Fig. 1—A, Displaced (angulated) two-part surgical neck fracture. B, After treatment with closed reduction and percutaneous pinning. Reproduced from Greene A: Proximal humeral fractures, in Norris T (ed): OKU: Shoulder and Elbow 2, Rosemont, IL, American Academy of Orthopaedic Surgeons, 2002, p. 210.


Published 6/1/2009
Peter Pollack

One fracture, five treatments

Shoulder surgeons offer varied approaches to proximal humerus fractures

Proximal humeral fracture treatment remains a matter of some debate. During the American Shoulder and Elbow Surgeons (ASES) Specialty Day, five noted orthopaedists presented five different philosophies and options—ranging from the least to the most invasive—for treating these fractures.

Selecting nonsurgical treatment
According to Charles A. Rockwood Jr., MD, the trend in recent years has focused on surgical options, although previous medical literature often promoted a nonsurgical approach.

Several studies examine the nonsurgical treatment of proximal humeral fractures, reporting results that include minimal pain, good functional recovery, and satisfied patients. In a study of 104 patients who had one-part fractures with less than 1 cm displacement and less than 45 degrees angulation, 90 percent of patients had no or minimum pain and 90 percent had a good functional recovery without surgery.

Although the nonsurgical approach can work with many patients, compliance with rehabilitation programs is very important.

“I teach patients how to do isometric arm exercises to compress the fracture together,” says Dr. Rockwood. “I ask them to do the exercises 100 times in the morning and 100 times in the afternoon. Each time they do the isometric contracture, they hold it for a count of 10. I also teach patients how to do some very gentle pendulum exercises—four or five times a day—and I suggest that they sleep in a comfortable lounge chair that has a foot rest.”

Dr. Rockwood recommends that patients gradually increase the number of pendulum and isometric exercises they perform. Depending on their symptoms and range of motion, he may recommend therapy with a 3-foot stick or a pulley.

As the fracture heals, patients take on more activities of daily living, at first with the “good arm” assisting the injured extremity, and later using the injured arm by itself.

Many patients with one-part or even two- or three-part fractures can be managed with a dedicated home-based rehab program. Patients who are heavy smokers, or those with conditions such as osteoporosis, severe metabolic or endocrine disturbances, or hyperparathyroidism may not be good surgical candidates. An individual approach is required for each patient, he says, and with a nonsurgical initial approach, if an issue develops, it can usually be managed later with a less invasive surgery.

Picking percutaneous pinning
Some patients are not ideal candidates for open surgery, and some fractures are not ideal for plates and rods. Given that, percutaneous pinning is a “nice thing to have up your sleeve, even if you prefer other techniques,” says Evan L. Flatow, MD.

“One of the problems with fixed-angle plate fixation,” explains Dr. Flatow, “is that if the head does collapse, the plate holds the screws and they can penetrate the collapsed head and damage the glenoid. With the percutaneous technique, if you take out the pins, there is no metal to penetrate if the head collapses. If osteonecrosis develops, the patient may tolerate it rather than undergo another operation.”

Dr. Flatow’s indications for pinning are two-part surgical neck fractures, impacted valgus fractures, and certain other conditions, depending on the experience of the surgeon (Fig. 1). A head-split fracture is not a good indication for percutaneous fixation, because it is difficult to get the rotation and the angulation lined up perfectly. He also avoids pinning in true four-part fracture dislocations and in cases in which metaphyseal comminutions make it difficult to perch the head on top of the shaft. In all cases, he advises, surgeons should be prepared to use a different technique if it appears the pinning won’t work.

Dr. Flatow notes that most surgeons favor cannulated screws rather than pins for the tuberosities because screws allow patients to perform some pendulum exercises and movement at an earlier stage. To avoid overtightening the screw that catches the medial shaft (which could over reduce the greater tuberosity), he prefers to insert that screw second, after the transverse screw.

One thing Dr. Flatow has noticed, however, is that the development of osteonecrosis seems to increase over time. The average diagnosis of osteonecrosis is almost 2 years postoperative and can be nearly 3 years. He suggests that researchers keep this in mind when reporting 6- or 12-month data for percutaneous fixation.

Opting for ORIF
According to David L. Helfet, MD, all proximal humeral fractures are potential candidates for open reduction and internal fixation (ORIF). Whether to perform ORIF depends not on the type of fracture, he says, but on other factors such as patient age and demands, degree of comminution, amount of fracture displacement, and osteoporosis.

Although historic outcomes for ORIF have not always been good, Dr. Helfet believes that this was related to comminution, severe osteoporosis, and an inability to obtain or maintain a reduction. The implants were also inadequate, and stable fixation was difficult to achieve. Patients were immobilized after surgery, and stiffness and osteonecrosis frequently developed.

Fig. 1—A, Displaced (angulated) two-part surgical neck fracture. B, After treatment with closed reduction and percutaneous pinning. Reproduced from Greene A: Proximal humeral fractures, in Norris T (ed): OKU: Shoulder and Elbow 2, Rosemont, IL, American Academy of Orthopaedic Surgeons, 2002, p. 210.
Fig. 2 Good imaging studies are important for placement of intramedullary nails.
Courtesy of Clifford B. Jones, MD
Fig. 3— A, True AP radiograph of a head-splitting three-part proximal humeral fracture in a 70-year-old woman. B, Postoperative radiograph after successful treatment with a humeral head replacement. Reproduced from Greene A: Proximal humeral fractures, in Norris T (ed): OKU: Shoulder and Elbow 2, Rosemont, IL, American Academy of Orthopaedic Surgeons, 2002, p. 213.

Over the last several decades, however, surgeons have come to better understand the vascularity of the shoulder, including the roles of the rotator cuff and the subscapularis, the need to obtain a stable calcar reduction, and the necessity of avoiding varus and posterior roll-off.

Additionally, technology has improved, with locking plates providing a better method of fixing poor bone, bone graft substitutes, tension banding of soft tissue, and the ability to mobilize the joint immediately to prevent stiffness.

Dr. Helfet’s contraindications for ORIF include patients with severe osteoporosis, steroid users, or those with stroke paralysis. Additionally, he does not recommend ORIF if a patient is unreliable or cannot comply with a rehabilitation program. He notes that true head split-fractures are probably best treated with arthroplasty.

Deciding on a nail
“I think surgeons need to know all these techniques,” says Clifford B. Jones, MD. “I think fractures with diaphyseal extension, those with an extensive amount of metaphyseal comminution, those fractures that have ipsilateral diaphyseal-type fractures are great candidates for potential treatment with intramedullary (IM) nails.”

Dr. Jones’ contraindications for IM nailing include head-split fractures or those with marginal impaction, fractures with isolated greater or lesser tuberosity tears, anatomically unstable head patterns with a vascularity, and potentially, nonintact lateral column injury patterns.

Dr. Jones says that good imaging from a variety of angles is important (Fig. 2), and he emphasizes the need to use joysticks throughout the procedure.

“Once you have those pins in position, derotated out,” he explains, “you make an incision through the midportion of the rotator cuff, confirmed on both the anteroposterior (AP) view and rollover views, so you’re not too posterior or anterior, and then use 2.5 mm Shantz pins to retract the rotator cuff and protect it during the procedure. You then use a center position as your guide position.”

If there is comminution and instability, Dr. Jones suggests taking a joystick distally with the diathesis, which can be used to translate and find the correct alignment. Pollar blocking pins or screws can be used for reaming or redirecting the nail, and potentially, those screws can be left in position.

Dr. Jones stresses the importance of not inserting the nail too far. In some cases, inserting a straight nail into a bone with a slight interior angulation can create a fracture. If that occurs, it can be treated intraoperatively with a hybrid plate/screw construct that will not limit range of motion.

Choosing joint replacement
Louis U. Bigliani, MD, discussed humeral head replacement (HHR), possibly the most definitive option. His indications for this procedure include four-part fracture, four-part fracture-dislocation, head-split fractures, and comminuted and three-part fractures where internal fixation is not possible. Poor bone quality, severe comminution and advanced age are factors that favor a prosthesis. According to Dr. Bigliani, a massive rotator cuff tear with a four-part fracture is an indication for a reverse prosthesis.

HHR is a very technical and demanding procedure involving prosthesis positioning and tuberosity reconstruction (Fig. 3). Outcomes tend to be the best when HHR is performed as the primary procedure.

When performing HHR, Dr. Bigliani uses a delto-pectoral approach. It is important to identify the long head of the biceps as this will orient you to the position of the tuberosities.

“It is important to place large traction sutures at the tendon bone interface of the lesser and greater tuberosity so that you can mobilize the soft tissues, disimpact the head, and then mobilize both the greater and lesser tuberosities to get a full view of the glenoid. You can’t get the tuberosities around the prosthesis unless you have them fully mobilized,” said Dr. Bigliani.

Dr. Bigliani will often cement the prosthesis, especially if the bone is soft. Taking radiographs of the uninjured shoulder for comparison can help properly position the implant in reference to height. The prosthesis can be accurately positioned using a fracture jig and an inline insertion device.

To date, studies show significant improvement in pain, but there is variability in functional results.

Disclosure information: Dr. Bigliani — Zimmer and Innomed; Dr. Flatow — Innomed, Zimmer, and Wyeth; Dr. Jones — AONA; Dr. Helfet — Synthes; Dr. Rockwood — DePuy, a Johnson & Johnson Company, Zimmer, and Tornier.

Peter Pollack is a staff writer for AAOS Now. He can be reached at