I have been in practice now for 17 years, and over the past 10 years I believe that I have gained a better appreciation about what I can offer patients who have had previous rotator cuff surgery who are referred to me for an opinion regarding their continued pain and dysfunction. A realization for me was that “stiff shoulders are painful shoulders,” and one key component to revision rotator cuff surgery is to “free up” the joint. This loss of motion can be subtle and you have to look for it, but it is real; if you examine the patient under anesthesia, you will also start to appreciate how real it is.

These patients often have a workers’ compensation claim, have not been able to work typically for more than a year, may have already had multiple surgeries, and are frustrated and discouraged; other physicians have either told them there is nothing wrong with them or that there is nothing that can be done for them. Subsequently, some of these patients can be a little hardened. However, it is important to listen to the patient and his or her story. First and foremost, the story is about how much pain they continue to have and that they can’t believe they will have to live the rest of their life with this amount of pain. Sometimes other surgeons have told them that the rotator cuff has healed, the magnetic resonance imaging (MRI) is fine, there is no recurrent tear, and there is no obvious explanation for their pain.

However, the question is whether or not someone has appropriately examined the patient before truly saying there is nothing objective to explain the pain; it may not be found on the MRI, but more often than not it can be found on the examination. It amazes me when patients report that previous physicians did not examine their shoulder before they gave them final treatment recommendations or told them there was nothing that could be done. The exam is critically important, and I fear it is becoming a lost art with all of the other tools we have now for diagnosing patients. Not infrequently these patients will have a shoulder contracture (i.e., loss of full shoulder motion). It can be subtle, but typically they will have lost the last 20 degrees or so of forward elevation and 10 to 20 degrees loss of full external rotation at their side. The loss of range of motion is not like a stiff shoulder seen with adhesive capsulitis of a native normal shoulder, but rather a postsurgical contracture of the shoulder. I also have the patient lie down in a lateral position while I look at the internal rotation with the arm in abduction (i.e., sleeper position) and look for the degree of posterior contracture. This may be the only motion loss that is found, and if the patient otherwise has a healed rotator cuff there is a possibility that with a focused rehabilitation program with a special focus on posterior capsular stretching, he or she could improve without any further surgery.

I believe that this overall entity of “postsurgical contracture” is one important source of patients’ continued pain. I ask about their previous postoperative rehabilitation protocol. Did they start range of motion right away, or how long were they immobilized? If they were immobilized in a standard sling hugging the arm for 2 months postoperatively before they started any motion therapy, then it is no surprise they now have loss of external rotation motion at their side with a rotator interval contracture. My impression is that once patients develop a chronic postsurgical rotator interval contracture, it is going to be very difficult for them to gain this motion back with therapy alone. I feel more reassured that a contracture release of the rotator interval would help with their motion. However, if they report that they started range of motion exercises right away and they are still stiff, I am less reassured that surgery will help and wonder about their compliance/motivation or if the technique of the surgery itself is a culprit. I pay close attention to the strength of the rotator cuff as well, but sometimes they are just too painful and aggravated that you cannot get a good examination, and maybe this is why other doctors have stopped with the examination. Certainly if they have weakness on examination, I am concerned about a retear, and the MRI can be very helpful to clarify this. The patient exam can be deceiving; the patient may have a much larger retear on MRI than appreciated with the examination or vice versa. Lastly, I look specifically for any palpable pain at the acromioclavicular joint; typically there is not unless they have had a previous distal clavicle resection. If that is true, I make sure to determine if the resection is complete. I have seen cases where the distal clavicle has been partially resected only with bone left behind, and I talk to the patient about adding a revision distal clavicle resection to the procedure. Like any referral patient who has had previous surgery, it is important to look at all of the imaging and past surgical records and to investigate for infection. When reviewing the surgical reports, I look for repair technique, whether the surgeon talks about releases and mobilization of the tendon for repair, the status of the biceps tendon, and the condition of the articular surfaces to see if arthritis is a potential reason for the shoulder stiffness.

As far as imaging, a standard MRI is adequate. I do not often get MRI arthrograms. I talk to the patient about any atrophy and chronic fatty changes, risks, and benefits, as you would with primary rotator cuff repair discussions. I use intra-articular or subacromial injections as diagnostic and therapeutic tools if I am unsure whether the pain is joint based (i.e., arthritis) or subacromial based (i.e., rotator cuff/adhesions).

Deciding on whether to pursue surgery is a shared decision-making process with the patient. These patients fall into a revision category, so I explain that we cannot make any guarantees about the success of the procedure but that the primary reason to have surgery is for pain control. If we can decrease the pain by 50% or more, the procedure would be considered a success. I talk to patients about the contracture release being part of the procedure and full or partial repair of the recurrent rotator cuff tear. I have not yet needed to entertain other surgical options such as the subacromial balloon, superior capsular reconstruction, and allograft augments. If I do not think it is worth trying a revision rotator cuff repair, which is fairly rare, then my thought process leans toward the feasibility of a tendon transfer procedure. Currently my preferred approach is a lower trapezius tendon transfer in addition to addressing the stiff shoulder. I explain that I cannot guarantee that the patient will get back to work. This surgery will not “fix their job,” but they may have a better idea of their ability to return to work by 6 months postoperatively. I tell them that I will only plan to do one more surgery for them for the rotator cuff and to gain motion.

Some of the key findings at the time of surgery are a dense rotator interval contracture, anterior capsular contracture, a biceps tendon scarred into the rotator interval providing a hard block to external rotation, and dense and abundant subdeltoid adhesions. When performing the contracture release in these patients, it is not one size fits all, so it requires close attention be paid to what appears to be normal and what is densely contracted. The rotator interval is often very thick and needs to be released to the base of the coracoid. The biceps tendon or the stump of the tendon sometimes is found scarred into the interval and needs to be released. The anterior capsule needs more close attention because it is not always completely contracted as would be seen in an adhesive capsulitis. Rather, it is typically seen contracted along the anterior glenoid to the 4 or 5 o’clock position of a right shoulder, but most often the inferior capsule has a normal compliance and laxity to it. Thus I stop my release when it looks to be a normal state. There is often scarring of the supraspinatus tendon to the superior labrum, so I am careful to re-create the normal “gutter” above the labrum, being careful to protect the suprascapular nerve. Usually the infraspinatus is not scarred to the labrum. In my experience the posterior capsule is not necessarily involved, and I believe that the loss of internal rotation is more often related to the dense subacromial adhesions and a thick band that runs from posterolateral to posteromedial. This band can be very thick, so much so that a normal shaver cannot be used to remove it and it needs to be cut surgically.

I have three goals with the subacromial work. The first is for a complete bursectomy and extra-articular release of adhesions to the rotator cuff. The bursa is resected to the base of the coracoid and the spine of the scapula. The second is for removal of subdeltoid adhesions. I am very aggressive in re-creating the normal “gutter” between the deltoid and the rotator cuff circumferentially. I do not want to see any adhesions pulling on the rotator cuff. I remove free-floating or prominent sutures, but if they are otherwise buried in the tendon, I leave them alone. Third is revision repair of the rotator cuff without too much tension on the cuff (i.e., let it lay where it wants to). For patients with a retear, I believe that the pain is likely coming from both the retear and the stiffness, so I treat both. Following both intra-articular and extra-articular work, I manipulate the shoulder to be sure I have gained full range of motion.

Postoperatively, for a revision rotator cuff repair with contracture release, I follow a standard rotator cuff protocol that I typically use for small rotator cuff primary repairs, even if it was a large or massive rotator cuff repair (i.e., full active assisted forward elevation and 30 degrees external rotation at the side with sling immobilization for 6 weeks). I err on the side of letting the patient move early and not get stiff again. I cannot say for sure whether the contracture release or the revision cuff repair has allowed for improvement of their pain. For a patient with no revision rotator cuff repair and contracture release, I use a contracture therapy protocol (i.e., full unrestricted active assisted to active range of motion with sling immobilization for 2 to 4 weeks for comfort only). I talk to patients preoperatively about arranging therapy every working day for the first 2 weeks following surgery. They understand that they are responsible for doing therapy on their own four to six times per day as well and that the therapist is there only to instruct and provide guidance. I believe that the emotional and social support of the therapist is important in the early phase. At the 2-week mark, they transition to therapy only two or three times a week depending on how much support they need. For postoperative pain modalities, I have only used oral pain medications and antiinflammatories. If they were on chronic pain medications, we restart their previous dose and make sure that the pain management specialist is on board to manage narcotic treatment.

We have been looking at my results over the past 5 years and, despite this being a difficult patient population, we are seeing significant improvement in pain scores for both patients who did or did not have a revision rotator cuff repair at the time of revision surgery, and modest improvements in their motion, which is not surprising because more often than not it is subtle at initial presentation. Pain reduction is better in patients who had fewer previous surgeries (i.e., one surgery vs. more than one previous surgery). Approximately half of my patients had a workers’ compensation claim, and these patients still had a significant reduction in pain. These patients can be very grateful despite not having a perfect overall result—they are so relieved to have less pain and they report that their shoulder is “less stuck.” I do not think that patients tolerate loss of external rotation at the side very well; if you think about it, they have to maneuver the shoulder in awkward ways to accommodate the contracture, which leads to poor mechanics and irritability of the shoulder.

In summary, I think these patients were more difficult to manage earlier in my career because I did not necessarily appreciate the issue around the contracture of the shoulder. Over the years, I have taken a more aggressive approach to the shoulder contracture. We owe it to these patients to provide a full clinical assessment, including the physical examination. Stiff shoulders are painful shoulders, and surgery after previous rotator cuff repair is not such a dismal outlook as I once believed when I first started practicing.