Pascal Boileau, MD, is Professor of Orthopaedic Surgery, Chairman of the Institute for Locomotion & Sports, and former Vice-Dean at Nice Medical University in France. He did two traveling fellowships in the United States, visiting the main shoulder centers in the 1990s. He is Past-President of the European Shoulder & Elbow Society and designer of several surgical techniques and shoulder implants, including the first anatomic shoulder prosthesis, the first shoulder fracture-prosthesis (anatomic and reverse), and the first humeral intramedullary nail with tuberosity interlocking screws. He is a pioneer in shoulder arthroscopy and developed many techniques, including the guided arthro-Latarjet procedure. He is involved in extensive clinical research and has published more than 200 scientific articles. Dr. Boileau is also the founder and organizer of the Nice Shoulder Course, one of the largest and most attended shoulder meetings, and has trained more than 100 shoulder fellows from all over the world.
Devin Lemmex, MD, is an assistant professor at the University of Manitoba in Winnipeg, Manitoba, Canada. He completed his medical school and orthopedic surgery residency at the University of Calgary. Dr. Lemmex completed his first fellowship in sports medicine at Duke University in Durham, North Carolina, and then traveled to Nice, France, where he completed his second fellowship in shoulder reconstruction with Professor Pascal Boileau.
The treatment of proximal humerus fractures (PHFs) is very challenging and has inspired a number of classifications and devices. From implants to rehabilitation, opinions still diverge concerning treatment options and best practice. A well-treated fracture is forgotten with time, whereas a mistreated fracture may lead to major disability.
Classification and indications
The Neer classification for PHFs, based on the four-segment classification of Codman, is very important to remember. It is our opinion that this is the best classification because it is founded on the pathophysiology of the musculotendinous units pulling on the greater and lesser tuberosities and the humeral diaphysis. This classification is also very important because it reminds us that 80% of PHFs are undisplaced, or minimally displaced, and do not warrant surgical intervention. In fact, there is some misunderstanding in the treatment of PHFs, mainly because surgeons focus on the humeral head and the possibility of osteonecrosis. Although this can be a problem, it is relatively easy to exchange a necrotic humeral head with a hemiarthroplasty (HA) or a reverse shoulder arthroplasty (RSA). The outcome will be very good if the greater tuberosity (GT) is intact and anatomically united. However, a posteriorly migrated and nonunited GT leaves the surgeon in an almost impossible situation and the patient will have an undoubtedly poor outcome.
A grossly displaced GT fracture is equivalent to a massive posterosuperior rotator cuff tear, in that the muscle and tendons retract very rapidly and definitively. In revision cases where the tuberosity was not originally reduced, it is almost impossible to salvage the fragment and obtain a functional shoulder. This is why the main reason to operate on a PHF is not for the humeral head but rather the GT fragment. All efforts should be made to achieve anatomic reduction of the GT to achieve normal rotator cuff function. In other words, the main reason for the surgeon to decide to operate a PHF is the risk of leaving a displaced GT and not the risk of necrosis of the humeral head. Although we know how to solve the problem of necrosed head, we do not know how to solve the problem of a cuff-deficient shoulder with tuberosity lysis, tendon resorption, and atrophied retracted cuff muscles.
Another important point is that there are two types of PHFs: those that occur in young patients with good bone stock from high-energy trauma, and those that occur from a ground-level fall in elderly patients with osteoporosis. Patients in the first group possess very good cortical bone, which allows rigid fixation to be achieved through various forms of osteosynthesis. However, in elderly patients the bone is often too weak to obtain good fixation, and the surgeon should be more inclined to perform arthroplasty in this setting. In other words, surgeons should avoid overindicating a shoulder replacement in the young with good bone quality and open reduction and internal fixation in the elderly with poor bone quality.
Imaging
The correct diagnosis and therefore the correct treatment cannot be determined without adequate radiographs. The best images to obtain are what has been described as the Neer series, which consists of a true anteroposterior view (where the glenohumeral joint line can be seen), a true lateral view, and a true axillary view.
Computed tomography scans can be helpful in difficult cases. However, surgeons should be aware that two-dimensional images may lead them to think that the bone fragments are more displaced than they are. Ultimately, this may lead to overtreatment. Ideally, three-dimensional images provide better and more realistic images that can help surgeons in the decision-making process.
Greater and lesser tuberosity fractures
Arthroscopic treatment is currently preferred in cases of isolated tuberosity displacement (>5 mm). If the surgeon does not have the skill to do this under arthroscopy, it can be done open. For lesser tuberosity (LT) fractures it is better to use a deltopectoral (DP) approach, whereas for GT fractures it is better to use an anterolateral transdeltoid approach.
It should be remembered that GT fractures are associated with anterior dislocations or subluxations. Very often there is an associated anterior labrum tear (i.e., a Bankart lesion). On the other hand, an LT fracture is very often the consequence of a posterior subluxation or dislocation. In this case, there may be a posterior labrum tear, and it is always better to treat labral lesions in the same surgical session. This is why arthroscopy is preferred in these isolated tuberosity fractures.
In addition, isolated tuberosity fractures are associated with an increased risk of stiffness. Thus it is very important to begin the rehabilitation process as soon as possible.
Mechanical considerations: Plate versus intramedullary nail
Intramedullary (IM) nailing of PHFs has been largely criticized in the past and abandoned in favor of anatomically contoured locked plating systems. Early-generation nail designs (first- and second-generation IM nails) were discarded because of their deleterious effects on the rotator cuff tendinous insertion and poor proximal fixation options for tuberosity stabilization. They were too big and bent, forcing the surgeon to enter the nail through the supraspinatus tendon. However, modern nail designs (third-generation, straight IM nails) allow the surgeon to preserve the superior cuff insertion and have a variety of proximal locking screw options to create a fixed-angle device with improved tuberosity fixation.
Modern (third-generation) IM nails are designed to fix the tuberosities instead of the humeral head fragment, which is stable once the tuberosities are reduced. The more anterior- and posterior-based proximal locking screw options allow for screws to be inserted perpendicular to the fracture line separating the tuberosities. This screw configuration allows counteracting the posterior and medial pull of the infraspinatus and teres minor muscles on the GT and anterior and medial pull of the subscapularis muscle on the LT. This is in contrast to the various plating systems that cluster the superior screws to support the humeral head and largely neglect the more important GT fragment.
Modern IM nails use locking technology to avoid the complications seen in the past, such as screw migration with nonunion or malunion. In addition, most modern nails are straight and do not damage the footprint with the insertion of the supraspinatus. Instead, straight nails pass through the muscular portion of the supraspinatus and enter through a relatively nonarticular section of cartilage. Such modern IM nails are not associated with iatrogenic shoulder pain and associated weakness, as seen in the past with first- and second-generation IM nails.
During percutaneous fixation, the unreamed design permits the surgeon to use the nail as a reduction tool and achieve anatomic fixation. Therefore the fracture pattern must be carefully scrutinized to select the most appropriate start point and approach (Neviaser or anterior acromioclavicular joint).
Surgeons should know that the morbidity of PHFs is overlooked in the elderly patient population. A sudden loss of function in an elderly patient’s arm can threaten independence and place significant strain on caregivers and family. This is particularly important for these patients when going from a seated to standing position. As people age, they no longer possess the hip flexibility, core strength, and lower extremity strength to easily go from a seated to standing position. As such, they use their upper extremities to push off the arms of a chair or assistive walking devices. This “push to stand” technique is not possible in the setting of a painful PHF treated nonoperatively or with a lateral locking plate. We have witnessed many patients comfortably bear weight through their upper extremities, some as soon as 1 week postoperatively, after the insertion of a load-sharing humeral nail and successfully maintain their independence at home.
Ultimately, the debate between the use of locking plates or IM nails for PHFs will persist in the orthopedic community. It is important to be able to recognize the benefits of both techniques and apply these to the specific needs of your patient population.
Surgical approaches
Two approaches can be used to fix a PHF or perform a shoulder replacement: the anterior DP and the anterosuperior (AS) lateral approach. Most textbooks recommend using the DP approach to fix a PHF. We believe that this is a tactical mistake. Knowing that the key piece of the puzzle is the GT, we believe that the AS approach should be used. We point out that no surgeon in the past would have repaired a massive posterosuperior cuff tear through an anterior DP approach! The recommended approach to fix a four-part fracture should be the same as the one to fix a posterosuperior cuff tear: the AS lateral approach. Furthermore, external rotation of the arm is needed to reduce the GT on the lateral aspect of the humerus. However, when using an anterior DP approach for a four-part fracture, the surgeon must put the arm in abduction and internal rotation to get access to the GT and place the locked plate on the lateral surface of the proximal humerus. By doing this, the surgeon leaves the GT posteriorly. The final result is a posterior malunion of the GT and a definitive nonfunctional stiff shoulder. In summary, it is totally illogical to use an anterior DP approach to fix a posterior problem (i.e., the posterior displacement of the GT). In addition, the difficulties encountered during this anterior approach lead to the destruction of the anterior deltoid and the distal branches of the axillary nerve. This is why so many patients have a flat anterior deltoid after an attempt to fix a four-part fracture through an anterior DP approach.
Two-part displaced surgical neck fractures
The pathophysiology of surgical neck fractures is defined by the medialization of the humeral diaphysis by the deforming forces of the latissimus dorsi, teres major, and pectoralis major muscles. The head fragment remains undisplaced because the external and internal rotators remain intact. Although the anterior and posterior cuff muscles are well balanced, the head can be tilted in either varus or valgus. Most surgeons currently prefer to treat these fractures with a locking plate construct. Our preference, whatever the age of the patients, is to treat these fractures through a percutaneous approach with a short, straight, statically locked IM Nail (Aequalis IM Nail). We use either the Neviaser approach (through the trapezius) in cases of varus deformity or an anterior acromioclavicular joint approach in cases of valgus or severe translational fractures.
When using a modern IM nail for a two-part surgical neck fracture, an important trick is to first statically lock the nail distally and then perform retrograde hammering (so-called slap-back technique) to put immediate compression on the fracture site. This should be done after having put the arm in neutral rotation to control the retroversion of the humerus. The procedure can then be easily completed by placing one or two screws proximally.
Three- and four-part displaced fractures
Our preferred treatment depends on the patient’s age:
- •
In patients younger than 60 years, we prefer a locked IM Nail. The nail is inserted through a superior transdeltoid approach in cases with an associated GT fracture. Alternatively, it is inserted through a DP approach in cases involving the lesser tuberosity and anterior dislocation of the humeral head.
- •
In patients younger than 70 years with severe fracture-dislocations, we prefer to perform an HA while ensuring anatomic and stable fixation of the tuberosities.
- •
In patients older than 70 years with displaced four-part fractures or fracture-dislocations, we prefer an RSA with tuberosity reattachment. This is done through an anterolateral approach, and the supraspinatus tendon is not removed.
Complications associated with treatment of proximal humerus fractures
The main and most devastating complication seen after a PHF is not humeral head necrosis, but the posterior migration of the GT. This is the equivalent of a massive posterior cuff tear, and a major cause of stiffness and pseudoparalysis of the shoulder. Osteonecrosis of the humeral head can happen but, again, this can be treated relatively easily with an HA or RSA.
Glenoid erosion is also a significant complication of PHFs, especially after locked plating or nailing. This is a particularly devastating complication in the young patient population because it forces the surgeon to insert an RSA as a salvage operation rather than an HA.
Finally, we have observed a specific trio of specific complications that can evolve after locked plating of a PHF:
- 1.
Posterior migration of the GT (massive irreparable posterior rotator cuff tear)
- 2.
Humeral head osteonecrosis (increased risk because of devascularization after plating)
- 3.
Glenoid erosion (secondary to screw cut out)
We have termed this the “unhappy triad after locked plating.” It remains an extremely challenging problem to solve because these patients typically have undergone multiple surgeries and remain with a very stiff and nonfunctional shoulder. Unfortunately, even revision surgery with an RSA does not provide satisfactory results, which we attribute to the absence of the external rotators that remain attached to an irreducible GT fragment.
Immobilization and rehabilitation
It is almost universally agreed upon that the shoulder joint must be mobilized as soon as possible after a fracture to decrease the risk of stiffness. However, a short period of immobilization is advisable to allow the inflammatory phase of the healing process to subside. We believe it is critical to avoid any devices that immobilize the arm in internal rotation. This position places unnecessary tension on the GT fragment and external rotators, possibly leading to an early failure of fixation or migration of a fracture that was not previously recognized. As a result, a neutral rotation sling should be used to help maintain reduction of the GT and posterosuperior rotator cuff attachment. Two or 3 weeks of immobilization of the arm in neutral rotation is enough to obtain fracture healing. After this delay, self-rehabilitation with pendulum exercises and swimming pool rehabilitation is strongly recommended. Surgeons should not immobilize these PHFs for 6 to 8 weeks because it leads to a definitive stiff shoulder,
In conclusion, PHFs and their sequelae remain complicated issues to treat, even by the most experienced surgeons. It is imperative that the treating surgeon understands the pathophysiology and deforming forces on the proximal humerus fragments to successfully treat these injuries. The GT should be treated as the main fracture fragment because malreduction or malunion of this structure creates an unsolvable problem for both the patient and surgeon. Currently, two-part displaced fractures should be fixed percutaneously with a straight IM nail. For three- and four-part fractures, surgeons should not be mistaken: the humeral head is not the problem. The main goal of the treatment should be to reduce and fix the GT in an anatomic (lateral) position, which is best done through an AS (lateral) approach. In summary, there are at least three good reasons to operate a three- or four-part fracture:
- 1.
The GT (because of the supraspinatus)
- 2.
The GT (because of the infraspinatus)
- 3.
The GT (because of the teres minor)
Reducing and fixing the GT in an anatomic position must be the main objective of operating on a three- or four-part fracture.
Bibliography
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
