Focal avascular necrosis of the humeral head after arthroscopic suprapectoral biceps tenodesis

Avascular necrosis (AVN) of the humeral head is a debilitating process resulting from loss of blood circulation to the bone leading to cell death, subchondral collapse, and articular degeneration.^8,18 Although AVN of the humeral head can be caused by trauma, it can also be atraumatic and multifactorial. As with AVN of other joints, the most common precipitators include long-term corticosteroid use, alcohol abuse, radiation, sickle cell disease, storage diseases, vasculitis, connective tissue disorders, or idiopathic.^13,18 Rapidly progressive, global humeral head AVN has been reported following arthroscopic rotator cuff repair (RCR) with both bioabsorbable suture and metal anchors, resulting in Cruess stage IV and V osteonecrosis within 12 months of surgery.^3,5,7,13 Notably, in each of these reports the biceps tendon was not tenodesed.^3,5,7,13 Thus far in the orthopaedic literature there have been no reports of focal humeral head AVN following arthroscopic surgery, specifically in relation to bicep tendon pathology or treatment.

Here we report a case of focal humeral head AVN following an arthroscopic suprapectoral biceps tenodesis in a patient with no other risk factors. To our knowledge, this is the first report of this rare complication in this setting.

Case report

Informed consent was obtained from the patient. This is a 49-year-old right hand dominant female who presented with left shoulder pain. Past medical history was notable for hypothyroidism and depression, with no history of metabolic disease, prior steroid use, vasculitides, or connective tissue disorders. Social history was negative for alcohol, tobacco, or drug use, and she worked as an optometry technician. She previously had gastric bypass surgery, a right shoulder RCR, and a C6-7 anterior cervical discectomy and fusion (ACDF), all of which occurred without complication.

On examination, there was bicipital groove and AC joint tenderness, with positive Speed, Yergason, and cross body adduction testing. Imaging was consistent with AC joint arthrosis, a superior labrum anterior and posterior tear and rotator cuff tendinopathy. She underwent a left shoulder arthroscopic suprapectoral biceps tenodesis, subacromial decompression, and distal clavicle excision in June 2017. Intraoperatively, the biceps tendon was noted to be bulbous in nature with inflammation and tearing at the insertion. The tendon was marked and released. Diagnostic arthroscopy was completed noting the humeral head and glenoid cartilage were healthy appearing and intact. An onlay biceps tenodesis was performed in the suprapectoral region with an Arthrex 4.75-mm BioComposite SwiveLock (Arthrex, Naples, FL, USA) anchor. Estimated blood loss was minimal.

The patient had an unremarkable postoperative course until 3-month follow-up. She reported sudden increase in pain while doing physical therapy with associated decreased mobility. Unfortunately, she lost to follow-up for her shoulder over the next year. In the interim, she underwent a revision C6-7 AC discectomy and fusion in February 2019 for neck pain and left upper extremity radicular symptoms, attributed to a pseudarthrosis. She then re-presented in December 2019 for evaluation of deep left posterior shoulder pain with an associated catching sensation and difficulty moving her arm. On examination, the patient had 140° active abduction, 140° forward flexion, 80° external rotation, internal rotation to the level of the gluteus, and was unable to tolerate further passive range of motion. Rotator cuff strength was 5/5, with pain and catching on Neer’s and Hawkins examination, pain with Speed and Yergason maneuvers, and a positive empty can test. Patient-reported outcomes were as follows: VAS pain score 8, American Shoulder and Elbow Surgeons score 22, Oxford Shoulder Score 14, and Simple Shoulder Test 8.3%. Imaging demonstrated focal AVN of the humeral head with a 2 × 2 cm area of subchondral collapse consistent with a Cruess stage II AVN (Figs. 1 and 2), visualization of the prior biceps tenodesis location, and no involvement of the glenoid or rotator cuff (Fig. 3). Given the patient's age, desired activity level, and persistent symptoms despite conservative treatment, the patient was indicated for surgery. Multiple treatment options were discussed including a biologic resurfacing with matched osteochondral allograft vs. a partial metallic humeral head resurfacing. Due to concerns regarding a potentially compromised ability to heal an osteochondral allograft in the setting of AVN, the decision was made to proceed with a partial metallic humeral head resurfacing.

Figure 1.

(A) Left shoulder coronal CT with 2.4-mm subchondral collapse. (B) Left shoulder axillary CT view of subchondral collapse. CT, computed tomography.

Figure 2.

3D CT imaging demonstrating 24.0-mm region of focal AVN and subchondral collapse of the left humeral head. 3D, 3-dimensional; CT, computed tomography; AVN, avascular necrosis.

Figure 3.

(A and B) Left shoulder coronal and axillary proton density and (C and D) T2 demonstrating subchondral collapse consistent with avascular necrosis.

Surgical intervention

The patient underwent left shoulder arthroscopy and resurfacing arthroplasty in 2020 which was a notably delayed timeline due to the COVID pandemic. Diagnostic arthroscopy confirmed a 25-mm area of AVN with subchondral collapse involving the anteromedial humeral head. Utilizing a deltopectoral approach with subscapularis tenotomy, a resurfacing arthroplasty of the humeral head was performed (Fig. 4). The final implant was placed—Arthrosurface Shoulder HemiCAP (Arthrosurface, Inc., Franklin, MA, USA) system 25-mm articular component with a 2.5 × 2.5 mm offset (Fig. 5). The procedure was tolerated well without complications. Postoperative imaging demonstrated a well-placed component (Fig. 6).

Figure 4.

(A) Lesion measuring approximately 25 × 25 mm. (B and C) A 25-mm sizer and central guide pin drilled centrally into the lesion. (D-F) Central screw placement, overreaming to prepare bed, and final preparation of defect.

Figure 5.

Final implant placement.

Figure 6.

Immediate postoperative imaging demonstrating final positioning of metallic resurfacing component.

Postoperative course

The patient’s postoperative course was uneventful. Physical therapy was started with a subscapularis repair protocol. By 6 months postop the patient had improvement in her deep shoulder pain and range of motion to 145° forward flexion and 120° abduction; however, she had a positional catching sensation and sensitivity to the anterior shoulder. She was referred to pain clinic due to concern for chronic regional pain syndrome. At 1 year postop the patient reported 0/10 pain, had active range of motion to 175° forward flexion and abduction, 60° external rotation, and internal rotation to L5 with 5/5 cuff strength and no pain with provocative maneuvers. Patient-reported outcome scores were as follows: visual analog scale (VAS) pain score 0, American Shoulder and Elbow Surgeons score 82, Oxford Shoulder Score 44, and Simple Shoulder Test 58.3%. Imaging was without evidence of hardware compromise (Fig. 7).

Figure 7.

Four views of the left shoulder at 1 year postoperatively demonstrating retained resurfacing component without compromise, with maintained alignment and positioning.

Discussion

AVN of the humeral head can be debilitating. Although the cause may vary, the result is a loss of blood circulation to the bone leading to cell death, subchondral collapse, and articular degeneration. Thus far there have been no prior reports of focal AVN of the humeral head following biceps tenodesis.

Tendinopathy of the long head of the biceps (LHB) tendon is a common cause of anterior shoulder pain and causes may be secondary to over-use, degenerative or inflammatory causes, or occur as a result of trauma.¹⁶ The tendon originates at the supraglenoid tubercle and superior labrum; it traverses the humeral head prior to coursing through the bicipital groove in the proximal humerus, becoming extraarticular.

Controversy exists surrounding the optimal location and method of tenodesis in biceps tendon pathology. Biceps tenodesis can be performed via multiple techniques including arthroscopic intra-articular, suprapectoral, or open subpectoral. Reported complications of biceps tenodesis include length-tendon mismatch, loss of fixation, deformity, persistent pain, stiffness, infection, hematoma, neurologic injury, vascular injury, proximal humerus fracture, groove pain, implant failure, and reflex sympathetic dystrophy.¹⁹ Those who promote distal fixation via the open subpectoral technique report that excising the proximal tendon and removing the tendon from the groove limits the potential for postoperative pain from residual tenosynovitis. Sanders et al¹⁷ reported a significantly higher revision rate associated with proximal tenodesis techniques compared to fixation distally. However, multiple studies have reported no significant difference in pain relief, clinical outcome scores, range of motion, strength, or occurrence of Popeye deformity between the 2 techniques.^6,20

The blood supply of the LHB tendon is primarily from branches of the anterior humeral circumflex artery (AHCA) which course along the bicipital groove.^1,2 Variations in the blood supply, course, and distribution of the humeral head have been described.^10,11 The main blood supply to the humeral head was previously thought to be derived from the anterolateral branch of the AHCA. However, more recent cadaveric studies have shown that the posterior humeral circumflex artery provides more blood to the humeral head—notably in the superior, lateral, and inferior quadrants.⁹ The AHCA and its branches have been reported to be adherent to the humeral head in contrast to the posterior humeral circumflex artery, potentially lending them to be at increased risk of damage.9, 10, 11 Based on a cadaveric study, 7 variations have been described for the course of the AHCA. In 36% of specimens, the AHCA ran posterior to the bicipital groove and posterior to the LHB tendon prior to splitting into its branches. Contrast was infused to evaluate the intraosseous supply, identifying the anterior-superior and superior-medial aspect of the humeral head as a high-risk zone for AVN.^10,11

In the suprapectoral technique, the biceps tendon is fixed just distal to the exit from the glenohumeral joint in the bicipital groove and proximal to the pectoralis major tendon insertion.¹⁴ In one cadaveric study, a vascular branch originating from the AHCA vessels crossed the suprapectoral zone medial to lateral and deep to the biceps tendon in all 12 cadavers and was approximately 1.49 cm proximal to the superior edge of the pectoralis major tendon.¹⁵ This was reported to be a reliable landmark for tensioning in arthroscopic biceps tenodesis fixation. Additionally, the ascending branch, arcuate derivative of the anterolateral branch of the AHCA, is reported to be at theoretical risk of being damaged in arthroscopic fixation, specifically during soft tissue release at the bicipital groove and when performing a biceps tenodesis or during placement of anchors for an RCR.10, 11, 12

In one case review of 8 patients, rapidly progressive AVN developed following arthroscopic rotator cuff single and double row repair, using bioabsorbable suture anchors within 12 months of surgery. In this study, there was no identifiable risk factor or cause.¹³ Prior to this study, there were only 3 other reports of AVN of the humeral head following arthroscopic surgery.^3,5,7 In 2 of these studies, AVN following the index procedure was thought to be due to placement of multiple metal suture anchors in the setting of RCR.^3,7 In the third study, 3 patients developed Cruess stage IV and V AVN following arthroscopic surgery; however, no anchors or implants were placed in 2 of the 3 cases.⁵ In all these reports, the biceps tendon was not tenodesed.^3,5,7,13

In the case of our patient, AVN was localized to the anteromedial aspect of the humeral head, which is consistent with disruption in branches from the AHCA,⁹ with no identifiable risk factors. Given the variations described in the course of the AHCA, this was theorized to occur secondary to groove preparation and anchor placement from the suprapectoral biceps tenodesis performed in 2017 leading to a compromise in the blood supply to this focal portion of the humeral head.

Treatment following AVN of the humeral head typically involves arthroplasty. Bixby et al⁴ reported that treatment of AVN with resurfacing for more focal lesions provides a number of advantages: preservation of native cartilage and biomechanics, minimal bone resection allowing for easy conversion to other implant designs, decreased blood loss and operative time, decreased rate of periprosthetic fracture, implant loosening, and implant wear, and low revision rates. In our case, by 1 year postoperatively the patient had improved range of motion with 0/10 pain, regained the ability to perform occupational duties, and improvement in patient-reported outcome scores. We would therefore conclude that this is a reasonable treatment option in this small subset of patients.

Conclusion

This is the first case reported in the literature of focal AVN after suprapectoral biceps tenodesis. Although we cannot conclude with absolute certainty that the index surgery caused AVN, this patient lacked any risk factors and developed pain and debilitation within 1 year of surgery. AVN has been reported following RCR. Anchor placement for the suprapectoral biceps tenodesis is in very close proximity to standard rotator cuff anchors. Given the variations in the course of the AHCA, it is possible that anchor placement resulted in disruption of a branch supplying this portion of the humeral head and was the most likely source for this pathology. In addition, this case also highlights the importance of taking care in groove preparation to avoid injury to vasculature.

The reported findings emphasize the importance of recognizing focal AVN as a potentially rare complication following treatment of suprapectoral biceps tenodesis. Patients should be counseled that this is a possible risk factor of arthroscopic procedures, and in the setting of acute changes in pain and function during the postoperative period there should be a low threshold to obtain further imaging to evaluate for AVN. Humeral head resurfacing can be considered a viable treatment option in carefully selected cases of AVN.

Disclaimers:

Funding: No funding was disclosed by the authors.Conflicts of interest: The authors, their immediate families, and any research foundations with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.Patient consent: Obtained.