Tennis elbow which is also known as humeral epicondylitis is named for its concentrated outbreak in grass tennis tournament. In fact, neither of these two naming methods can accurately express the pathophysiological process of the disease. Possible pathological changes include the epicondylitis of humerus, the rupture of joint capsule at the lateral epicondyle of humerus, the compression of lateral synovial fold, extensor tendon disease, cartilage injury of brachioradial joint, nerve compression or a mixture of the above. There is no definitive conclusion in current clinical study. In recent years, research has tended to interpret it as one type of tendinopathies near the insertion of the tendon. As extensor tendon extensor carpi radialis brevis (ECRB) is the most common and major cause of this disease, many clinical treatments are aiming at clearing ECRB and obtain good clinical result.Anatomical study has found that ECRB is close to joint capsule, which opens up the possibility of clearing ECRB under elbow arthroscopy. More and more centers are beginning to use this technique to treat the disease. However, the vascular nerves around elbow joint are rich, and the elbow joint cavity is not large with extremely irregular structure. Also, the elbow joint is prone to postoperative activity limitation, and the elbow arthroscopy learning curve is long. The reasons above limit the development of elbow arthroscopy technique. How to establish a safe, fast and accurate way is the key.

Ⅰ. General information. This study provided a retrospective analysis of 28 cases (11 males and 17 females) of refractory tennis elbow in our hospital from January 2008 to December 2016. The age ranged from 35-61 years with an average of 45.6 years. 16 cases had the right elbow affected, and 12 cases had the left elbow affected. There were 6 cases with the history of clear sports injury or heavy weight carrying. The clinical manifestation was the lateral pain of elbow joint. 19 patients had the pain radiated to the middle of anterior humerus; 13 patients developed rest pain and nocturnal pain. When they fisted and stretched their wrist, they could induce or exaggerate the pain. Daily activities including towels twisting, floor sweeping and goods stolen are all greatly affected. All patients underwent the conservative treatments of anti-inflammatory analgesic, physiotherapy, etc. for more than half a year, and all of them were treated with multiple injections (≥3 times) on pain points. The number of treatment ranged from 3-12 times with an average of 4.8 times. The time from onset to surgery was >6 months with an average of 11.5 months. Physical examination: All patients had tenderness on the lateral side of elbow. The pain point was located in lateral humerus, humeral head or the area between the two; the MILL'S sign was positive; 4 patients had slight restriction in elbow joint; 10 patients showed clicking during elbow flexion Ⅱ. Operation steps. 1. Tools: 2.7 mm or 4.0 mm of 30° or 70° arthroscopic lens, 3.5 mm or 3.0 mm synovial planer, 2.3 mm of 35° short bevel plasma radio head, micro-fracture device, exchange bar 2, common perfusion Liquid suspension gravity perfusion. 2. Position and anesthesia: with 90° side lying and axillary pad protection, the upper side of the healthy side was placed in front of abdomen, and the restraining band was fixed; with the affected limb on top, the hemostasis band was placed on upper arm root before disinfection, and the elbow was bent to 90° to be placed on the upper arm supporter with cushion. The support position was on the distal end of tourniquet. Brachial plexus block anesthesia was used. 3. Anatomic marks: humeral ectocondyle, humeral entepicondyle, lateral crest of distal humerus, ulnar olecranon, ulnar crest, capitulum radii and ulnar nerve were marked clearly. After the anatomy was marked, three waypoints were marked further. Iodine was used to fix all the markers. 4. Sequential approach: The primary task of arthroscopy was to enter joint cavity safely, accurately and easily, so the first approach was designed for this purpose. This approach was named as proximal anterolateral portal, which was located 2 cm proximal to the lateral epicondyle of humerus and 1 cm anterior to the lateral crest of humerus. As the sharp knife only cut skin, the arthroscopic sheath puncture slide into joint cavity while staying close to anterior cortical bone of the humerus and aligning with the front of elbow joint. During the entering of male joint cavity, there was a sense of breakthrough. For females, however, there was usually no. Meanwhile, the number of times required for successful puncture was recorded. The arthroscope was placed, and the handle was locked in parallel with forearm to observe the joint capsule in front of trochlea humeri. The out-side-in technique was used to pierce lumbar needle through the body surface marker. The second approach was proximal anteromedial portal. The surface marker was located 1 cm in front of the condylus medialis humeri and 2 cm above. The lumbar puncture needle point within joint is located on the plane line of coronoid process. With ulnar side skin cut open, the first exchange rod was inserted in the direction of the needle, and the lens were taken out of the first portal. With the arthroscopic sheath kept in elbow joint cavity, the second exchange rod was inserted. Then, the arthroscopic sheath was withdrawn. The first exchange rod on the ulnar side of arthroscopic sheath was placed in joint. At this time, the second portal (the ulnar side observation portal) has been established. Under arthroscopic observation, the outside-in technique was still used to establish the third portal (anterolateral portal) . The body surface marker points were located 1 cm in front of and 2 cm below the humeral ectocondyle. The needle point was located in front of the the capitulum radii articular surface, and the needle insertion direction was from anterolateraland pointed slightly to rear. With skin cut open, straight tong was extended into joint cavity in the direction of needle, which was used to bluntly dilate the capsule and lateral soft tissue. 5. Arthroscopic examination and arthroscopic processing: The first portal can be used to observe ulnar coronoid process, trochlea humeri, coronal fossa, the front of humeroulnar joint, capitulum radii, capitulum humeri and humeroradial joint. The second portal was the main observation portal, which can be used for the arthroscopic examination of capitulum humeri, capitulum radii, the front clearance and radial capsule of humeroradial joint. In this group of patients, there were 10 cases of radial capsule rupture, 9 cases of lateral humeroradial joint fold and 5 cases of obvious synovitis. The planer was placed in the third portal, and the lateral joint capsule was cleaned to reveal the ECRB. At this time, the indwelling exchange rod of the first approach was placed in the muscle of the front of ECRB, and soft tissue was cut forward to fully reveal the ECRB. The color of ECRB was dull, or only a small amount of shiny tendon stripe was scattered. The ECRB was cleaned with planer and radiofrequency. The range of cleaning was up to external epicondyle of humerus and to the midline of capitulum radii. Attention should be paid to avoid the damage to lateral collateral ligament. After the clean-up, the observation approach was switched to the first portal, and radiofrequency was further used to clean external humeral epicondyle. Planer was used to uncover fresh cortical bone, and microfracture operation was carried out. The microfracture roles were longitudinally arranged on humerus. Ⅲ. Postoperative rehabilitation. In the first stage after surgery, the upper limbs are suspended for 3-5 days, and a small range of activities are performed in a painless state, in conjunction with hand and wrist muscle strength and activity training. The second stage is the training of the range of active and passive joints after 2 weeks. After 4 weeks, the intensive activity training is started until 6 weeks to reach the normal flexion and extension angle. Slight resistance exercise began 4-6 weeks after surgery, and resistance muscle training began at 8 weeks. 10-12 weeks before returning to work or exercise. Daily stretching of the elbow and wrist stretching contributes to the improvement of function and symptoms. Ⅳ. Follow-up and evaluation. The patients were followed up for 12-38 months. The VAS was used to evaluate the degree of pain in the lateral elbow. The hypalgesia and motion pain were recorded respectively (average scores were obtained in two cases, and 0: no pain was scored as 1 point; 3 or less: slight pain, can bear to be recorded as 2 points; 4-6: pain and affect sleep, can still bear to be recorded as 3 points; 7-10: patients have progressively strong pain, pain is unbearable, affecting appetite and sleep is recorded as 4 points.) ; QUICK-DASH function score to evaluate elbow joint function, elbow function without difficulty 1 point (≤25 points) ; difficulty 2 (25-75) points; extremely difficult 3 points (>75 points subjective satisfaction survey was used to evaluate patient satisfaction, and the satisfaction at the last follow-up was recorded. The satisfaction was divided into four levels: 1 point for satisfaction, 2 points for satisfaction, 3 points for general satisfaction, and 4 points for dissatisfaction; The muscle strength was evaluated by grip strength, with 5 points (same as the healthy side) being 1 point, 4 levels (slightly lower than the healthy side) being 2 points, and grade 3 and below (significantly lower than the healthy side) being 3 points. Based on the above four aspects, the total score is: excellent, 4-5 points; good, 6-7 points; medium, 8-9 points; poor, 10-14 points. The preoperative course, operation time, number of successful first punctures, and postoperative complications were recorded. Ⅴ. Statistical analysis. The obtained data of this study were analyzed by SPSS 20 software. Numerical variables are expressed in terms of mean ± standard deviation. VAS, muscle strength scores, and QUICK-DASH were all analyzed by paired t test. The mean of the number of times required for the first 10 cases and the last 18 cases of the first approach puncture was compared by an independent t test, which was statistically significant at P <0.05.

1. The comparison of evaluation result between preoperative and last follow-ups: The VAS score ( hypalgesia) of 28 patients with tennis elbow was (4.97±2.62) before the operation, and the score during last follow-up was (0.75±1.08) . The VAS score (active pain) was (6.43±1.45) before the operation, and the score during last follow-up was (1.00±1.28) . Before the operation, the muscle strength score was (2.61±0.50) , and the score during last follow-up was (1.18±0.39) . The QUICK-DASH score was (71.95±11.87) before the operation, and the score during last follow-up was (14.04±11.70) . The VAS scores (night pain and active pain) , muscle strength scores and QUICK-DASH scores before and after final follow-up were statistically significant (P <0.01) . Before the operation, there were 17 people with preoperative muscle strength below grade 3 and 11 people with muscle strength at grade 4. During the final follow-up, there were 23 patients with muscle strength at grade 5 and 5 patients with muscle strength at grade 4. According to the scoring criteria above, there were 18 patients with excellent follow-up score, 6 patients with good score, 4 patients with medium score and no patient with poor score . The excellent rate of three-way technique in the treatment of tennis elbow was 85.7%. 2. Other circumstances: In 28 tennis elbow patients who were treated with elbow arthroscopy through anterior three-portal technique, the average time required for first-pass puncture success of first approach was (1.96±1.27) , and the average operation time was (48.57±10.03) minutes. The mean time required for first-pass puncture success of first approach in the first 10 patients was (3.30±0.95) , and the mean time required for first-pass puncture success of first approach in the last 18 patients was (1.22±0.43) . There was statistical significance in between (P <0.01) . With the increase in the number of surgical procedure, the time required for successful first-stage puncture showed a downward trend, which became stable after 10 cases of operation. There was no complication such as nerve damage in this group.

The anterior three-portal sequential technique in the arthroscopic surgical treatment of recalcitrant Tennis Elbow is feasible, safe and effective.