Intermediate Dynamic Compression and Decreased Posterior Tilt With Interlocked Pins in Femoral Neck Fixation in Synthetic Bone

Abstract

Fixation failure with resulting non-union is the key complication after femoral neck fixation. It can be avoided by permitting dynamic compression and reducing rotation and posterior tilt of the femoral head. To achieve this, a novel implant that features an interlocking plate with three hook-pins (The Hansson Pinloc^® System) was developed from the original two hook-pins. Only an enhanced torsional fixation by the implant modification is reported. The purpose was to compare the biomechanical compressive and bending stability of the original and modified implant in femoral neck fixation. To analyze the contribution of both modified components, three individual pins were included, although not in regular use. Forty-eight synthetic femurs with mid-cervical wedge osteotomies were fixated by two pins or identical triangular pin patterns with or without the plate. Eight specimens of each group were loaded cyclically in compression with an inferior wedge to simulate stance and anteroposterior bending with a posterior wedge to imitate sitting down. The clinically relevant stability measurements were stiffness and deformation. Fissure formation defined failure. The novel implant improved bending stability by 30% increased stiffness, 44% reduced deformation, and less frequent posterior neck fissure formation (p < 0.001) while increased compressive stability was only evident with 25% reduced deformation and less frequent inferior neck fissures (p < 0.001). These impacts were mainly mediated by the third pin, while the plate prevented a lateral fissure in compression (p < 0.001). The clinical stability was improved by dynamic compression and decreased posterior tilt by implant modification.

Introduction

High revision rates of 11–27% persist with internal fixation in non-displaced and subgroups of displaced femoral neck fractures in patients between 55 and 70 years [1,2].

Fixation failure with resulting non-union is the key complication after femoral neck fixation. It can be avoided through a reduced femoral head compression, rotation, and posterior tilt achieved by the common use of pins or screws [3–5]. No clear conclusion has been reached on which traditional fixation is superior in this setting [6].

Applying locking plates is the latest clinical development with this fracture [7–9]. This was based on the advantageous medial hold of multiple screws and lateral hold of a sliding hip screw device combined in locking plates, which have been reported to enhance fixation and restrict fracture site motion ex vivo [10–14]. Disappointingly, altered failure patterns of implant failure or cut-out through the femoral head was reported with such implants [7]. This identified the importance of allowing intermediate dynamic compression with locking plates, to reduce complication and reoperation rates without changing failure patterns [8]. As a dynamic locking plate with individually telescoping screws also may cause implant cut-out, potentially by mechanical jamming inside the femoral neck [9], newer plate systems have to be critically explored [15].

To achieve the three biomechanical goals of permitting dynamic compression combined with reducing rotation and posterior tilt of the femoral head, a novel implant was designed (The Hansson Pinloc System, Swemac Innovations AB, Linköping, Sweden). This novel femoral neck plate interlocking three smooth titanium hook-pins was developed from the two Hansson hook-pins in steel. The Hansson Pinloc System introduced in 2013 represents an innovative concept where the locking plate itself is not fixated laterally to the femoral cortex (Fig. 1). Principally, dynamic compression with sliding of the implant as one single unit is possible, while the rotations and posterior tilting of the femoral head may be counteracted. So far, only increased torsional fixation stability has been demonstrated in a custom-made test model [16]. Following its clinical introduction, concerns on biomechanical performance have been raised, as improved fixation stability may come at cost of increased stress medially or laterally [17]. No evaluation of the theoretically improved compressive and bending fixation stability, requiring an anatomical test model, has been made.

Fig. 1.

The Hansson Pinloc System with three hook-pins interlocking in a plate fixating a femoral neck fracture. The intention of the implant design to (a) ensure dynamic compression and (b) prevent posterior tilting is illustrated on corresponding anterior and lateral view by arrows indicating force vectors of compression and distraction (Illustration from brochure supplied by Swemac Innovation AB, Linköping, Sweden). The concept is theoretical and has been insufficiently evaluated so far.

The natural next step is revealing the novel implant's strengths and weaknesses by bending and compressive testing. The choice of two pins as a natural comparator was based on two Hansson pins being the most prevalently studied pin configuration in randomized clinical trials [6]. The choice of three interlocked pins as an optimum test group was based on an earlier experimental report [16].

The primary aim was to investigate if and how the modified implant improved bending and compressive stability parameters in femoral neck fixation biomechanically. To analyze the role of both modified components, a group of three isolated pins was included in the comparison, despite not commonly used clinically with the Hansson pins. The secondary aim was to study any adverse effects by these changes. We hypothesized increased fixation stability by the modification, as intended with the implant design.

Material and Methods

Model Preparation.

Forty-eight large synthetic femurs (model #3406, Fourth Generation Composite Bone; Sawbones, Pacific Research Laboratories, Vashon, WA) were cut and the proximal 15 cm kept for investigation.

A milling machine (Fadal 4020, Control System Fadal 88 HS; Fadal, Brea, CA) standardized pin channels with a 6.7 mm diameter drill bit. In accordance with the neck dimensions and the best fitting plate, channels were drilled in a top-down triangle. The center-to-center distance between the distal and proximal holes was 14.5 mm and 12.0 mm, respectively, between the proximal holes. The depths of the channels were 105 mm proximally and 115 mm distally, ending subcortically in the head. The holes were drilled at an angle of 125 deg with the shaft, in parallel with the neck. To aid reduction, drilling preceded the osteotomy.

To simulate a cervical fracture, the neck was cut perpendicularly to the channels 53 mm from the head surface in a jig. This created a fracture 55 deg to the horizontal, i.e., midways between a Pauwels’ type 2 and 3 fracture (Fig. 2) [18]. To mimic the situation where complication rates with other locking plates were increased by fracture angulation and displacement [7,9], comminution was simulated by removing a standardized wedge. As unfavorable comminution of the calcar has been reported to predict healing disturbances [19], a varus wedge was removed from subcapitally inferiorly with an 18 deg angle to the first cut in half of the specimens. In the other half, a posterior wedge was created at the same angle from posterior to simulate an unstable fracture pattern predictive of reoperation [5]. Both the incomplete wedges permitted partial cortical bone contact between the proximal and distal fragment, which created a semi-stable situation.

Fig. 2.

The fracture model with fixation methods: The fixated composite bones with cervical osteotomies simulating a semi-stable femoral neck fracture 55 deg to the horizontal. To the upper left an additional 18 deg varus wedge, while the posterior wedge is seen to the upper right. In the lower half the fixation methods A–C from left to right; the original linear two pin configuration (a), the three individual pins (b) and the same triangular configuration with the plate (c).

All implants were chosen from the Hansson Pinloc System (Swemac Innovations AB, Linköping, Sweden). For all configurations, titanium pins (Ti6Al4V) were used with a shank diameter of 6.5 mm, enforced to 6.7 mm at the base, and lengths of 105 mm and 115 mm. A medium-sized plate with 8 mm between the distal and proximal pins was applied. The same surgeon (JEB) inserted all implants in a standardized manner following surgical technique instructions. Osteotomies were fixated by either the original configuration of two pins (Method A), the altered configuration with 3 pins (Method B), or the same triangular pattern as in B, interlocked in the plate (Method C) (Fig. 2). Three-pointed cortical support of each pin was guided by two pins introduced in the lateral entry holes, along the calcar and posteriorly until subcortically in the head (A) and correspondingly by the third pin introduced in the additional anterior channel (B + C). Before introducing the pins' hook into the head, the assembly torque was preset to 5 N·m with a torque wrench when interlocking the pins' threaded base into the plate (C). Following this, eight standardized specimens were prepared for testing of each fixation method, both with the varus wedge in compression (n = 24) and with the posterior wedge in bending (n = 24).

Test Procedure.

The specimens were mounted in a testing machine (MiniBionix 858 MTS Systems, Eden Prairie, MN) with a load cell with following axial characteristics; capacity 10 kN, resolution 1 N, displacement 1 μm, accuracy 0.5%, sensing 0.01 s. The load cell measured force and displacement of the actuator, data being recorded by a computer. The specimens were press-fit inserted distally into a channeled steel-tube. To avoid accumulation of shear forces, the actuator transferred compression through a low-friction piston.

In the compression test, specimens with a varus wedge were mounted with 7 deg adduction to compress the head with the direction of the hip contact force vector during one-leg stance (Fig. 3(a)) [20]. In the bending test, the specimens with a posterior wedge were oriented horizontally and loaded on the anterior aspects of the head to simulate the bending when sitting down [20]. In bending, a support was placed just beneath the minor trochanter to isolate the displacement to the osteotomy (Fig. 3(b)). Cyclic loading was performed with 20,000 and 1000 cycles in compression and bending, respectively, with loadings approximated to measurements in vivo [20]. The number of cycles intended to reflect the time until fracture consolidation in compression [13], while a recommended smaller number of cycles was considered in bending [20]. The loadings were applied dynamically with a sinusoidal motion pattern using load control (rate 1 Hz, maximum compressive load 1900 N, maximum bending load 1300 N, preload 90 N). The applied loads simulated relevant full weight-bearing approximated to the estimated respective joint reaction force in one-leg stance phase and during sitting down of a 92 kg Caucasian male, which the large composite bones are modeled after [21].

Fig. 3.

The test set-ups: (a) set-up for the compression test of a proximal femur with a varus wedge in an upright position and (b) set-up for the bending test of specimens with a posterior wedge in a seated position

Regarding stability parameters, the slope of the best line fit of the load-deformation curve's linear elastic portion defined initial stiffness of the test model during the first three cycles and was used to obtain an average value. Deformation was measured during the loaded phase of the last cycle to reflect deformation by weightbearing in the simulated fracture healing period [13,20]. Formation of fissures defined initial fixation failure, while loosening and destruction of the bone-implant construct defined complete failure.

Statistical Analysis.

Data were processed with matlab software (MathWorks Inc., Natick, MA). Categorical variables were expressed as proportions with variance, crosstabs were computed, and Fisher's exact test applied. Continuous variables were expressed as arithmetic means and dispersion as standard deviations. Due to normality of the data evaluated by Q-Q plots, the groups were compared using parametric testing. For comparisons of continuous parameters across the three groups, one-way analyses of variance (ANOVA) were conducted using IBM SPSS Statistics (version 25 for Windows; SPSS Inc., Chicago, IL). Level of significance was set to P < 0.05. Post hoc multiple comparisons were made with Bonferroni correction. A Pearson product–moment correlation was performed to determine the relationship between outcomes within each test. The correlation coefficients were interpreted as low, moderate, and high respective of their size in intervals between: 0.00, 0.30, 0.50, and 1.00. A post hoc sample-size calculation was performed to ensure that there were enough specimens in each group to detect all statistical differences that may be present, i.e., avoid a type II error. This detected the number needed in each group to demonstrate a difference in parameters with 80% power, ensuring a good study design (STATA, SE 14.1 for Windows, College Station, TX).

Results

Stability Parameters.

Stiffnesses and deformations, along with proportions of failure patterns with fixation methods and their comparisons, are presented in Tables 1 and 2. The mean compressive stiffness ranged from 404 to 467 N/mm and mean bending stiffness from 184 to 240 N/mm. Mean compressive deformation ranged from 5.6 to 10.7 mm and mean bending deformation from 6.5 to 8.7 mm.

Table 1.

The biomechanical test results for all fixation methods

Biomechanical parameter	Compressive fixation stability				Bending fixation stability
Fixation method	Stiffness (N/mm)	Deformation (mm)	Medial failure (n/N)	Lateral failure (n/N)	Stiffness (N/mm)	Deformation (mm)	Posterior failure (n/N)
A:2 isolated pins	404 (52) a	10.7 (2.5) a	7/8 a (0.1)	7/8 a (0.1)	184 (7) a	8.7 (0.4) a	8/8 a (0.0)
B:3 isolated pins	467 (34) a	5.6 (1.1) b	0/8 b (0.0)	6/8 a (0.2)	237 (26) b	6.6 (0.7) b	0/8 b (0.0)
C:3 interlocked pins	439 (60) a	6.0 (1.0) b	1/8 b (0.1)	0/8 b (0.0)	240 (14) b	6.5 (0.6) b	0/8 b (0.0)

Each column expresses axial biomechanical parameters; either mean values or incidence rates of failure modes with respective standard deviation or variance in parentheses. The use of the lower-case letters “a” and “b” indicate statistically significant differences between fixation methods in the same column (p < 0.05).

Table 2.

The comparison between biomechanical test results for all fixation methods

Biomechanical parameter	Compressive fixation stability				Bending fixation stability
Fixation method	Stiffness	Deformation	Medial failure	Lateral failure	Stiffness	Deformation	Posterior failure
Adding pin (B/A)	1.16	0.52a	0.00a	0.86	1.29a	0.76a	0.00a
Adding plate (C/B)	0.94	1.07	—	0.00a (0.0)	1.01	0.98	—
Adding Pinloc (C/A)	1.09	0.56a	0.14a (0.0)	0.00a (0.0)	1.30a	0.75a	0.00a (0.0)

^aComparisons within the same column showing statistically significant difference between fixation methods (p < 0.05).

Each column expresses comparison of axial biomechanical parameters between fixation methods; either Ratio (Y/X) = Mean Fixation method Y/Mean Fixation method X or Risk Y/Risk X for categorical variables.

- Not calculable, formally explained as “undefined “due to zero in the denominator.

The three pins without or within a plate showed a significant improvement in pairwise comparison with two pins, (B/A) and (C/A), respectively, with reduced mean deformation by a factor of 0.52 and 0.56 in compression and 0.76 and 0.75 in bending (p < 0.001). Correspondingly, initial stiffness was increased by a factor of 1.29 and 1.30 in bending (p < 0.001), while the tendency to an increased stiffness in compression (B/A) (p = 0.06) was statistically insignificant comparing the original and modified implant (C/A) (p = 0.5). The post hoc sample-size calculation by comparing the original and modified implant (C/A), revealed that each fixation group should include 48 specimens to detect an impact on compressive stiffness. When comparing the 3 pins within or without plate (C/B), no statistical difference was detected with these parameters (p > 0.8).

Failure Pattern.

Only cortical cracks occurred during compression, either laterally from the boundaries of the holes and spreading in-between (Fig. 4a), and/or medially at the osteotomy inferiorly to the distal pin and developed distally (Fig. 4b). The lateral failure pattern was prevented in the plate group (C = 0/8 versus B = 6/8, P = 0.007 and B = 6/8 versus A = 7/8, p = 1.0). Triangular pin configuration per se reduced the proportion of the medial fracture line (C = 1/8 versus A = 7/8, P = 0.01 and B = 0/8 versus A = 7/8, p = 0.001).

Fig. 4.

The failure patterns: (a) The lateral fissures between pin insertion holes present in two or three individual pins after compression, (b) The medial fissure inferiorly along the calcar was more common in compression with two pins, and (c) The medial fissure along the posterior neck was only identified in the bending test with two pins

In bending, similar medial fissure formation occurred in the posterior neck under the juxta-positioned pin (Fig. 4(c)). The triangular configurations prevented this failure pattern (C or B = 0/8 versus A = 8/8, p < 0.001). Notably, no permanent damage of implants was detected.

The correlation analysis between the outcomes in each test is presented in Table 3. The compressive stability parameters revealed moderate to high correlation coefficients (r = 0.46–0.77, n = 24, p = 0.05) (absolute values), with the comparison between stiffness and lateral failure as the only exception (r = −0.24, n = 24, p = 0.3). The bending test's parameters were highly correlated (r = 0.84–0.96, n = 24, p = 0.01) (absolute values).

Table 3.

The Pearson correlation of biomechanical outcomes within each test

Test	Comparison	Correlation coefficient
Compression	Stiffness versus deformation	−0.67a
	Stiffness versus medial failure	−0.49b
	Stiffness versus lateral failure	−0.24
	Deformation versus medial failure	0.77a
	Deformation versus lateral failure	0.46b
	Medial failure versus lateral failure	0.47b
Bending	Stiffness versus deformation	−0.96a
	Stiffness versus posterior failure	−0.84a
	Deformation versus posterior failure	0.88a

^aCorrelation significant at the 0.01 level.

^bCorrelation significant at the 0.05 level.

Discussion

Interpretation.

The interlocking system improved bending and compressive stability parameters compared to the forerunner configuration. The enhanced bending stability by increased stiffness with reduced deformation and no fixation failure reflects the implant design of intended posterior tilt counteraction. Only a trend toward increased compressive stiffness without significant correlation to the lateral failure pattern was detected. To detect a possible minor impact of the novel implant in compressive stiffness, six times the number of specimens were needed, which questions its importance. This indicated that fracture site compression was allowed as with the original pins. The enhanced dynamic compressive fixation stability according to the reduced deformation and failure initiation in longer lasting test procedures was interpreted as intermediate dynamic compression with the novel interlocked pins.

Physical Explanation.

As expected, both the third pin and the plate contributed to improved fixation.

The only finding exclusively mediated by the plate was the prevention of the compressive lateral failure pattern. This is explained by offloading the lateral cortex between pins, which identifies the plate as a load transmitter building bridges between the pins not only in torsion [16], but also in compression. However, the greatest improvement was caused by the third pin. The impact of triangularity is reasoned by a comparison between moments of inertia for three versus two pins in the testing directions (Fig. 5). A triangular prism (three pins) has similar moments of inertia in both orientations, while a rectangular prism (two pins) has a definite decrease from the standing to sitting position. This load direction-dependent finding revealed by biplanar loading explained the more consistent impact by the third pin on bending fixation stability. The deviation between calculated and real ratios of moment of inertia is most likely attributed to the rough approximations regarding the prims' form and orientation and the theory should be valid. Improved fixation according to the moment of inertia is a common concept, but has previously only been reported in torsional testing of triangular configurations with locking plates in this setting [16].

Fig. 5.

Physical explanation by strength of materials. The formula of the moment of inertia (I) is given for two and three pins interpreted as a rectangular and equilateral triangular prism with “t” as the width and “b” as the height in the situation for vertical (I_Z) and horizontal (I_Y) loading. The ratio of moments of inertia between a triangular and rectangular prism (B/A) is 1.1 in vertical loading and 8.4 in horizontal loading indicating maximum superiority of a third pin confined to seated position. Force vectors (F) indicate the function of the plate as a load transmitter.

Review of the Literature.

While conventional implants allow compression, locking plates differ regarding the amount of interfragmentary compression being permitted. True locking plates principally restrict fracture motion by non-parallel screws [10,11,13,14] or by fully threaded parallel screws [12]. In combination with sliding screws, locking plates may permit intermediate compression [22–25]. Interlocked screws have been the only other available alternative to achieve intermediate compression with locking plate technology [26–28]. With the reported findings in the current study, this observation also may apply to interlocked pins.

A definite impact on stability by true locking plates has only been reported after compressive testing [10,11], but occasionally the results are reported as indifferent from common methods [14], while an impact also has been reported after torsional testing [13]. Previous ex vivo studies of interlocked screws compared a triangular screw configuration with or without an interlocking plate in human femurs with Pauwel's type three fractures in combined compressive and torsional testing. The reported advantages of interlocked screws involve parameters of increased compressive and torsional fixation stability [27–29]. No direct comparison to findings in studies with different set-ups is possible, but the magnitude of improved stability parameters with interlocking seem less striking than with true locking plates [10–14]. The controlled dynamic compression and preserved micromotions reported as being unable to resist unwanted deformation with screws interlocked in parallel support that fracture motion is allowed with interlocked screws [26,27]. This corresponds well with the reduced stability of the bone-implant construct when locking plates permit motion in between the components [29].

To our knowledge, this study is the first to compare interlocked pins with a concurrent change in pin configuration, including evaluation of not only compressive, but also bending stability. We are not aware of other studies in this setting, which interpret improved compressive and bending stability ex vivo as the clinical important ability to permit dynamic compression and counteract posterior tilt. The prominent impact by the third pin in bending and compression contrasted the most prominent impact by the interlocking plate in torsion [16]. This agrees with the findings of slightly, but significantly, reduced total migration of the femoral head fragment, which may be due to the reduced micromotions about the femoral neck axis with interlocking [26,27]. Hence, the results in this study correspond well with biomechanical comparisons reporting increased bending and compressive fixation stability in favor of three versus two screws [30]. In addition, the results in the present study elaborate further, as a definite biomechanical impact without negative adverse effects by modifying the original concept of independent hook-pins was assessed with clinically relevant and more systematic testing in bending and compression, supplementing the findings in torsion [16].

Clinical Relevance.

The initial case series with Pinloc revealed no unforeseen events, while a recent multicenter randomized trial reported similar complication and reoperation rates when compared to the 2 Hansson pins [17,31]. However, concerns related to biomechanical performance [17] identified the need of more preclinical data prior to widespread use.

This complies well with the new EU regulations, which will increase the need of technical documentation when introducing orthopedic devices [32].

Delivery of patient safety requirements by preclinical testing is an important step during product development. Regarding the short-term safety, the failure patterns by implant fatigue or cut-out through the femoral head with true locking plates are less likely with interlocked pins [7], as our results indicate that dynamic compression at the fracture site may be safely guided by interlocked parallel pins sliding en bloc while posterior tilt is prevented. The findings in the current study agree well with no altered failure patterns reported with Pinloc so far [17,31] and contribute to the novel implant's short-term safety declarations by providing at least equivalence to a well-known implant and by systematic technical evaluation of the implant as one unit and each of the components [32,33].

Considering long-term performance expectations requiring bone healing, the key question is whether the relatively increased stability parameters by the interlocking system may be too low to reach clinical importance. Most improvements were caused by a third pin in our study, while only sparse evidence supports a third implant clinically [6]. Interlocking three pins may be more decisive, but are yet not verified clinically [17]. A definite conclusion on the effect of femoral neck interlocking plates may be premature as results are preliminary and biases of a learning curve and multicenter trials cannot be excluded.

Limitations.

The elements of biomechanical study design regarding implant, bone, and test factors may affect the clinical relevance of our findings. Regarding the choice of test groups, only titanium pins were applied to correct for different material properties by using different alloys. Only the medium-sized plate was evaluated and results may differ with plate size [16]. No clear conclusion can be drawn between the clinical results of pins and screws in this setting [6]. Correspondingly, our findings with two or three individual hook-pins may be expected within the ranges of the common three screws. Most improvements being explained by the third pin in the current tests, questions the clinical importance of the new implant, but to answer this question, comparison to other relevant fixation methods is necessary.

Bone quality is a key variable in the effectiveness of fracture fixation ex vivo [34]. Clinically, osteoporosis more often precedes reoperation in femoral neck fixation [35]. Hence, failure pattern evaluations with human femurs are recommended. The clinical relevance of failure patterns in the present study is uncertain, as the appearance may be attributed the more brittle composite femurs [21]. However, the prevention of fissures was utilized to explain each component of the novel fixation as a stress reducer. Rather than detection relevant absolute values with the gold standard of paired human femurs ex vivo [21], standardized synthetic femurs enabled comparison of relative differences. To identify and explain a stepwise effect by both modified implant components were possible only by comparing multiple fixations. Despite neglecting the bone quality variation, the synthetic replicas perform within the range of healthy bone [36]. The advantage of more reproducibility makes replicas preferable in such relative comparisons.

The osteotomies replicated fractures that are known to predict healing complications due to fixation instability. The need of increased fixation stability may have accentuated difference between test groups. Standardized reduction and pin positioning represented the ideal cases. With an acceptable displacement, the fragments' contribution to the fractures inherent stability is not essentially reduced and positioning of the pins will only be tolerably changed. Harmoniously, our findings should be transferable to the setting of less successful implant positioning and mild displacement.

The simplified tests of full weight-bearing in prolonged intervals facilitated the findings of failure patterns in our study. While no load-to-failure evaluation was performed, the dynamic testing evaluated overloading by fatigue to deliver safety declarations. Regarding a more relevant partial weight-bearing, this would reduce stability requirements by the implants, but we argue differences would persist as they are revealed in relative comparisons. The important tradeoff, that enhanced fixation stability may result in heightened contact stresses have been highlighted in this setting [17]. However, the latter parameter was not directly evaluated in this study.

The significant correlations between outcomes within each load direction, question the overlap between outcomes. We reason that the varying magnitudes of the coefficients express different physical aspects of fixation stability, which add to the argumentation of including both static and cyclic testing in systematic evaluations. In this study, such an approach led to the detection of the dynamic compression feature by the novel implant.

We are aware that a calculation of the “observed power,” i.e., a statistical power calculation based on the observed effect, gives no additional information to the reported p-value. However, we argue that the large post hoc sample-size reported with compressive stiffness was neither feasible nor indicative of an important difference to detect.

Despite potentially finding different absolute values from advanced physiological loading in human femurs, neither the artificial bone, the fractures nor the pins' material properties explain the relative differences between the fixation groups in the present study. Improved biomechanics were therefore most likely caused by the geometry of the novel interlocked triangular pin configuration.

An inherent value of biomechanical studies is identified as our findings and interpretations agree with and validate the physical calculations of the implant design, even if clinical superiority remains to be documented. This challenges the conception of biomechanical studies being of no more value than to suggest implants for clinical testing, which was based on the poor relationship between findings ex vivo and in vivo with locking plates not allowing fracture site compression [37].

Conclusions

Femoral neck fixation by the interlocked pins allows intermediate dynamic compression at the fracture site and decreases posterior tilting. Most improvements were mediated by the third pin with an additional impact by the plate when compared to the original implant. Superiority ex vivo to its forerunner delivers the short-term safety requirements and justifies implant introduction. Only a limited biomechanical impact regarding long-term safety declarations is consistent with clinical results of similar non-union rates as with the original two hook-pins so far.

Funding Data

This research did not receive any grants from funding agencies.