Proximal humerus fractures are of common occurrence in daily trauma practice. Their prognosis depends not only on the pattern of the fracture but also on the quality of the bone in which they occur, the age and medical condition of the patient and the treatment option applied.
The treatment of complex proximal humerus fractures has always been a challenge for orthopaedic surgeons for decades. Although most authors agree that successful outcomes are more likely when the tuberosities are anatomically reduced [7, 10, 11] and the shoulder is mobilized early, there is no consensus on the most successful surgical procedure to achieve both goals [19, 20]. This resulted in a myriad of procedures including percutaneous pinning, intramedullary nailing, hemiarthroplasty, and open reduction and internal fixation (ORIF) using a multitude of implants, which have all failed to yield consistently good or excellent clinical results [7, 10, 21]. These techniques have each their own strengths and weaknesses.
Amongst all available options, open reduction and internal fixation offers the best chances of an anatomical reduction and stable fixation allowing for early range of motion [8, 10, 11], however this method requires an extensive soft tissue stripping to obtain both adequate exposure and rigid fixation. In fact, the greater tuberosity fragment is usually externally rotated (superior and posterior displacement) under the pull of the rotator cuff muscles [22, 23], this makes achieving reduction and fixation of this fracture difficult through a classic anterior delto-pectoral approach and extensive soft tissue release is sometimes mandatory to obtain adequate exposure. In some cases, detaching the deltoid origin or insertion [22, 23, 24] and/or the pectoralis major insertion [2, 22, 23, 25] at least partially is necessary to access this fragment.
Gallo et al. [21] went as far as to associate a supero-lateral approach to the classic anterior delto-pectoral approach in open reduction and internal fixation of complex proximal humerus fractures to improve the reduction of the greater tuberosity fragment.
It is well established that, with the advent of fixed-angle locked proximal humerus plates, ORIF with locking anatomical plates is associated with the best radiologic and functional results and is more adequate for fractures fixation in osteoporotic bone [26]. These advantages are counterbalanced by the risks of joint screw penetration, loss of reduction and joint stiffness due to large soft tissue dissection.
Percutaneous pinning, although minimizing soft tissue disruption, requires a longer period of shoulder immobilization, a good bone stock for stability of the fracture fixation and exposes to loss of reduction and pin migration [6]. Furthermore, fluoroscopically guided percutaneous reduction of the greater tuberosity can only be indicated in minimally displaced fractures as suggested by Jaberg et al. [6]
Barnes et al described the surgical technique of close reduction and percutaneous pinning of proximal humerus fractures reporting the possibility of percutaneous lag screwsand concluded that “Reduction can be difficult with this limited incision technique” [27].
We believe that our surgical management of the Neer three-part fractures using a combined open lag screw fixation and PRIP offers an original compromise between the two procedures and to the best of our knowledge this is the first series describing the association of these two techniques reported in the English literature.
The first surgical step consists in an open reduction of the greater tuberosity fragment using a classic supero-lateral approach. In our experience, this approach greatly facilitates the reduction of the greater tuberosity under direct visual control, which is viewed by some as the key to a successful outcome [8, 10, 11]. After this first step is achieved, the three-part fracture is technically transformed into a two-part fracture. At this stage and under fluoroscopic guidance, three 20 to 30/10th pins are inserted in a retrograde manner into the medullary canal and through the neck of humerus fracture after close reduction is obtained.
This percutaneous intramedullary retrograde pinning (PRIP) technique was first described by Hacketal et al in 1961 [12] to treat midshaft fractures of the humerus. De la Caffinière [13] in 1988 described a modification to the original technique by changing the entry to a point 3 to 4cm above the lateral condyle and that was the technique we used in our study. This surgical technique has the advantage to preserve the humeral head vascular supply by minimizing soft tissue dissection and provide solid fixation by the means of a diverging anchoring of the pins into the sub-chondral bone [28, 29] allowing for a shorter immobilization period with some authors prescribing immediate mobilization.
In our study, we postponed shoulder passive range of motion to four to six weeks post operatively because of the predominance of old aged patients with signs of osteopenia to avoid possible secondary displacement.
The two mostly reported complications of the PRIP were per-operative fractures at the pins’ entry point and pin migration. In a series of 136 Neer two-part neck of humerus fractures treated with PRIP, Putz et al. [30] reported two per-operative fractures (1.5%) and 14 intra-articular pin migrations (10%).
In our study, one case of pin migration was reported and resulted in the removal of the hardware at three months post-operatively. No per-operative fractures and no elbow post-operative stiffness occurred.
At the latest fallow-up, our results were overall satisfactory with good radiologic and functional results and little complications.
The main strength of this study is that it reports an original surgical technique in the treatment of Neer three-part proximal humerus fractures. In the context of our developing countries where locked anatomical plates and intramedullary nails are not readily available for obvious economic reasons, this technique combination can be an interesting alternative.
The limitations, however, are the small sample size and a short follow-up period allowing us to report only preliminary results of this technique. In fact, AVN rates increase considerably with follow-up. Harrison et al. [1] reported that AVN of the humeral head occurred in 4% of the patients after an average follow-up of 35 months, this same rate in the same series of patients climbed as high as 26% of the patients after an average 84 months’ follow-up. For this reason, a larger number of patients and a longer follow-up period are needed to draw more definite conclusions.