The patient history is critical to selecting the right candidates for revision rotator cuff repair surgery. The presence or absence of pain and its location, character, intensity, and timing all need to be assessed. Periscapular and cervical spine pain, which may be reported by patients as “shoulder pain,” are often indicative of inadequate
strengthening, poor posture, or cervical spine pathology as opposed to pain related to a failed rotator cuff repair that usually localizes to the anterior or lateral deltoid. Sharp, intermittent pain with a mechanical component, pain that is worse with activities, and pain at night are other characteristics associated with a structural rotator cuff repair failure.

It is critical to understand the patients’ own perception of their function and their ability to perform activities when assessing the role for revision rotator cuff repair. It is not uncommon to find patients with significant, objective functional deficits following rotator cuff surgery, who are nevertheless very satisfied with their ability to perform daily living and recreational activities. Others may have full active range of motion and strength in the shoulder but feel functionally limited because of persistent pain, perceived weakness, or lack of endurance.

The chronicity of the symptoms and timing of the initial repair will give some insight into the quality of the tendon tissue, likelihood of rotator cuff muscle degenerative changes, and healing potential of the repair at the time of the first surgery. If these details surrounding the first surgery suggest a very poor-quality tendon associated with preexisting atrophy and fatty infiltration of the rotator cuff muscles at the time of the first surgery, then another attempt at repairing the rotator cuff tear may be futile.

Finally, understanding how the postoperative rehabilitation was performed with the first surgery may give insight into the cause of the failure. Prior use of an abduction brace, the timing of initiation of passive stretching and strengthening exercises, and the report of a specific traumatic event in the postoperative period may all give insight into what could be modified or avoided the next time to prevent a second failed surgery.

The physical exam for patients being evaluated for possible revision rotator cuff repair begins with a visual inspection of the shoulder girdle. Surgical scars and increased warmth, swelling, or induration around the operative site should be noted. If a prior open rotator cuff repair has been performed, the deltoid origin should be carefully assessed by inspection and palpation looking for a possible deltoid detachment, which could be a source of pain or weakness (7). Particular attention should be paid to any asymmetry or muscle wasting in the infraspinatus fossa as this is usually indicative of a significantly atrophied infraspinatus muscle that is associated with a high likelihood of revision rotator cuff repair failure (2). Fixed humeral head superior migration or superior escape of the shoulder is important to identify at this stage, since this physical exam finding is also predictive of rotator cuff repair failure in the setting of either primary or revision surgery.

Active and passive range of motion should also be assessed and limited active motion due to weakness or pain must be distinguished from limited passive motion due to stiffness. Pain that is limiting assessment of motion may be alleviated by a subacromial lidocaine injection. This will facilitate the assessment of active motion for the patient by eliminating pain-related inhibition of function and facilitate the assessment of passive motion by the physician by limiting patient guarding. Weakness in the anterior and posterior rotator cuff can be assessed by performing resistance muscle strength testing with the belly-press and lift-off tests for the subscapularis (8), resisted external rotation with the elbow in adduction for the infraspinatus (9), or resisted external rotation with the arm in 90 degrees abduction for the teres minor (10). Positive belly-press and lift-off tests, or external rotation lag signs, represent profound weakness and almost complete dysfunction of the rotator cuff muscle involved. In the largest tears, the patient may be unable to maintain the position of the arm that has been passively rotated externally, and the affected arm “drifts” back to the neutral position. These findings may be due to the location of the tear but may alternatively be related to degenerative muscle changes and/or neurologic damage within the surrounding, intact rotator cuff musculature. In such cases where the lag sign appears to be due more to the degenerative muscle changes than due to the location of the tear itself, repair of the tear is unlikely to improve the existing lag sign and is an indication of a higher likelihood of a poor outcome with revision rotator cuff repair (1).

Nicotine and poorly controlled diabetes have both been associated with impaired healing in general and worse outcomes following rotator cuff surgery (3, 4, 5, 6). Therefore, an attempt at revision surgery to repair a previously failed rotator cuff repair should be strongly discouraged in anyone who is actively smoking or currently using any nicotine product including nicotine patches. For diabetic patients, pre- and postoperative blood glucose levels should be tightly controlled with a hemoglobin A1C level of 6.0 or less (4).

Numerous immunomodulatory medications, used in patients with rheumatologic or chronic systemic inflammatory diseases, and patients who have undergone organ transplantation, have the potential to suppress the natural and desirable postoperative healing inflammatory response required to achieve successful healing. These medications can also increase the chances for infection that may have contributed to failure of their initial shoulder surgery or predispose them to complications with their revision. Due to the recent, rapid increase in the number and types of these immunomodulatory medications available, a complete understanding of how
these medications may alter the healing process is not yet well defined. Additionally, the risks associated with discontinuation of these medications perioperatively may be greater than the possible benefits. The best course of action is clear communication with the prescribing doctor to minimize the number of immunosuppressive medications being taken, while continuing to effectively manage their other associated medical conditions.

Any preoperative evaluation for possible revision rotator cuff repair should start with plain radiographs. While further imaging studies are likely to be needed, basic radiographs may direct the surgeon’s choice of subsequent studies or immediately rule out revision rotator cuff repair as a viable option all together, such as in patients with bone changes suspicious for chronic osteomyolitis, an arthritic glenohumeral joint, or patients with fixed superior migration of the humeral head (Fig. 27-1).

MRI is the most commonly obtained follow-up imaging study to evaluate the shoulder girdle when considering revision surgery. Concern is often given to the ability to adequately visualize the rotator cuff when an MRI is performed in the setting of retained metallic suture anchors in the humeral head. While metal artifact often obscures the bone immediately adjacent to the anchor(s), it has not been our experience that retained metal anchors in the humeral head significantly compromise the ability to evaluate the essential characteristics of a failed rotator cuff repair, and for that reason, this is still our imaging modality of choice (Fig. 27-2). The tear location should be assessed to help the surgeon to develop a surgical strategy (open vs. arthroscopic approach; likelihood of the need for soft-tissue releases; fixation strategies; etc.) and correlated with physical exam findings. Greater degrees of rotator cuff muscle degeneration have been demonstrated to correlate with lower structural healing rates and worse functional outcomes following surgery (1, 2); therefore, these characteristics should be specifically and carefully evaluated on MRI in order to accurately estimate the chances for success if revision repair is to be undertaken. The magnitude of medial tendon retraction seen on MRI should also be assessed. Although it has not been shown to correlate directly with repair success, a retracted tendon is more likely to require extensive soft-tissue releases to mobilize the tendon prior to repair and is more likely to result in a repair under higher tension, which is believed to be one of the primary causes of postoperative repair failure (11).

It is important to note that superior migration of the humeral and the acromiohumeral distance are not characteristics well evaluated using MRI or CT scan since superior translation of the humeral head may occur in patients when they lie down and gravity is no longer pulling the humeral head inferiorly. Such dynamic superior migration is not known to be associated with lower rotator cuff repair success rates and is distinctly different from static or “fixed” superior migration of the humeral head, which is most effectively identified on plain radiographs and will be present even when the patient is standing/sitting.

Computed tomography (CT) arthrogram is another commonly used imaging modality to evaluate the size and location of a rotator cuff tear when planning revision surgery. In addition to being faster to obtain and less expensive, it is also an easy alternative for people who are unable to obtain MRIs for other reasons (claustrophobia, retained intracranial clips, cardiac pacemakers, etc.). This modality is very effective for determining the size, location, and tendon retraction associated with a rotator cuff tear. However, it can
suffer from problems with metal artifact, and additionally, several studies have suggested that it is more difficult to assess degenerative muscle changes with this modality than with MRI (12). Although rotator cuff tendon integrity can be assessed with this imaging method, some attributes of the tendon such as the presence of tendonosis, intratendonous delamination, and tendon thickness can be difficult if not impossible to assess.

Ultrasound is another less expensive imaging option for evaluating rotator cuff tears that has become the study of choice for some institutions. Performed by a skilled ultrasonographer, this modality can be used to determine tear size, location, the magnitude of tendon retraction, and even the degree of rotator cuff muscle degeneration and intrasubstance changes (tendonosis) within the tendon itself. The greatest limiting factor to use of this modality is that it is highly user dependent. Detailed and accurate information about the tear characteristics requires skilled and experienced group of people performing the study and interpreting the findings. Most people experienced with using this modality emphasize the importance of evaluating dynamic imaging— previously recorded video or preferably real-time images as opposed to static pictures—in order to accurately assess the tear. Body habitus (very muscular or morbidly obese patients) or limited motion of the shoulder can also make adequate imaging of the rotator cuff difficult if not impossible with this modality.

In addition to the standard lab work needed for general preoperative medical clearance, a C-reactive protein and erythrocyte sedimentation rate as well as white blood cell count should be obtained to rule out occult infection on any patient planned for revision rotator cuff repair. Patients with known diabetes should also obtain a hemoglobin A1C to confirm that their blood glucose level is well controlled. For those patients with a recent history of smoking, consideration should be given to checking the blood for a nicotine level so that the true smoking status of the patient at the time of surgery is known.

Many aspects of the surgical techniques and postoperative rehabilitation protocols employed are actively debated even among people who perform complex primary and revision rotator cuff repairs on a regular basis. Unfortunately, very little level I, II, or III evidence is available to support one surgeon’s preferences over any others’. The following paragraphs outline some of the more frequently debated aspects surrounding how to achieve success with a complex primary or revision rotator cuff repair. Regardless of the approach and techniques used to achieve a revision rotator cuff repair, we believe that several general objectives should be considered and balanced against each other to optimize the chance for successful repair healing (Table 27-3). As long as the general repair objectives are followed, the choice of approach, fixation method, repair construct, etc. can be equally accomplished by a number of different methods.

The current literature, as well as our personal experience, demonstrates no difference in outcomes of rotator cuff repairs when performed open versus arthroscopically (13). The decision of arthroscopic versus open revision rotator cuff repair should be based on an accurate preoperative assessment of what may need to be done at the time of surgery (direct repair, graft augmentation, tendon transfer, etc.) and an honest assessment by the surgeon of which type of approach provides him or her with the best opportunity to effectively accomplish these goals.

We do not specifically distinguish between a mini-open approach (without detaching the deltoid origin) and a full open approach in which the deltoid origin is released from the acromion. Since open approaches are more often required when graft augmentation or tendon transfers are anticipated, we feel that a full view and easy access to the entire surgical field are of top priority. As such, we do not hesitate to detach the deltoid origin as needed, recognizing that the risk of deltoid repair failure is quite low if meticulous attention is paid to proper deltoid release during the approach and proper deltoid repair during the closure, followed by a reasonable rehabilitation protocol in the early postoperative period.

No clinical difference in outcomes has ever been demonstrated when comparing one type of fixation (various anchors or bone tunnels) to another. Most importantly, a number of biomechanical studies (11, 14, 15) and clinical observations (16) have demonstrated that when rotator cuff repair failure occurs, it almost always occurs at the suture-tendon interface and not at the bone fixation interface. Therefore, the relatively small differences in failure loads that may be demonstrated in vitro with different types of fixation are likely to be clinically irrelevant. For this reason, we do not have any preferred type of fixation or anchor.

Single-row repair, double-row repair, suture bridge/transosseous equivalent repair, and tension band repair constructs have all been described and each has its proponents and detractors. Many studies exist comparing various aspects of failure mode, fixation strength, and footprint contact forces that occur with each different type of repair configuration. A global review of the literature cannot support the use of any one repair construct over another, and clinical outcome studies have not shown improved outcomes with any one type of construct.

We do not strictly adhere to the use of one repair construct over another but rather make a choice about the repair construct based on the tendon quality, bone quality, repair tension, and footprint coverage that can be achieved at the time of repair.

Most rotator cuff repair failures (primary and revision) can be attributed to a combination of two things: (a) insufficient repair strength and (b) an insufficient biologic healing response. In the failed rotator cuff repair, the combination of these two factors results in the rotator cuff repair pulling apart before the healing tissues have enough mechanical integrity to take over the load-bearing function that is required from them. With this in mind, revision rotator cuff repairs would seem the perfect indication for use of a mechanical and/or biological augmentation device. At this time, there are a number of mechanical augmentation devices on the market for reinforcing the rotator cuff repair construct. Despite FDA approval, at the time of this writing, no device has demonstrated reduced rotator cuff repair failure rates, improved healing, or improved clinical outcomes in human application as compared to a nonaugmented repair construct (17). Table 27-4 lists the nonsynthetic extracellular matrix (ECM) devices currently on the market in the United States that are FDA approved for rotator cuff repair augmentation.

In the absence of available clinical outcome data in human application, we use the available data from in vitro and in vivo literature to guide our decision making about when to use a mechanical augmentation device. This literature suggests that if applied properly, graft augmentation can off-load the underlying tendon-bone repair site by approximately 25% to 45% and that these benefits are greatest when the strength of the underlying tendon repair is relatively weak (18). Published cadaveric models investigating Dermal ECMs (14) and synthetic graft materials (15), as well as additional ongoing work by our group investigating alternate natural/synthetic composite graft materials, have demonstrated 10% to 20% increases in the load to failure strength for mechanically augmented rotator cuff repairs and a reduced rate of repair failure at the tendon-suture interface. Additional in-vivo studies of a synthetic poly-L-lactate acid device (X-Repair^tm, Synthasome Inc., San Diego, CA) have demonstrated reduced gap formation at the site of rotator cuff repairs when used to augment repairs in
a canine model (19). Most of the devices have shown that they are well tolerated in limited human application, although early generation of the porcine-derived small intestine submucosa (SIS) device, Restore (Depuy Inc.), was associated with significant sterile inflammatory reactions in approximately 20% to 30% of patients (20, 21).