18 Interventional Pain Treatment Options


18 Interventional Pain Treatment Options

Eric A.K. Mayer, Ryan Zate


In 2016, there were approximately 46 million Americans aged 65 and over. This number is expected to double by 2060, representing an increase from 15% of the population to 24%. CMS Data reports that lifetime prevalence of low back pain has been estimated to be as high as 84%. 1 According to the Medicare database, epidural injection utilization in Medicare populations rose 271–420% and use of facet joint injections increased 231–571% from 1995 to 2016. 2 Along with the growth of the aging population, a commensurate surge in medical utilization is expected. Procedural utilization has outpaced population growth logarithmically. Medical professionals will be tasked with treating a different pool of diagnoses with different outcome expectation (low likelihood for “return to work” metrics) in this older adult demographic. The goal of this chapter is a broad survey of common pathologies encountered in the aging population, of the continuum of interventional pain treatment options, and of the evidence underlying efficacious, individualized treatment recommendations.

Key Points

  • Determining the correct diagnosis is critical to performing successful interventional pain procedures.

  • There is evidence that epidural steroid injections improve pain and function in spinal stenosis WITH neurogenic claudication.

  • Indications for intervention include spinal stenosis, facet arthropathy and sacroiliac joint dysfunction.

  • Interventional treatments rarely succeed without the concomitant use of physical therapy and exercise.

  • Remember that time spent assessing for masqueraders of spine pain results in better outcomes and complication avoidance.

18.1 Indications and Contraindications

18.1.1 Lumbar Stenosis

Lumbar spinal stenosis (LSS) clinically presents with symptoms of neurogenic claudication (in the limbs). LSS is ALWAYS a clinical rather than a radiological diagnosis. This constellation of symptoms involves reduced walking distance tolerance and is difficult to differentiate from Peripheral Arterial Disease (PAD). Too often, patients with nonspecific, axial, mechanical chronic low back pain are misdiagnosed with LSS (though they have neither limitation in walking distance nor claudication symptoms in lower limbs). Surgical fusion MCID-adjusted success for nonspecific low back pain from (wrongly diagnosed LSS) has been found to benefit approximately 1 in 6 individuals. 3 This erroneous radiological diagnosis may lead to ineffective treatment plans. When a physician takes time to listen to a patient with LSS, they usually describe clinical symptoms of ”intermittent,” “dull or achy” quality pain in a nondermatomal pain pattern starting in the buttocks/legs occurring ONLY during standing or walking and that follows a crescendo-type pattern (worsening as the patient stands and walks) that we term the ”claudicant character” of pain. Most importantly, the patient’s pain resolves rapidly (seconds to minutes) with a position change of leaning forward, sitting, or lying. The subjective quality of the pain may vary in description: aching, cramping, burning, or simply “heavy” limb fatigability. Care must be taken in differentiating neurogenic claudication from vascular claudication (SeeTable 18‑1). Patients often note relief of symptoms shortly after sitting or with forward-flexed lumbar spine during ambulation (shopping cart sign). The pathophysiology of claudication is not fully understood, but the mechanism of repeated first-degree (neurapraxia), ischemic irritation of the intracanal nerve roots is hypothesized, with electrophysiological results to support this hypothesis. 4 Others hypothesize pathophysiological neuro-mechanical mechanisms related to narrowing around the nerve from ossified ligaments, joints, or discs consistent with a reduced dynamic cross-sectional area due to physiological bony arthritis in the spine.

Table 18.1

Neurogenic Claudication

Vascular claudication

May be proximal or distal muscles

Distal >> Proximal Muscles

Symptoms not improved by standing still

Symptoms improve slightly by standing

Improved with flexion at waist

Cool extremities +/– pulses

Improved performance with biking/or activities with flexion at the waist

Symptoms are time-related and fatigue of legs happens at the same time as other positions.

For Lumbosacral stenosis, an assessment of ambulatory distance is a requisite part of the examination. The reflexes may be diminished, but most cases present with variable focal neurologic deficits. Having the patient stand in prolonged extension of the lumbar spine may be difficult in the geriatric population but can be helpful in reproducing “claudicant” symptoms.

Interventional treatment options include the interlaminar or transforaminal epidural steroid injection. There is no clear evidence supporting one approach over the other for this indication. Studies have demonstrated that epidural steroid injections are reasonable, ‘short-to-medium term’ treatment in patients with neurogenic claudication to improve function, walking distance, and pain. 5 Studies have demonstrated that 64% of patients deemed to be surgical candidates had subjective improvement in symptoms 1 year after epidural injection, with fewer patients in the treatment group electing for surgery. 6 More recent studies have demonstrated improvements in both numerical pain scores of pain and disability. 7

Interestingly, this data also parsed response to anesthetic alone (vs. traditional combination of corticosteroid and anesthetic) and found no significant differences in response to treatment. Flexion-based exercise protocols may be recommended for patients with neurogenic claudication. Although active lifestyle and aerobic conditioning should be encouraged in all patients, there is inconsistent evidence that specific rehabilitation programs are effective in completely eliminating the return of symptoms. Given enough time in the best prospective cross-sectional study, nearly 7/10 patients randomly assigned to the nonsurgical group will opt to “cross-over” for surgical intervention to alleviate claudicatory symptoms within 5 years. 8 Different interventional approaches to delivering medication to the epidural space remain debated amongst specialists. In truth, comparing transforaminal vs. interlaminar epidural injection techniques has not reliably shown either approach to be significantly superior for treatment durability in symptomatic lumbosacral stenosis.

18.1.2 Facet Arthropathy

Facet arthropathy is anatomically ubiquitous in the aging population. Simultaneously, the prevalence of axial back pain is slightly greater than the prevalence in the general population, with the facet joint proposed to be a very common cause of axial back pain. 9 Seventy to 80% of the population will endure a “significant episode” of neck or back pain during adulthood. 10 Facet-mediated pain has been described as early as 1911 by Goldthwaitt and has been estimated to account for up to 45% of chronic spine pain. Clinically, painful facet arthropathy can be difficult to diagnose, as no single pathognomonic test, study, complaint, or physical exam maneuver has been reliably validated. Moreover, a number of twin-twin studies implicate genetics rather than stress/load-exposure as proximate cause of radiological changes. 11 Research has established that occupational spine-loading fails to demonstrate a causal link to intensity of spine pain or degree of radiological change observed by a physician. 12 The most rigorous evidence shows “activity-discordant” twin-twin studies in several countries. This body of evidence shows that the more “active” twin (presumed to have greater spine loading) had significantly lower risk of developing disabling spine-related pain in both short- and long-duration follow-up over multiple years. 13 Although diagnosing cervical or lumbar facet-mediated pain relies on the assimilation of clinical presentation and possibly invasive testing; treatment is incomplete if it fails to include physical activity. Moderately strong evidence shows that actively-resisted exercise with “directional preference” shows the greatest evidence for efficacy. 14

Patients presenting with painful facet arthritis tend to present with axial back or neck pain that is worsened with prolonged static position and may be exacerbated during changing positions or weight bearing (standing, sitting in place, etc). There may be “pseudo-radicular” referral patterns into the proximal legs and buttocks that have been described and validated via intra-articular injection or electronic stimulation of respective sensory nerves in the lumbar spine. 15 Referral patterns in the neck have also been demonstrated through similar means, with referral patterns into the occiput, periscapular area, shoulder, posterior/lateral neck, and upper thoracic regions. 16

As with the lumbar spine, cervical spondylosis is very common with increasing age and presence on imaging is found in a large number of asymptomatic patients. In the chronic neck pain population cervical facet pain prevalence is estimated to range from 49–61%. Presentation often includes referral patterns into the occipital region, traps, or proximal shoulders. 17

Pain is typically described as aching, deep, dull, and exacerbated by movement. Examination is non-diagnostic and must be taken into context with presentation and imaging, though no focal neurologic deficits should be apparent.

As is seen in the lumbar spine exercise and therapy focusing on posture, flexibility, and strength is the mainstay of treatment. Interventional treatment for cervical facet pain may be performed as an intra-articular injection (into the joint capsule), or by blocking the cervical medial branch nerves. Since the C1 nerve root exits cephalad to the atlas (C1), the medial branches in the cervical spine each correlate to their respective level (C3 and C4 medial branch innervated the C3–4 facet joint). Radiofrequency ablation (RFA) for the facet-mediated cervical spine pain has demonstrated excellent outcomes when ‘double block’ diagnostic paradigm (described below) is followed. 18

18.1.3 Sacro-Iliac Joint Pain

Like the above-mentioned facet joints, the sacro-iliac joint (SIJ) is an accepted source of low back, buttock, and proximal leg pain. Nevertheless, the etiology, epidemiology, and diagnostic criteria for SIJ pain is still debated among experts. Though the SI joint is diarthrodial like the facet joints; the SI afferent innervation is complex and continues to be a disputed topic with the lateral and ventral rami of the sacral spinal nerves both contributing to nociceptive and possibly neurogenic character of pain. Evidence supporting intra-articular mechanoreceptors within the joint complicates effective treatment.

Patients with SIJ-mediated pain often present with low back/buttock pain overlying the visible landmark of the posterior superior iliac spine (PSIS) and endorse referral patterns down the ipsilateral lateral and/or posterior-lateral thigh. Provocative joint injections reliably recreate concordant sclerotomal referral pattern that mimics patients usual pattern of exacerbation. 19 Numerous physical exam maneuvers exist. No particular maneuver possesses either high sensitivity or specificity, but a combination of several maneuvers are commonly employed in standard clinical practice to diagnose of SIJ mediated pain. 20 Some studies have suggested that if multiple exam maneuvers are positive (Gaenslens, distraction, compression, Patricks, Gillet’s) the positive-predictive value increases, but this topic remains unsettled.

Currently, intra-articular delivery of medication into the joint is the favored diagnostic method. The ‘gold standard’ to diagnose SIJ pain is still intra-articular injection. The specificity of this procedure varies widely in literature (40–100%). 20 The addition of good physical therapy intervention aimed at lengthening/stretching the iliopsoas muscle, lengthening the adductors, and strengthening the gluteus and paraspinals are the “mainstay” of treatment. Radiofrequency ablation of the SIJ has mixed results in the literature and is not performed with nearly the frequency of facet RFA, likely due to incomplete understanding or access to the innervation of the joint.

18.1.4 Sagittal Balance

Kyphosis and Degenerative Scoliosis

Older patient’s are often particularly sensitive to the appearance of kyphotic deformity “Dowager’s Hump.” The change in appearance may or may not have a painful, nociceptive component, but the cosmetic appearance can be a common reason that patients present to a spine specialist. It is important to remember that kyphosis and adult degenerative scoliosis are different entities from juvenile idiopathic scoliosis. The old biomechanical aphorism: “kyphosis begets kyphosis” remains true. At its most basic, physical training and physical fitness emphasizing extensor-muscle strength and endurance remains the ‘core’ to keep patient at full function. ‘Perfect’ radiological balance is less important than a patient’s ability to perform advanced ADL’s and maintain their independence. Often when degenerative scoliosis causes foraminal stenosis, judicious use of selective (transforaminal) nerve blocks and occasional micro-foraminotomies are safer than a large multi-level deformity surgery. A trans-disciplinary approach to deformity that starts with a shared functional goal with input from physical therapy, physical trainer, surgeon and interventional spine specialist can achieve greater functional outcomes with less harm to the patient.

18.2 Contraindications

The contraindications to interventional pain treatments can be divided into systemic and diagnostic. The systemic contraindications are as follows: allergy to iodinated contrast, or any other medication used in the intervention, kidney disease and poorly controlled diabetes. The diagnostic contraindications are treating the patient’s pain without a clear understanding of the diagnosis from a clinical perspective. Extreme care must be taken to avoid interventions on a patient that do not match the patient’s complaints. These so called “masqueraders” are discussed later in the “Pitfalls, Complications and Avoidance“ section of this chapter. Additionally, no interventional pain technique should be employed without the addition of adjunctive physical therapy treatments to strengthen and improve the range of motion of the surrounding tissues.

18.3 Technique(s) Description

18.3.1 Midline Epidural Steroid Injection

Lumbar interlaminar injections are often safely performed but even in patients with no surgical history there are a number of anatomical considerations that must take place. The target for these injections is in the epidural space (fat pad) that can be seen on sagittal and axial imaging in T1 sequence (See example). Although this space is usually present at each level in the lumbar spine, it is not uncommon for it to be absent or exceedingly small. It is best avoided in patients with prior laminectomies. In the cervical spine the ligamentum flavum (often used to give tactile feedback when performing the injection) becomes discontinuous cephalad to the C7-T1 space. In addition, the epidural fat pad is also less common as you move cephalad. For these reasons, it is uncommon to perform interlaminar epidurals above the C7-T1 level. Despite these limitations coverage of the entire epidural space has been demonstrated with as little as 3 mL of solution. Cervical interlaminar injections are the preferred method of delivering medicine into the epidural space and at their core are similar in technique to lumbar epidural steroid injections (Fig. 18‑1).

Fig. 18.1 Cervical Interlaminar Epidural Steroid Injection.

For either cervical or lumbar epidural steroid injections, the patient is placed in a prone position and the target level is visualized under fluoroscopy. Frequently, the intensifier (X-ray) is obliqued slightly to the side to avoid passing through the interspinous ligaments. The needle is then placed in line with the beam and carefully directed back towards the midline. A lateral or “contralateral oblique” view is best taken well before the needle nears the ligamentum flavum to allow for a reference of depth. On this view the needle tip will be near the epidural space and ligamentum flavum when it is at the spinolaminar line (convergence of the spinous process and lamina). As this landmark is approached one usually notes a distinct sensation from the needle that a “fascial or firm” layer is breached. Contrast should be utilized looking for a thin line running cephalad to caudad along the epidural space in lateral or contralateral view. On the AP view a “bubbly” appearance is indicative of epidural placement.

18.3.2 Transforaminal Epidural Steroid Injection

Transforaminal epidural injections are another delivery method into the epidural space with studies by Derby and Bogduk suggesting it delivers more medication into the ventral epidural space. 21 This technique classically involves placing the tip of the needle into the superior foramina just above the exiting nerve root. Care should be taken when planning this procedure as severe foraminal narrowing or prior fusion surgeries can make it technically challenging or nearly impossible. Approaching this injection requires obtaining a more oblique view in which the superior articular process and pedicle are clearly visualized. If you recall looking for the “scotty-dog” on oblique X-ray in the lumbar spine you want to place the needle “under the dog’s chin.” The needle should be advanced towards the target and both lateral and AP views are helpful in precisely locating the needle. On the AP view you should avoid placing the needle tip medial to the pedicle at risk of puncturing the dura, with the lateral view allowing you to gauge depth and ensure that the needle tip is within the intervertebral foramen. Once satisfied with the needle placement contrast is used and should demonstrate flow along the exiting nerve and superomedially past the pedicle into the epidural space (Figs. 18‑2,Fig. 18‑3) .

Fig. 18.2 Illustrated is an L4 transforaminal epidural steroid injection, AP view (a) and lateral view (b).
Fig. 18.3 Illustrated is an S1 transforaminal epidural steroid injection AP view (a) and lateral view (b).

18.3.3 Cervical Spine

Recently, the safety of transforaminal approach in the cervical spine has come into question by multiple case reports of catastrophic neurologic injury following cervical transforaminal epidurals. 22 For this reason, the authors do not recommend this approach in the cervical spine.

18.3.4 Intraarticular Facet Injections

Intra-articular facet injections (IAF) involves breaching the joint capsule with a needle and placing medicine (intra-articular) within the joint, using a small volume of medication to anesthetize the joint. Intra-articular facet injections are very technically demanding, 23 have not reliably demonstrated long term superiority, and demonstrate higher false positive outcomes than medial-branch facet blocks adhering to strict protocols. 24

The medial branches originate from the dorsal ramus of the spinal nerves and innervate the joint from above and below (Fig. 18‑4). Due to the presence of the C8 nerve in the cervical spine, the thoracic and lumbar joint levels do not correlate with their respective medial branches (eg The L3 and L4 medial branches supply the L4–5 facet joint). The diagnostic value of a single intra-articular facet (IAF) injections and single facet joint blocks consistently demonstrates a high false positive rate because of multiple potential nociceptive sources (pain generators) near the facet itself, prompting work on more reliable diagnostic protocols to ensure efficacious treatments. Because of this diagnostic inaccuracy the so called “double block” paradigm has been described to improve diagnostic specificity. The technique requires the use of low injection volume, precise placement during facet blocks, 25 and use of contrast material during two separate sessions (separated in time). The “double block” paradigm has been demonstrated to vastly improve diagnostic specificity and provide commensurately better therapeutic outcomes. 26 If a patient demonstrates marked improvement in pain scores and function using the methods noted above they are eligible for radiofrequency ablation in which a lesion is created along the medial branch (or dorsal ramus) allowing for a more extended effect from the procedure. Careful technique during the RFA and adherence to the “double block” paradigm mentioned above have demonstrated better outcomes with 60% of the patients obtaining at least 90% relief of pain at 12 months, and 87% obtained at least 60% relief at the same time interval. 27 It is the authors recommendation that RFA only be pursued with strict adherence to the “double-block” paradigm, and reported >80% improvement during the half-life of anesthetic with anesthetic medial branch block of facet(s).

Fig. 18.4 Medial-branch facet block, AP view.

18.3.5 Facet Radiofrequency Ablation

Radiofrequency ablation (RFA) can be performed in all segments of the spine, but is most commonly performed in the lumbar and cervical and will be discussed below only in these segments for the sake of brevity. The numbering system is discussed in the chapter above, but it is crucial to recall that below the cervical spine each joint receives half of its sensory innervation from the cephalad level (L3 and L4 medial branches supply the L4–5 joint.)

18.3.6 Lumbar Spine

The advent of new needle technology has allowed for differing techniques to ablate the nerve. The classic approach involves attempting to lie the needle as parallel as possible along the medial branches, in order to impart the largest ablation of the nerve. This is accomplished by lying the patient prone and tilting the fluoroscope caudal and oblique towards the side of interest. Using this approach the inferior to superior angle allows the needle to lie across the nerve imparting a larger lesion. The oblique trajectory is ideal when the view between the superior articular process (SAP) and transverse process (‘scotty dog’ ear and head respectively) is no longer acute ensuring the facet will not impede the trajectory of the needle. The needle tip is then directed just inferior to confluence between the SAP and the transverse process. Care should be taken not to slide over the target, towards the exiting spinal nerve. Placement of the needle for ablation of the L5 dorsal ramus is slightly different. Ideally, you will approach by again tilting the fluoroscope caudally, but the oblique angle is often limited by the ipsilateral iliac crest. The target for needle placement is the junction of the SAP and sacral ala.

Proper placement is ensured by checking the AP view to see that the needle terminates at the junction of the SAP and transverse process. On lateral view the needle should be seen near the midpoint of the SAP, and NOT inside the intervertebral foramen. Once the needles are in proper position motor stimulation is used to demonstrate mutlifidis firing (needles will twitch/bounce) and ensure no activation of the spinal nerve (movement/contraction of the leg). After proper and safe placement is confirmed radiofrequency ablation can commence.

18.3.7 Cervical Spine

The technique for cervical RFA typically involves two separate lesions as medial branches in the cervical spine are more variable than in the lumbar spine. 28 Generally speaking, the corresponding medial branch is targeted as it courses along the waist of the articular pillar (or just cephalad) just prior to innervating the corresponding joints. The patient is lying prone on the table with the fluoroscope tilted slightly caudally and slightly oblique towards the ipsilateral side. In this view the “waist” of each articular pillar should be visible and the needle placed in line with the beam, advancing so that it will contact at the waist. The needle should contact bone at this structure, with a true AP view demonstrating the needle contacting bone at or just above lateral margin of the lateral mass. If the needle tip is seen medial to the osseous structure it may be too anterior or posterior. On a true lateral view the needle should be seen near the midpoint of the lateral mass in terms of depth, perhaps most critically the needle should not be too far anteriorly as the vertebral artery or spinal nerve could be contacted. Again motor stimulation is used once the needle tip is contacting bone and is in a safe position on multiple views. Although generally less dramatic than in the lumbar spine, the needles should be seen moving slightly with muscle contraction. No motor stimulation into the arm should be observed. After proper and safe placement is ensured, radiofrequency ablation can commence. Given the slight variability of the cervical medial branches a second lesion is frequently performed by slightly varying the position of the needle.

18.3.8 SI Joint

Successful Injection of the sacro-iliac joint (SIJ) can be one of the more demanding procedures for an interventionist as the posterior joint has variable anatomy. For the purposes of this chapter we will discuss the recommended starting approach to access the SIJ. When looking at an AP view of the pelvis, both the anterior (lateral on radiograph) and posterior (medial on radiograph) joint are seen however the inferior portion of the joint will demonstrate overlap. In most cases successful cannulation of the space can occur by slightly positioning the image intensifier in an oblique position looking for a “hyperlucent” region within the SI joint with sharp demarcation of bony landmarks. The needle is then placed in line with the beam and advanced into the joint, often times minor adjustments are required to access the space. A lateral view should demonstrate that the needle is ventral to the posterior sacrum, if the needle is not beyond the periosteum of the sacrum, it is not within the joint. The primary risk during this procedure is advancing the needle ventrally (or inferiorly) through the joint space into the pelvic cavity, which although unlikely, has been reported.

18.4 Benefits and Risks

Benefits of pursuing injection based therapy in appropriately selected patients can be seen as they are relatively low risk, have minimal to no recovery time, and possible reduction in use of daily analgesic medications/polypharmacy in the elderly cohort. Injections should also be performed in conjunction with an adequate physical therapy regimen to show durable benefit.

Serious complications from interventions can be minimized by thorough preparation (obtaining and personally reviewing advanced imaging) and use of multiview fluoroscopy to pinpoint the location of the needle at all times. Perhaps the most common intraoperative complication is vasovagal reaction, reported between 1–2% in multiple studies. 28 , 29 , 30 In these same studies minor complications such as increased localized pain and headaches with cervical injections are reported in under 6% of cases. Catastrophic complications are thankfully rare, and with good technique can be minimized but never fully eliminated. Epidural injections have slightly increased risk over facet or SI joint injections in regards for risk of dural puncture, CSF leak, or nerve damage. With studies of over 43,000 facet blocks no dural puncture or CSF leak was noted. 29 Dural puncture is more often seen with Interlaminar approaches, but still is frequently reported as less than 1% across all epidural approaches in the lumbar and cervical spine. 29 , 30 , 31 Though exceedingly rare, case reports of spinal cord injury or CVA from direct puncture or particulate steroid injection into the artery of Adamkiewicz (thoracolumbar) or cervical vasculature supply the brain have prompted recommendations by multiple societies to eliminate deep sedation during procedures and to strongly consider non particulate steroids for epidural injections. Another potential safeguard in the cervical spine is employed by initially injecting lidocaine only after obtaining an epidurogram and waiting 30–60 seconds. If inadvertent vascular uptake to the brain is present, the patient may complain of perioral numbness/metallic taste, or seizures could be seen. This should prompt abortion of the procedure prior to introduction of any particulate (or “non-particulate”) steroid. The development of an epidural hematoma or infection is also a concern, but is again reported in less than 1% of injections. Clinically significant epidural hematoma is estimated to be as low as 1 in 190,000. 32 Finally, case reports of dropped head syndrome (cervical kyphosis) have been reported after cervical RFA. 33 A true incidence is hard to obtain, but this complication has led to requiring surgical stabilization and though exceedingly rare should be considered. The authors would recommend avoiding multiple levels, or bilateral procedures if possible to mitigate risk.

18.5 Outcomes/Evidence

18.5.1 Lumbar Stenosis

Studies have demonstrated epidural steroid injections are reasonable, ‘short-to-medium term’ treatment in patients with neurogenic claudication to improve function, walking distance and pain. 5 Studies have demonstrated 64% of patients deemed surgical candidates had subjective improvement in symptoms one-year after epidural injection and fewer patients electing for surgery in the treatment group. 6 More recent studies have demonstrated improvements in both numerical pain scores of pain and disability. 7

Interestingly, this data also parsed response to anesthetic alone (vs traditional combination of corticosteroid and anesthetic) and found no significant differences in response to treatment. Flexion-based exercise protocols may be recommended for patients with neurogenic claudication. Although active lifestyle and aerobic conditioning should be encouraged in all patients, there is inconsistent evidence that specific rehabilitation programs are effective in completely eliminating the return of symptoms. Given enough time in the best prospective cross-sectional study, nearly 7/10 patients randomly assigned to non-surgical group will opt to ‘cross-over’ for surgery to alleviate claudication symptoms. 8 Different interventional approaches to delivering medication to the epidural space remain debated amongst specialists. Comparing transforaminal vs interlaminar epidural injection techniques has not reliably shown either approach to be significantly superior for treatment durability in symptomatic lumbosacral stenosis.

Only gold members can continue reading. Log In or Register to continue

Aug 1, 2020 | Posted by in NEUROSURGERY | Comments Off on 18 Interventional Pain Treatment Options
Premium Wordpress Themes by UFO Themes