Fracture healing after reamed and unreamed intramedullary nailing in sheep tibia

Injury, Int. J. Care Injured 42 (2011) 667–674

Contents lists available at ScienceDirect

Injury journal homepage: www.elsevier.com/locate/injury

Fracture healing after reamed and unreamed intramedullary nailing in sheep tibia F. Ho¨gel a,*, U. Schlegel a, N. Su¨dkamp b, C. Mu¨ller c a

AO-Research Institute Davos, Clavadelerstrasse 8, Davos, CH-7270, Switzerland Department Orthopa¨die und Traumatologie, Universita¨tsklinik Freiburg i. Brsg. Hugstetter Str. 55, 79106 Freiburg, Germany c Klinik fu¨r Unfall-, Hand- und orthopa¨dische Chirurgie, Sta¨dtisches Klinikum Karlsruhe, Moltke Str. 90, 76133 Karlsruhe, Germany b

A R T I C L E I N F O

A B S T R A C T

Article history: Accepted 13 October 2010

Intramedullary nailing is a well-established method for stabilisation of long-bone shaft fractures. It is still a controversy as to whether the procedure should be done by an unreamed or reamed technique. In the present animal study, 24 sheep were treated with intramedullary nailing. Midshaft fractures (Arbeitsgemeinschaft fu¨r Osteosynthese (AO) type 42-A2/3) were created. Eight sheep were treated with an unreamed nailing technique (UN), a further eight sheep underwent tibia nailing by the reamed technique using the conventional AO reaming system (RC) and in a further eight sheep, reamed nailing was performed using an experimental reaming system (RE). Intra-operatively, the intramedullary pressure was measured and, during a healing time of 10 weeks, the growth of callus formation was labelled with fluorescence markers after 4 and 6 weeks. After 10 weeks, the animals were euthanised and the quality of fracture healing was determined by recording stiffness in torsion, antero-posterior and mediolateral bending and the load at yield. In addition, the callus formation at the fracture zone was evaluated by fluorescence microscopy and macroradiographs. The results showed a decrease of intramedullary pressure when reamed nailing was performed with the RE (72.5 mmHg) system compared with the conventional AO reaming system (227 mmHg). Mechanical testing did not reveal any significant differences either for torsional or bending stiffness or for load at yield for any of the three procedures. Histological evaluation showed a similar callus formation for the UN group and the RE group. Callus formation in the UN (65 mm2) and RE (63 mm2) groups showed a higher increase during the first 6 weeks than those treated with the conventional AO reaming system (27 mm2). This means that, especially during the first weeks of fracture healing, damage to the bone by the reaming process can be reduced by reaming with a reaming device with lowered cutting flutes and smaller drive-shaft diameter. Intramedullary pressure can be significantly reduced by using reaming systems with reduced driveshaft diameters and deepened cutting flutes. In the early phase of fracture healing, callus formation can be influenced positively when using the RE system. ß 2010 Elsevier Ltd. All rights reserved.

Keywords: Fracture healing Intramedullary nailing Tibia fracture Callus formation

Introduction Intramedullary nailing has been well established as a standard treatment procedure for diaphyseal fractures of the long bones for many years even though the negative effects of intramedullary nailing, such as aseptic endosteal necrosis and systemic fat embolism have been known for years and have been the subject of much controversy.14,34,35

* Corresponding author at: Berufsgenossenschaftliche Unfallklinik Murnau, Prof. Ku¨ntscher Str. 8, 82418 Murnau, Germany. Tel.: +49 08841 48 43 88; fax: +49 08841 48 46 17. E-mail addresses: fl[email protected] (F. Ho¨gel), [email protected] (N. Su¨dkamp), [email protected] (C. Mu¨ller). 0020–1383/$ – see front matter ß 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.injury.2010.10.004

In the tibia, damage to the blood supply is the greatest problem, whereas systemic embolism is more relevant to the treatment of femoral fractures. Trauma caused by the reaming procedure has been made responsible for the negative effects of intramedullary nailing. For this reason, unreamed nailing, a method used even by Kuentscher, was further developed in the 1980s and enjoyed widespread clinical application not only in the treatment of closed, but also in the treatment of open fractures.11,19,20,15 Depending on the surgical school of thought and the geographical location, the two procedures, that is, unreamed versus reamed nailing, were the subject of lively debate. It has been consistently demonstrated that the reaming procedure damages the endosteal blood supply, which consequently leads to avascular endosteal necrosis. If healing is uneventful, these necrotic zones will be remodelled in the normal course of healing.3–5,13,33,24,28,29 If reaming is not

F. Ho¨gel et al. / Injury, Int. J. Care Injured 42 (2011) 667–674

668

performed, primary endosteal necrosis can be reduced from 70% to 31%.12 This accelerates healing by approximately 2 weeks for simple fractures in sheep.28,29 Reichert was able to demonstrate a positive effect of reaming, namely, that the blood supply to the periosteum was increased sixfold after reaming.26 What consequences the increased blood supply has for fracture healing and how long the effect lasts is unknown because the measurement is only momentary. A clinical study of the Canadian Orthopaedic Trauma Society, published in 2003, stated the other side of the coin. This multicentre study showed a significant higher rate of nonunions in femoral fractures.2 Another positive effect on reaming the medullary canal is the internal bone grafting. Ho¨gel et al.10 were able to show that osteoblasts are still vital in the bone debris after correctly performed reaming. A study by Froelke et al.7 showed the positive effect of reaming debris on the healing of bone after being placed in a defect in the sheep tibia. They also showed that there is no significant difference between reaming debris or iliac crest. This study should clarify the influence of intramedullary pressure on callus development and quality of callus formation when intramedullary nailing is performed by an unreamed or reamed nailing technique. We hypothesised that lower intramedullary pressure will increase the speed of callus formation and the quality of the fracture callus during the healing process of tibia-shaft fractures. Materials and methods For this study, 27 adult Swiss mountain sheep were used (Animal trial adoption: GR 2/1997, 7/12/13/1998). They were all female with an average age of 5 years, and a weight ranging from 65 to 70 kg. First, the diameter of the medullary cavity of the left tibia was measured by X-rays in the lateral view. Whilst reaming was performed up to 9.0 mm, sheep with a cavity diameter ranging from 8.0 to 9.0 mm were selected to be treated with the unreamed technique and sheep with a diameter of 7.0–8.0 mm were chosen to be treated with reamed intramedullary nailing. The magnification was 1.15. For this experiment, three groups were made. In the first group, intramedullary reaming was performed with an experimental reaming system (RE) and in the second group, the conventional Arbeitsgemeinschaft fu¨r Osteosynthese (AO) reaming system (RC) was used. Sheep with a cavity diameter of 7.0–8.0 mm were randomised. Eight animals were treated with the unreamed nailing technique (UN) (Table 1). The operative procedure was performed under anaesthesia with isoflurane. After correct positioning of the sheep, the left leg was shaved and disinfected. To produce standardised fractures, a custom-made fracture device was used (ADI, Davos, Switzerland) as described in the venia legendi of Prof. Christof Mu¨ller. This fracture device worked on the principles of four-point bending. After incision of the skin, an osteotomy of the cortex of one third was performed and a bending plate was positioned, and by Schanz screws, which were placed bi-cortically, guided through the bending plate, prestress was applied. A2/A3 fractures were

Table 1 Groups due to the diameter of the medullary cavity. Method

n

Medullary cavity diameter

Experimental reaming system (RE) Conventional AO reaming system (RC) Unreamed nailing (UN)

8 8 8

7–8 mm 7–8 mm 8–9 mm

Fig. 1. Reaming systems with different reamer heads. The experimental reaming system contains a reamer head with deepened cutting flutes and a reduced drive shaft.

produced by the impulse of a guided weight in the midshaft of the tibia, which was constantly 11 cm proximal from the medial malleolus. After performing the fracture, the soft tissues were closed and the fracture was reduced. For fracture stabilisation, a solid nail (UHN, Synthes, Paoli, USA) with a length of 190 mm and a diameter of 7.5 mm was used. Interlocking was performed with two proximal and two distal 3.9-mm screws. The entry point for intramedullary nailing of the sheep tibia was placed anterior to the anterior cruciate ligament, which is different from the entry point in humans. As reaming systems, we used the RC system (Synthes1) containing four front-cutting reaming edges with relatively small chip flutes, which were adapted on flexible two- to three-layered spring band steel drive shafts. The experimental reaming system was based on the geometry of the AO reaming system, but the chip flutes were deepened. All reamer heads were fixed on 3.3-mm steel shafts (Custom-made product, Mathys1, Medical AG, Bettlach, Switzerland). All reamer heads were brand-new and of 7.0-, 8.0-, 8.5- and 9.0-mm diameter (Fig. 1). For measuring pressure 3 cm from the distal epiphysis of the tibia, the pressure tube, filled with sterile Ringer solution, was fixed in the anterior cortex. The piezoresistant pressure sensor (Type: PR-11-2; Keller Druckmesstechnik AG, Winterthur, FS: 0.5%) was connected with the computer system. After opening of the medullary canal, the reaming procedure was done, starting with the 7.0-mm reamer heads, followed by the 8.0-mm, 8.5-mm up to 9.0mm. In the unreamed group, the nail was inserted directly after opening the medullary canal. Intra-operatively, the intramedullary pressure was measured during the reaming process and the nail insertion (Fig. 2). For the examination, the highest pressure values during the reaming and nailing procedures were used. The animals were euthanised after 10 weeks and the tibiae explanted. The quality of fracture healing was determined after implant removal using a uniaxial testing device (MTS, Bionix; resolution of 1%, accuracy 0.1 mm). After embedding the proximal and distal ends of the bone in acrylate (BeracrylTM), torsional testing was performed via a universal joint set for inner rotation at a speed of 18 min 1. Loading was carried out within the elastic range, whereby measuring was discontinued at an applied force of approximately 1000 Nm. Next, bending stiffness in the antero-posterior plane within the range of elasticity was determined in four-point bending, whereby the anterior cortex was placed under tension.

F. Ho¨gel et al. / Injury, Int. J. Care Injured 42 (2011) 667–674

669

Pressure values of the reamer heads pressure [mmHg]

1200 900 600 300 0 7,0 mm 7,0 mm 8,0 mm 8,0 mm 8,5 mm 8,5 mm 9,0 mm 9,0 mm (AO) (Ex) (AO) (Ex) (AO) (Ex) (AO) (Ex)

Fig. 3. Intramedullary pressure of reaming the fractured tibia shows a significant reduction of pressure values using the experimental reaming system.

Statistical analysis was performed using the Kruskal–Wallis test for more than two uncombined groups by the Institute for Biomathematics of the University of Freiburg, Germany. The Kruskal–Wallis test was chosen because no normal distribution was given. As statistics program, the SAS 6.12 (Statistical Analysis System) package was used. Results Intramedullary pressure increase

Finally, bending stiffness in the mediolateral plane was investigated, whereby the medial cortex was placed under tension. In this plane, testing continued until refracture. Polychrome labelling was performed subcutaneously with Calcein green (10 mg kg 1 body weight) after 4 weeks and Xylenol orange (80 mg kg 1 body weight) after 6 weeks.22,27 The fluorochrome label showed an enhancement during new bone formation for 2 weeks; hence, the time for measurement of the specific labels are at 6 weeks for Calcein green, at 8 weeks for Xylenol orange and for the macroradiographies at 10 weeks. After 10 weeks, the sheep were sacrificed and the tibiae were explanted. For histological reprocessing, the specimens were fixed in 4% buffered formalin. Afterwards, they were dehydrated with an ascending alcohol series and embedded. The hardened cylinders were cut in the sagittal plane with slices of 200 mm. For further examination, one slice from the lateral quarter of the tibia was chosen and ground to 100 mm. Before fixation of the chosen slices, macroradiography was performed. The ground slice was fixed on a microscope slide. The embedded slices were examined with the fluorescence microscope (Zeiss ICM 405, Jena, Germany) and innervated with the respective fluorescence filter. For Calcein green labelling, a filter combination of an exciter filter BP 485, a colour diviser FT510, a blocking filter P 520 and KP 540 were used. For Xylenol orange, an exciter filter BP 546, a colour diviser FT 580 and a blocking filter 590 were used. Macroradiographies were chosen for the sizing of the callus formation after 10 weeks. Photographs were taken with a Canon F1 and a Zeiss zoom lens (Luminar 40/1:4.5, Fuji Film 400 ASA). On each photofilm, a 1-cm2 area was copied for the calibration of the analysis program. The slides were digitalised with a diascanner (RFS-3570 Film Scanner, Kodak Professional, Stuttgart, Germany). The examination of increase of the callus was performed with the Scion-Image1 image processing program. First, the program was calibrated. Then, the size of callus formation was measured by marking the illuminated areas of the slices in square millimetres for each labelling and the data were added to Excel1 files.

A highly significant (p < 0.002) reduction in pressure values was achieved with the RE system for all sizes of reamer head compared with the RC system. The differences in pressure were as follows and are given as median values (Minima–Maxima) (Fig. 3): 7.0-mm reamer 184) mmHg; 8.0-mm reamer 112) mmHg; 8.5-mm reamer 236) mmHg; 9.0-mm reamer 143) mmHg.

head: RC 348 (227–476) versus RE 75 (4– head: RC 209 (172–293) versus RE 64 (0– head: RC 123 (51–205) versus RE 25 (0– head: RC 348 (85–553) versus RE 56 (9–

During UN implantation, the highest median pressure value recorded was 576 (143–1065) mmHg (p = 0.0008). During insertion of the nail after prior reaming, clearly lower pressure values were measured compared with unreamed nailing: RC 106 (52– 161), RE 200 (0–377) mmHg. The pressure values for nail insertion after reaming with either of the two reaming systems (RC/RE) did not differ significantly from each other (Fig. 4). Pressure development over time To obtain a measure for pressure values in relation to time of application, the integral of the intramedullary pressure value was determined dependent on the time of application. It was found that

Intramedullary Pressure of the Nail Pressure [mmHg]

Fig. 2. A pressure tube is fixed in press fit technique to the anterior cortex 3 cm proximally of the medial malleolus.

1200 800 400 0 UN

Rc

Re

Fig. 4. There are no significant values of pressure increase during the insertion of 7.5 mm nails in the reamed groups (RE, RC), whilst the intramedullary pressure increases significantly in the unreamed group (UN) in comparison to the RE and RC groups.

F. Ho¨gel et al. / Injury, Int. J. Care Injured 42 (2011) 667–674

Pressure/ time unit [mmHg/s]

670

Addition of Pressure/ time unit values 12000 9000 6000 3000 0 UN

Rc

Re

Fig. 5. Intramedullary pressure over time of intramedullary instrumentation shows no significant differences for the UN and RE groups. The intramedullary pressure over time of instrumentation of the RC group is significantly higher than in the UN and RE groups.

Reaming times Analysis of the reaming times showed a significantly different value only for the 8.5-mm reaming head (p < 0.05). The median (Minima–Maxima) value for reaming with the RE system was 15 (13–37) s compared with 26 (16–45) s for the RC system. None of the other sizes of reamer head produced significant differences. Insertion force Although the reaming times hardly differed, the RE system with the 7.0- and 9.0-mm reaming heads required significantly (p < 0.05) less axial force to complete the reaming procedure in the same time as compared with the RC system. Median (Minima–Maxima): 7.0-mm reamer head: RC 20 (12–81) versus RE 2.5 (1–44) N; 9.0-mm reamer head: RC 36 (25–44) versus RE 17 (1–46) N.

Fig. 6. Callus formation between 4th und 6th week. Considerable periosteal callus formation. The endosteal callus formation is low.

histological slides, resulted from the biomechanical testing and was not the original fracture gap. Histological examination of the period between 4 and 6 weeks The examination of the callus formation (mm2) shows a significantly lower callus size in the RC group. With respect to median, it was 27 mm2 in comparison to the UN group (median: 65 mm2 (p < 0.004)) and the RE group (63 mm2 (p < 0.003)). The callus areas of the UN group and the RE group are much greater even though there are no significant differences between these two groups (Figs. 6 and 7).

Calcein-Grün (4. - 6. week) 150 callus area [mm2]

the unit for pressure/time for all reamer sizes was significantly lower for the RE system compared with the AO reaming system. When the data for all the different steps in the procedure, that is, all reaming procedures and nail insertion, were added together, it became apparent that the average pressure/time unit for the UN group did not differ significantly from the RE group. However, the pressure/time unit for the group with the RC system was significantly higher (p = 0.001) (Fig. 5).

100

*

50 0 RC

RE

UN

Fig. 7. Callus formation between the 4th and 6th week. Significantly lower callus formation in the RC group and no significant differences between the UN and RE group.

Mechanical testing Remarkably, neither the data for torsional and bending stiffness nor the data for load at yield differed significantly from each other across the three groups. It can only be stated that there was wider scatter for the values for the unreamed group compared with the reamed groups. Torsion: UN 35 (23–45); RC 35 (22–43); RE 34 (21–46) Nm/8; Antero-posterior bending: UN 47 (37–57); RC 45 (40–52); RE 46 (30–53) Nm/8; Mediolateral bending: UN 38 (20–42); RC 39 (31–43); RE 37 (33–42) Nm/8; Load at yield: UN 65 (36–99); RC 69 (49–79); RE 69 (48– 83) Nm/8.

Histological investigation Before the histological examination, biomechanical testing was performed until refracture. The gap, which is visible in the

Fig. 8. Callus formation between the 6th and 8th week. The gap relates to biomechanical testing and is not related to the original fracture gap. The periosteal callus shows a trabecular length alignment and shows a higher bone mineral density than the callus formation between the 4th and 6th week.

F. Ho¨gel et al. / Injury, Int. J. Care Injured 42 (2011) 667–674 Table 2 Median values of callus area.

callus area [mm2]

Xylenol-Orange (6. - 8. week) 300 200 100 0 RC

RE

671

UN

Fig. 9. Callus area between the 6th and 8th week. No significant differences between the three groups could be detected.

Histological examination of the period between 6 and 8 weeks By the excitation for Xylenol orange, a strong callus formation with a trabecular alignment of the periosteal callus was detected. The callus area of the UN group was 130 mm2 in median and 137 mm2 for the RE group. The callus area of the RC group was 110 mm2 in median. The analysis of the callus area showed no significant differences between the three testing groups. This means that between the 6th and the 8th week, the speed of fracture healing is of no difference between the different methods of intramedullary nailing (Figs. 8 and 9). Histological examination of the period between 8 and 10 weeks On the sagittal slices of the contact radiographies, the balance of the callus tissue which formed until the 10th week appears. The mature callus with the strong trabecular and linear alignment is impressive. The callus area of the UN group was 271 mm2 in median and 314 mm2 for the RE group. The callus area of the RC group was 279 mm2 in median. There is no significant difference

Area [mm2]

UN

RE

RC

4–6 weeks 6–8 weeks 8–10 weeks

65 130 271

63 137 314

27 110 279

Table 3 Level of significance p of the callus area between the 4th and 6th week. Sig. p<

UN

RE

RC

UN RE RC

X n.s. 0.0032

n.s. X 0.0029

0.0032 0.0029 X

between the three groups concerning the callus area after 10 weeks (Figs. 10 and 11). Chronological sequence of fracture healing In summary, there is a nearly linear callus increase until the 8th postoperative week concerning the increase of callus formation. Between the 8th and 10th week postoperatively, there is a disproportionate increase in callus formation. During the first 8 weeks, there is a parallel curve progression of the UN and RE groups, whilst, in the beginning, there is a significantly lower increase of callus formation in the RC group. Regarding the chronological sequence of callus development given in Table 2, the RE group shows a tendency towards more callus formation than the others (Fig. 12) (Tables 2 and 3). Discussion

Fig. 10. Callus formation after 10 weeks. The gap relates to former biomechanical testing and is not of the original fracture gap. A strong trabecular callus is visible. The original fracture gap is consolidated and there is a sharp boundary on the endosteal callus where the intramedullary nail had taken place.

The intramedullary pressure values showed a significantly lower pressure increase during the reaming process with the RE system. The pressure over time showed no significant difference whether nailing was performed with the unreamed technique or the reamed technique, using the RE system. Biomechanical investigation showed that no significant differences were recorded for any of the three groups with regard to torsion or four-point bending in the antero-posterior and mediolateral planes. Nor were any significant differences found for load at yield. This means that after 10 weeks, all simple transverse fractures of the sheep tibia had consolidated. The evaluation of fluorochrome labelling in 24 Swiss alpine sheep after reamed and unreamed intramedullary nailing showed a significant increase in callus formation during the first weeks when unreamed nailing and reamed nailing with the experimental reamer were performed. After 8 weeks, the fracture was healed with a strong trabecular callus formation and, until the 10th week, callus formation increased clearly. progression of callus increase

400

callus area [mm2]

callus area [mm2]

Contact radiography (8. - 10. Woche) 600 400 200

300

UN RE RC

200 100 0

0

OP

RC

RE

UN

Fig. 11. Callus area after 10 weeks. No significant differences in the three different testing groups.

4.-6. Woche

6. -8. Woche

8. -10. Woche

Fig. 12. Progression of callus increase. Low callus increase of RC until 8th week. No significant differences of callus increase from 8th to 10th week. More callus formation in RE group by trend.

672

F. Ho¨gel et al. / Injury, Int. J. Care Injured 42 (2011) 667–674

For this experiment, Swiss alpine sheep were chosen because this model has earlier been proven in questions of osteosynthesis and fracture healing.7,23,22 There are some reasons for this: first, the sheep tibia fits to the load axis like in human beings. Further, its structure and shape are similar to the human tibia. Concerning Wolff’s law, it is assumed that if two bones consist of the same shape a similar load initiation takes place. This was verified by Lanyon et al.18 in 1975. He performed in vivo investigations for the comparability of sheep and human tibiae. Furthermore, it is possible to use implants and intramedullary reamers like in human beings31; hence, the biological effects of conventional implants can be examined. A positive effect of standardised fractures of the AO type A2/A3 was that influences of intramedullary pressure increase, vascularity and fat intravasation during the reaming and nailing process regarding different fracture morphologies could be neglected. By preoperatively performed radiographs, the dimensions of the sheep tibiae were detected; hence, similar test conditions for all groups were created concerning the reaming rate and implant size. Since then, we are not aware of former experimental investigations regarding the reaming rate, which is an important factor for vascular damage of the bone cortex. Reaming and nailing were always performed by the same surgeon, because every surgeon has his own reaming and nailing technique, which influences the comparability of the results. The intramedullary pressure could be significantly reduced with the RE system in comparison to the RC system. In this study, the highest pressures in reamed intramedullary nailing were also recorded for the first stages of reaming. Intramedullary pressure fell continuously as the size of the reamer increased. This is attributed to the increased throughput due to the enlarged flutes and the reduction of the diameter of the flexible shaft. The importance of deepened cutting flutes for pressure increase was described by Stu¨rmer in 1980 in the form of the so-called gap formula.32 The RE system with its deep flutes and thin reamer shaft provided the conditions for a high throughput of medullary content, and thus led to lower pressure values. Based on these findings, the Synream (Synthes, USA) was developed. The RC system was characterised by plain cutting flutes and a large drive-shaft diameter. The high pressure values in the unreamed group can be attributed to the resistance of the medullary fat because an implant with a diameter of 7.5 mm was inserted into a medullary cavity completely filled with fat, which is then displaced in one go. Calculation of the pressure/time surface area integral meant that statements could be made regarding how long and at what level the intramedullary pressure was in effect. It became apparent that after adding the data for all steps in the reaming procedure, including nail insertion, no difference was found between the RE group and the unreamed nailing group. However, the pressure/time unit for the RC system was significantly higher compared with the other two groups. It is remarkable that the highest pressure values were obtained during insertion of the unreamed nail. These findings can be explained by the continuous removal of intramedullary content during the reaming process using the RE system. In this group, each reamer size showed significantly lower pressure values than the RC group. The high pressure/time unit of the RC group results from a constantly high pressure during reaming, whilst these intramedullary pressure values consist of the reamer and driveshaft geometry. Biomechanical investigation showed that no significant differences were recorded for any of the three groups with regard to torsion or four-point bending in the antero-posterior and mediolateral planes. Nor were any significant differences found for load at yield. This means that, after 10 weeks, all simple transverse fractures of the sheep tibia had consolidated.

Thus, it can be assumed that the histological differences, which were described in the literature, are irrelevant for the quality of fracture healing after 10 weeks. At this time, no differences of any kind were found in the quality of fracture healing, whether a reamed or unreamed technique of intramedullary nailing had been applied, nor as a result of the significantly lower pressure values found for the RE system or the significantly higher pressure levels recorded for the RC system. Fluorochrome labelling was performed with the proven stains Calcein green and Xylenol orange.25 To gain an objective assessment method for the examination, we decided to choose a computer-assisted picture analysis system. As a representative dimension unit of the callus size, the area (mm2) of the periosteal and interfragmentary callus was chosen. Kro¨pfl et al.16 also examined the endosteal callus formation, which was not sensible to us, because the endosteal callus formation depends on the space between the cortex and the nail.6 Therefore, the endosteal callus only seems to be relevant when small diameter nails are used. The examination of sagittal planes was performed because the largest area of callus formation was seen on the dorsal cortex in the radiological follow-up. Slices of the lateral cortex were chosen because we wanted to examine the callus formation and the remodelling process of the fracture and not out of the osteotomy, which was performed on the medial cortex. Because it is not possible to gain a continuous serum speculum of the applied stains and the labelling contains a mineralisation period, an exact determination of the area of mineralisation at a certain point is not possible. Therefore, a description of a temporal basis is not possible, but, by using the same application scheme, it is possible to show differences in the mineralisation between the groups on a certain point. During fracture healing, callus formation starts from the periosteum with a distance of some millimetres proximally and distally from the fracture gap.12,36 This was shown in the sagittal slices of the early phase in excitation by Calcein green (Fig. 4). The callus area in the examined sagittal slides, between the 4th and 6th week, is relatively small with a median value of 27–65 mm2, but a bridging of the fracture gap can already be seen in all groups. Regarding callus formation during the early phase of fracture healing, significant differences were detected between the testing groups. The significantly lowest callus increase was seen in the RC group with 27 mm2 (p < 0.004). The UN group instead showed a 2.4-fold larger callus area (65 mm2) and the RE group a 2.3-fold callus development (63 mm2). The differences in further healing progress became similar, so that, between the 6th and 8th week, no significant differences were seen between the testing groups. The evaluation of the macroradiographies after 10 weeks also showed no significant differences between the testing groups. The UN and RE groups showed a nearly linear increase of fracture callus until the 8th week, but, after 8 weeks, the RE group tended to develop a larger callus area (Fig. 7). The curve of the RC group showed a significantly lower increase of callus formation until the 6th week compared with the other groups, but between the 6th and 8th week, a fourfold callus increase could be detected. Between the 8th and 10th week, a 2.5-fold callus increase could be shown in the RC group, so that no significant differences could be shown in the late phase of fracture healing between the three testing groups. Runkel et al.30,28 examined the callus increase in the sheep tibia after reamed and unreamed nailing. Whilst the study was not performed in a fracture model, but in an osteotomy model, they detected a 1.6-fold callus increase (155 mm2) of the sheep treated with an unreamed nail. The animals treated in a reamed nailing technique only developed a callus area of 95 mm2. In the unreamed nailing technique, the callus formation after 4 weeks was larger

F. Ho¨gel et al. / Injury, Int. J. Care Injured 42 (2011) 667–674

than in the reamed technique after 6 weeks. The higher amount of callus in the unreamed and RE group compared with the conventionally reamed group results from the relatively instability in the unreamed group and the lower fat occlusion of the transcortical vessels. The relative instability in the unreamed group results from the low intramedullary guidance of the nail, which was described by Augat et al.1 Another argument for reamed intramedullary nailing can be the increase of growth factors, such as transforming growth factor-beta (TGF-b) and vascular endothelial growth factor (VEGF), after reaming the medullary cavity.8 Basically, these results corresponding to those of the present study. After a period of 4 weeks, we even found a 2.4-fold callus area in the UN group than in the RC group. The absolute differences in the size of the fracture callus are not comparable to ours because different sheep races were used for testing. A further point against the direct comparison of the studies is that Runkel et al. did not determine the rate of reaming preoperatively. According to a study of Mueller et al.,21 the fat occlusion of transcortical vessels was significantly reduced by reamed nailing using a reaming device with lowered cutting flutes and reduced drive-shaft diameter to 1.6% compared with unreamed nailing with a vessel occlusion of 5.7%. Kro¨pfl et al.17 compared the callus increase after reamed and unreamed nailing in femur osteotomies of Papio ursinus. The endosteal callus formation in the unreamed group was 28 mm2 in mean and in the reamed group, 11.5 mm2 (p < 0.001). The periosteal callus formation in the unreamed group was 238 mm2 and 142 mm2 for the reamed group (p < 0.001). However, the comparability of the two groups is difficult because the reaming rate was not mentioned. Whilst a 9-mm implant was used in the reamed group and a 7mm nail was used for the unreamed group, a number of variables were implemented, which lead to the assumption that the groups are not comparable. An excessive reaming, as is visible in the published radiographs, leads to a massive damage to the vascularisation of the bone and, therefore, to a delayed callus increase. The higher rigidity of the 9-mm nail also leads to a high influence on the callus increase leading to the conclusion that the callus formations of the two groups are not comparable. In addition, the determination of the endosteal callus is not sensible because the different nail diameters lead to a different space between the nail and the bone and is therefore of disadvantage for the reamed nailing technique. For these various reasons, we did not differentiate between endosteal and periosteal callus formations, but calculated the whole callus formation. In addition, we founded groups of comparable diameters of the medullary cavity and used the same implant size to make a comparison of the groups possible. Therefore, we paid attention to the certainty that a higher stability of a bigger implant size leads to a decrease in callus formation.9,23 Conclusion From the findings of this study, it can be stated that intramedullary pressure can be significantly decreased during intramedullary reaming with a reaming system of deepened cutting flutes and a reduced drive-shaft diameter. During fracture healing, this leads to a faster increase in callus formation during the first weeks. Hence, we can recommend that the negative effects of intramedullary reaming can be reduced. Conflict of interest All authors listed consented to the submission and all data were acquired according to the Swiss legislation by the ethical review board (No.: GR2/97, 7/12/13/98). No conflict of interest was stated.

673

References 1. Augat P, Penzkofer R, Nolte A, et al. Interfragmentary movement in diaphyseal tibia fractures fixed with locked intramedullary nails. J Orthop Trauma 2008;22(1):30–6. 2. Canadian Orthopaedic Trauma Society. Nonunion following intramedullary nailing of the femur with and without reaming. J Bone Joint Surg 2003;85A:2093–6. 3. Danckwardt-Lilliestroem G. Reaming of the medullary cavity and its effect on diaphyseal bone. A fluorochromic, microangiographic and histologic study on the rabbit tibia and dog femur. Acta Orthop Scand 1969;128(Suppl.):1–153. 4. Danckwardt-Lilliestroem G, Lorenzi GL, Olerud S. Intramedullary nailing after reaming. An investigation on the healing process in osteotomized rabbit tibias. Acta Orthop Scand 1970;134(Suppl.):1–78. 5. Danckwardt-Lilliestroem G, Lorenzi L, Olerud S. Intracortical circulation after intramedullary reaming with reduction of pressure in the medullary cavity. J Bone Joint Surg Am 1970;52(7):1390–4. 6. Friedl W, Anmerkung zum Beitrag von, Kro¨pfl A, et al. Kallusbildung nach aufgebohrter und unaufgebohrter Femurverriegelungsnagleung. Unfallchirurg. 2001;104:675–76. 7. Froelke JP, Bakker FC, Patka P, Haarman HJ. Reaming debris in osteotomized sheep tibiae. Injury 2001;50:65–9. 8. Giannoudis PV, Pountos I, Morley J, et al. Growth factor release following femoral nailing. Bone 2008;42(April (4)):751–7. 9. Hente R, Cheal EJ, Perren SM. Die Dehnungstheorie als Erkla¨rungsgrundlage des Erfolges der biologischen Osteosynthese. Hefte zu ‘‘Der Unfallchirurg’’ 1993;232:445–7. 10. Ho¨gel F, Mu¨ller CA, Peter R, et al. Bone debris—dead matter or vital osteoblasts. J Trauma 2004;56(February (2)):363–7. 11. Karlstro¨m G, Olerud S. Fractures of the tibial shaft. A critical evaluation of treatment alternatives. Clin Orthop Rel Res 1974;82–115. 12. Kessler SB, Hallfeldt KKJ, Perren SM, Schweiberer L. The effect of reaming and intramedullary nailing on fracture healing. Clin Orthop 1986;212: 18–25. 13. Klein MPM, Rahn BA, Frigg R, et al. Reaming versus non-reaming in medullary nailing: interference with cortical circulation of the canine tibia. Arch Orthop Trauma Surg 1990;109(6):314–6. 14. Krettek C, Schandelmaier P, Rudolf J, Tscherne H. Aktueller Stand der operativen Technik fu¨r die unaufgebohrte Nagelung von Tibiaschaftfrakturen mit dem UTN. Unfallchirurg 1994;97:575–99. 15. Krettek C, Schandelmaier P, Tscherne H. Nonreamed interlocking nailing of closed tibial fractures with severe soft tissue injury. Clin Orthop Rel Res 1995;315:34–47. 16. Kro¨pfl A, Davies J, Berger U, et al. Intramedullary pressure and bone marrow fat extravasation in reamed and unreamed femoral nailing. J Orthop Res 1999;17:261–8. 17. Kro¨pfl A, Hertz H, Redl H, Schlag G. Kallusbildung nach aufgebohrter und unaufgebohrter Femurverriegelungsnagelung. Unfallchirurg 2001;104: 41–9. 18. Lanyon LE, Hampson WG, Goodship AE, Shah JS. Bone deformation recorded in vivo from strain gauges attached to the human tibial shaft. Acta Orthop Scand 1975;46:256. 19. Mu¨ller CA, Morakis Ph, Strohm P, Pfister U. Intramedullary nailing of the tibia: current status of primary unreamed intramedullary nailing. Part 2: results for open fractures. Injury 1999;30(Suppl. 3):44–54. 20. Mu¨ller CA, Strohm P, Morakis Ph, Pfister U. Intramedullary nailing of the tibia: current status of primary unreamed nailing. Part 1: results of closed fractures. Injury 1999;30(Suppl. 3):37–43. 21. Mueller CA, Schlegel V, Hoegel F, et al. Cortical perfusion and local fat occlusion after intramedullary nailing of the ovine tibia—comparison of different surgical procedures. Injury 2009;40(July (7)):760–6. 22. Neudeck F, Obertacke U, Wozasek G, et al. Experimentelle Untersuchung zur intramedulla¨ren Druckentwicklung und Fettembolisation bei der gebohrten und ungebohrten Marknagelung versus Plattenosteosynthese am Schafsfemur Hefte zu ‘‘Der Unfallchirurg’’ K E Rehm(Hrgb), vol. 249. Berlin, Heidelberg: Springer-Verlag; 1995. p. 243–4. 23. Perren SM, Ganz R, Cordey J. Wissenschaftliche Grundlage zur biologischen Osteosynthese: Dehnungstheorie, vaskula¨re Elemente. Hefte zu ‘‘Der Unfallchirurg’’ 1993;230:777–82. 24. Pfister U, Rahn BA, Perren SM, Weller S. Vaskularita¨t und Knochenumbau nach Marknagelung langer Ro¨hrenknochen. Akt Traumatol 1979;9:191–5. 25. Rahn BA. Die polychrome Sequenzmarkierung. Habilitationsschrift. AlbertLudwigs-Universita¨t Freiburg i Br; 1976. 26. Reichert IL, Mc Carthy ID, Hughes SPF. The acute vascular response to intramedullary reaming. J Bone Joint Surg Br 1995;77-B(3):490–3. 27. Rhinelander FW. The normal microcirculation of diaphyseal cortex and its response to fracture. J Bone Joint Surg Am 1968;50-A(4):784–800. 28. Runkel M, Wenda K, Ritter G, et al. Knochenheilung nach unaufgebohrter Marknagelung. Unfallchirurg 1994;97:1–7. 29. Runkel M, Wenda K, Stelzig A, et al. Knochenumbau nach aufgebohrter und ungebohrter Marknagelung. Eine histomorphometrische Studie. Unfallchirurg 1994;97:385–90. 30. Runkel M, Wenda K, Degreif J, Blum J. Ergebnisse nach prima¨rer ungebohrter Tibiamarknagelung von Unterschenkelfrakturen mit schwerem offenem oder geschlossenem Weichteilschaden. Unfallchirurg 1996;771–7.

674

F. Ho¨gel et al. / Injury, Int. J. Care Injured 42 (2011) 667–674

31. Stu¨rmer KM, Schuchardt W. Neue Aspekte der gedeckten Marknagelung und des Aufbohrens der Markho¨hle im Tierexperiment. I. Die Schafstibia als Tiermodell fu¨r die Marknagelung. Unfallheilkunde 1980;83:346–52. 32. Stu¨rmer KM, Schuchardt W. New aspects of closed intramedullary nailing and marrow cavity reaming in animal experiments. II: intramedullary pressure in marrow cavity reaming. Unfallchirurg 1980;83:346. 33. Trueta J, Cavadias AX. A study of the blood supply of the long bones. Surg Gynecol Obstet 1964;118(3/3):485–98.

34. Weller S. Internal fixation of fractures by intramedullary nailing: introduction, historical review and present status. Injury 1993;24(Suppl. 3):1–6. 35. Wenda K, Henrichs KJ, Biegler M, Erbel R. Nachweis von Markembolien wa¨hrend Oberschenkelmarknagelungen mittels transo¨sophagealer Echokardiographie. Unfallchirurgie 1989;15(2):73–6. 36. Wilke HJ, Claes L, Fredrich H, Gall H. Mechanische und histomorphologische Untersuchungen zur Kallusbildung. Hefte zu der Unfallchirurg 1993; 230:1127–30.