Biomechanical Strength of Retrograde Fixation in Proximal Third Scaphoid Fractures

Abstract

Background: Current techniques for fixation of proximal pole scaphoid fractures utilize antegrade fixation via a dorsal approach endangering the delicate vascular supply of the dorsal scaphoid. Volar and dorsal approaches demonstrate equivalent clinical outcomes in scaphoid wrist fractures, but no study has evaluated the biomechanical strength for fractures of the proximal pole. This study compares biomechanical strength of antegrade and retrograde fixation for fractures of the proximal pole of the scaphoid. Methods: A simulated proximal pole scaphoid fracture was produced in 22 matched cadaveric scaphoids, which were then assigned randomly to either antegrade or retrograde fixation with a cannulated headless compression screw. Cyclic loading and load to failure testing were performed and screw length, number of cycles, and maximum load sustained were recorded. Results: There were no significant differences in average screw length (25.5 mm vs 25.6 mm, P = .934), average number of cyclic loading cycles (3738 vs 3847, P = .552), average load to failure (348 N vs 371 N, P = .357), and number of catastrophic failures observed between the antegrade and retrograde fixation groups (3 in each). Practical equivalence between the 2 groups was calculated and the 2 groups were demonstrated to be practically equivalent (upper threshold P = .010). Conclusions: For this model of proximal pole scaphoid wrist fractures, antegrade and retrograde screw configuration have been proven to be equivalent in terms of biomechanical strength. With further clinical study, we hope surgeons will be able to make their decision for fixation technique based on approaches to bone grafting, concern for tenuous blood supply, and surgeon preference without fear of poor biomechanical properties.

Introduction

Proximal third scaphoid fractures are now nearly uniformly treated with surgical fixation. The scaphoid is the most commonly fractured carpal bone, occurring in greater than 20 000 individuals each year in the US alone. Classification of scaphoid fractures is typically based on location within the scaphoid: tuberosity (17%), proximal pole (6%), waist (66%), and distal pole (11%).⁹ This classification is critical because it holds implications in treatment options, vascularity of the fracture fragment, healing potential of the fracture site, and necessity for surgical treatment.⁹ The vascularity of the scaphoid is unique in that it relies on retrograde flow from the palmar and dorsal branches of the radial artery. The dorsal branch of the radial artery is solely responsible for the vascularity of the proximal pole through branches entering the dorsal ridge of the scaphoid.^5,14

Proximal pole scaphoid fractures continue to suffer from less-than-optimal outcomes despite recent innovations in bone grafting techniques and improved fixation.² In a study of 23 proximal pole fractures treated operatively via a dorsal approach only 43% showed evidence of union at 14 weeks post injury.² Traditionally, proximal pole scaphoid fractures are treated dorsally with fixation inserted in an antegrade fashion, as there are often concerns for fixation strength in the small proximal pole fracture fragment with retrograde screw fixation. In cases of nonunion without avascular necrosis (AVN), minimization of dorsal soft tissue (particularly microvascular) trauma is key in preserving the tenuous blood supply to the proximal pole.

Current techniques uniformly utilize antegrade fixation with a dorsal approach. This dorsal approach may disrupt the delicate vascular supply entering at the dorsal ridge of the scaphoid. Particularly in cases of nonunion, dorsally based curettage and bone grafting can further put the sole blood supply to the proximal pole of the scaphoid at risk. Proximal pole scaphoid fractures are typically oriented in an oblique fashion with the fracture line exiting the cortex much more proximally dorsally than volarly; this results in difficulty achieving a central position in the proximal fragment from a dorsal approach³ (Figure 1). Unlike the dorsal approach, which requires flexion of the wrist, the volar approach to the scaphoid is facilitated by extension at the wrist, which aids in reduction of any residual flexion deformity of the scaphoid. Volar (retrograde) and dorsal (antegrade) approaches have both been validated and demonstrate equivalent clinical outcomes in prior studies focusing on scaphoid wrist fractures.^6,8,11,15 There has been a significant amount of research into biomechanical strength of fixation of various scaphoid fracture morphologies, but surprisingly, to our knowledge no study has specifically evaluated the biomechanical strength of fixation of fractures of the proximal pole.^4,6,8,11,15

Figure 1.

Typical orientation of the proximal pole scaphoid fracture. Radial and dorsal views demonstrate location of the dorsal ridge of the scaphoid in relation to the fracture line.

The primary goal of this study is to compare the biomechanical strength of both antegrade and retrograde fixation of fractures of the proximal pole of the scaphoid. Our hypothesis is that, with screws placed centrally in the proximal pole fragment, retrograde fixation will be similar to antegrade fixation of the scaphoid in cyclic load testing and load to failure.

Methods

Twenty-two (left and right) matched cadaveric scaphoids were obtained. The average age of the specimens was 68 years. Twelve were male and 10 were female. The specimens were dissected free of soft tissue attachments, and each specimen underwent an osteotomy at a location defined by previous study of the anatomy of acute scaphoid fractures to best represent the average proximal pole scaphoid fracture.³ These fractures were produced with an osteotome 2 mm proximal to the dorsal ridge laterally, ending 0.5 mm proximal to the apex of the dorsal ridge medially, and extending at a 30° angle from dorso-ulnar to proximo-radial.³

Each fracture was anatomically reduced utilizing a bony tenaculum clamp. A single guidewire was placed in a central fashion maximized to achieve a central position within the proximal pole of the scaphoid. This guidewire was placed in either an antegrade or retrograde fashion into one of each matched pair based on computerized randomization. Careful attention was given to achieving a central position within the proximal pole fragment. A depth gauge was used to determine appropriate screw length. A cannulated drill was then passed over the guidewire and driven bicortically and through the osteotomy site.

Definitive scaphoid fracture fixation was performed utilizing a single, centrally placed 2.7 to 3.0 mm fully threaded variable pitch headless compression screw (Acumed Acutrak 2; Figure 2). Appropriate screw length was confirmed using visual inspection. Due to the tapered, variable pitch design of the screw, more threads but with lower pitch and less thread depth were present when the screw was placed antegrade, with the converse true during retrograde insertion (Figure 2). The scaphoid was then potted in 10-cm diameter polyvinyl chloride end cap utilizing dental cement (Heroes Kulzer Inc, South Bend, Indiana). A Material Testing Machine (MTS 858 Mini-Bionix Test System, Minneapolis, Minnesota, USA) was utilized to apply a cyclic load and load to failure. Each specimen was oriented such that a cantilever force could be applied from dorsal to volar and underwent a standardized cyclic load from 80 N to 120 N at 1 Hz which has previously been demonstrated to best represent the primary physiologic load on the scaphoid (Figure 3).^10,12 Testing was carried out until 2 mm of fracture displacement occurred or 4000 cycles was reached. The testing protocol and displacement of 2 mm were used in previous studies of similar design.^4,10 The specimens that reached the 4000-cycle limit were then loaded to failure, and the maximal force applied was recorded. Failure was defined as screw cutout or fracture of the distal pole of the scaphoid creating at least 2 mm of fracture displacement. Mechanism of failure for each specimen was recorded.

Figure 2.

Radiographs of antegrade and retrograde screw fixation of proximal pole fractures of the scaphoid.

Figure 3.

Testing apparatus demonstrating simulated proximal pole fracture and antegrade fixation.

The paired t test was utilized to compare load to failure, number of cycles during cyclic loading, and screw length for screws placed in an antegrade and retrograde fashion. We then calculated both Pearson and Spearman correlation coefficients between screw length and load to failure. Finally, we calculated the practical equivalence between antegrade and retrograde screw configurations with a specified practical difference threshold of 75 N. Equivalence testing was utilized as it is more statistically strenuous than proving no statistically significant difference between the 2 means, instead placing the burden of proof on proving that the 2 means are equivalent within a range of acceptable values for the difference. In noninferiority testing, common practice is to set the value of the practical difference threshold to a fraction of the lower limit of a confidence interval of the difference between the current therapy and the placebo.¹⁶ When the outcome is mortality, the Food and Drug Administration has suggested that fraction be 0.5.¹⁶ In this case, we do not have a suitable placebo, but the lower limit of the confidence interval for force sustained by the scaphoid was 236 N; 75 N (approximately 16.9 lbs) was chosen as it is far less than one half of the difference between the lower limit of the confidence interval and those forces observed across the radiocarpal joint during maximal physiologic loading.^1,7

Results

The average screw length was 25.5 mm for the antegrade group and 25.6 mm for the retrograde group, which was not statistically significant (P = .935). Right-sided specimen screw lengths averaged 26 mm and left-sided 25 mm (P = .654). The average number of cycles that right-sided specimen withstood prior to failure was 3754 cycles and left-sided specimen withstood 3820 cycles (P = .713). The load to failure on right- and left-sided specimen averaged 358.5 and 358.6 N respectively (P = .996) (Figure 4). Load to failure was compared with screw length with a correlation coefficient equal to 0.2198. Antegrade screws on average sustained 3738 cycles and retrograde screws 3847 cycles (P = .553). In all, 6 specimens did not go on to the load to failure phase of testing because they catastrophically failed during the cyclic loading phase. Screw cutout was the method of failure for each of the specimens, both those that failed during cyclic loading and those that failed during the load to failure testing. Three of these failures occurred in the antegrade group (at 2529, 3109, and 3212 cycles) and 3 failures occurring in the retrograde group (at 2900, 3700, and 3870 cycles). There were no significant differences in number of cyclic loading cycles that were sustained between the antegrade and retrograde fixation groups (P = .553). The average load to failure was similar between antegrade and retrograde screw fixation (348 N vs 371 N) (P = .358). Practical equivalence between the 2 groups was calculated and the 2 groups were demonstrated to be practically equivalent with an upper threshold P value of .010.

Figure 4.

Average load to failure in antegrade and retrograde fixation.

Discussion

If bone stock is adequate, surgical fixation is clearly the treatment of choice. Despite advances in surgical techniques and implants, treatment of the proximal pole of the scaphoid continues to be mired in difficulty both technically and in those patients who develop nonunion and avascular necrosis. The intrasynovial nature of the entirety of the proximal pole of the scaphoid alone predisposes this fracture to delayed healing. The tenuous vascularity of the proximal pole of the scaphoid, which relies solely on vascularity from the dorsal branches of the radial artery entering through the dorsal ridge of the scaphoid, further increases the likelihood of nonunion at this location.

Surgical intervention has successfully decreased the rate of AVN but up to one-third of these fractures may result in nonunion even with appropriate treatment.¹³ While yet to be studied, surgical intervention likely results in some degree of injury to the dorsal arterial branches and may potentiate the avascularity of the proximal pole. This is of particular concern in cases of delayed treatment and revision fixation for avascular necrosis or nonunion of the proximal pole fracture. Dissection through a scarred wound bed only increases the risk of devascularization by injury to the microcirculation supplying the proximal pole of the scaphoid. In these cases, even the most meticulous dissection yields difficulty in preserving vascularity through the dorsal ridge of the scaphoid while providing adequate visualization for anatomic reduction of the fracture fragments or introduction of bone graft.

For fixation of the proximal pole of the scaphoid, the strength of retrograde screw fixation is statistically equivalent to antegrade. We have demonstrated that with further study surgical approach and technique for these fractures may not have to be limited by fear of poor biomechanical fixation. Screw length is similar between the groups and has been previously well correlated with scaphoid fracture fixation strength; however, it is of the authors’ opinion that for small proximal pole fractures of the scaphoid, achieving a central position in the proximal fragment is of the utmost importance.

While in this ex vivo testing scenario fixation strength was similar, this has yet to be proven in vivo or with clinical outcomes. Without exposure of the proximal pole of the scaphoid, critical examination of radiographs is required to ensure central position of the screw and guidewire within the proximal pole of the scaphoid. Due to the results discussed in this study, we believe that with further study surgeons may be able to make their decision for fixation technique based on approaches to bone grafting, concern for tenuous blood supply, and surgeon preference without fear of poor biomechanical properties.