6 Extracorporeal Shock Wave Application in the Treatment of Chronic Tennis Elbow Introduction The causes of tennis elbow remain unclear (Chard and Hazleman 1989, Pienimäki et al. 1998). While most authors attribute pain at the lateral epicondyle to overstrain of the insertion of the m. extensor carpi radialis bre- vis and consequent local angiofibroblastic tendinosis (Finestone and Helfenstein 1994, Foley 1993, Kraushaar and Nirschl 1999, Roe- tert et al. 1995), there are reports suggesting a reflex chain between intervertebral joint dys- function and peripherally localized soft-tissue pain syndromes (Sutter 1995, Waldis 1989, Wanivenhaus 1986, Wyke 1979). Wright et al. (1994) write that neuronal changes within the spinalcordmightbemoreimportantthan peripheral nociceptor sensitization in the development of chronic musculoskeletal dis- orders such as tennis elbow. This is in accor- dance with reports of an association between lateral epicondylitis and a dysfunction in the cervical spine and at the cervicothoracic junc- tion in more than 80% of patients (Ehmer 1998). Cyriax (1982), however, argues that at the age of 40-60 years it is hightly probably that any patient suffering from chronic lateral epicondylitis would have radiographical evi- dence of cervical spondylosis as well. He denies that pain in the elbow provoked by wrist movements could have the neck as its origin. Labelle et al. (1992) conclude that con- servative procedures in tennis elbow lacked any scientific basis. Boyer and Hastings (1999) also find no conclusive studies on operative and nonoperative treatment concepts. Newtreatmentmethodshavesincebeen undergoing trials. The finding that physical stimuli are capable of activating endogenous nociceptive control systems has led to the use of shock waves in the treatment of persistent tennis elbow (Rompe et al. 1996a). Extracor- poreal shock wave therapy (ESWT) was said to fulfill major properties of hyperstimulation analgesia, but the exact mechanism of pain reduction produced by ESWT is still unknown (Melzack 1994). Nevertheless, success rates of 50% were achieved in prospective, con- trolled studies on recalcitrant tennis elbow and plantar fasciitis (Heller and Niethard 1998). The current study reports a comparison of patients treated with shock waves and manual therapy of the cervical spine with age- matched cases receiving only shock wave therapy (SWT) in a single unit (Rompe et al. 2001a). Fig. 6.1 Treatments received before ESWT. LA: Local anesthetic. Materials and Methods Included in this study were patients present- ing with chronic lateral epicondylitis of the elbow in our university hospital. Inclusion Criteria Inclusion criteria were: pain over the lateral epicondyle for more than 6 months; unsuc- cessful conservative therapy during the 6 months prior to referral to our hospital; at least 10 physical therapy visits (electrotherapy, ion- tophoresis,cryotherapy,orultrasound)(Fig. 6.1); at least three local injections (steroid and/ or local anesthetic) (Fig. 6.2); at least two of the following provocation tests positive: 1. palpa- tion of the lateral epicondyle; 2. resisted wrist extension(Thomsentest:withtheshoulder flexed to 60°, the elbow extended, the forearm pronated, and the wrist extended about 30°, pressure is applied to the dorsum of the second and third metacarpal bones in the direction of flexion and ulnar deviation to stress the m. extensor carpi radialis brevis and longus); 3. resisted finger extension (with the shoulder flexed to 60°, the elbow extended, the forearm pronated, and the fingers extended the middle finger is actively extended against resistance); 4. Chair test (with the shoulder flexed to 60°, and the elbow extended the patient attempts to lift a chair weighing 3.5kg); signs of cervical dysfunction with persistent pressure pain at the C4/C5 and/or C5/C6 level, protraction posi- tion of the head. Fig. 6.2 Atrophy of the skin after multiple corticoste- roid injections. 6 Extracorporeal Shock Wave Application in the Treatment of Chronic Tennis Elbow40 Exclusion Criteria Exclusion criteria were: age less than 18 years; pregnancy; previous operations on the lateral epicondyle; previous manual therapy to the cervical spine; bilateral epicondylitis; osteoarthritis of the elbow joints; pathologi- cal, neurological, and/or vascular findings of the upper extremity; provocation of pain in the lateral elbow when examining the cervical spine; local infection of the upper extremity; tumorous disease of the upper extremity; rheumatoid arthritis; coagulopathy; other treatments or drugs used in the 6 weeks beforethetrialsbeganandduringthefirst3 months after ESWT. Group I Group I comprised 30 patients suffering from lateral elbow pain and neck pain. All patients were referred to our shock wave unit for treat- ment of a recalcitrant epicondylitis. These patients received SWT and manual therapy to the cervical spine. There were 16 women, 14 men, with a mean age of 47 years (range: 35–65 years) and a mean duration of pain of 38 months (range: 12–180 months). There were 27 right-handed individuals; the dominant side was affected in 25 cases. A mean of 5.4 ± 1.5 conservative treatment procedures had been carried out without success. The patients were not able to workforanaverageof4.4±7.2weeksbefore the first ESWT. Group II For each elbow studied, a control matched by age (3-year band) and sex at first conservative treatment was drawn at random from a series of 146 patients who had undergone a mono- therapy with low-energy shock waves in the same unit in the past 3 years (group II). There were 16 women and 14 men, with a mean age of 48 years (range: 37–68 years) and ameandurationofpainof40months(range: 12–208 months). There were 26 right-handed patients; the dominant arm was affected in 27 patients. An average of 5.5 ± 1.8 unsuccessful conservative therapy procedures had been carried out. The mean period of inability to work was 4.5 ± 8.9 weeks before the first ESWT started. Method of Treatment The ESWT was applied by an easily maneuver- able therapy unit especially designed for orthopedic use (Sonocur Plus, Siemens AG, Erlangen, Germany), with the shock wave head suspended from an articulating arm for flexible movement of the head in three planes. The shock wave head was equipped with an electromagnetic shock wave emitter. Shock wave focus guidance was established by in- line integration of an ultrasound probe—a 7.5 MHz sector scanner—in the shock head. According to the consensus report (Wess et al. 1997) the features of the device, measured with a fiberoptic hydrophone, are listed in Table 6.1. Both groups were treated under the same conditions and the patients were treated sin- gly to avoid influencing one another. Three times, at weekly intervals, 1000 impulses of 0.16 mJ/mm 2 were administered at the ante- rior aspect of the lateral epicondyle at a fre- quency of 4 Hz (Fig. 6.3). No local anesthesia was applied to the treated area, although the treatment is moderately painful. After the last ESWT, patients from group I were referred to physiotherapists certified for manual therapy who had been instructed to perform soft mobilization therapy of the cer- vical spine and of the cervicothoracic junction to relieve pain in the C4/C5 and C5/6 motion segments, and to correct the observed pro- traction of the head due to an increased kyphosis of the neck (Butler 1995). Therefore, extension mobilization of the cervicothoracic Method of Treatment 41 Table 6.1 Fiberoptic data on the shock wave device 1 Physical Value Unit Energy level 1 (Minimum value) Energy level 2 (Mean value) Energy level 3 (Maximal value) Peak positive pressure P + Mpa 5.5 14.2 25.6 –6 dB focal extend in x,y,z direction f x(-6dB) f y(-6dB) f z(-6dB) mm mm mm 6.0 6.0 58 5.2 5.2 55 4.8 4.8 49 5MPafocal extent,lateral f x(5 Mpa) f y(5 Mpa) mm mm 2.2 2.2 7.8 7.8 19 19 Positive energy flux density ED + mJ/mm 2 0.016 0.09 0.22 Total energy flux density ED mJ/mm 2 0.04 0.24 0.56 Positive energy of –6dB focus E +(-6 dB) mJ 0.38 1.6 3.5 Total energy of –6 dB focus E (-6 dB) mJ 1.1 4.0 9.0 Positive energy of 5MPa focus E +(5 Mpa) mJ 0.5 1.8 9.2 Total energy of 5 MPa focus E (5 Mpa) mJ 1.8 4.8 24 Positive energy of 5mm focal area E +(5 mm) mJ 0.24 1.26 3.5 Total energy of 5mm focal area E (5 mm) mJ 0.63 3.0 9.0 Rise time t r ns 750 615 481 Pulse width t p+ ns 1380 1160 920 1 The Sonocur Plus provides eight user-selectable energy levels. The physical data listed in the table are typical values for the minimum energy, medium energy, and maximum energy. The shock wave param- eters are described according to the consensus meet- ing in February 1997 (Wess et al. 1997). All measure- ments were made using a fiberoptic hydrophone. Fig. 6.3 ESWT with the ultrasound-guided Sonocur Plus device. Fig. 6.4 Mobilization of the cervical spine. 6 Extracorporeal Shock Wave Application in the Treatment of Chronic Tennis Elbow42 Fig. 6.5 Evaluation of grip strength with the JAMAR hand dynamometer. junction (e.g., Maitland (1991) grade IV) (Fig. 6.4) and flexion mobilization of the high cervical joints was recommended (e.g., Mait- land grade IV). For the intermediate cervical segments traction and glide movements were suggested (e.g., Maitland grade II or III minus). No therapeutic procedures were to be admin- istered to the lateral elbow. Ten treatment sessions were held after the last ESWT. Follow-ups, by an independent observer, were done 3 months and 12 months after the last application of the extracorporeal shock waves. Method of Evaluation At all follow-ups the patients were asked abouttheirpainassessmentcomparedtopre- treatment conditions. The total outcome was rated following Roles and Maudsley (1972): —Excellent: No pain, full movement, full activity; —Good: Occasional discomfort, full movement, full activity; —Acceptable: Some discomfort after pro- longed activities; —Poor: Pain limiting activities. The Roles and Maudsley outcome score at the 12-month follow-up was defined as the main outcome measure. The extent of pain was specified using a Visual Analogue Scale (VAS) ranging from 0, i.e., no pain, to 10, i.e., maxi- mal pain. The examination was carried out independently of the treating physician and comprised the same four diagnostic tests that determined entry into the study. Additionally, grip strength was measured bilaterally at the extended and pronated forearm with a vigori- meter (Jamar Dynamometer, Preston Health- care,Jackson,UnitedStates),thepressure being registered in kp/cm 2 . Reduction of pain andgripstrengthcomparedtotheunaffected side were regarded as secondary outcome parameters (Fig. 6.5). Statistics For statistical analysis, the Wilcoxon–Mann– Whitney test for two independent samples, the t-test for the normally distributed vari- ables of the vigorimeter measures, and the Fisher exact test and its extension to r× c tables were used to compare the two groups. Comparison between different examinations were done by means of the Wilcoxon test and t-test, respectively, for dependent samples, and the McNemar test. The level of signifi- cancewassetat95%.Testedcomparisons with p-values 5%wereconsideredtobesig- nificantly different. Method of Evaluation 43 Table 6.2 Total outcome according to Roles and Maudsley 0 months 3-month follow-up 12-month follow-up Group I (n=30) Group II (n=30) Group I (n=30) Group II (n=30) Group I (n=27) Group II (n=25) Excellent – – 13.3% 23.3% 22.3% 24.0% Good – – 26.7% 26.7% 33.3% 36.0% Acceptable – – 43.3% 33.3% 33.3% 20.0% Poor 100% 100% 16.7% 16.7% 11.1% 20.05 Table 6.3 Pain rating on the Visual Analogue Scale (VAS) 1 0 months 3 months 12 months Group I Group II p-value Group I Group II p-value Group I Group II p-value Pressure pain 6.21± 2.65 6.15 ± 2.43 0.28 3.99 ± 2.74 3.59 ±2.29 0.65 2.27 ± 2.59 1.97± 2.05 0.82 Thomsen test 6.18± 1.72 6.24 ± 1.74 0.60 3.69 ± 2.52 3.86 ± 2.28 0.55 1.93 ±1.97 2.09 ± 2.01 0.71 Resisted finger extension 4.62 ± 3.29 4.97 ± 2.84 0.18 2.77 ± 2.29 3.01 ± 3.32 0.58 1.45± 1.84 1.66 ± 1.79 0.57 Chair test 5.46 ± 2.11 5.59 ± 2.13 0.54 2.98 ± 2.46 3.00± 2.40 0.81 1.91± 2.51 1.97 ± 2.27 0.76 1 mean ± standard deviation Results There was no difference between the groups concerning the affected side, sex, age, period of pain, period of inability to work, and num- ber of conservative treatment procedures. All the patients were examined at 3 months. Twelve (40%) patients in group I and 15 (50%) patients in group II reached an “Excellent” or “Good” result. Additionally, the patients were asked to estimate the improvement of pain in percent, 0% meaning no relief, 100% meaning complete relief of symptoms. Mean subjective improvement of the symptoms was rated at 62 ± 27% in group I and at 60 ± 34% in group II. Mean grip strength increased to 91% of theunaffectedsideingroupIandto94%in group II. After12monthsweevaluated27patients from group I and 25 from group II. At this point in time we observed 15 (55.5%) “Excel- lent” or “Good” results in group I and 15 (60%) “Good” or “Excellent” results in group II (Table 6.2). Mean subjective improvement was 75 ± 23% in group I and 72 ± 33% in group II. Mean grip strength compared to the contralateral side now amounted to 100% in group I and to 101 % in group II. The data concerning VAS ratings for 3 monthsand12monthsaresummarizedin Table 6.3. With the numbers available there was no statistically significant difference between group I and group II concerning the Roles and Maudsley score (extended Fisher test) and the VAS rating (Wilcoxon test for independent samples). Within the two groups there was a highly significant improvement in the VAS and of the Roles and Maudsley out- come score at both follow-ups compared with pretreatment conditions (p 0.0001). Additional Treatment No patient reported additional treatment at the assessment of results at 12 weeks. Between 3 and 12 months, three patients in group I and two patients in group II had undergone a release operation. In group I two patients required occasional pain medication; 6 Extracorporeal Shock Wave Application in the Treatment of Chronic Tennis Elbow44 Fig. 6.6 Hematoma after low-energy ESWT of the lateral elbow. three regularly. In group II four patients took pain medication on a regular basis; three occasionally. With the numbers available there was no statistically significant differ- ence between group I and group II with regard to additional treatment. Complications Besides petechial bleeding (Fig. 6.6), which occurred in 7 out of 60 patients, no adverse effects were recorded. Discussion The biological working mechanism of shock wave application (SWA) is poorly understood. Haake et al. (2001) fail to demonstrate any spi- nal response to low-energy SWA on the endog- enous opioid systems in rats. Schmitz (2001) reports an investigation on alterations of neu- ropeptides after applying 1500 shock waves of an energy flux density of 0.9 mJ/mm 2 to the intact rabbit femur. When measuring the con- centrations of substance P eluated from the femurperiosteumcomparedtotheuntreated contralateral limb, release of substance P had increased 6 hours and 24 hours after SWA, but had decreased 6 weeks after SWT. Remarkably, there was a close relationship between the time course of the release of substance P and thewell-knownclinicaltimecourseofinitial pain occurrence and subsequent pain relief after SWA for tendon diseases. According to a review of the literature by Heller and Niethard (1998), the first prospec- tive controlled study on the effectiveness of extracorporeal shock waves for the treatment of chronic tennis elbow was published in 1996.Atthe6-monthfollow-up48%“Good” or “Excellent” outcomes in the treatment group compared to 6% in the control group were reported according to the Roles and Maudsley score (Rompe et al. 1996a). Krischek et al. (1999) prospectively compare the anal- gesic effects of ESWT in patients with recalci- trant lateral or medial epicondylitis. With regard to the Verhaar score (Verhaar et al. 1993) they report 62% “Good/Excellent” out- comes in patients with tennis elbow after 1 year compared to 28% in patients with golfer elbow. Perlick et al. (1999a) prospectively compare the outcome after ESWT and after Discussion 45 surgery in 60 patients with chronic tennis elbow. They describe “Good” or “Excellent” results according to the Roles and Maudsley score in 43% and 73% at the 12-month follow- up. Twenty-three percent versus 10% of the patients reported no improvement at all. Haake and Boeddeker (2001) analyze early results of a prospective placebo-controlled, double-blinded, multicenter trial on 272 patients. Group I was treated three times, at weekly intervals, with 2000 low-energy impulses under local anesthesia; group II received sham therapy. After 3 months an identical successful outcome was observed in only25%ofpatientsinbothgroups.vomDorp et al. (2001) report preliminary results of 40 out of 114 patients involved in a randomized, placebo-controlled, double-blinded trial. Three months after three treatments, at weekly intervals, with 2100 low-energy impulses without local anesthesia, a reduc- tion in pain of more than 50% on the VAS was observed in 60% of patients compared with 15% of patients after placebo therapy. Besides these comparative studies there are numerous uncontrolled retrospective reports on the efficacious use of shock waves in the treatment of tennis elbow (Auersperg 1998, Boxberg et al. 1996, Brunner et al. 1997, Göbel et al. 1997, Lohrer et al. 1998, Tsironis et al. 1997, Wolf and Breitenfelder 1996). As inclu- sion criteria, treatment procedures, and out- come measurements were not standardized the success rates of these studies—58–85%— have to be viewed with caution. Besides local hematomas, no shock wave–related complica- tions have been reported (Sistermann and Katthagen 1998). The present study confirmed former results, leading to “Good/Excellent” results in 56% and 60% of the patients at the 1-year follow-up. This is comparable with results after local cor- ticosteroid injection (Day et al. 1978, Verhaar et al. 1996), but in our patients a mean of 2.1 (range: 1–5) steroid infiltrations had led to no improvement of the symptoms prior to ESWT. Although the procedure is approved by manual therapists, the quality of the literature concerning manual therapy of the cervical spine in the treatment concept of lateral epi- condylitis is poor. In a Medline search we found only five matches for cervical spine and tennis elbow between 1976 and 1998 (Wani- venhaus 1986, Waldis 1989, De Marco et al. 1998, Gunn and Milbrandt 1976, Vicenzino et al. 1996). This is surprising because following our experience of now more than 160 patients with chronic tennis elbow hardly any patient has no signs of cervical dysfunction such as localized pressure pain at the lower cervical spine, limited range of motion, and protrac- tion of the head. Radiologically there were of course signs of cervical spondylarthrosis due to the age of our patients. Patients with neu- rological deficits or pathological conditons of the spinal canal in computed tomography (CT) or in magnetic resonance imaging (MRI) were excluded from this study. Gunn and Milbrandt (1976) discuss a reflex localization of pain from radiculopathy at the cervical spine in patients with therapy- resistant tennis elbow who had presented with hypomobility of the lower cervical motion segments. Maitland (1991) finds that mobilization, traction, isometric exercises and heat, and/or ultrasound, applied to the cervi- cal spine, improved the signs and symptoms of lateral epicondylitis. Maigne (1988) reports complete healing of symptomatic tennis elbow after exclusively manual therapy for the cervical dysfunction in 51 out of 92 patients, and significant improvement in another 29 patients. Only two patients required surgery. However, inclusion criteria, outcome assessment, and follow-up were not described. Huguenin (1988) treated 49 patients with chronic tennis elbow with an ipsilateral cervical segmental dysfunction. All patients reported neck pain, an induration of the autochthonous musculature, and limita- tion of the joint play was described. The type of manual treatment applied was not explained. No results were specified, never- theless the author stated that his good results documented the connection between seg- mental dysfunction and peripheral muscular symptoms. In his opinion a treatment success could not be expected before 4–8 weeks after 6 Extracorporeal Shock Wave Application in the Treatment of Chronic Tennis Elbow46 manual therapy. de Branche (1988) analyzed 58 cases with local epicondylalgia and a cervi- cal spine pathology. All the patients received one to four manipulations of the cervical spine at weekly intervals. In 28.4% of the patients a significant improvement was achieved for 2–4 days; 43.1% for a longer, not exactly specified, period. Only 15.5% remained painfree and were rated a success. The author was not able to establish selection criteria for manual treatment of the cervical spine. Vincenzino et al. (1996) focuses on the immediacy with which manipulative therapy may initiate improvement in pain and func- tion. They report a randomized, double blind, placebo-controlled, repeated measures design to study the initial effects of a cervical spine treatment in a group of 15 patients with lat- eral epicondylitis. All the subjects received treatment, placebo, and control conditions. The treatment condition (contralateral lateral glide treatment technique for the cervical spine) produced significant improvement in the pressure pain threshold, painfree grip strength, neurodynamics, and pain scores rel- ative to placebo and control conditions. The authors conclude that manipulative therapy of the cervical spine is capable of eliciting a rapid hypoalgesic effect. In their opinion impairment of lateral epicondylagia is pro- jected from the hypomobile cervical spine motion segments and that the improvements gained following application of the contralat- eral gliding technique result from treating the source of the referred pain. Moreover, mobili- zation of the lower cervical spine is discussed as being capable of producing a sufficient sen- sory input to recruit and activate descending pain inhibitory systems which exert a portion or all of the pain relieving effects (Bogduk 1994, Grieve 1994). In the current study, the focus was on possi- ble additive effects of cervical spine manual therapy on patients treated with ESWT for chronic tennis elbow. With the numbers avail- able we failed to demonstrate a positive effect of a standardized manual therapy to the cervi- cal spine. Though, statistically, our treatment groups did not differ with regard to epidemio- logical data, it must be made clear that the patients for both procedures were not ran- domized, therefore selection and information bias cannot be ruled out. Our data underline the value of low-energy ESWT in chronic lateral epicondylalgia and question the usefulness of additional cervical spine manual treatment in these patients. Fur- ther studies are mandatory to establish the optimum treatment regime with ESWT for patients with recalcitrant tennis elbow and to clarify what role manual therapy of the cervi- cal spine may play in the treatment of this enthesiopathy. The mechanisms by which ESWT or cervical spine manual therapy achieve improvements in pain and function are yet to be elucidated and must form the basis for ongoing research. Discussion 47 Page intentionally left blank . level 1 (Minimum value) Energy level 2 (Mean value) Energy level 3 (Maximal value) Peak positive pressure P + Mpa 5.5 14.2 25 .6 6 dB focal extend in x,y,z direction f x ( -6 dB) f y ( -6 dB) f z ( -6 dB) mm mm mm 6. 0 6. 0 58 5.2 5.2 55 4.8 4.8 49 5MPafocal. Group II p-value Group I Group II p-value Group I Group II p-value Pressure pain 6. 21± 2 .65 6. 15 ± 2.43 0.28 3.99 ± 2.74 3.59 ±2.29 0 .65 2.27 ± 2.59 1.97± 2.05 0.82 Thomsen test 6. 18± 1.72 6. 24 ±. cervical spine. Vincenzino et al. (19 96) focuses on the immediacy with which manipulative therapy may initiate improvement in pain and func- tion. They report a randomized, double blind, placebo-controlled,