FOOT AND ANKLE: RESEARCH
Screw fixation of medial malleolar fractures A CADAVERIC BIOMECHANICAL STUDY CHALLENGING THE CURRENT AO PHILOSOPHY
L. Parker, N. Garlick, I. McCarthy, S. Grechenig, W. Grechenig, P. Smitham From The Royal Free Hospital, London, United Kingdom
L. Parker, BM, MRCS(Eng), FRCS(Tr & Orth), Specialist Registrar Trauma and Orthopaedic Surgery N. Garlick, FRCS(Orth), Consultant Orthopaedic Surgeon The Royal Free Hospital, Pond Street, London NW3 2QG, UK. I. McCarthy, BSc, PhD, FIPEM, Principal Research Fellow P. Smitham, PhD, MRCS(Eng), Clinical Lecturer, Specialist Registrar Institute of Orthopaedics and Musculoskeletal Sciences, The Royal National Orthopaedic Hospital, Brockley Hill, Stanmore, Middlesex HA7 4LP, UK. S. Grechenig, Dr.med.Univ., Doctor University Hospital Regensburg, Department of Traumatology, Franz-JosefStrauss Allee, Regensburg, Germany. W. Grechenig, Prim.Prof.Dr., Professor AUVA Trauma Hospital, Graz, Gostingerstrasse 24, 8020, Graz, Austria. Correspondence should be sent to L. Parker; e-mail: [email protected] ©2013 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.95B12. 30498 $2.00 Bone Joint J 2013;95-B:1662–6. Received 20 July 2012; Accepted after revision 24 May 2013
1662
The AO Foundation advocates the use of partially threaded lag screws in the fixation of fractures of the medial malleolus. However, their threads often bypass the radiodense physeal scar of the distal tibia, possibly failing to obtain more secure purchase and better compression of the fracture. We therefore hypothesised that the partially threaded screws commonly used to fix a medial malleolar fracture often provide suboptimal compression as a result of bypassing the physeal scar, and proposed that better compression of the fracture may be achieved with shorter partially threaded screws or fully threaded screws whose threads engage the physeal scar. We analysed compression at the fracture site in human cadaver medial malleoli treated with either 30 mm or 45 mm long partially threaded screws or 45 mm fully threaded screws. The median compression at the fracture site achieved with 30 mm partially threaded screws (0.95 kg/cm2 (interquartile range (IQR) 0.8 to 1.2) and 45 mm fully threaded screws (1.0 kg/cm2 (IQR 0.7 to 2.8)) was significantly higher than that achieved with 45 mm partially threaded screws (0.6 kg/cm2 (IQR 0.2 to 0.9)) (p = 0.04 and p < 0.001, respectively). The fully threaded screws and the 30mm partially threaded screws were seen to engage the physeal scar under an image intensifier in each case. The results support the use of 30 mm partially threaded or 45 mm fully threaded screws that engage the physeal scar rather than longer partially threaded screws that do not. A 45 mm fully threaded screw may in practice offer additional benefit over 30 mm partially threaded screws in increasing the thread count in the denser paraphyseal region. Cite this article: Bone Joint J 2013;95-B:1662–6.
Fractures through dense cortical bone may be treated by fixation with a lag screw, where the threads nearest the tip of the screw engage the cortex beyond the fracture firmly while those nearest the head of the screw are free to slide in the fragment to be secured. This is achieved by drilling a hole in the displaced fragment to the outer diameter of the screw threads; alternatively, a partially threaded screw can be used, so that when the screw is tightened, the fracture site is compressed.1-3 This technique is used in the treatment of fractures of the medial malleolus where anatomical reduction and rigid internal fixation is desirable.4 The traditional fixation recommended by the AO Foundation involves two parallel partially threaded 4.0 mm diameter cancellous screws, augmented with a washer in osteoporotic bone.5 During insertion of a screw, tightening is limited to a maximum torque and beyond this, either the screw or the bone fails. Provided the screw remains intact the relationship between insertion torque and axial compression is
linear while the bone trabeculae are intact. Further tightening beyond approximately 89% of the insertion torque results in microfracturing of the bone trabeculae that significantly reduces the axial compression at the fracture site.3 In our experience, the commonly used long partially threaded screws are more prone to lose purchase if over-tightened, resulting in reduced compression of the fracture, especially in osteoporotic bone. We hypothesised that superior grip might be obtained within the physeal scar of the distal tibia, where bone may be denser and that the threads of a screw obtaining purchase in this region in human cadavers would obtain better compression and more secure fixation of the fracture than the traditionally used longer partially threaded screws (Fig. 1). We are unaware of any studies examining compression at the fracture site and relating it to the position of the screw threads in the radiodense distal tibial physeal scar. Previous studies have focused on the design characterisTHE BONE & JOINT JOURNAL
SCREW FIXATION OF MEDIAL MALLEOLAR FRACTURES
1663
Fig. 2
Fig. 1 Plain anteroposterior radiograph showing the radiodense bone of the distal tibial physeal scar located approximately 1 cm to 2 cm above the tibial plafond.
tics of the screw that optimise the pull-out strength and compressive force, such as the geometry of the threads and the effects of tapping and cannulation.6-8 The density of the bone also clearly influences the performance of the screw.9,10 Our null hypothesis was that no difference would be found in the compression at the fracture site that is achieved when long partially threaded screws are compared with shorter partially threaded screws or fully threaded screws that engage the physeal scar.
Materials and Methods A comparison of the compression at the malleolar fracture site achieved in human cadavers was conducted at the Anatomical Institute, University of Graz, Austria. Experimental subgroups were designed to investigate the compression that was achieved with long and short partially threaded and fully threaded screws, noting the location of the screw threads on an image intensifier. The experiments were performed using 21 randomised unpaired adult cadaver lower limbs preserved with the method of Thiel.11 The embalming method preserves the colour and consistency of the tissues as well as allowing an almost full range of movement of the joints. None of the limbs used in this study had signs of previous injury, abnormality or disease. The mean age of the donors was 71 years (44 to 84) at the time of death and the gender distribution was 1:1. The deltoid ligament was detached from the medial malleolus. In each ankle a typical fracture of the medial malleolus, level with the distal tibial plafond, was marked out and a 2.5 mm hole pre-drilled with an AO drill-bit perpendicular to the plane of the fracture to optimise VOL. 95-B, No. 11, DECEMBER 2013
Photograph of the experimental set-up. The cadaver ankle specimen has a simulated medial malleolar fracture at the level of the tibial plafond. The pressure transducer is inserted into the fracture site and held temporarily with reduction forceps, and the screw is then inserted into the pre-drilled pilot hole.
compression. The medial malleolus was divided with a 25 mm wide oscillating saw blade to simulate the fracture. Compression of the fracture was measured using a Tekscan pressure transducer (Tekscan Inc., South Boston, Massachusetts) calibrated to display pressure in kg/cm2 and inserted into the fracture site. The pressure cells in the paper-thin transducer produce a mean value for the entire area within the fracture site, omitting the area that is occupied by the screw. Thus, in order to maximise the area available for pressure measurement, only one screw at a time was used to secure the malleolus in every case. The malleolar fragment was aligned and held with reduction clamps while the AO screws were inserted, and the pressure transducer was compressed until the screw achieved maximum compression. The transducer was activated after the screw was tightened, so that pressure exerted by the investigator was removed as a source of error and the mean pressure in kg/cm2 over 20 seconds was recorded (Fig. 2). The compression at the fracture site was tested using two paired experimental groups. The main group of interest was group A with B (n = 13, numbered one to 13) and then groups C and D (n = 8, numbered 14 to 21). Therefore, ankle 1 had the fixation performed with a long partially threaded screw (group A; 45 mm long AO 4.0 mm diameter partially threaded cancellous screws (15 mm thread length) with a washer) and a measurement taken; this was then repeated by drilling a separate hole away from the long partially threaded screw and inserting a fully threaded screw (group B; 45 mm long AO 4.0 mm diameter fully threaded cancellous screws without a washer) in a similar trajectory; this was confirmed with an image intensifier and the measurement was then repeated. The study was undertaken in this manner to rationalise the cadaver material and to allow direct comparisons to be made between compressions obtained at the same fracture site (Fig. 3). For groups A and
1664
L. PARKER, N. GARLICK, I. MCCARTHY, S. GRECHENIG, W. GRECHENIG, P. SMITHAM
A
2.5 mm pilot hole
B
C
D
+ A: 4.0 × 45 × 15 mm partially-threaded cancellous screw + B: 4.0 × 45 mm fully-threaded cancellous screw + C: 4.0 × 45 mm fully-threaded cancellous screw with 4.0 mm glide hole + D: 4.0 × 30 × 16 mm partially-threaded cancellous screw Fig. 3 Diagram of the four experimental groups.
B this step was repeated 13 times in cadaver ankles 1 to 13, using the respective screw types. In group C, ankles, 14 to 21, we examined the effect of over-drilling the near fragment to 4.0 mm in order to provide a glide-hole and achieve a true lag effect using a 4.0 mm AO fully threaded cancellous screw 45mm in length with a washer, and repeated it eight times. In group D ankles we examined the effect of using a shorter (30 mm) partially threaded cancellous AO screw with a washer while maintaining a thread length of 15 mm (ankles 14 to 21). In order to confirm a similar trajectory of the screw in each fixation and to correlate the measurements of the compression at the fracture site to the position of the screw threads in relation to the physeal scar, a mini C-arm image intensifier was used. DEXA scanning was also performed in a sample of the ankles (1 to 5) to examine their relative bone densities and to ensure a degree of uniformity of the specimens used. Statistical analysis. Data were analysed using Stata/IC version 12.1 (StataCorp, College Station, Texas), and a p-value < 0.05 was considered statistically significant. The Shapiro–Wilk test for normality confirmed that the mean compression measurements (in kg/cm2 over 20 s) from the human cadaver experiment were not normally distributed. The non-parametric Wilcoxon’s signed ranks test was
therefore used to compare the paired measurements and the Mann–Whitney U test was used to compare the unpaired measurements.
Results In all 21 ankles the image intensifier confirmed that the screw threads of the 45 mm long partially threaded screws passed beyond the physeal scar. However, the threads of the fully threaded screws and those of the shorter, 30 mm partially threaded screws fully engaged the physeal scar. The median compression at the fracture site generated over 20 s was significantly greater in fractures treated with fully threaded, 45 mm long AO 4.0 mm cancellous screws without a washer (Group B: 1.0 kg/cm2 (interquartile range (IQR) 0.7 to 2.8)) than with partially threaded 45 mm long AO 4.0 mm cancellous screws with a washer (group A: 0.6 kg/cm2 (IQR 0.2 to 0.9)) (p < 0.001, Wilcoxon’s signed ranks test) (Fig. 4). The median compression at the fracture site generated over 20 s for those ankles treated with fully threaded 45 mm long AO 4.0 mm cancellous screws with a washer, where the near fragment of the fracture was over-drilled to facilitate a lag effect (Group C: 0.25 kg/cm2 (IQR 0.2 to 0.5)), was significantly reduced compared with group B (fractures treated with fully threaded 45 mm long THE BONE & JOINT JOURNAL
SCREW FIXATION OF MEDIAL MALLEOLAR FRACTURES
1665
Table I. Results of dual energy x-ray absorptiometry of five ankles in the experimental series
Fig. 4 Box and whisker plot illustrating the median average fracture compression in each group. These boxes represent the interquartile range.
AO 4.0 mm cancellous screws without a washer and without over-drilling) (p = 0.002, Mann–Whitney U test) (Fig. 4). Short (30 mm) partially threaded AO screws with a washer and engaging the physeal scar (Group D) generated over 20 s a median compression at the fracture site of 0.95 kg/cm2 (IQR 0.8 to 1.2). This was significantly greater than Group A (longer partially threaded screws with a washer) (p = 0.04, Mann–Whitney U test). There was no significant difference between the median compression using the short partially threaded screws (Group D) and that using the long fully threaded screws (Group B) (p = 0.56, Mann–Whitney U test). In a sample of ankles (one to five) taken from experimental groups A and B, DEXA scanning showed that the bone densities of the cadaveric specimens were similar; however, the sample size was too small to examine a relationship between compression at the fracture site and bone density (Table I).
Discussion The AO Foundation advise that fixation of medial malleolar fractures be carried out with partially threaded screws, supplemented in osteoporotic bone with a washer.5 Poor purchase of the threads of a long partially threaded screw will cause poor compression of the fracture and possible failure of fixation. The results of the human cadaver study showed that significantly higher compression at the fracture site can be generated with the use of long (45 mm) fully threaded screws compared with long (45 mm) partially threaded screws (p < 0.001). Imaging confirmed that fully threaded screws obtain purchase within the radiodense bone in the physeal scar, whereas partially threaded screws of an equivalent length bypass this bone. The radiodense physeal scar of the distal tibia clearly offers VOL. 95-B, No. 11, DECEMBER 2013
Ankle
Density (g/cm2)
1 2 3 4 5
0.231 0.483 0.789 0.762 0.524
the best purchase for the threads of the screw and the most stable fixation. As the physeal scar is the main determinant of fixation, as we have shown, and the near fragment is of low density with minimal thread purchase, compression of the fracture is still achieved with a lag effect when fully threaded screws are used. We assume the bone of the near fragment is of such low density that the threads of the fullythreaded screw are free to slide as the screw is tightened. This experiment suggests that there is no advantage to inserting long screws with a thread lying well beyond the physeal scar, as the thread lies in low-density bone. An obvious weakness of this study is that the explanation for the difference in the behaviour of the screws being due to the purchase of the threads in the physeal scar requires 3D analysis of bone density in the human distal tibia with quantitative DEXA (pQCT scanning). Further research using either cadaveric distal tibiae or preferably on patients following the required ethical approval is needed to confirm that the physeal scar is of significantly greater density than the bone of the mid-metaphyseal region where threads of the long partially-threaded screw come to rest after insertion. Our results show that the compression of the fracture site that is achieved by long fully threaded screws can also be achieved with far shorter (30 mm) partially threaded screws. However, with the short partially-threaded screw, it may be that not enough screw would pass the fracture site to resist shear and torsional forces, which were not analysed in this study. The implication is that fully-threaded screws, despite achieving similar compression of the fracture may offer an additional advantage over the 30 mm partially threaded screw in maximising the thread count beyond the fracture site. We noted that the beneficial effect on compression at the fracture site was removed when fully threaded screws were inserted with a near fragment glide hole (group C ankles), suggesting that the removal of bony material is detrimental to achieving secure fixation in osteoporotic bone. It has previously been shown that tapping removes material in a similar way, enlarging the hole so that the screw fails by increasing the shear area for the threads.7 It remains to be seen exactly which malleolar fractures are suitable for treatment with fully threaded screws. This study was undertaken on a small sample size in osteoporotic human cadaver bone, and it is possible that in younger
1666
L. PARKER, N. GARLICK, I. MCCARTHY, S. GRECHENIG, W. GRECHENIG, P. SMITHAM
patients without osteoporosis a fully threaded screw would fail to achieve a lag effect and may instead act as a positional screw, preventing compression at the fracture site. In this situation a shorter, partially threaded screw may be more suitable. From this study, it is not possible to determine whether purchase within the physeal scar is crucial in all medial malleolar fractures, although in osteoporotic bone it seems to improve the purchase of the screw threads and compression at the fracture site. In conclusion, the significant improvement in stability of fixation of a medial malleolar fracture conferred by fully threaded screws in this study challenges the AO doctrine advocating the use of partially threaded screws. The more secure fixation is likely to be as a result of purchase of the screw threads within the radiodense physeal scar, which can be achieved with either fully threaded screws or shorter, partially threaded screws. This purchase is desirable, particularly in osteoporotic bone. The fully threaded screw may offer additional advantages over the short 30 mm partially threaded screw, by maximising the thread count beyond the fracture, even though they both purchase the physeal scar and produce similar compression at the fracture site. We therefore recommend that fully threaded screws be used instead of long partially threaded screws when treating medial malleolar fractures in osteoporotic bone. Supplementary material Two tables detailing the results of the studies of the compression at the fracture site for all four groups are available alongside the electronic version of this article on our website www.bjj.boneandjoint.org.uk
The authors would like to thank Professor A. Weinberg, Department of Paediatric and Adolescent Surgery, Medical University of Graz, Austria, P. Ferlic and H. Dimai from Klinische Abteilung für Endokrinologie und Stoffwechsel, Graz, Austria for DEXA analysis, and S. Cro, Clinical Trials Unit, Medical Research Council, UK for statistical analysis. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. This article was primary edited by J. Scott and first-proof edited by G. Scott.
References 1. Solomon L, Warwick DJ, Nayagam S. Apley’s system of orthopaedics and fractures. Eighth ed. London: Hodder Arnold, 2001:553–554. 2. Wood GW. General principles of fracture treatment. In: Campbell’s operative orthopaedics. Vol. 3. Eleventh ed. Philadelphia: Mosby-Elsevier, 2008:3053–3054. 3. Krettek C, Gosling T. Principles of internal fixation. In: Rockwood and Green’s fractures in adults. Vol. 1. Sixth ed. Philadelphia Lippincott Williams and Wilkins, 2006:223–226. 4. Hak DJ, Lee MA. Ankle fractures: open reduction internal fixation. In Master techniques in orthopaedic surgery-fractures. Second ed. Philadelphia: Lippincott Williams and Wilkins, 2006:556–557. 5. Barbosa P, Bonnaire F, Kojima K. AO Foundation. Malleoli 44-C2 ORIF for medial malleolus; reduction and fixation; medial lag screws (transverse fracture). https:// www2.aofoundation.org/wps/portal/surgery/?showPage=redfix&bone=Tibia&segment=Malleoli&classification=44-C2&treatment=&method=ORIF+for+medial+malleolus&implantstype=Lag+screw&redfix_url= (date last accessed 29 May 2013). 6. DeCoster TA, Heetderks DB, Downey DJ, Ferries JS, Jones W. Optimizing bone screw pullout force. J Orthop Trauma 1990;4:169–174. 7. Chapman JR, Harrington RM, Lee KM, et al. Factors affecting the pullout strength of cancellous bone screws. J Biomech Eng 1996;118:391–399. 8. Brown GA, McCarthy T, Bourgeault CA, Callahan DJ. Mechanical performance of standard and cannulated 4.0-mm cancellous bone screws. J Orthop Res 2000;18:307–312. 9. Faran KJ, Ichioka N, Trzeciak MA, et al. Effect of bone quality of the forces generated by compression screws. J Biomech 1999;32:861–864. 10. Alonso-Vázquez A, Lauge-Pedersen H, Lidgren L, Taylor M. The effect of bone quality on the stability of ankle arthrodesis: a finite element study. Foot Ankle Int 2004;25:840–850. 11. Thiel W. The preservation of the whole corpse with natural color. Ann Anat 1992;174:189–195 (in German).
THE BONE & JOINT JOURNAL
Screw fixation of medial malleolar fractures A CADAVERIC BIOMECHANICAL STUDY CHALLENGING THE CURRENT AO PHILOSOPHY
L. Parker, N. Garlick, I. McCarthy, S. Grechenig, W. Grechenig, P. Smitham From The Royal Free Hospital, London, United Kingdom
L. Parker, BM, MRCS(Eng), FRCS(Tr & Orth), Specialist Registrar Trauma and Orthopaedic Surgery N. Garlick, FRCS(Orth), Consultant Orthopaedic Surgeon The Royal Free Hospital, Pond Street, London NW3 2QG, UK. I. McCarthy, BSc, PhD, FIPEM, Principal Research Fellow P. Smitham, PhD, MRCS(Eng), Clinical Lecturer, Specialist Registrar Institute of Orthopaedics and Musculoskeletal Sciences, The Royal National Orthopaedic Hospital, Brockley Hill, Stanmore, Middlesex HA7 4LP, UK. S. Grechenig, Dr.med.Univ., Doctor University Hospital Regensburg, Department of Traumatology, Franz-JosefStrauss Allee, Regensburg, Germany. W. Grechenig, Prim.Prof.Dr., Professor AUVA Trauma Hospital, Graz, Gostingerstrasse 24, 8020, Graz, Austria. Correspondence should be sent to L. Parker; e-mail: [email protected] ©2013 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.95B12. 30498 $2.00 Bone Joint J 2013;95-B:1662–6. Received 20 July 2012; Accepted after revision 24 May 2013
1662
The AO Foundation advocates the use of partially threaded lag screws in the fixation of fractures of the medial malleolus. However, their threads often bypass the radiodense physeal scar of the distal tibia, possibly failing to obtain more secure purchase and better compression of the fracture. We therefore hypothesised that the partially threaded screws commonly used to fix a medial malleolar fracture often provide suboptimal compression as a result of bypassing the physeal scar, and proposed that better compression of the fracture may be achieved with shorter partially threaded screws or fully threaded screws whose threads engage the physeal scar. We analysed compression at the fracture site in human cadaver medial malleoli treated with either 30 mm or 45 mm long partially threaded screws or 45 mm fully threaded screws. The median compression at the fracture site achieved with 30 mm partially threaded screws (0.95 kg/cm2 (interquartile range (IQR) 0.8 to 1.2) and 45 mm fully threaded screws (1.0 kg/cm2 (IQR 0.7 to 2.8)) was significantly higher than that achieved with 45 mm partially threaded screws (0.6 kg/cm2 (IQR 0.2 to 0.9)) (p = 0.04 and p < 0.001, respectively). The fully threaded screws and the 30mm partially threaded screws were seen to engage the physeal scar under an image intensifier in each case. The results support the use of 30 mm partially threaded or 45 mm fully threaded screws that engage the physeal scar rather than longer partially threaded screws that do not. A 45 mm fully threaded screw may in practice offer additional benefit over 30 mm partially threaded screws in increasing the thread count in the denser paraphyseal region. Cite this article: Bone Joint J 2013;95-B:1662–6.
Fractures through dense cortical bone may be treated by fixation with a lag screw, where the threads nearest the tip of the screw engage the cortex beyond the fracture firmly while those nearest the head of the screw are free to slide in the fragment to be secured. This is achieved by drilling a hole in the displaced fragment to the outer diameter of the screw threads; alternatively, a partially threaded screw can be used, so that when the screw is tightened, the fracture site is compressed.1-3 This technique is used in the treatment of fractures of the medial malleolus where anatomical reduction and rigid internal fixation is desirable.4 The traditional fixation recommended by the AO Foundation involves two parallel partially threaded 4.0 mm diameter cancellous screws, augmented with a washer in osteoporotic bone.5 During insertion of a screw, tightening is limited to a maximum torque and beyond this, either the screw or the bone fails. Provided the screw remains intact the relationship between insertion torque and axial compression is
linear while the bone trabeculae are intact. Further tightening beyond approximately 89% of the insertion torque results in microfracturing of the bone trabeculae that significantly reduces the axial compression at the fracture site.3 In our experience, the commonly used long partially threaded screws are more prone to lose purchase if over-tightened, resulting in reduced compression of the fracture, especially in osteoporotic bone. We hypothesised that superior grip might be obtained within the physeal scar of the distal tibia, where bone may be denser and that the threads of a screw obtaining purchase in this region in human cadavers would obtain better compression and more secure fixation of the fracture than the traditionally used longer partially threaded screws (Fig. 1). We are unaware of any studies examining compression at the fracture site and relating it to the position of the screw threads in the radiodense distal tibial physeal scar. Previous studies have focused on the design characterisTHE BONE & JOINT JOURNAL
SCREW FIXATION OF MEDIAL MALLEOLAR FRACTURES
1663
Fig. 2
Fig. 1 Plain anteroposterior radiograph showing the radiodense bone of the distal tibial physeal scar located approximately 1 cm to 2 cm above the tibial plafond.
tics of the screw that optimise the pull-out strength and compressive force, such as the geometry of the threads and the effects of tapping and cannulation.6-8 The density of the bone also clearly influences the performance of the screw.9,10 Our null hypothesis was that no difference would be found in the compression at the fracture site that is achieved when long partially threaded screws are compared with shorter partially threaded screws or fully threaded screws that engage the physeal scar.
Materials and Methods A comparison of the compression at the malleolar fracture site achieved in human cadavers was conducted at the Anatomical Institute, University of Graz, Austria. Experimental subgroups were designed to investigate the compression that was achieved with long and short partially threaded and fully threaded screws, noting the location of the screw threads on an image intensifier. The experiments were performed using 21 randomised unpaired adult cadaver lower limbs preserved with the method of Thiel.11 The embalming method preserves the colour and consistency of the tissues as well as allowing an almost full range of movement of the joints. None of the limbs used in this study had signs of previous injury, abnormality or disease. The mean age of the donors was 71 years (44 to 84) at the time of death and the gender distribution was 1:1. The deltoid ligament was detached from the medial malleolus. In each ankle a typical fracture of the medial malleolus, level with the distal tibial plafond, was marked out and a 2.5 mm hole pre-drilled with an AO drill-bit perpendicular to the plane of the fracture to optimise VOL. 95-B, No. 11, DECEMBER 2013
Photograph of the experimental set-up. The cadaver ankle specimen has a simulated medial malleolar fracture at the level of the tibial plafond. The pressure transducer is inserted into the fracture site and held temporarily with reduction forceps, and the screw is then inserted into the pre-drilled pilot hole.
compression. The medial malleolus was divided with a 25 mm wide oscillating saw blade to simulate the fracture. Compression of the fracture was measured using a Tekscan pressure transducer (Tekscan Inc., South Boston, Massachusetts) calibrated to display pressure in kg/cm2 and inserted into the fracture site. The pressure cells in the paper-thin transducer produce a mean value for the entire area within the fracture site, omitting the area that is occupied by the screw. Thus, in order to maximise the area available for pressure measurement, only one screw at a time was used to secure the malleolus in every case. The malleolar fragment was aligned and held with reduction clamps while the AO screws were inserted, and the pressure transducer was compressed until the screw achieved maximum compression. The transducer was activated after the screw was tightened, so that pressure exerted by the investigator was removed as a source of error and the mean pressure in kg/cm2 over 20 seconds was recorded (Fig. 2). The compression at the fracture site was tested using two paired experimental groups. The main group of interest was group A with B (n = 13, numbered one to 13) and then groups C and D (n = 8, numbered 14 to 21). Therefore, ankle 1 had the fixation performed with a long partially threaded screw (group A; 45 mm long AO 4.0 mm diameter partially threaded cancellous screws (15 mm thread length) with a washer) and a measurement taken; this was then repeated by drilling a separate hole away from the long partially threaded screw and inserting a fully threaded screw (group B; 45 mm long AO 4.0 mm diameter fully threaded cancellous screws without a washer) in a similar trajectory; this was confirmed with an image intensifier and the measurement was then repeated. The study was undertaken in this manner to rationalise the cadaver material and to allow direct comparisons to be made between compressions obtained at the same fracture site (Fig. 3). For groups A and
1664
L. PARKER, N. GARLICK, I. MCCARTHY, S. GRECHENIG, W. GRECHENIG, P. SMITHAM
A
2.5 mm pilot hole
B
C
D
+ A: 4.0 × 45 × 15 mm partially-threaded cancellous screw + B: 4.0 × 45 mm fully-threaded cancellous screw + C: 4.0 × 45 mm fully-threaded cancellous screw with 4.0 mm glide hole + D: 4.0 × 30 × 16 mm partially-threaded cancellous screw Fig. 3 Diagram of the four experimental groups.
B this step was repeated 13 times in cadaver ankles 1 to 13, using the respective screw types. In group C, ankles, 14 to 21, we examined the effect of over-drilling the near fragment to 4.0 mm in order to provide a glide-hole and achieve a true lag effect using a 4.0 mm AO fully threaded cancellous screw 45mm in length with a washer, and repeated it eight times. In group D ankles we examined the effect of using a shorter (30 mm) partially threaded cancellous AO screw with a washer while maintaining a thread length of 15 mm (ankles 14 to 21). In order to confirm a similar trajectory of the screw in each fixation and to correlate the measurements of the compression at the fracture site to the position of the screw threads in relation to the physeal scar, a mini C-arm image intensifier was used. DEXA scanning was also performed in a sample of the ankles (1 to 5) to examine their relative bone densities and to ensure a degree of uniformity of the specimens used. Statistical analysis. Data were analysed using Stata/IC version 12.1 (StataCorp, College Station, Texas), and a p-value < 0.05 was considered statistically significant. The Shapiro–Wilk test for normality confirmed that the mean compression measurements (in kg/cm2 over 20 s) from the human cadaver experiment were not normally distributed. The non-parametric Wilcoxon’s signed ranks test was
therefore used to compare the paired measurements and the Mann–Whitney U test was used to compare the unpaired measurements.
Results In all 21 ankles the image intensifier confirmed that the screw threads of the 45 mm long partially threaded screws passed beyond the physeal scar. However, the threads of the fully threaded screws and those of the shorter, 30 mm partially threaded screws fully engaged the physeal scar. The median compression at the fracture site generated over 20 s was significantly greater in fractures treated with fully threaded, 45 mm long AO 4.0 mm cancellous screws without a washer (Group B: 1.0 kg/cm2 (interquartile range (IQR) 0.7 to 2.8)) than with partially threaded 45 mm long AO 4.0 mm cancellous screws with a washer (group A: 0.6 kg/cm2 (IQR 0.2 to 0.9)) (p < 0.001, Wilcoxon’s signed ranks test) (Fig. 4). The median compression at the fracture site generated over 20 s for those ankles treated with fully threaded 45 mm long AO 4.0 mm cancellous screws with a washer, where the near fragment of the fracture was over-drilled to facilitate a lag effect (Group C: 0.25 kg/cm2 (IQR 0.2 to 0.5)), was significantly reduced compared with group B (fractures treated with fully threaded 45 mm long THE BONE & JOINT JOURNAL
SCREW FIXATION OF MEDIAL MALLEOLAR FRACTURES
1665
Table I. Results of dual energy x-ray absorptiometry of five ankles in the experimental series
Fig. 4 Box and whisker plot illustrating the median average fracture compression in each group. These boxes represent the interquartile range.
AO 4.0 mm cancellous screws without a washer and without over-drilling) (p = 0.002, Mann–Whitney U test) (Fig. 4). Short (30 mm) partially threaded AO screws with a washer and engaging the physeal scar (Group D) generated over 20 s a median compression at the fracture site of 0.95 kg/cm2 (IQR 0.8 to 1.2). This was significantly greater than Group A (longer partially threaded screws with a washer) (p = 0.04, Mann–Whitney U test). There was no significant difference between the median compression using the short partially threaded screws (Group D) and that using the long fully threaded screws (Group B) (p = 0.56, Mann–Whitney U test). In a sample of ankles (one to five) taken from experimental groups A and B, DEXA scanning showed that the bone densities of the cadaveric specimens were similar; however, the sample size was too small to examine a relationship between compression at the fracture site and bone density (Table I).
Discussion The AO Foundation advise that fixation of medial malleolar fractures be carried out with partially threaded screws, supplemented in osteoporotic bone with a washer.5 Poor purchase of the threads of a long partially threaded screw will cause poor compression of the fracture and possible failure of fixation. The results of the human cadaver study showed that significantly higher compression at the fracture site can be generated with the use of long (45 mm) fully threaded screws compared with long (45 mm) partially threaded screws (p < 0.001). Imaging confirmed that fully threaded screws obtain purchase within the radiodense bone in the physeal scar, whereas partially threaded screws of an equivalent length bypass this bone. The radiodense physeal scar of the distal tibia clearly offers VOL. 95-B, No. 11, DECEMBER 2013
Ankle
Density (g/cm2)
1 2 3 4 5
0.231 0.483 0.789 0.762 0.524
the best purchase for the threads of the screw and the most stable fixation. As the physeal scar is the main determinant of fixation, as we have shown, and the near fragment is of low density with minimal thread purchase, compression of the fracture is still achieved with a lag effect when fully threaded screws are used. We assume the bone of the near fragment is of such low density that the threads of the fullythreaded screw are free to slide as the screw is tightened. This experiment suggests that there is no advantage to inserting long screws with a thread lying well beyond the physeal scar, as the thread lies in low-density bone. An obvious weakness of this study is that the explanation for the difference in the behaviour of the screws being due to the purchase of the threads in the physeal scar requires 3D analysis of bone density in the human distal tibia with quantitative DEXA (pQCT scanning). Further research using either cadaveric distal tibiae or preferably on patients following the required ethical approval is needed to confirm that the physeal scar is of significantly greater density than the bone of the mid-metaphyseal region where threads of the long partially-threaded screw come to rest after insertion. Our results show that the compression of the fracture site that is achieved by long fully threaded screws can also be achieved with far shorter (30 mm) partially threaded screws. However, with the short partially-threaded screw, it may be that not enough screw would pass the fracture site to resist shear and torsional forces, which were not analysed in this study. The implication is that fully-threaded screws, despite achieving similar compression of the fracture may offer an additional advantage over the 30 mm partially threaded screw in maximising the thread count beyond the fracture site. We noted that the beneficial effect on compression at the fracture site was removed when fully threaded screws were inserted with a near fragment glide hole (group C ankles), suggesting that the removal of bony material is detrimental to achieving secure fixation in osteoporotic bone. It has previously been shown that tapping removes material in a similar way, enlarging the hole so that the screw fails by increasing the shear area for the threads.7 It remains to be seen exactly which malleolar fractures are suitable for treatment with fully threaded screws. This study was undertaken on a small sample size in osteoporotic human cadaver bone, and it is possible that in younger
1666
L. PARKER, N. GARLICK, I. MCCARTHY, S. GRECHENIG, W. GRECHENIG, P. SMITHAM
patients without osteoporosis a fully threaded screw would fail to achieve a lag effect and may instead act as a positional screw, preventing compression at the fracture site. In this situation a shorter, partially threaded screw may be more suitable. From this study, it is not possible to determine whether purchase within the physeal scar is crucial in all medial malleolar fractures, although in osteoporotic bone it seems to improve the purchase of the screw threads and compression at the fracture site. In conclusion, the significant improvement in stability of fixation of a medial malleolar fracture conferred by fully threaded screws in this study challenges the AO doctrine advocating the use of partially threaded screws. The more secure fixation is likely to be as a result of purchase of the screw threads within the radiodense physeal scar, which can be achieved with either fully threaded screws or shorter, partially threaded screws. This purchase is desirable, particularly in osteoporotic bone. The fully threaded screw may offer additional advantages over the short 30 mm partially threaded screw, by maximising the thread count beyond the fracture, even though they both purchase the physeal scar and produce similar compression at the fracture site. We therefore recommend that fully threaded screws be used instead of long partially threaded screws when treating medial malleolar fractures in osteoporotic bone. Supplementary material Two tables detailing the results of the studies of the compression at the fracture site for all four groups are available alongside the electronic version of this article on our website www.bjj.boneandjoint.org.uk
The authors would like to thank Professor A. Weinberg, Department of Paediatric and Adolescent Surgery, Medical University of Graz, Austria, P. Ferlic and H. Dimai from Klinische Abteilung für Endokrinologie und Stoffwechsel, Graz, Austria for DEXA analysis, and S. Cro, Clinical Trials Unit, Medical Research Council, UK for statistical analysis. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. This article was primary edited by J. Scott and first-proof edited by G. Scott.
References 1. Solomon L, Warwick DJ, Nayagam S. Apley’s system of orthopaedics and fractures. Eighth ed. London: Hodder Arnold, 2001:553–554. 2. Wood GW. General principles of fracture treatment. In: Campbell’s operative orthopaedics. Vol. 3. Eleventh ed. Philadelphia: Mosby-Elsevier, 2008:3053–3054. 3. Krettek C, Gosling T. Principles of internal fixation. In: Rockwood and Green’s fractures in adults. Vol. 1. Sixth ed. Philadelphia Lippincott Williams and Wilkins, 2006:223–226. 4. Hak DJ, Lee MA. Ankle fractures: open reduction internal fixation. In Master techniques in orthopaedic surgery-fractures. Second ed. Philadelphia: Lippincott Williams and Wilkins, 2006:556–557. 5. Barbosa P, Bonnaire F, Kojima K. AO Foundation. Malleoli 44-C2 ORIF for medial malleolus; reduction and fixation; medial lag screws (transverse fracture). https:// www2.aofoundation.org/wps/portal/surgery/?showPage=redfix&bone=Tibia&segment=Malleoli&classification=44-C2&treatment=&method=ORIF+for+medial+malleolus&implantstype=Lag+screw&redfix_url= (date last accessed 29 May 2013). 6. DeCoster TA, Heetderks DB, Downey DJ, Ferries JS, Jones W. Optimizing bone screw pullout force. J Orthop Trauma 1990;4:169–174. 7. Chapman JR, Harrington RM, Lee KM, et al. Factors affecting the pullout strength of cancellous bone screws. J Biomech Eng 1996;118:391–399. 8. Brown GA, McCarthy T, Bourgeault CA, Callahan DJ. Mechanical performance of standard and cannulated 4.0-mm cancellous bone screws. J Orthop Res 2000;18:307–312. 9. Faran KJ, Ichioka N, Trzeciak MA, et al. Effect of bone quality of the forces generated by compression screws. J Biomech 1999;32:861–864. 10. Alonso-Vázquez A, Lauge-Pedersen H, Lidgren L, Taylor M. The effect of bone quality on the stability of ankle arthrodesis: a finite element study. Foot Ankle Int 2004;25:840–850. 11. Thiel W. The preservation of the whole corpse with natural color. Ann Anat 1992;174:189–195 (in German).
THE BONE & JOINT JOURNAL
