Rationale and Basic Science Pertinent to the Procedure

Classically, it was considered that consolidation of scaphoid fractures could be achieved without surgery. For many years since, however, open reduction and internal fixation has been the recommended and well-accepted treatment for displaced and unstable intra-articular fractures.

The complex morphology and the small size of the scaphoid bone resulted in the development of numerous sophisticated techniques to achieve an anatomical and stable fixation. In 1984, Herbert and Fischer¹ reported their experience using a cannulated screw, which originally was not developed for fixation of scaphoid fractures. In the early 1990s, the first article was published² describing inserting cannulated screws with a minimally invasive technique. The main principle was to preserve the surrounding ligaments of the carpal bones to avoid a destabilization of the reduction and to protect the fragile vascularization of the scaphoid bone.³

Meanwhile, patients and their referring physicians became more and more demanding. The surgical indication was expanded because of the inconvenience of conservative treatment with its unpredictable economic consequences owing to the long duration of immobilization.

Whipple^4,5 first presented a method with percutaneous screw fixation using a modified Herbert screw combined with image intensifier control and arthroscopic examination of the wrist. This method allowed the surgeon to control the reduction and to assess potential associated lesions.

In the treatment of scaphoid fractures by surgical reduction and internal fixation, some rules have to be respected, as follows: verify the exact fracture reduction, avoid an intra-articular penetration of the screw, maintain the fixation under compression, and allow an early return to activities of daily living. We report our more recent experience of 38 scaphoid fractures treated with an arthroscopically assisted percutaneous screw fixation technique using a cannulated Herbert screw.

Indications

The disabling long cast immobilization in this mostly young and active patient population and the risk of nonunion or malunion favor surgical reduction and internal fixation of scaphoid fractures. The aim is stable fracture fixation allowing early mobilization without compromising consolidation. It is important to neutralize the fracture forces, while compressing the fracture. The minimally invasive technique, and hence limited operative trauma, allows early functional rehabilitation. Arthroscopy helps to reduce the fragments; to control the quality of the reduction; and to assess the position of the screw, especially with regard to the radiocarpal joint.

The ideal candidate for surgery is a patient motivated to return to sports or business activities as soon as possible (e.g., a professional swimmer on the French national team who sustained a scaphoid fracture and was determined to qualify for the Olympic Games at Athens 2004 15 days later). The typical patient is young (mean age in our series 31.5 years), understands the aim of the treatment, and understands its risks and benefits. The delay between fracture and surgery should be as short as possible and not more than 1 month. One should delay immediate surgery if the conditions are not optimal (e.g., cutaneous lesions). In these situations, the wrist should be immobilized until the conditions are perfect.

Contraindications

In uncooperative patients who do not understand the risks and benefits, this method is contraindicated. In comminuted fractures, it is not realistic to achieve an anatomical reduction with this technique. Advanced age, cutaneous lesions, suboptimal operative conditions (e.g., incomplete surgical equipment), and severe associated lesions (e.g., severe scapholunate dissociation) are relative contraindications that can delay the intervention or preclude this type of surgery.

Operative Technique

Under ambulatory conditions, the operation is performed with locoregional anesthesia. The patient is placed in the supine position on a special arm table with a tourniquet on the arm applied as proximal as possible. During the critical parts of the operation, the forearm can be extended using a pad underneath the wrist. Another possibility is to put the wrist under traction with a traction device, which is placed outside the arm table still allowing positioning the image intensifier. A retrograde (from distal to proximal) screw fixation is aimed. First, the fracture is visualized under arthroscopy using standard portals, leaving the forearm free on the table. Next, a 1-mm pin is placed through a small (5-mm) incision to the distal tuberosity of the scaphoid in a retrograde fashion (Fig. 44-1). Then the wrist is put under traction allowing arthroscopic control to verify the exact reduction of the scaphoid.

FIGURE 44-1 Percutaneous retrograde pinning of the scaphoid through a small approach centered on the distal tuberosity.

The arthroscope is introduced through a radial midcarpal portal through which the fracture can be assessed easily. If necessary, a débridement of the articulation can be done with the shaver while cleaning the medial surface of the scaphoid. If the fracture is displaced, reduction of the fragments is possible with a little retractor introduced through the STT midcarpal portal. Under arthroscopic control, the fracture fixation pin is slightly pulled back beyond the fracture line, then the fracture is reduced, and the pin is replaced into the proximal fragment (Figs. 44-2, 44-3, and 44-4). As soon as a satisfactory reduction is achieved, the hand is removed from the traction device, and the wrist is positioned on a pad on the arm table.

FIGURE 44-2 Mediocarpal arthroscopic view of a displaced scaphoid fracture.

FIGURE 44-3 Mediocarpal arthroscopic view of the same fracture during reduction with a probe after the pin was pulled back distally to the fracture line.

FIGURE 44-4 Mediocarpal arthroscopic view of the same fracture after reduction and stabilization.

Under fluoroscopic control, the hole for the screw is tapped. If a resorbable screw is chosen, a double-sized trephine, drilling proximal with a 3-mm part and distal with a 3.5-mm part at the same time, is used. The two different screw threads are separately prepared mechanically to avoid the phenomenon of blockage in torsion with a resorbable screw. This avoids the risk of implant fracture (Fig. 44-5). The screw is inserted over the guidewire. Again under arthroscopic control, the radiocarpal compartment is visualized through the 3,4 portal; this allows the surgeon to verify the absence of an intra-articular penetration of the screw head at the level of the proximal pole (Fig. 44-6).

FIGURE 44-5 A, Tapping of the proximal part of the scaphoid. B, Tapping of the distal part of the scaphoid.

FIGURE 44-6 Screw fixation under arthroscopic control.

The entire radiocarpal compartment is inspected to assess potential associated lesions. Midcarpal exploration allows the inspection of the fracture line to the medial articular surface of the scaphoid and the assessment of the reduction quality. In case of insufficient compression, the screw can be redrilled while visualizing the compressive effect. The scaphotrapeziotrapezoid articulation remains untouched. The resorbable screw has only one size with a long distal part, which allows adaptation to different scaphoid sizes. The surgeon must cut the part that is outside of the bone short at the scaphoid with a mini-oscillating saw (Fig. 44-7). The incisions are closed using Steri-strips. Postoperatively, the wrist is left unprotected; a simple anterior splint can be applied after the first dressing to help with postoperative pain.

FIGURE 44-7 Placement of the resorbable screw before to cut it easily at the level of the bone.

Important aspects of the operative technique are as follows:

The risks of the procedure are as follows:

Complications

Results

Between 2001 and 2005, 36 patients with 38 isolated scaphoid fractures (two bilateral fractures) underwent arthroscopically assisted percutaneous screw fixation. Thirty patients were men, and six patients were women. The mean age was 34 years (range 17 to 52 years). The dominant side was involved in 75% of the cases. One patient with bilateral fractures was operated the same day; in the other patient, the two operations were performed within 5 days.

All fractures were acute (type B according to the classification of Herbert and Fischer¹); the mean delay from trauma to surgery was 9 days (range 2 to 30 days). There were mainly waist fractures (type III and IV according to Schernberg’s classification⁶); fractures of the proximal pole were not included (Table 44-1). The fracture was nondisplaced in 22 cases.

In 16 cases, the reduction could be held with intra-articular arthroscopic maneuvers. Eight resorbable screws and 30 Herbert screws were inserted retrograde from distal to proximal. No remarkable difference in the use of the two screws was noticed. Arthroscopic control was used systematically independent of the perioperative fluoroscopic control. Although there was a satisfying intraoperative fluoroscopic result, wrist arthroscopy revealed an overlength of the screw tip owing to an intra-articular breakout at the proximal pole of the scaphoid in three cases. The screws had to be changed to shorter ones. This observation confirms the interest in the arthroscopic control during this procedure. The guiding pin was broken in the radiocarpal compartment in one patient. Pin removal could be realized under arthroscopic control without any difficulties.

In our series, we have found just a few associated lesions. In one patient, a central perforation of the triangular fibrocartilage complex was débrided with the shaver. In another case, a lesion of the anterior part of the scapholunate ligament without dynamic instability was observed. In this 52-year-old patient, it did not have any therapeutic consequences.

The mean duration of surgery was 32 minutes (range 15 to 70 minutes). The last case had the shortest duration, documenting very well the learning curve of this technique. All patients were reviewed by one examiner independent of the operating surgeon. Postoperative fracture consolidation was assessed radiographically with four x-rays.

The mean follow-up was 29 months (range 6 to 56 months). All fractures healed primarily; nonunion or malunion was not observed. The mean duration of consolidation was 62 days (range 45 to 80 days). The mean duration in originally displaced fractures was 70 days, whereas it was 55 days in originally nondisplaced fractures.

All patients were very satisfied or satisfied with the result. None of the patients regretted choosing this method. The main reason for this high satisfaction rate was the fast functional recovery and the absence of postoperative immobilization. Patients appreciated the small scars, an observation regularly made after most of the endoscopic and arthroscopic procedures.

Return to work was very fast. Twenty patients returned to work immediately. Rapid return to work facilitated patients’ decision to choose this technique. Most patients either had an independent occupation or were professional high-level athletes. The mean duration of return to their activities was 21 days (range 0 to 92 days).

The results in terms of pain were excellent. Only five patients had intermittent slight pain. All the other patients were completely pain-free. Four patients with nonresorbable screws have had slight anterior pain in the area of the scaphotrapeziotrapezoid joint. Screw removal solved this problem in three patients, whereas in the fourth patient slight pain persisted.

At final follow-up, 31 of the 34 wrists (two cases with bilateral fractures excluded) reached 90% of the mobility compared with the contralateral side. Comparison of the grip strength measured with a Jamar Dynamometer (Sammons Preston, Inc., Bollingbrook, IL) also confirmed the quality of the recovery (91% of the strength of the healthy contralateral side).

Discussion and Review of Literature

Numerous more recent studies have shown the capability of a percutaneous fixation of scaphoid fractures using cannulated screws.^2,7–9 The various cannulated screw models underline the interest in this method and create competition with the classic conservative method of forearm immobilization for 3 months. Several studies confirm the increased rate of fracture consolidation with this method.^7–10 The time to consolidation in nondisplaced fractures seems to be shorter with percutaneous screw fixation. Shin and coworkers¹⁰ reported in their randomized study (percutaneous screw fixations versus conservative treatment) a consolidation time of 4 to 5 weeks after percutaneous screw fixation. With our results with an average radiological consolidation of less than 2 months in nondisplaced fractures, we can confirm this statement.

In various series, return to professional activities was earlier after screw fixation.^{8,10,11,14,21} In our study, the functional recovery also was exceptionally fast. This also might be due to a patient selection bias because many of our patients have chosen this method with regard to their professional and personal duties. It seems to be more logical to propose the percutaneous screw fixation to a motivated and well-informed patient even more when conservative treatment has the risk to fail (e.g., in unstable fractures). The failure rate in terms of consolidation can be 15% after cast immobilization of 3 months.^12,13 In our series, there were no nonunions.

Wrist arthroscopy combined with percutaneous screw fixation allows avoiding certain complications that are relatively frequent in fracture fixation of the scaphoid. Filan and Herbert¹³ found 14 intra-articular (Herbert) screw penetrations in their series of 431 patients. In our series, after final arthroscopic control of the radiocarpal joint, we had to change three screws because of breakout of the screw tip out of the scaphoid. Arthroscopic midcarpal examination also allows assessing the quality of fracture reduction after screw fixation.

We agree with Whipple¹⁶ that direct visual examination of the reduction quality is much more efficient than fluoroscopic evaluation. Direct visual control of fracture compression is an additional security for the surgeon. The possibility to diagnose and treat associated lesions with arthroscopic exploration of the wrist was described by many authors.^15,16 Shin and coworkers¹⁰ found 11 intracarpal lesions during arthroscopic exploration in a series of 15 displaced scaphoid fractures that were treated with arthroscopic reduction and percutaneous fixation. Most of them were minor lesions, but the authors also found two complex scapholunate lesions that could be treated with reduction and pinning.

Because of the need for reduction, displaced scaphoid fractures usually required classic open reduction.¹⁷ The possibility to maintain the reduction by external maneuvers justified the use of percutaneous screw fixation.¹⁸ If one could not maintain the reduction, change to an open procedure was indicated.¹⁹ In our series we did not have any problems with the reduction, and all the displaced scaphoid fractures could be reduced and maintained under arthroscopic control (Fig. 44-8).

FIGURE 44-8 A, Displaced scaphoid fracture. B, Mediocarpal arthroscopic view showing the rotational displacement between the two fragments. C, Mediocarpal arthroscopic view showing the reduction after disimpaction of the two fragments with the help of a small curet. D, Drilling with fluoroscopic control. E, Radiographic control after 60 days showing the consolidated fracture.

Wrist arthroscopy has found its place for this indication allowing direct and indirect reduction maneuvers. Slade and colleagues²⁰ described in detail their technique of percutaneous arthroscopically assisted fracture fixation from proximal to distal. The guiding pin is placed with the wrist in hyperflexion and hyperpronation after reduction of the geometric axis of the scaphoid under fluoroscopic control. The screws that were used (Acutrak LLC, Hillsboro, OR) are voluminous, and in our opinion penetration of the screw at the proximal pole may cause a significant cartilage lesion at the scaphoid head. In the case of a fracture type I according to Schernberg, the small-sized proximal fragment has to be fixed with a standard screw. Apart from these specific cases, the technique of screw fixation through a volar approach with the assistance of a small elevator through an STT midcarpal portal to align the fragments always allowed us to verify the reduction without major difficulties. The classic disadvantages of the scaphotrapeziotrapezoid articulation with a volar approach have led to a systematic removal of the screw after 1 year to avoid secondary osteoarthritis.²⁰ The advent of resorbable screws may have solved this particular problem.²² We have performed screw removal only in the case of persistent anterior wrist pain. A longer follow-up is necessary to evaluate the potential development of secondary osteoarthritis.