1 Introduction

The first scientific paper describing the condition now commonly named tennis elbow (TE) was probably that of Dr. Runge published in 1873, though he referred to the condition as writer's cramp (“Schreibekrampfes”).¹ We consider the synonym, lateral epicondylitis, to be a misnomer, as an inflammation is generally not present.² This is the likely reason why injections of corticosteroids do not alter the long-term course of the disorder. In the population at large, TE has a prevalence of 1–3%, affecting men and women at the same rate, and occurring most frequently in age group 40–60 years.^3–7 Manual workers are predisposed for the condition with a prevalence of more than 10% in some studies.⁸ Thus, while the name “tennis elbow” suggests a relation to sports, most patients attribute the problem to their work.² In concordance with the high prevalence in the working population, the reported risk factors (of TE) include repetitive and forceful manual activities.^2,4 The reason is probably that such mechanical load of the elbow leads to an overuse injury of the common extensor tendon (CET) insertion and, probably most importantly, the extensor carpi radialis brevis (ECRB) tendon part of the CET.^2,4 The commonly accepted description of the pathogenesis is an ongoing repetitive microtrauma that surpasses the tissue's capability for restoration, eventually causing degeneration. Histologically, this process is characterized by disruption of tendon fibres, disorganized collagen, fibroblast invasion, and vascular hyperplasia.^9,10

This overuse insertion tendinopathy will eventually be evident on MRI scans and by ultrasound (US) examination, complimenting the history and clinical examination, which unfortunately may not be as accurate (in leading to the correct diagnosis) as clinicians like to believe.¹¹ MRI scans of TE typically show swelling of the proximal part of the CET with intensified fluid signal. In more severe cases, partial ruptures will start to appear.¹² The US examination has the benefit of being dynamic and easily complimenting the clinical examination (by a US competent clinician). The US probe becomes an extra palpating finger localising the punctum maximum for pain, and an extra eye with the ability to look deep into tissues (and see the cause of pain and illness). Using a linear probe and musculoskeletal settings, the common findings in TE include tendon (CET) thickening with loss of normal fibrillar pattern and hypoechoic areas. As with MRI scans, partial tears will be appearing in more severe cases. Using colour doppler, increased vascularity of the CET is a typical sign (of TE).¹³

Several (older) studies have concluded that conservative care, focused on rehabilitation, leads to complete recovery in most TE patients in a year or two,^9,10,14 and that surgery is unnecessary in 90% of the cases.¹⁰ However, authors of some newer studies have found poorer prognosis in untreated or conservatively-treated TE patients.^15,16 Thus, Bot and colleagues found that although most of the TE patients (n ​= ​181) reported at least some progress after one year of conservative care, only a third reported a full recovery.¹⁵ Further, Nilsson and colleagues found, in nearly 300 patients, that after two years of non-surgical treatment (of different types), half of the study population still experienced pain and/or reduced function.¹⁶

Thus, surgery might be indicated in cases that are resistant to conservative (non-surgical) treatment, including a rehabilitation program overlooked by a physiotherapist. In our own surgical experience, until 2005, we treated patients with TE (that resisted conservative care) with an open division of the CET insertion as described by Hohmann.¹⁷ In 2011, we reported on the median 4-year outcome in 80 consecutive TEs (in 77 patients) operated with an open procedure, and we examined possible risk factors for a poor result.² In 2005, we altered the surgical procedure in TE from the open Hohmann technique to an inside-out arthroscopically assisted, selective release of the ECRB insertion. The technique is similar to that advocated by Baker and colleagues¹⁸ and by Owens and colleagues.¹⁹ In 2013, we compared the outcome after open versus arthroscopic tenotomy, of the CET and the ECRB, respectively. The latter method resulted in better outcome.⁷ Based on the study by Budoff²⁰ about the scratch-test, we asked ourselves if tenotomy (of the ECRB) was a necessary part of the arthroscopic procedure, or if debridement of degenerative tissue only would lead to similar outcomes for our patients. Thus, in 2016, we presented a clinical study comparing the outcome of the two methods, finding that debridement of ECRB without tenotomy resulted in a quicker back-to-work period.²¹ Since then, debridement alone has been our standard method for treating recalcitrant TE.

The purpose of current paper is to describe, thoroughly and step by step, two different arthroscopic techniques of treating TE, tenotomy or debridement (only) of the ECRB insertion, accompanied with surgical images of our own. Further, we will try to share some surgical pearls, as well as potential pitfalls, and discuss the results in light of recent literature.

2 Techniques

2.1

2.1 Indications and red flags

We consider patients eligible for arthroscopic TE surgery to include grown-up individuals with: (1) a typical history of pain located in front of the lateral epicondyle (LE) that is related to activity, hampers normal daily activities, and has resisted non-surgical care that includes a rehabilitation program led by a physiotherapist for at least 3–6 months; (2) a positive physical examination that includes local pain on palpation at the CET insertion at the LE, pain in the same area provoked by extension of the wrist and the third digit against resistance, normal ROM, normal neurological function, and normal joint laxity tests; and (3) no radiographic joint disorder.

Further, we generally require a positive MRI scan and/or ultrasound (US) examination to confirm the diagnosis (Figs. 1 and 2). We discourage performing TE surgery in patients with a negative MRI/US examination because we believe that it is unlikely that a patient has TE in need of surgery without clear positive images of (CET/ECRB) tendinopathy and/or partial tears. We are strong advocates of using ultrasound examinations in orthopaedic clinical practice, especially now that small lightweight units are available that can be coupled to the screen of a smartphone or a tablet by Bluetooth or Wi-Fi technology. Orthopaedic surgeons performing arthroscopy will enjoy a rapid (steep) learning curve by being able to confirm the accuracy of their US diagnosis (Fig. 2) by performing an arthroscopic examination immediately afterwards (Fig. 3).

Close

Fig. 1 MRI scan of right-sided tennis elbow. Coronal fat-suppressed FSE T2-weighted image showing (arrow) swelling and intensified fluid signal affecting large parts of the common extensor tendon (CET) proximal insertion. Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 2 US examination (with a linear probe and musculoskeletal settings) of right-sided tennis elbow of the same patient as in Fig. 1. Tendon thickening with loss of normal fibrillar patter and tears of the CET (arrow) proximal insertion. Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 3 Arthroscopy of right-sided tennis elbow of the same patient as in Figs. 1 and 2. Arrow points to areas with macroscopic changes of the CET including capsule tears. Viewing portal is the proximal anteromedial portal (PAMP). Private photography. Copyright by the authors.

Export to PPT

We generally do not recommend TE surgery when one or more of the following “red flags” is present: mechanically restricted ROM (many TE patients experience pain in full extension, but not true mechanical restriction), neurologic deficits, transposed n. ulnaris (care should be taken when naturally anteriorly dislodged, but this is not a contraindication), ligamentous instability, sequelae after fracture, articular cartilage defects, osteoarthritis, or loose bodies. Further, Solheim and colleagues found a significant linear relationship between residual symptoms at the final follow-up after open TE surgery and the following baseline factors: severe preoperative complaints, acute manifestation of complaints, lengthy history of complaints, female sex and young individuals.² Thus, caution should thus be exercised when recommending surgery in these situations, especially when more than one of these factors are present. In the same study, having the dominant arm affected, the patient judging that the cause of TE was work-related, or having a work involving performing strenuous activities (for the arm) did not seem to affect the outcome.²

2.2

2.2 Elbow arthroscopy

We will start this technical review with a description of elbow arthroscopy itself. The procedure is most often performed in a surgical day care unit under a combination of general total intravenous anaesthesia (TIVA) and local anaesthesia placed subcutaneously at the portals and into the joint. If the patient is not fit for TIVA, the procedure may be performed with just local anaesthesia (injected at the portals and into the joint), but some discomfort for the patient should then be expected and discussed with the patient beforehand. The patient places him- or herself in the lateral decubitus position (before the TIVA is commenced) with the elbow at a 90-degree angle (Figs. 4 and 5). Neither traction nor tourniquet is needed. Viewing is performed with a standard 30° 4.0-mm arthroscope with a video camera connected to a monitor. Joint distention is maintained with an arthroscopy pump. An outflow cannula is needed for unrestricted circulation of fluid to avoid overdistention and soft tissue oedema and, when using a radiofrequency (RF) system, to flush out bubbles (that impedes visibility) and to prohibit overheating.

Close

Fig. 4 Operating table for arthroscopy of the right elbow. The armrest marked TIVA (total intravenous anaesthesia) is where we place the opposite arm to establish a secure venous access including for providing the drugs necessary for general anaesthesia. Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 5 The patient has been scrubbed and draped in the lateral decubitus position on the table shown in Fig. 4. Private photography. Copyright by the authors.

Export to PPT

We recommend marking important anatomical landmarks and the portals planned to be used on the skin with a waterproof marker before starting the procedure (Fig. 6). The localization of the ulnar nerve at the elbow level should be established by palpation and/or the use of US before the start of the procedure. Special care is warranted when the nerve is anteriorly luxated or surgically transposed.

Close

Fig. 6 To avoid nerve injuries, we recommend marking (on the skin) important landmarks, including the medial epicondyle (ME), olecranon (OC) and the lateral epicondyle (LE) as well as the localization of the (safe) portals intended to be used beforehand. Only the skin is cut by moving the skin and holding the knife steady (against it) with light pressure. Private photography. Copyright by the authors.

Export to PPT

The joint may be extended by injecting approximately 10 ml of liquid (local anaesthetic or saline), to ease the entrance to the joint. The proximal anteromedial portal (PAMP) is usually established first (Fig. 7). Only the superficial skin is cut for a length of 4–5 ​mm (Fig. 6) to allow the entrance of the trocar (sheet) with a fitted blunt metal rod inside. The blunt trocar is directed at the centre of the joint (Fig. 8), sliding in front of the humerus (trochlea). A pop is generally felt when entering the joint. A slight resistance is normally felt thereafter when passing from the anteromedial to the lateral compartment. Minimal force should be used when entering the joint to not inflict injuries to the articular surface. If in doubt about being inside the joint, we recommend pulling the trochar back 1–2 cm and try again, staying close to bone and using minimal force. We advise against passing through the skin multiple times as this enhances the risk of nerve injuries and, possibly, deep infection. If saline or local anaesthetic has been deposited in the joint before the arthroscopy, it should be seen spilling back out of the trochar (Fig. 9) when the blunt metal rod is replaced with the arthroscope.

Close

Fig. 7 The proximal anteromedial portal (PAMP) is usually established first as a viewing portal. The portal is located 2 cm proximal to the medial epicondyle (ME) and immediately anterior to the intermuscular septum. Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 8 The trochar with a blunt metal rod (awl) is slid along the front of the humerus (trochlea humeri) close to bone. A “pop” is usually felt when the trochar penetrates the joint capsule and enters the interior of the joint. Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 9 If the joint has been distended with local anaesthetic or saline beforehand (as we recommend easing the entrance to the joint), when the trochar is inside the joint, the liquid will be seen to run back through the trochar when the blunt rod (awl) inside is removed (arrow). This confirms the correct position. Private photography. Copyright by the authors.

Export to PPT

For a complete examination of the various compartments of the joint, several portals are needed, at least one at each location, anteromedially (Fig. 10), anterolaterally (Fig. 11) and posteriorly (Fig. 12). Placing the arthroscope in an anteromedial, an anterolateral and a posterior portal will provide a complete examination of the inside of the elbow joint (Fig. 13, Fig. 14, Fig. 15, and Fig. 16). For arthroscopic TE surgery, only three portals are needed: an anteromedial viewing portal, an anterolateral working portal (close to the proximal LE) and a posterior (cannula) portal for the outflow of water (Fig. 17). In TE, the arthroscopic picture of capsule and the CET origin at the LE varies from minor derangements (Fig. 18) to tears of capsule and tendon (Fig. 19). Sharp instruments should never be entered into the joint without full visibility of their entrance. When switching the anterior viewing- and working-portal, some surgeons find it helpful to use a switching rod.

Close

Fig. 10 The most used anteromedial portals (of the right elbow). Proximal anteromedial portal (PAMP) is located 2 cm proximal to the medial epicondyle (ME) and immediately anterior to the intermuscular septum. The PAMP is the safest and most used anteromedial portal. Mid-anteromedial portal (MAMP) is located 1 cm proximal and 1 cm anterior to the ME. (Distal) Anteromedial portal (AMP) is located 2 cm anterior and 2 cm distal to the ME. Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 11 The most used anterolateral portals (of the right elbow). Proximal anterolateral portal (PALP) is located 1–2 cm proximal to the lateral epicondyle (LE) and just anterior to the humerus. Tennis elbow portal (TEP) is used for treating TE. It is located very close to the proximal part of the LE to ease debridement or tenotomy of the extensor carpi radialis brevis. Mid-anterolateral portal (MALP) is located directly anterior to the radio-capitellar joint. RH marks the location of the radial head. The MALP is the most used anterolateral portal (except when treating TE). (Distal) Anterolateral portal (ALP) is generally not used anymore due to its proximity to the radial nerve and is only mentioned here for historical reasons. Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 12 The most used posterior portals (of the right elbow). They are all safe, far away from important nerves passing the elbow. Direct posterior portal (DPP) is located 3 cm proximal to tip of the olecranon (OC) and pierces through the tendinous portion of the triceps. It provides a direct access to the olecranon fossa and is usually the ideal working portal for removing loose bodies (and/or osteophytes) in the posterior compartment. Posterolateral portal (PLP) is located 1 cm proximal to a midpoint of a line connecting OC and the lateral epicondyle (LE). The DPP and PLP are the most used posterior portals. Direct lateral portal (DLP) is located at the soft-spot close to the radio-capitellar joint, and helpful for addressing posterior parts of this joint. This portal may be used for distending the joint with liquid prior to the procedure as well as the placement of the outflow cannula. Distal ulnar portal (DUP) is located 3-4 cm distal to the radio-capitellar joint along the border of the ulna. The portal is helpful for treating osteochondral lesions of the capitellum. Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 13 This is the view from the proximal anteromedial portal (PAMP) in the right elbow (Fig. 10). We are looking through the anterior compartment towards the opposite side, seeing the capitellum, radial head, the annular ligament (AL) and the anterolateral capsule (which may show tears in cases of TE). Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 14 The radio-ulnar joint of the right elbow viewed from the proximal anteromedial portal (PAMP) (Fig. 10). Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 15 The view from the mid-anterolateral portal (MALP) in the right elbow (Fig. 11). We are looking through the anterior compartment towards the opposite side, seeing the trochlea, coronoid process, and the anteromedial capsule (which may show pathology in cases of golfer's elbow). Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 16 The view of the posterior compartment viewed from the posterolateral portal (PLP) in the right elbow (Fig. 12), with the olecranon, trochlea, and olecranon fossa (where any loose bodies in the posterior compartment are located). Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 17 For arthroscopic TE surgery, only three portals are needed: An anteromedial viewing portal, e.g., the proximal anteromedial portal (PAMP); an anterolateral working portal close to the proximal lateral epicondyle (LE), i.e., the tennis elbow portal (TEP); and a posterior cannula portal, e.g., the posterolateral portal (PLP) for the outflow of water. Still, we encourage the surgeon to do a complete examination of the joint before addressing the TE. ME marks the location of the medial epicondyle. Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 18 Arthroscopic view of moderate right-sided tennis elbow viewed through the PAMP. We can see ECRB tendinopathy (dotted line) and possibly some calcifications of the capsule. Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 19 Arthroscopic view of severe right-sided tennis elbow viewed through the PAMP. We can see a capsular tear (dotted line) and degenerative changes and partial tears in the ECRB behind the torn capsule. Private photography. Copyright by the authors.

Export to PPT

We recommend asking the patient to move his/her fingers and wrist as soon as he or she awakes from the general anaesthesia to confirm that no permanent harm has been inflicted to the motoric nerves passing by the elbow (as many will experience temporary numbness and reduced function for the next hours while the injected local anaesthesia is in full effect). If any nerve damage is suspected, a new examination should be conducted the next day. Surgically damaging the deep radial nerve, the median nerve or the ulnar nerve could lead to significant and permanent handicap (in the worst case, a drop-hand).

2.3

2.3 Tenotomy

A working TE portal (TEP) is established just in front of the proximal border of the LE to facilitate the tenotomy of the ECRB (Fig. 17). A bipolar RF electrode is introduced through the described TEP and passed bluntly until it can be recognized directly underneath the anterolateral capsule. We currently use a hook electrode (Dyonics RF Hook 30° probe, Smith ​+ ​Nephew, London, UK) but have used others previously.

With the RF probe in vaporization mode, the capsule is divided from the proximal border of the capitellum all the way down to the annular ligament. The effect is applied in short bursts to avoid overheating and creation of air bubbles that obstruct visibility (Figs. 20 and 21). The release should be made cautiously to not damage the articular cartilage.

Close

Fig. 20 Arthroscopic view of the first step of selective tenotomy of the ECRB in a right-sided tennis elbow viewed through the PAMP. A bipolar RF hook electrode has been bluntly introduced through the modified anterolateral TE portal at the upper border of the capitellum. Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 21 Arthroscopic view of the same patient as in Fig. 20. The capsular release is extended (by the bipolar RF hook electrode) close to the LE distally to the annular ligament. Private photography. Copyright by the authors.

Export to PPT

As the capsule is being released, the ECRB tendon is being revealed (Fig. 22). Starting proximally, working downwards, the ECRB is completely freed from its bony insertion by the RF probe (Fig. 23). When fully cut, the tendon moves distally, approximately 6–10 mm, leaving its footprint on the LE visible (Fig. 24). A 3.5mm soft tissue resector is used to debride the insertion sites of the capsule and ECRB. We do not recommend using a burr for this purpose as it may cause unnecessary bleeding, swelling, pain and prolonged rehabilitation.

Close

Fig. 22 Arthroscopic view of the same patient as in Figs. 20 and 21. The capsular release has been completed down to the annular ligament resulting in the ECRB tendon insertion coming into view. Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 23 Arthroscopic view of the same patient as in Figs. 20, Fig. 21, and Fig. 22. A tenotomy of the ECRB is performed at its proximal insertion at the LE. Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 24 Arthroscopic view of the same patient as in Fig. 20, Fig. 21, Fig. 22, and Fig. 23. The tenotomy of the ECRB has been completed. Private photography. Copyright by the authors.

Export to PPT

2.4

2.4 Debridement

A 3.5mm shaver is positioned at the TEP and pushed bluntly into interior of the joint close to the proximal border of the ECRB (Figs. 25 and 26). An RF probe may be used to divide an intact capsule (only) that is difficult to penetrate bluntly, and thereby ease the access, but this enhances the cost of the procedure and may be an unnecessary step. The overlaying capsule is removed by the shaver to uncover the ECRB insertion (Fig. 27). Using the tissue resector, a debridement of the proximal ECRB tendon is commenced to remove deteriorated tendon tissue. The debridement is based on both visual (arthroscopic) information and the “Scratch Test” principle.^20,21 The debridement is performed by moving the tip of the oscillating shaver 90° along the path of the tendon with a very light compression (Fig. 28). Sometimes, parts of the ECRB insertion are fragile and/or or already inflicted with transverse rupture resulting in a partial tenotomy after the debridement. We have not found this situation to lead to a worse long-term outcome.

Close

Fig. 25 Arthroscopic view of severe right-sided tennis elbow viewed through the PAMP. We can see clear macroscopic signs of degenerative changes and tears in both capsule and ECRB. Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 26 Arthroscopic view of the same patient as in Fig. 25. A 3.5mm shaver is being bluntly introduced (into the joint) through tears of the ECRB and capsule. Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 27 Arthroscopic view of the same patient as in Figs. 25 and 26. The 3.5mm shaver is being used for removing pathologic capsule revealing the ECRB (with clear signs of longitudinal tears). Private photography. Copyright by the authors.

Export to PPT

Close

Fig. 28 Arthroscopic view of the same patient as in Fig. 25, Fig. 26, and Fig. 27. Using the shaver, a debridement of the proximal ECRB tendon is undertaken to remove degenerated tissue based on both visual information and the “Scratch Test” principle by moving the tip of the oscillating resector perpendicular along the tendon. Private photography. Copyright by the authors.

Export to PPT

3 Discussion

The purpose of this paper is to walk the reader, step by step, through the techniques of arthroscopic treatment of TE. The paper is founded on our experience with more than 2000 such surgeries since 2005. Further, we offer our own photographs as illustrations of the procedures, and we try to highlight some surgical pearls and potential pitfalls of the procedures. The paper is intended to be a technical review of the two procedures (and is not a clinical study). Probably unnecessary to mention, but proficiency in elbow arthroscopy is a precondition for starting using the described techniques. In this section, we will discuss the advantages and possible restrictions of arthroscopic treatment of TE and compare the outcome to that of open tenotomy, based on the current literature.

In elbow surgery, as in that of knee or shoulder, e.g., in rotator cuff repair, a gradual evolution from open surgery towards arthroscopically-assisted procedures has been happening in recent years.²² Arthroscopy generally permits better view of all the interior parts of the joint, the rehabilitation is often speedier and the risk of postoperative infection is lower. Further, in our experience, every patient favours an arthroscopy to an open procedure when given the chance to choose. However, the shift towards arthroscopic techniques in the elbow has been lengthier compared to that in the knee and shoulder due to the greater intricacy of the joint and the nearness to important (motoric) nerves.⁷

As in most non-acute orthopaedic maladies, a conservative/non-surgical treatment of TE is the prime choice. As most patients relate their TE to their work situation,² the reasonable first step is to make changes in their work to relieve the strain on the arm. This can include assigning the employee to a type of work that places less stress on the arm, and ideally, more varied tasks throughout the work day.²³ Unfortunately, this is often not possible, especially not for blue-collar workers, who comprise a large percentage of TE patients.² They are generally employed for their particular manual skills, e.g., as hair-dressers, electricians, plumbers or carpenters, and may not be useful to their employer for other types of work. Secondly, most patients will benefit from a structured rehabilitation program led by a competent physiotherapist. It should be stressed that the exercises must be continued for not just weeks, but years, because the benefits soon wear off. A professional athlete does not train for 6 weeks, but for his or her entire career.

Thirdly, many patients find that the use of orthosis (of various kinds) reduces pain and enhances function.²⁴ Other commonly used treatments like injection of corticosteroids or platelet enriched plasma (PRP) and the use of extracorporeal shockwave treatment (ECSWT) have, in our opinion, modest scientific support and we do not recommend using them until there is more evidence of their efficacy.

When conservative treatment fails, surgery is often considered to be an option in cases of patients suffering considerable pain and disability, e.g., that inflict on their ability to work. Most techniques aim to relieve the tension at the tendon insertion at the LE by a release of the CET origin or part of it, such as the ECRB insertion^2,17,25 and/or removal of the degenerative tissue.¹⁰ Especially in open TE surgery, such as the Nirschl procedure, removal of degenerative tendon tissue is a main part of the procedure.¹⁰ The unhealthy tissue has been described as grey/dull, disorganized, friable and oedematous compared to the off-white, shiny, firm and linearly-organized appearance of normal tendon tissue.²⁰

However, sometimes, it is difficult to judge (visually) which parts of the tendon(s) that are pathologic and how much tissue that must be removed. Macroscopically, the TE pathology that is easiest to spot (openly or through an arthroscope) are partial tendon tears at the CET/ECRB insertion, but they are often not present. It must be remembered that much of the described pathology is microscopically and only visible on histologic slides, MRI scans and in US examinations. As the friable/soft pathologic tendon tissue in tendinopathy, such as in TE, are more easily removed, Nirschl and Budoff have advocated using a so-called “Scratch Test” (or, rather, a scratch method) to excise pathological tissue without removing healthy tendon tissue.²⁰ The method consists of scraping the tendon (suffering from a tendinopathy) with the blade of a knife perpendicular to the tendon. The procedure continues until the scratching becomes unproductive leaving only healthy tissue in place.²⁰ In arthroscopic surgery, we have suggested that the scalpel blade may be substituted by a small diameter soft tissue resector used in a similar way, by moving the oscillating resector perpendicular along the tendon with a light pressure.²¹ Thus, the resection is based on both macroscopic appearance and the arthroscopic soft tissue resector “Scratch Test”.²¹

In the beginning of the 1990s, several innovative elbow surgeons who mastered the arthroscope started using endoscopic techniques for treating TE. Possibly first to describe an arthroscopic procedure, Jürgen Krämer and colleagues reported on a technique for an endoscopic extensor release similar to that of the open Hohmann procedure¹⁷ with the arthroscope and instruments in the subcutaneous space and working outside-in.^26,27 A year later, Victor Tseng described a similar extensor tendons release, but with the arthroscope being inside the joint, performing the release inside-out with a hooked knife, after the capsule in the area had been removed with a soft-tissue resector. Nine out of 11 patients had excellent results in the follow-up period that lasted from 3 to 11 months.²⁸

In 1996, Stapleton and Baker briefly described (as an abstract) a selective release of the ECRB insertion (N ​= ​5) arthroscopically by a soft-tissue resector (shaver).²⁹ After the tenotomy, the footprint of the ECRB was decorticated with a burr in an effort to promote healing. The outcome of the arthroscopic technique was compared to a control group patients treated with an open Nirschl procedure (N ​= ​10). They reported a failure rate of 20% in both groups. The surgical costs were higher in the arthroscopic group. However, the rate complications was less, and return to work and sports was speedier compared to open surgery.²⁹

Baker and colleagues reported in 2000 on the arthroscopic treatment of TE in 42 elbows.¹⁸ A soft tissue resector or a RF probe was used to release the ECRB insertion. Generally, the LE was decorticated using a shaver or a burr. Thirty-seven out of 39 elbows were rated as “better” or “much better” at a minimum of 2-year follow-up. Return to work occurred in an average of 2.2 weeks. The grip strength was found to be on average 96% that of the unaffected limb. The authors noticed concomitant lesions in more than 2/3 of the cases, including synovitis, bone spurs and loose bodies.

In 2001, Owens et al. published on the outcome after arthroscopic ECRB release in 16 patients with TE resistant to non-surgical care.¹⁹ Follow-up data were obtained in 12 patients at an average length of 24 months. Improvement compared to baseline was found in all patients. The patients returned to unrestricted work in average 6 days. The favourable results reported in arthroscopic division of ECRB in TE by Stapleton and Baker, Baker et al. and Owens et al. has been confirmed both in short- and medium-term^30–32 as well as long-term³³ follow-up case studies.

To the best of our knowledge, no randomized studies have been published comparing arthroscopic and open techniques in recalcitrant TE. However, some case-control studies (in addition to that of Stapleton and Baker already mentioned) have been published comparing different techniques. Peart et al. reviewed retrospectively 87 cases of surgically treated TE, openly (N ​= ​54) or arthroscopically (N ​= ​33).³⁴ They found no significant difference in outcome, with approximately 70% of the patients having a good or excellent result in both groups. However, patients treated arthroscopically returned to work significantly earlier.³⁴

Szabo and colleagues compared 109 patients treated for TE by a percutaneous (N ​= ​24), arthroscopic (N ​= ​44) or open (N ​= ​41) approach.³⁵ No significant differences between the three groups were found regarding the rate of complications and failures or in the mean figures of different PROMs.³⁵ Othman examined the outcome of after arthroscopic (N ​= ​14) versus percutaneous (N ​= ​19) tenotomy.³⁶ No significant difference in DASH score or VAS between the groups could be found at the final follow-up. Further, the average time from surgery to return to normal activity was 3 weeks in both groups.³⁶

Solheim and colleagues found a better clinical outcome in a group operated with an arthroscopically assisted release of the ECRB tendon (N ​= ​225) compared to that of a control group undergoing an open tendon release (N ​= ​80) as evaluated by the mean QuickDASH score and the percentage of excellent outcomes defined as QuickDASH <20. The age of the patients, gender distribution, length of follow-up period and baseline QuickDASH scores were not significantly different between the arthroscopic and open groups.⁷ The better outcome by the arthroscopic technique may be explained by constituting a less surgical trauma. We can only speculate about why similar previous studies have failed to reveal any significant difference in the outcome between arthroscopic and open techniques. A possible explanation is that most previous studies have included too few patients for inhabiting sufficient statistical power to detect smaller differences in the outcome. Further, some previous studies have included decortication of the ECRB insertion site. This step seems to be unnecessary for achieving repair³⁷ and may lead to more postoperative haematoma, stiffness and pain.^38–40

Another clinical study²¹ compared arthroscopic release (N ​= ​204) to arthroscopic debridement ad modum the “Scratch Test”²⁰ (N ​= ​79) at a minimum of 4-year follow-up. Both methods resulted in significant improvement from baseline as evaluated by various PROMs. The failure rate, 8% and 10%, respectively, and the excellent outcome rate (QuickDASH <20 points) of 78% and 79%, respectively, were not significantly different. However, the patients in the debridement group reported sick-leave (workers' compensation) periods that were a mean of 2 weeks shorter than the tenotomy group. This could represent an economic benefit to both the employers and the taxpayers, depending on the country's social security system. The reason for the lengthier recovery (back to work) period in the tenotomy group could be that it constitutes a more extensive procedure resulting in more pain and loss of strength during the first weeks after surgery. This is probably of special importance in the typical group of TE patients that consists largely of manual workers. Further, as the use of a single-use RF probe is omitted, the surgical procedure (in debridement only or “Scratch Test”) is quicker (a few minutes shorter surgical duration) and less expensive.²¹

Sareni and coworkers,⁴¹ reported on the midterm outcome in 40 patients with recalcitrant TE using a technique similar to that described by Solheim et al.,²¹ i.e., arthroscopic debridement of capsule and ECRB tendon insertion site. At an average follow-up period of 42 months the authors reported improved VAS and QuickDASH scores. Further, they found no significant change in grip strength between operated and non-operated arm.⁴¹

Baker and colleagues introduced a system for classifying the arthroscopic findings of TE into three groups.¹⁸ In 43 arthroscopic releases for TE, they found 15 type I lesions (intact capsule), 15 type II lesions (linear capsular tear), and 12 type III lesions (complete capsular tear). Unfortunately, there seems to be little correlation between the Baker grade and the outcome after surgery.^18,30,42 A possible explanation may be that at the end of the procedure (tenotomy) all patients have a surgically-inflicted Baker III lesion (transverse release of the capsule, as well of the ECRB tendon) regardless of the preoperative macroscopic lesions. Further, if the most important pathology is indeed situated in the tendinous insertion of the ECRB,¹⁰ the relevance of grading concomitant lesions of the capsule is questionable. Lastly, the most important lesions may be microscopic rather than macroscopic. Solheim and colleagues found that most cases of TE have minimal degenerative macroscopic findings in capsule and tendons and, accordingly, 75% of the elbows were rated as Baker type I.⁷ This observation seems to be in concordance with that of Cohen and Romeo who stated that: "Most commonly, the capsule is intact, although small linear tears can be present".⁴³

The reported mean time to return to work after an arthroscopic TE procedure varies from one week¹⁹ to 19 weeks.³⁰ Oki and colleagues found that the results of various elbow PROMs as well as the grip strength improved continuously during the first three months after arthroscopic TE surgery and discouraged letting manual workers returning to work too early.⁴⁴ Solheim and colleagues found that the mean time to return to work was 7 weeks after arthroscopic tenotomy and 5 weeks after arthroscopic debridement.²¹ In our experience, 2–4 weeks may be appropriate for office employees with various tasks, whereas heavy manual workers usually need 8–12 weeks, and sometimes more, based on the progress of the rehabilitation and the nature of the patient's work.

4 Conclusion

This technical review provides a step-by step description of the surgical technique of arthroscopic treatment of tennis elbow and highlights some pearls of wisdom and potential pitfalls, based on our experience. Proficiency in elbow arthroscopy is a prerequisite. For an orthopaedic surgeon already familiar with arthroscopy of this joint, the techniques presented should be easily incorporated into his or her surgical armamentarium. We present two techniques, a selective release of the ECRB or just debridement. Both techniques offer similar outcome as evaluated by PROMs, but the latter with somewhat shorter rehabilitation.

Financial support and sponsorship

No financial support or sponsorship was requested or received for this manuscript.

Authors contribution

E. S. gathered the information, including clinical experience, regarding the reported technique. J. H. was responsible for the collection of patient data presented in the manuscript. E. S. drafted the manuscript. C. G. and E. I. performed critical revision of the manuscript. All the authors read and approved the submitted final version of the manuscript.

Declaration of competing interest

The authors declare that they have no competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Ethical considerations

This paper, being a review of the surgical technique based on clinical experience, and not a clinical study reporting patient data, did not warrant institutional ethical approval. All patients depicted in the Figures gave their informed consent to having their intra-operative pictures published anonymously.