Ultrasound-Guided First Annular Pulley Injection for Trigger Finger

Transcription

1 Article Ultrasound-Guided First Annular Pulley Injection for Trigger Finger Marko Bodor, MD, Tiffany Flossman Objective. The purpose of this study was to develop an ultrasound-guided first annular (A1) pulley injection technique for trigger finger with documentation of outcomes at 1 year. Methods. We performed a short-axis injection into a triangle bordered by the A1 pulley, the flexor digitorum superficialis and profundus tendons and volar plate, and the distal metacarpal bone with a 10-mg median dose of triamcinolone acetonide and 2% lidocaine. This was a prospective study of 50 of 52 consecutive trigger fingers from 24 patients recruited from a physical medicine and rehabilitation private practice. Results. All patients were available for follow-up, with 94% (47 of 50) of fingers having complete resolution of symptoms at 6 months, 90% (37 of 41) at 1 year, 65% (17 of 26) at 18 months, and 71% (12 of 17) at 3 years after a single injection. Conclusions. Our ultrasound-guided A1 pulley injection technique is a highly effective and minimally invasive treatment option for trigger finger with a 90% success rate at 1 year for complete resolution of symptoms after a single injection. Assuming similar patient populations, our results were statistically significant (P <.01) compared with the 56% to 57% success rates recently reported for blind injections. Key words: first annular pulley; injection; triamcin - olone acetonide; trigger finger; ultrasound. Abbreviations A1, first annular; FDP, flexor digitorum profundus; FDS, flexor digitorum superficialis; MCP, metacarpophalangeal Received January 29, 2009, from Queen of the Valley Medical Center, Napa, California USA. Revision requested February 3, Revised manuscript accepted for publication February 24, We thank Rahman Azari, PhD, from the Department of Statistics, University of California, Davis, California, for statistical analysis, Tish Green, medical librarian at Queen of the Valley Medical Center, for references, and Daniel Bodor, MD, for review. Results of the first 18 trigger finger injections were presented by Albert Retodo, MD, and Marko Bodor, MD, as a peer-reviewed poster at the American Academy of Physical Medicine and Rehabilitation 67th Annual Assembly; 2006; Honolulu, Hawaii. Address correspondence to Marko Bodor, MD, Queen of the Valley Medical Center, 3421 Villa Ln, Suite 2B, Napa, CA USA. mbodormd@sbcglobal.net Trigger finger is a common hand problem, with a lifetime prevalence of 2.6% in the general population and 10% among those with diabetes. 1 Symptoms range from a vague sense of tightness in the fingers or pain in the palm of the hand to overt locking and triggering. Tenderness is usually present at the first annular (A1) pulley, where there is an increase in friction or a mismatch in size between the flexor tendons and pulley. The A1 pulley consists of annular bands of connective tissue located at and proximal to the metacarpophalangeal (MCP) joint and contiguous with the tendon sheath (Figure 1). 2 The mean length of the A1 pulley is 12 mm for the adult index, middle, and ring fingers and 10 mm for the little finger. 2 Ultrasound imaging findings of trigger finger may include diffuse hypoechoic thickening of the A1 pulley, swelling of the tendons, synovial sheath effusion, and dynamic changes in the shape of the synovial sheath during flexion and extension by the American Institute of Ultrasound in Medicine J Ultrasound Med 2009; 28: /09/$3.50

2 Ultrasound-Guided First Annular Pulley Injection for Trigger Finger Corticosteroid injections are considered firstline treatment for trigger finger. 4 These are typically done by inserting a needle through the skin and A1 pulley and piercing or making contact with the flexor tendons before injecting a mixture of a corticosteroid and lidocaine. 5,6 In their evidence-based review, Fleisch et al 4 combined the results of 4 prospective randomized trials 6 9 and found that a single blind trigger finger injection had a 57% success rate at improving or resolving trigger finger or thumb symptoms at 1 to 4 months to 1 year or more compared with 17% for a placebo. Similarly, Peters-Veluthamaningal et al 10 found that 1 or 2 blind 10-mg injections of triamcinolone acetonide resulted in complete resolution of pain and triggering in 56% of fingers at 1 year compared with 20% at 1 week for a placebo. We have not found any studies assessing the efficacy of ultrasound-guided corticosteroid injections. Godey et al 11 described a long-axis technique but did not publish outcomes. We raised the hypothesis that ultrasound-guided trigger finger injections would have a higher success rate than blind injections. Two studies have shown that blind injections miss their intended target, the flexor tendon sheath within the A1 pulley, at least 50% of the time. 9,12 Taras et al 9 claimed, however, that the location of the injection did not matter, citing no statistically significant difference in success between injections given under the A1 pulley or subcutaneously. Figure 1. Annular and cruciate (C) pulleys of the index finger. The proximal aspect of the A1 pulley is at the level of the distal metacarpal bone. Reprinted with permission from Chase RA. Atlas of Hand Surgery. Vol II. Philadelphia, PA: WB Saunders Co; 1984:145. We thus endeavored to develop an ultrasoundguided A1 pulley trigger finger injection technique with the goals of accuracy, minimal invasiveness as defined by avoidance of piercing the tendons, and high efficacy. Materials and Methods Technique On axial ultrasound views, the A1 pulley is hypoechoic and shaped like an inverted parabola overlying the flexor digitorum superficialis (FDS) and flexor digitorum profundus (FDP) tendons and volar plate (Figure 2). In thumbs, the A1 pulley has a more circular shape because of only 1 tendon present, the flexor pollicis longus. Our target for injection is a triangle under the A1 pulley whose borders consist of the FDS and FDP tendons and volar plate, the distal metacarpal bone, and the pulley (Figure 2). We typically identify the flexor tendons in an axial view at the level of the proximal phalanx and scan proximally, noting how the underlying concave surface of the proximal phalanx gives way to the convex surface of the distal metacarpal bone as we cross the MCP joint. At this location, we see the A1 pulley, noting any abnormalities such as thickening of the pulley or tendon nodules, and assess its function dynamically by having the patient flex and extend the finger. Figure 2. Short-axis view of the normal A1 pulley in the right middle finger. Open arrow indicates the A1 pulley and needle trajectory; asterisk, middle of the target triangle and ideal location for the needle tip for injection; dotted arrows, digital nerves; L, lumbrical muscle; M, metacarpal; and VP, volar plate. The bifurcation of the common digital artery is shown as 2 small hypoechoic dots below the digital nerve on the left. The common digital artery is shown as the larger hypoechoic dot below the bifurcating digital nerve on the right. The image is from an unaffected volunteer not in the study. 738 J Ultrasound Med 2009; 28:

3 Bodor and Flossman We then identify the target triangle and place it in the center of the screen or slightly to the left of center for someone injecting with the right hand. It does not matter whether the triangle on the radial or ulnar side of the tendons is selected. We prepare the skin distal to the transducer with 70% isopropyl alcohol. We are careful to maintain a small but sterile field just distal to the edge of the transducer. Because we do not use a sterile transducer cover or gel, the needle and puncture site are never allowed to contact the ultrasound transducer or transmission gel. We use a distal-to-proximal short-axis or outof-plane approach. We plan a trajectory to the hypotenuse of the triangle via an approximately 70 angle to horizontal in the axial plane and a 45 angle in the sagittal plane (see angle of open arrow in Figure 2, photographs in Figure 3, and angles of arrows in Figures 4 and 5). We use a 30- gauge needle and 0.25 ml of 2% lidocaine to anesthetize the skin and then insert a 27- or 30- gauge needle and direct it to the triangle using real-time ultrasound visualization. When the tip of the needle is within the target triangle (Figures 4 and 5), we inject approximately 1 ml of 10-mg triamcinolone acetonide with 2% lidocaine, making sure to visualize flow under the A1 pulley. If we obtain flow outside the pulley or there is no flow, we adjust the needle until flow is obtained. Sometimes we note initially high resistance to outflow followed by a steep drop in resistance accompanied by simultaneous distention of the pulley (Figure 6). The pulley can be tough to penetrate, and the needle may clog, requiring insertion of another, possibly larger-gauge, needle. Afterward, a small sterile adhesive dressing is applied, and the patient may resume activities as tolerated. Figure 3. First annular pulley injection of the right middle finger: front (A) and side (B) views. The tip of the needle does not touch the transducer or gel. A Figure 4. Needle within the target triangle. The tip of a 27- gauge needle (arrow) is shown as a bright hyperechoic oval entering the top of the target triangle. The angle of the arrow reflects the angle of the needle in the coronal plane. The image is from a patient in the study. B Figure 5. Needle within the target triangle. The tip of a 30- gauge needle (arrow) is shown as a hyperechoic dot next to the FDP tendon and volar plate. The angle of the arrow reflects the angle of the needle in the coronal plane. The image is from a patient not in the study. J Ultrasound Med 2009; 28:

4 Ultrasound-Guided First Annular Pulley Injection for Trigger Finger Figure 6. First annular pulley before (A) and after (B) injection with triamcinolone acetonide and lidocaine. Distention of the pulley is noted (arrows). The images are from a patient not in the study. A B We find it technically more difficult to inject the thumb because of its oblique axis, orientation toward the palm of the hand, and smaller and rounder profile of the pulley, making it harder to visualize and approach the target. We find it easier if an assistant holds the thumb in extension. We used an 8- to 16-MHz transducer (Diasus; Dynamic Imaging Ltd, Livingston, Scotland) or a 10- to 15-MHz transducer (MyLab 25; Biosound Esaote, Florence, Italy) for our study patients. Since completing our recruitment of patients for this study, we have used a 5- to 17-MHz transducer (iu22 ultrasound system; Philips Healthcare, Bothell, WA). We think that the critical anatomy pertinent to the procedure is best illustrated with this equipment. Therefore, several of our images (Figures 2, 5, and 6) were obtained with this equipment from patients not enrolled in the study. We have found that the use of a transducer with a narrower head and less tapered (more square than V- or U-shaped) side profile improves our ability to accurately place the needle. Statistical analysis was performed with a 2- group continuity-corrected χ 2 test with a 2-sided significance level of P <.05 having 99% power to detect the difference between a group 1 proportion of 0.9 and a group 2 proportion of 0.56 (odds ratio, 7.071), when the sample size in group 1 was 41 and that in group 2 was 137, and 80% power to detect the difference between a group 1 proportion of 0.9 and a group 2 proportion of 0.56 (odds ratio, 7.071), when the sample size in group 1 was 41 and that in group 2 was 32. Patients and Setting The setting of our study was a physical medicine and rehabilitation, sports, and electrodiagnostic medicine private practice. The study was approved by a local community Ethics and Human Research Committee. The procedure was explained in detail to each patient, and informed consent was obtained. Our primary outcome measure was complete resolution of pain, stiffness, and triggering at 1 year. The investigator not performing the injections (T.F.) called all of the patients to assess treatment responses at 6 months, 1 year, 18 months, and 3 years. The patients were asked how their finger was and whether they had any pain, stiffness, or triggering. A positive response to any of the latter was considered a treatment failure. Completeness of follow-up was facilitated by establishing a separate telephone line for patients to call. Results Starting in 2003 until the middle of 2008, we prospectively studied 50 of 52 consecutive trigger fingers in 24 of 26 patients, 14 men and 10 women ranging in age from 29 to 92 years (mean, 66 years; SD, 15.6 years) with pain, uneven movement, triggering, or locking. Two fingers were not included in the study because 1 patient declined the injection and the other had previous trigger finger surgery, our only exclusion criteria. Using the Quinnell scale, 13 (0, normal movement; 1, uneven movement; 2, actively correctable locking; 3, passively correctible locking; and 4, fixed deformity of the digit), 34% (17 of 50) of our trigger fingers were grade 1, 56% (28 of 50) grade 2, 10% (5 of 50) grade 3, and 0% (0 of 50) grade 4. Resolved trigger fingers that had recurrent symptoms after 1 year were eligible for another injection and to be included again in the study: J Ultrasound Med 2009; 28:

5 Bodor and Flossman patients each had 1 middle finger injected a second time; 1 had a little finger injected 3 times; and 1 had a little finger injected 4 times. Resolved trigger thumbs were eligible for a second injection after 6 months: 3 thumbs were injected twice, and 1 was injected 3 times. Concurrent diagnoses included carpal tunnel syndrome (58%), spinal stenosis or degenerative disk disease (42%), osteoarthritis (36%), hypertension (36%), shoulder dysfunction (22%), gastrointestinal reflux disease (19%), hypothyroidism (16%), hypercholesterolemia (13%), tendonitis (11%), coronary artery disease (11%), asthma (8%), diabetes mellitus (8%), neuropathy (8%), prostate enlargement (5%), and gout (2%). Trigger finger was the chief condition in 72% of fingers and the secondary condition in 10% of those with carpal tunnel syndrome and 6% each of those with wrist and shoulder pain and spinal stenosis. Some patients had multiple fingers injected on the same day and others months or years apart. Each finger received a single injection of triamcinolone acetonide (Kenalog-40; Bristol-Myers Squibb Company, New York, NY) and 2% lidocaine HCl (Hospira, Inc, Lake Forest, IL) from the lead investigator (M.B.) or one of his fellows. The median triamcinolone dose was 10 mg, with a mean of 9 mg (SD, 4.8 mg) and a range of 2.5 to 20 mg. We used, for example, 2.5 mg per finger in a diabetic man with multiple finger involvement and 20 mg in a 188-cm, 100-kg man with a large finger. We were able to visualize good flow under the A1 pulley and around the flexor tendons in all fingers. In a number of our patients, the initial flow pattern was above the pulley, so some of the medication would have been deposited there. Also, while attempting to get flow within the target triangle, it was possible that some medication could have been deposited within the substance of the pulley. We also injected thumbs but did not obtain good flow in all of them and will present those results separately. We were able to obtain follow-up in 100% (24 of 24) of our patients. We noted complete resolution of symptoms in 94% (47 of 50) of fingers at 6 months, 90% (37 of 41) at 1 year, 65% (17 of 26) at 18 months, and 71% (12 of 17) at 3 years. We also injected 24 consecutive trigger thumbs in 15 patients, 8 men and 7 women ranging in age from 45 to 95 years (mean, 64 years; SD, 15.5 years), and obtained follow-up in 14 of them. We noted complete resolution of symptoms in 78% (18 of 23) of thumbs at 6 months, 55% (12 of 22) at 1 year, and 31% (4 of 13) at 18 months. Our 90% success rate in 41 trigger fingers at 1 year was statistically significant (P <.01) compared with the 57% success rate in 137 fingers described by Fleisch et al 4 and the 56% success rate in 32 fingers noted by Peters-Veluthamaningal et al. 10 We noted no complications in any of our patients or reports of excessive pain or discomfort during or after the procedures. Discussion We achieved our goal of an accurate, minimally invasive, and effective ultrasound-guided A1 pulley trigger finger injection technique. We used fine needles and a relatively low dose of corticosteroid. We did not puncture the tendons, as is typically done to confirm placement when doing the procedure blindly. 5,6 Our 94% success rate at 6 months and 90% at 1 year were statistically significant compared with the 57% rate described by Fleisch et al 4 and the 56% rate noted by Peters-Veluthamaningal et al, 10 assuming similar patient populations. We did not note any effect of concurrent diagnoses on treatment outcome, although our number of failures was too low to achieve adequate power to do so. Our trigger thumb success rates were lower at 79% at 6 months and 55% at 1 year. This may have been due to the greater difficulty of obtaining good flow under the A1 pulley of the thumb, too low of a medication dose, or intrinsic differences in anatomy and abnormalities. In a comparison of our results with those of previous studies, the study by Taras et al 9 presents the greatest contrast, challenging the hypothesis that ultrasound-guided trigger finger injections should be more effective than blind injections. They noted a 70% (45 of 64) success rate for blind unconfirmed subcutaneous injections, 50% (12 of 24) for blind fluoroscopically confirmed mixed subcutaneous and intrasheath injections, and 47% (9 of 19) for blind fluoroscopically confirmed intrasheath injections. 9 J Ultrasound Med 2009; 28:

6 Ultrasound-Guided First Annular Pulley Injection for Trigger Finger Although their results were not statistically significant, they said that subcutaneous injections might be more effective by concentrating medication closer to the site of the abnormality. Interestingly, in their study, no patient changed categories (good, fair, or poor) between 2 and 4 weeks and 6 months after their injections. Their final follow-up ranged from 10 to 60 months, complicating direct comparisons with our study. By contrast, we noted a gradual decline in the effect of treatment with time such that 94% of our fingers were asymptomatic at 6 months, 90% at 1 year, 65% at 18 months, and 71% at 3 years. We used an average corticosteroid dose that was one-fourth the equivalent used by Taras et al. 9 Another interesting comparison is with the double-blind randomized controlled study of Peters-Veluthamaningal et al. 10 They provided 1 or 2 injections of 10-mg triamcinolone acetonide and, citing Taras et al, 9 did not attempt to place the medication directly under the A1 pulley, and they noted a 56% (18 of 32) success rate at 1 year. The main limitation of our study was the lack of certainty as to how similar our patients were compared with those of other studies, 4 10 which did not include information on the Quinnell scale. 13 Approximately one-third of our fingers were Quinnell grade 1, characterized by uneven movement, pain, or stiffness as well as tenderness at the A1 pulley but not locking, and none were Quinnell grade 4, characterized by fixed deformity. We are unaware of any studies, however, showing a correlation between the severity of triggering and the efficacy of injections. Ideally, we would have included a blind injection cohort; however, we did not do so because the number of trigger finger injections in our practice was relatively low ( 1 per month), and for ethical reasons and our reputation in the community, we wanted to provide each patient with the best possible injection and at the same time gain as much experience with the technique as possible. Additional limitations included the lack of verification of needle placement via an independent means, such as injection of a contrast agent followed by fluoroscopy or injection of methylene blue in cadavers followed by dissection. Furthermore, our procedures were either performed or supervised by a single person, thus precluding evaluation of interobserver differences. Finally, the number of patients in our study was only half that of a number of studies 8 10 and about the same as in others or their active arms. 6,7 Histologically, the normal A1 pulley consists of 2 layers, an outer loose connective tissue layer and an inner dense connective tissue layer. In trigger fingers, a third layer is present, consisting of an innermost irregular connective tissue layer formed by small collagen fibers, abundant extracellular matrix, and chondroid-metaplasia. 14 The mechanism of action of corticosteroid injections for trigger finger is unknown. We hypothesize that corticosteroids may weaken the A1 pulley such that ongoing mechanical forces result in stretching and lengthening of the pulley, degrade chondroid metaplasia within the pulley, reduce focal swelling within the tendons, reduce friction by altering the surface properties of the tendons and pulley, or a combination thereof. Our injections targeted the proximal aspect of the A1 pulley at the level of the MCP joint and distal metacarpal bone. We hypothesize this to be the tighter, more stenotic side, given that triggering occurs as the FDS and FDP tendons pass via the proximal aspect of the pulley as the finger is moved from flexion to extension. Although our success rate for trigger fingers at 1 year was 90%, at 18 months to 3 years it dropped to 60% to 70%. We are considering how to improve our long-term success rate. We might increase the triamcinolone dose or inject on both the inner and outer sides of the pulley or within its substance to accentuate the effect of the medication. For recurrent trigger finger, there is always the possibility of a second ultrasound-guided injection. For multiple recurrences, we hope to develop an ultrasound-guided proximal A1 pulley release in the future. In conclusion, our ultrasound-guided A1 pulley injection technique resulted in a 90% success rate for complete resolution of pain, stiffness, and triggering in 41 trigger fingers 1 year after a single injection. Our results compared favorably with the 56% to 57% success rates reported by Fleisch et al 4 and Peters-Veluthamaningal et al 10 for blind injections and were statistically significant assuming a similar patient population. In the future, we would like to see a double-blind randomized controlled trial comparing ultrasound-guided and blind trigger finger injections. 742 J Ultrasound Med 2009; 28:

7 Bodor and Flossman References 1. Akhtar S, Bradley MJ, Quinton DN, Burke FD. Management and referral for trigger finger/thumb. BMJ 2005; 331: Wilhelmi BJ, Snyder N IV, Verbesey JE, Ganchi PA, Lee WP. Trigger finger release with hand surface landmark ratios: an anatomic and clinical study. Plast Reconstr Surg 2001; 108: Serafini G, Derchi LE, Quadri P, et al. High resolution sonography of the flexor tendons in trigger fingers. J Ultrasound Med 1996; 15: Fleisch SB, Spindler KP, Lee DH. Corticosteroid injections in the treatment of trigger finger: a level I and II systematic review. J Am Acad Orthop Surg 2007; 15: Freiberg A, Mulholland RS, Levine R. Nonoperative treatment of trigger fingers and thumbs. J Hand Surg [Am] 1989; 14: Lambert MA, Morton RJ, Sloan JP. Controlled study of the use of local steroid injection in the treatment of trigger finger and thumb. J Hand Surg [Br] 1992; 17: Murphy D, Failla JM, Koniuch MP. Steroid versus placebo injection for trigger finger (published erratum appears in J Hand Surg [Am] 1995; 20:1075). J Hand Surg [Am] 1995; 20: Maneerit J, Sriworakun C, Budhraja N, Nagavajara P. Trigger thumb: results of a prospective randomised study of percutaneous release with steroid injection versus steroid injection alone. J Hand Surg [Br] 2003; 28: Taras JS, Raphael JS, Pan WT, Movagharnia F, Sotereanos DG. Corticosteroid injections for trigger digits: is intrasheath injection necessary? J Hand Surg [Am] 1998; 23: Peters-Veluthamaningal C, Winters JC, Groenier KH, Meyboom-de Jong B. Corticosteroid injections effective for trigger finger in adults in general practice: a double-blinded randomized placebo controlled trial. Ann Rheum Dis 2008; 67: Godey SK, Bhatti WA, Watson JS, Bayat A. A technique for accurate and safe injection of steroid in trigger digits using ultrasound guidance. Acta Orthop Belg 2006; 72: Kamhin M, Engel J, Heim M. The fate of injected trigger fingers. Hand 1983; 15: Quinnell RC. Conservative management of trigger finger. Practitioner 1980; 24: Sbernardori MC, Bandiera P. Histopathology of the A1 pulley in adult trigger fingers. J Hand Surg Eur Vol 2007; 32: J Ultrasound Med 2009; 28: