root compression causing focal neurological symptoms. Patients were referred from the departments of traumatology, orthopedic surgery, neurosurgery, and neurology. While clinical evaluation was the main indication, some of the patients also underwent MRI prior to our intervention.

We recorded the patients’ age, weight, and height as well as the BMI (defined as body weight [kg] / body height [m²]) as an approximation of the absorption mass. All lumbar interventions between August 2015 and October 2016 were included.

Procedure

All periradicular infiltrations were performed on an 80-slice CT scanner (Toshiba Aquilion PRIME, Toshiba Medical Systems, Ottawara, Japan) by a board-certified radiologist with 13 years of experience. Tube current (mAs), tube voltage (kV), the CTDI_vol (CT dose index, unit: mGy), and its derivate DLP (CTDI_vol*scan length, unit: mGy*cm) of the complete interventional series were evaluated. CTDI_vol and DLP both serve as index for radiation output of a CT scanner.

CT protocol

After obtaining a lateral scout image the correct position for the intervention was determined by acquiring a scout-guided spot scan (four sections of 1 mm thickness, 100 kV, 20 mAs determined from the lateral scout). The interventional setup used a protocol of a non-enhanced image acquisition with 100 kV, 5 mAs, 0.5 s rotation time, 32 cm field of view, and a single slice acquisition of 8 mm to allow visualization of the needle even with minor deviation in z-axis. Images were reconstructed using a sharp (FC19) kernel and then examined using “bone window” setting. For all imaging, we used a specific convolution filter in the reconstruction process called Adaptive Iterative Dose Reduction 3D (AIDR 3D, Toshiba Medical Systems). The iterative reconstruction level was chosen medium (AIDR 3D standard).

To reduce overall dose, we used intermittent CT fluoroscopy (“step-and-shoot technique,” one single slice acquisition) with the physician staying next to the patient. We used a 19.5 G spinal cannula and the transforaminal access route for every patient. Criteria for safe treatment were met when both the needle tip with its artifact and the landing area were visualized clearly (Fig. 1) at any time. In case of insufficient identification of these structures (Fig. 2) the radiologist increased the tube current-time product in steps of 5 mAs until image quality was rated sufficient (Fig. 3). Every acquisition was recorded and counted to the interventional dose.

Fig. 1.

Axial CT, bone window. Male patient, 42 years old, BMI 29 kg/m². For patient safety, both the needle tip (white arrow), anatomic landmarks to pass (dotted arrow) and the target area (white circle) needed to be visualized. Setup: 100 kV, 5 mAs.

Fig. 2.

Axial CT, bone window. Female patient, 62 years old, BMI 32 kg/m². This image (setup: 100 kV, 5 mAs) exemplifies insufficient image quality (needle tip and neuroforamen not clearly visualized) which was followed by the decision to raise the tube current time product to 10 mAs).

Fig. 3.

Axial CT, bone window. Same patient as in Fig. 2. This image (setup: 100 kV, mAs raised to 10) shows the effect of achieving sufficient image quality by raise of mAs.

Prior to the therapeutic injection (2 mL Bupivacain 0.5% and 1 mL Triamcinolonacetonid) a test injection containing contrast medium was performed (Fig. 4) to proof and document correct medical deposition.

Fig. 4.

Axial CT, bone window. Male patient, 42 years old, BMI 29 kg/m². Test injection with contrast medium approves reach of the spinal nerve. Setup: 100 kV, 5 mAs.

The calculated effective dose can be estimated from the calculated dose (i.e. DLP) using conversion coefficients. Although those coefficients are influenced by many variables such as scanner type, tube current, weight, and age, we used the generally accepted conversion factor of 0.020 for abdominal CT imaging as most recently published (23).

Data analysis

Statistical analysis was performed and graphics were generated using R (version 3.2.3, R Foundation for Statistical Computing, Vienna, Austria), a GNU project with copyright by the R Foundation. After ruling out normal distribution via Shapiro–Wilk test, we used Spearman’s rank correlation as a non-parametric measure of statistical dependence between DLP and BMI as well as number of acquisitions and BMI. The non-parametric Wilcoxon rank sum test was applied to test for dose-related differences between groups of BMI. Statistical significance was assumed for P values < 0.05.

According to Cohen (24), two parameters are interpreted to show small correlation at a correlation coefficient level of lower than 0.3, medium correlation at lower than 0.5, and large correlation at 0.5 and higher.

Results

A total of 183 CT-guided interventions in 79 patients were included in this study (38 women, 41 men; median age, 56 years; median weight, 75.0 kg; median BMI, 27.0 kg/m²). Seventy patients received injections of the nerve root L5, seven at L4, and one each at L1 and L3.

Mean calculated DLP for planning purposes was 1.92 mGy*cm (median, 0.8 mGy*cm). The mean number of acquisitions including all acquired images were 7.8 on average (median, 7.33).

The datasets were divided into three groups according to BMI using the accepted international classification of the World Health Organization: lower than 25 (group 1), 25 to lower than 30 (group 2), 30 and above (group 3) (25).

DLP and BMI showed a correlation of r = 0.11 (P value = 0.56) for group 1, r = 0.11 (P value = 0.54) for group 2, and r = 0.61 (P value = 0.02) for group 3. The corresponding dot plot is presented in Fig. 5.

Fig. 5.

Dot plot showing the correlation of BMI and DLP with regression line for each of the three BMI groups investigated. Group 1 is represented by black circles with crosses (dotted regression line), group 2 by gray triangles (dashed regression line), and group 3 by black points (regression line drawn through).

The number of acquisitions and BMI showed a correlation of r = –0.04 (P value = 0.33).

Analysis of mean calculated DLP during interventions in the different BMI groups revealed no differences between group 1 (DLP 2.41 mGy*cm) and group 2 (2.31 mGy*cm). Group 3 showed a statistically significantly (P value = 0.002) higher mean DLP of 3.37 mGy*cm (calculated effective dose = 0.067 mSv) compared to groups 1 and 2 (DLP 2.36 mGy*cm, calculated effective dose = 0.047 mSv). The results are presented in Tables 1 and 2, the corresponding box plot in Fig. 6.

Table 1.

Characteristics of the three groups according to the BMI.

Parameter/ Group	Group 1: BMI < 25	Group 2: BMI < 30	Group 3: BMI ≥ 30
n	30	34	15
Mean BMI	22.8	27.76	33.4
Mean DLP	2.41	2.31	3.37
Mean calculated effective dose	0.048	0.046	0.067

Table 2.

Comparison of DLP means between the three groups according to the BMI.

Groups compared	Mean A	Mean B	P (Wilcoxon rank sum test)
1 vs. 2	2.41	2.31	0.701
1 vs. 3	2.41	3.37	0.011
2 vs. 3	2.31	3.37	0.002
1 + 2 vs. 3	2.36	3.37	0.002

Fig. 6.

Box plot showing the DLP for each of the three BMI groups investigated. Black points indicate the median, boxes the 25th and 75th percentiles. Dashed lines represent the most extreme data points, outliers defined as outside the 25th and 75th percentiles are indicated by gray points. The Wilcoxon rank sum test revealed statistical significance (P = 0.002) when comparing groups 1 and 2 taken together with group 3.

The median calculated effective dose of all patients was 0.045 mSv (DLP of 2.27 mGy*cm, CTDI_vol of 2.80 mGy). The corresponding mean calculated effective dose was 0.051 mSv (DLP of 2.55 mGy*cm, CTDI_vol of 3.11 mGy).

Protocol modifications were not required in groups 1 and 2, while a single increase in tube current from 5 to 10 mAs was required in 7/15 patients in group 3 to keep sufficient image quality (average BMI, 34 kg/m²; range, 31–38 kg/m²). The maximum calculated dose applied here was a DLP of 6.13 mGy*cm (0.123 mSv) in one patient with a BMI of 37. In the remaining eight patients of group 3, tube current of 5 mAs was sufficient.

Overall, the new ultra-low-dose protocol presented here allowed safe treatment without protocol changes in 72 (91.1%) patients.

Discussion

Achieving low-dose in CT musculoskeletal imaging is a topic of high demand. Besides interventional procedures, evident advances have been made especially in diagnostic imaging of the spine. Alshamari et al. published a conclusion based on a phantom and sequentially a clinical study showing that CT of the lumbar spine can be performed at about 1 mSv low-dose versus about 0.7 mSv for plain radiography. Hereby the diagnostic information of the CT was still significantly superior to radiography (26,27). Recently, also the implementation of iterative reconstruction software and its contribution to achieve low-dose in diagnostic CT imaging of the spine has been positively evaluated (28).

CT-guided periradicular infiltration is still an elective procedure so it is imperative to keep elaborating all options for patients’ radiation protection.

Amrhein et al. recently presented a study mainly focusing on the the planning CT to lower the overall dose (29). Their interventional part counted for a mean DLP of 66.3 mGy*cm, 1.326 mSv converted to calculated effective dose. They measured the anteroposterior diameter of the patient to select either 100 mAs or 50 mAs with 120 kV fixed.

Also, Paik targeted the planning CT by replacing it with a spot scan and stayed with a fixed setup of 120 kV and 30 mAs (also 30 mAs, rotation time set at 1.0 s) during the interventional part where they acquired single images of 5 mm slice thickness. He stated a median DLP of 3.71 mGy*cm, 0.056 mSv converted to calculated effective dose (30). It has to be noted that Paik chose a conversion factor of 0.018. Recalculation with the updated conversion factor of 0.020 results in a calculated effective dose of 0.074 mSv.

The application of our interventional protocol resulted in even lower doses compared to literature. At the same time, even patients with a high BMI were exposed to lower doses than recently published for all patients. In comparison to other studies we have demonstrated that especially our significantly reduced tube current-time product had potential of further dose reduction without compromising the interventional procedure because of insufficient image quality. The excellent contrast-to-noise ratio between bone/needle and soft tissue facilitated dose lowering.

As the dose resulting from the interventional part also depends on the number of scans taken during the intervention, an experienced examiner might need fewer scans and thus achieve lower dose levels (31).

The results of our correlation analysis suggest that patients with a BMI of 30 kg/m² and higher require higher radiation doses. In terms of DLP, we found an increase from 2.36 (groups 1 + 2) to 3.37 (group 3), corresponding to a 42.7% increase. In terms of calculated effective dose, this corresponds to an average rise from 0.047 mSv to 0.067 mSv. Moreover, the calculated effective dose tends to be overestimated in overweight patients and underestimated in underweight patients (32).

In this study, we focused on the interventional part where dose is applied on the same millimeters of the body several times. Here, dose reduction will affect cell biology the most although interventional planning contributes to the overall dose as well (29,33). Therefore, we present an ultra-low-dose protocol for the interventional procedure with detailed statistical analysis. With doses levels presented here planning dose might come into the focus to achieve even lesser overall dose. These considerations should be examined in future studies. Our results might not only be interesting in sense of adopting the presented protocol but also for reevaluating other interventional protocols in terms of specific BMI groups.

Our study has some limitations including its retrospective design. AIDR 3D significantly reduces noise (34 –36) and contributes to improved spatial resolution (37). It still has to be studied, if implementation of iterative reconstruction software to interventional imaging is essential regarding dose reduction. We decided to set AIDR 3D standard strength (mild, standard, and strong can be selected) because in our opinion this provides best proportion of image noise reduction to adversely smoothing sharp edges of anatomical structures.

We divided patients into three BMI groups. While the BMI provides some measure of a patient’s constitution, it contains no information on body mass distribution, which may be inhomogeneous. Anyway, a bottom-up strategy starting with the lowest achievable dose will allow for the lowest dose needed. As an alternative scout-based tube modulation at high noise levels could propose an alternative starting dose.

Although effective dose estimation using conversion coefficients has proven to be an acceptable way of calculating the effective dose, one has to be reminded that the inaccuracy increases the more the BMI differs from normal range. Calculating the size corrected dose estimates (SSDE) would provide higher accuracy. According to published conversion factors to calculate SSDE from CTDI_vol patients with higher effective diameter will receive even lower dose estimations relatively to those with a lower diameter (38). Therefore, using DLP, we might have overestimated dose of patients with higher BMI.

When imaging the neuroforamen, nerve root L5 might request a higher dose as the bony pelvis will be part of the image as well. In our study, even here the minimum possible dose was rated sufficient with no need for dose adaption. The next step could be reducing the voltage to 80 kV on levels of L4 and higher.

According to our results, we can recommend use featuring a basic setup of 100 kV and 5 mAs for periradicular infiltration of the lumbar spine. All patients with a BMI lower than 30 kg/m² were treated safely using this protocol. For patients with a higher BMI (obesity grade 1, BMI > 30), increasing the tube current-time product in steps of 5 mAs should be considered to maintain adequate image quality.

In conclusion, the interventional ultra-low-dose protocol presented here allows safe and efficient treatment of patients suffering from low back pain with a median calculated effective dose of 0.045 mSv.

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.

References

Airaksinen

Brox

Cedraschi

et al.

Chapter 4. European guidelines for the management of chronic nonspecific low back pain. Eur Spine J 2006; 15: S192–300. Suppl. 2.

Manchikanti

Singh

Falco

et al.

Epidemiology of low back pain in adults. Neuromodulation 2014; 17: 3–10. Suppl. 2.

Martin

Deyo

Mirza

et al.

Expenditures and health status among adults with back and neck problems. JAMA 2008; 299: 656–664.

Oder

Thurnher

. [Periradicular infiltration therapy: Clinical indications, technique and results]. Radiologe 2015; 55: 833–839.

Lee

Lin

Hwang

et al.

Transforaminal periradicular infiltration guided by CT for unilateral sciatica–an outcome study. Clin Imaging 2005; 29: 211–214.

Vad

Bhat

Lutz

et al.

Transforaminal epidural steroid injections in lumbosacral radiculopathy: a prospective randomized study. Spine (Phila Pa 1976) 2002; 27: 11–16.

Andreisek

Jenni

Klingler

et al.

Access routes and reported decision criteria for lumbar epidural drug injections: a systematic literature review. Skeletal Radiol 2013; 42: 1683–1692.

Wagner

. CT fluoroscopy-guided epidural injections: technique and results. Am J Neuroradiol 2004; 25: 1821–1823.

Streitparth

De Bucourt

Hartwig

et al.

Real-time MR-guided lumbosacral periradicular injection therapy using an open 1.0-T MRI system: an outcome study. Invest Radiol 2013; 48: 471–476.

10.

Kramer

Blettner

Hammer

. Image-guided injection therapy in the lumbar spine. Dtsch Arztebl Int 2008; 105: 596–598.

11.

Maurer

Schreiter

de Bucourt

et al.

Cost comparison of nerve root infiltration of the lumbar spine under MRI and CT guidance. Eur Radiol 2013; 23: 1487–1494.

12.

Daly

Templeton

. Real-time CT fluoroscopy: evolution of an interventional tool. Radiology 1999; 211: 309–315.

13.

Brenner

Hall

. Computed tomography–an increasing source of radiation exposure. N Engl J Med 2007; 357: 2277–2284.

14.

Rogits

Jungnickel

Lowenthal

et al.

Prospective evaluation of the radiologist’s hand dose in CT-guided interventions. Rofo 2013; 185: 1081–1088.

15.

Seals

Lee

Cagnon

et al.

Radiation-induced cataractogenesis: a critical literature review for the interventional radiologist. Cardiovasc Intervent Radiol 2016; 39: 151–160.

16.

Wagner

Eifel

Geise

. Potential biological effects following high X-ray dose interventional procedures. J Vasc Interv Radiol 1994; 5: 71–84.

17.

Kong

Liu

et al.

Biological effects of low-dose radiation from computed tomography scanning. Int J Radiat Biol 2013; 89: 326–333.

18.

Geisel

Zimmermann

Rief

et al.

DNA double-strand breaks as potential indicators for the biological effects of ionising radiation exposure from cardiac CT and conventional coronary angiography: a randomised, controlled study. Eur Radiol 2012; 22: 1641–1650.

19.

Artner

Cakir

Weckbach

et al.

Radiation dose reduction in CT-guided periradicular injections in lumbar spine: Feasibility of a new institutional protocol for improved patient safety. Patient Saf Surg 2012; 6: 19–19.

20.

Hoang

Yoshizumi

Toncheva

et al.

Radiation dose exposure for lumbar spine epidural steroid injections: a comparison of conventional fluoroscopy data and CT fluoroscopy techniques. Am J Roentgenol 2011; 197: 778–782.

21.

Lutze

Stendel

Vesper

et al.

Periradicular therapy in lumbar radicular syndromes: methodology and results. Acta Neurochir (Wien) 1997; 139: 719–724.

22.

Malik

Willett

. Global obesity: trends, risk factors and policy implications. Nat Rev Endocrinol 2013; 9: 13–27.

23.

Shrimpton

Jansen

Harrison

. Updated estimates of typical effective doses for common CT examinations in the UK following the 2011 national review. Br J Radiol 2016; 89: 20150346–20150346.

24.

Cohen

. Statistical power analysis for the behavorial sciences , 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates, 1988.

25.

World Health Organization. Obesity: Preventing and Managing the Global Epidemic, Geneva: WHO, 2000.

26.

Alshamari

Geijer

Norrman

et al.

Low-dose computed tomography of the lumbar spine: a phantom study on imaging parameters and image quality. Acta Radiol 2014; 55: 824–832.

27.

Alshamari

Geijer

Norrman

et al.

Low dose CT of the lumbar spine compared with radiography: a study on image quality with implications for clinical practice. Acta Radiol 2016; 57: 602–611.

28.

Yang

Chiou

et al.

Imaging quality and diagnostic reliability of low-dose computed tomography lumbar spine for evaluating patients with spinal disorders. Spine J 2014; 14: 2682–2690.

29.

Amrhein

Schauberger

Kranz

et al.

Reducing Patient Radiation Exposure From CT Fluoroscopy-Guided Lumbar Spine Pain Injections by Targeting the Planning CT. Am J Roentgenol 2016; 206: 390–394.

30.

Paik

. Radiation dose reduction in CT fluoroscopy-guided lumbar interlaminar epidural steroid injection by minimizing preliminary planning imaging. Eur Radiol 2014; 24: 2109–2117.

31.

Schmid

Schmitz

Borchardt

et al.

Effective dose of CT- and fluoroscopy-guided perineural/epidural injections of the lumbar spine: a comparative study. Cardiovasc Intervent Radiol 2006; 29: 84–91.

32.

Prokop

. [Radiation dose in computed tomography. Risks and challenges]. Radiologe 2008; 48: 229–242.

33.

Paik

. Radiation Dose reduction in CT fluoroscopy-guided cervical transforaminal epidural steroid injection by modifying scout and planning steps. Cardiovasc Intervent Radiol 2016; 39: 591–599.

34.

Schindera

Odedra

Raza

et al.

Iterative reconstruction algorithm for CT: can radiation dose be decreased while low-contrast detectability is preserved?

Radiology 2013; 269: 511–518.

35.

Gervaise

Osemont

Lecocq

et al.

CT image quality improvement using adaptive iterative dose reduction with wide-volume acquisition on 320-detector CT. Eur Radiol 2012; 22: 295–301.

36.

Gervaise

Osemont

Louis

et al.

Standard dose versus low-dose abdominal and pelvic CT: comparison between filtered back projection versus adaptive iterative dose reduction 3D. Diagn Interv Imaging 2014; 95: 47–53.

37.

Irwan R, Nakanishi S and Blum A. AIDR 3D – reduces dose and simultaneously improves image quality. Toshiba Medical Systems Europe White Paper. The Netherlands: Zoetermeer, 2011.

38.

American Association of Physicists in Medicine. Size Specific Dose Estimates (SSDE) in pediatric and adult body CT examinations. Report of the AAPM Task Group 204, College Park, MD, 2011.