Gait & Posture 39 (2014) 750–755
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Loads on a vertebral body replacement during locomotion measured in vivo A. Rohlmann *, M. Dreischarf, T. Zander, F. Graichen, G. Bergmann Julius Wolff Institute, Charite´ – Universita¨tsmedizin Berlin, Augustenburger Platz 1, 13353 Berlin, Germany
A R T I C L E I N F O
A B S T R A C T
Article history: Received 27 May 2013 Received in revised form 10 October 2013 Accepted 12 October 2013
Walking is one of the most important activities in daily life, and walking exposes the spine to a high number of loading cycles. Little is known about the spinal loads during walking. Telemeterized spinal implants can provide data about their loading during different activities. The aim of this study was to measure the loads on a vertebral body replacement (VBR) during level and staircase walking and to determine the effects of walking speed and using walking aids. Telemeterized VBRs were implanted in five patients suffering from compression fractures of the L1 or L3 lumbar vertebral body. The implant allows measurements of three force and three moment components. The resultant force on the VBR was measured during level and staircase walking, when walking on a treadmill at different speeds, and when using a wheeled invalid walker or crutches. On average, the resultant force on the VBR for level walking was 171% of the value for standing. This force value increased to 265% of the standing force when ascending stairs and to 225% when descending stairs. Walking speed had a strong effect on the implant force. Using a walker during ambulation on level ground reduced the force on the implant to 62% of standing forces, whereas using two crutches had only a minor effect. Walking causes much higher forces on the VBR than standing. A strong force reduction can be achieved by using a walker. ß 2013 Elsevier B.V. All rights reserved.
Keywords: Walking Load measurement Vertebral body replacement Telemetry Spinal loads
1. Introduction Walking is one of the most important activities of daily living. During each step, the discs and vertebrae of the lumbar spine are loaded with a large force [1,2]. However, few in vivo data exist on the spinal loads experienced during level and staircase walking. Furthermore, almost no in vivo measured data exist on the impact of walking speed and walking aids on the loads. For preclinical implant tests, however, and for finite element studies, realistic loads are mandatory to obtain relevant results. In addition, knowledge about spinal loads may help to advice patients regarding using of walking aids. Several groups used electromyography, force plate and body kinematic data to calculate spine loading during walking [3–6]. For the validation of their models in vivo measured data are required. The intradiscal pressure during level walking has been measured in a few subjects [2,7]. Wilke et al. measured a value in one subject that was 30% higher than the value measured for standing [2,8]. For slow walking, Nachemson and Elfstro¨m [7]
* Corresponding author. Tel.: +49 30 2093 46128; fax: +49 30 2093 46001. E-mail addresses: [email protected], [email protected] (A. Rohlmann). 0966-6362/$ – see front matter ß 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gaitpost.2013.10.010
measured an increase of intradiscal pressure of 15% in four patients compared with standing. These measurements were performed only during one measuring session per subject. The loads in internal spinal fixation devices during level walking were measured in 10 patients by Rohlmann et al. [1,8]. On average, they determined a maximum implant load during walking that was 28% higher than that for standing. For two patients [1], they reported only a small influence of walking speed on implant load. The normal use of crutches led to a small reduction of the implant loads, while a wheeled invalid walker reduced the loads on the internal spinal fixators by approximately 25%. Staircase walking caused higher intradiscal pressure values than level walking did [2,8]. Ascending stairs led to a maximum pressure in the disc that was 40% higher than for walking, while the value for descending stairs was only 20% higher [2]. Similar load values were measured in spinal fixators [1], with the loads usually being higher for ascending than descending stairs. However, no in vivo measured data exists about the loads on an anterior spinal implant during level and staircase walking. The telemeterized vertebral body replacement (VBR) allows the in vivo measurement of the loads acting in the anterior column of the lumbar spine. Such an implant has already been used to quantify the loads for several activities, including sitting, standing, lifting up and laying down a weight, and whole body vibration
A. Rohlmann et al. / Gait & Posture 39 (2014) 750–755
[9–12]. The force values varied strongly from patient to patient, while the measured moments were usually small. The aims of the present study were (1) to measure in vivo the resultant force on a vertebral body replacement during level walking; (2) to determine the influence of walking speed on implant loads; (3) to measure the resultant implant force during ascending and descending stairs; and (4) to determine the effects of walking aids such as a wheeled invalid walker and crutches on implant loads.
2. Materials and methods 2.1. Telemeterized vertebral body replacement A modified VBR (Synex, Synthes Inc., Bettlach, Switzerland) was used to measure the loading in vivo. Six semiconductor strain gauges (Type KSP 1-350-E4, Kyowa, Tokyo, Japan) were glued to the inner wall of the hollow implant, and they served as load sensors. A nine-channel telemetry unit and a coil for the inductive power supply of the electronic circuits were also integrated in the hermetically sealed implant. The instrumented implant allows the in vivo measurement of the three force and three moment components at a frequency of approximately 125 Hz. The resultant force is the geometric sum of the three force components. Prior to implantation, the VBRs were calibrated in the laboratory. The average measuring errors were within 2% for the force and 5% for the moment components as related to the maximum calibration values of 3000 N and 20 Nm, respectively. The sensitivity of the measuring implant is smaller than 1 N and 0.01 Nm. The telemeterized implant and the measuring system have been described in detail elsewhere [13,14]. 2.2. Patients Telemeterized VBRs were implanted into five patients who had suffered from an A3 type compression fracture [15] of a lumbar vertebral body. Four patients (WP1–WP4) had a fracture of the L1 vertebral body, and one patient (WP5) had a fracture of the L3 vertebral body. The fractures were first stabilized with pediclescrew-based internal fixation devices, implanted from the posterior. In a second surgery, parts of the fractured vertebral body and the adjacent discs were removed, and the instrumented VBR was inserted into the corpectomy defect. To enhance fusion of the adjacent segments, autologous bone material was added to the VBR. Table 1 provides data on patients and surgical procedures. The Ethics Committee of our hospital approved implantation of the modified implant in patients (Registry number 213-01/22520). The procedure was explained to the patients before surgery, and they gave their written consent for implantation of the telemeterized VBR, participation in measurements, and publication of their images.
751
2.3. Measurements Measurements were taken with an inductive power coil placed around the patient’s trunk at the level of the VBR and a small loop antenna fixed to the patient’s back with a harness. The patients were videotaped during the measurements, and the load-dependent telemetric signals were recorded on the same videotape. This allowed a detailed analysis of implant loads and activity at a later state without the patient having to be present. Parallel to this, the implant forces and moments were calculated from these signals and displayed online on a monitor [13]. Measurements were taken during the following activities: Level walking at a self-chosen walking speed (approximately 3– 4 km/h) in five patients during a total of 66 measuring sessions (9–17 per patient) up to 65 months postoperatively; Walking on a treadmill at different speeds in three sessions with patient WP1 and in one session each with patients WP2, WP4, and WP5. The speed of the treadmill was varied between 1 and 5 km/h in increments of 1 km/h. However, not all patients agreed to have measurements taken at all speeds. The patients did not use the handrail during the measurements; Walking with a wheeled invalid walker with arm support in four patients (WP1–WP4) during at least one session and walking with two crutches in all five patients in up to four sessions. The crutches were loaded alternatingly at the contralateral side of the supporting leg; Ascending and descending stairs in all five patients in up to six measuring sessions. The standard stairs had six treads. During some trials the patients used the handrail for safety; A total of about 600 trials, each consisting of approximately 5 steps, were evaluated for these four basic activities. These trials were performed during 66 measuring sessions. The patients reported no pain during the measuring sessions. 2.4. Evaluation During walking, the measured load components changed periodically. The peak values of a step were evaluated in this study. For each activity and session the median value of the peak resultant forces on the implant was determined. From these medians, the median value was determined for each activity and patient to obtain a representative value. The ranges of the peak resultant force values from several measuring sessions were also calculated for each patient. In order to determine the load ranges during walking the differences between the maximum and minimum values of the resultant force (max–min difference of each measurement) were calculated as well. The implant loads for standing relaxed were measured on the average nine times during each session. The median resultant force
Table 1 Data on patients and surgical procedures. Parameter
Gender Age at the time of surgery (y) Height (cm) Body mass (kg) Fractured vertebra Level of internal fixation device Bone material added Implantation date (mo/y)
Patient WP1
WP2
WP3
WP4
WP5
Male 62 168 66 L1 T12-L2 Yes 09/2006
Male 71 169 74 L1 T12-L2 Yes 11/2006
Female 69 168 64 L1 T11-L3 Yes 03/2007
Male 63 170 60 L1 T11-L3 Yes 01/2008
Male 66 180 63 L3 L2–L4 Yes 07/2008
752
A. Rohlmann et al. / Gait & Posture 39 (2014) 750–755
Fig. 1. Load components and force vectors during walking on a treadmill at 4 km/h. The results from five double steps are shown. The patient is shown at the instant of force maximum (vertical black line in the diagram). The nonsymmetric load in the frontal plane is mainly caused by nonsymmetric implantation of the anterior and posterior implants.
during standing was determined for each session and set to 100%. The force values for the other activities were related to this force value during standing, which was measured on the same day. Only descriptive statistics could be applied because no more than five patients had a telemeterized VBR and not all of them agreed to perform all activities. 3. Results 3.1. Level walking The peak load during each step usually occurred during the stance phase of the leg, directly after toe-off of the contralateral leg (Fig. 1). The peak force acted in a nearly axial direction of the VBR. The peak forces and moments acted simultaneously. The direction of the force vectors on the VBR varied only slightly during walking (Fig. 1).
The median peak resultant force on the VBR for 9–17 measuring sessions per patient was between 120% and 183% the values for standing (%STG) (Fig. 2). The median value for the five patients was 171 %STG. The average change of the resultant force during walking for the five patients was 105 %STG (range 47–118 %STG) (Fig. 2). There was a large variation in the median peak forces and the ranges of the resultant force for the different measuring sessions. Compared with the peak force value during walking, the median max-min difference was 56% (range 30–63%). 3.2. Effect of walking speed The peak resultant force on the implant generally increased with the walking speed when walking on a treadmill (Fig. 3). For patient WP1, the forces differed strongly for the three different measuring sessions. After linear regression, the force on the implant increases by approximately 22 N per 1 km/h speed increase and has a value of approximately 150 N for a speed of zero. Normal level walking, measured on the same day, corresponded to a walking speed on a treadmill of approximately 1 km/h, except for patient WP5, for whom it corresponded to a speed of approximately 3 km/h. 3.3. Ascending and descending stairs Ascending stairs generally caused higher forces than descending stairs (Fig. 4). The values for level walking, measured on the same day, were similar to or lower than those for descending stairs. The force value for going up the staircase varied strongly between the five patients. The force varied between 150 and 430 %STG with a median value of 265 %STG.
Fig. 2. Median values and ranges of peak resultant forces on the vertebral body replacement and difference between the maximum and minimum resultant force measured during level walking in 9–17 measuring sessions per patient. The forces are relative to the value for standing (=100%) measured on the same day. %STG = percentage of value for standing.
3.4. Effect of walking aids Using a wheeled invalid walker strongly reduced the forces on the VBR. Walking without a walking aid caused a median peak
A. Rohlmann et al. / Gait & Posture 39 (2014) 750–755
753
Fig. 3. Effect of walking speed on the median peak resultant force measured during a single session. The forces are relative to the value for standing measured on the same day.
Fig. 4. Comparison of peak resultant forces while ascending stairs, descending stairs and level walking. Median values and ranges of the forces from several measuring sessions relative to the value for standing.
force for the five patients which was approximately 170 %STG. Walking with a walker decreased the implant force to 62 %STG while walking with two crutches and loading each crutch contralaterally to the supporting leg had only a minor influence on the resultant force on the VBR. 3.5. Measured moments in the VBR The measured bending moments during level walking were usually less than 2 Nm, and the torsional moments were less than 1.5 Nm. High lateral bending moments (up to 4.6 Nm) were measured for stair climbing. Flexion bending moments of up to 4 Nm were measured when climbing stairs. This corresponds to a shift of the point of load application of less than 6 mm in the anterior direction. High bending moments were accompanied by high axial and resultant forces. The highest torsional moments (up to 2.1 Nm) were measured again for staircase walking. 4. Discussion The loads acting on a VBR during level and staircase walking were measured in five patients. The effects of walking speed and walking aids were investigated. The resultant force on the implant
was on the average approximately 170 %STG during level walking and more than 250 %STG for stair climbing. The walking speed had a strong influence on the resultant force. The force on the VBR was strongly reduced when a wheeled invalid walker was used, but not when using crutches. Level walking caused implant loads that were approximately 170 %STG (Fig. 2). For most patients, even the lowest relative median value from any of the measuring sessions was higher than the relative median values described in the literature for the intradiscal pressure [2,7] and for the loads in the internal fixators [1,8]. In our patients, the spinal load is shared by the VBR, the internal fixators, and the bone around the implant. Due to the long, pliable pedicle screws, the axial stiffness of the posterior internal spinal fixation device is much smaller than that of the anterior column at the implant level. The load changes in the fixators are therefore small [16,17]. Thus, the main portion of the spinal load is transferred by the VBR, and this has been confirmed by a finite element study [18]. Possible reasons for the differences compared with previous studies are the higher age of the patients in the present study and therefore the associated differences in shape and position of the spine and pelvis [19]. The walking speed had a considerable influence on the resultant implant force (Fig. 3). In previous studies [1,2,8], the walking speed
754
A. Rohlmann et al. / Gait & Posture 39 (2014) 750–755
had only a marginal effect on the measured loads. Again, in contrast to the subjects in previous studies, the patients of the present study were older, and most of them were not used to walking on a treadmill. In order not to fall backward while walking on a treadmill, the patients bent their upper body forward. We were able to measure the flexion angle of the upper body in the videos and found that it increased with increasing walking speed. This is in agreement with the findings of Frigo et al. [20] who also measured a forward inclination of the trunk during walking. Flexion of the upper body is accompanied by shifting the center of gravity anteriorly. This requires higher back muscle forces to balance the upper body, which in turn leads to higher spinal forces [11,12]. During level walking, the risk of falling backward is much lower than during treadmill walking. Thus, the forward inclination angle of the upper body is lower during level walking, and the measured loads are therefore lower than at the same speed on the treadmill. Cappozzo [3] used a biomechanical model of the trunk and a photogrammetric technique to obtain three-dimensional kinematic data and predicted an increase of spinal loads by the factor of 2.5 when doubling the speed from 4 to 8 km/h. This is much higher than what we measured here. However, his subjects were young and healthy, and the speeds were higher than in the present study. Ascending stairs caused higher forces on the VBR than descending stairs or level walking. This is in agreement with previous measurements by Wilke et al. [2] and Rohlmann et al. [1,8]. However, the load increase was generally much higher in the present study. This too may be due to the increased age of our patients. The upper body is usually more bend forward during ascending stairs than during descending stairs, especially in elder patients. The corresponding shift of the center of gravity explains the differences in the implant forces [11,12]. Using a wheeled invalid walker strongly reduced the force on the VBR. Placing one’s arms on the walker causes most of the upper body weight to be supported by the walker, and this reduces the force on the spine. The small load differences between the four patients are surprising and show the effectiveness of this walking aid. In three patients, the measurements with the walker were performed within three weeks after surgery. The load reduction on spinal fixators by using a walker without arm support, measured in two patients, was only 25% [1]. This is less than that observed in the present study. Walking with two crutches and loading them alternatingly did not reduce the force on the VBR. Previous measurements with telemeterized spinal fixation devices also showed only a small reduction from the normal use of the two crutches [1]. The main benefit of crutches is not a spinal load reduction, but an increase in walking safety and a reduction in the risk of stumbling and falling. During walking, changes of the bending moments in the sagittal and the frontal plane are most likely caused mainly by inclination changes of the upper body in the different planes [20]. Bending moments in the VBR are due to a shift of the point of load application on the VBR. This shift also depends on the axial force component. The maximum shift was less than 6 mm because high bending moments were accompanied by high axial forces. Due to the high stiffness of the VBR, a high bending moment has only a minor effect on the intersegmental rotation at the implant level. A high bending moment will, however, strongly increase the stresses in the VBR. The variation of the torsional moments measured during walking depends mainly on the rotation of the spine. Frigo et al. [20] found a total shoulder–pelvis rotation of 138 in the horizontal plane for a double step. This study has some limitations: Unique telemeterized VBRs were implanted in a small cohort of only five patients, and thus, only descriptive statistics could be applied. All patients were older than 60 years at the time of surgery. Not all patients agreed to
perform all activities described in this paper. The number of measurements was in some cases small, e.g., when walking with a wheeled invalid walker or with crutches. The number of measurements of an activity often differed between the various patients. The postoperative time of the measurements varied, and the resultant implant force for standing differed for the various measuring sessions. All results were therefore related to the values during standing, which were measured several times in each measuring session. In conclusion, the resultant forces on the VBR during walking were considerably higher than during relaxed standing. The large force variations between patients and from one measuring session to the next should be noted. The implant force can be lowered by reducing the walking speed and by using a wheeled invalid walker, but not by using crutches. Staircase walking, especially climbing upward, caused higher forces on the VBR than level walking did. Patients trying to prevent high loads should avoid or reduce this activity. Acknowledgments The authors greatly appreciate the friendly cooperation of their patients. They thank Dr. A. Bender, J. Dymke, J. Schleusener, and H. Srbinoska for technical assistance and Dr. U. Weber, Dr. C. Heyde, and Dr. R. Kayser for their medical support. Conflict of interest statement: The authors do not have any conflict of interest. Funding: This study was supported financially by the Deutsche Forschungsgemeinschaft, Bonn, Germany, (Ro 581/18-1) and by the Deutsche Arthrose-Hilfe, Frankfurt, Germany. Sponsors role: None. References [1] Rohlmann A, Bergmann G, Graichen F. Loads on an internal spinal fixation device during walking. J Biomech 1997;30:41–7. [2] Wilke H-J, Neef P, Caimi M, Hoogland T, Claes LE. New in vivo measurements of pressures in the intervertebral disc in daily life. Spine 1999;24:755–62. [3] Cappozzo A. Compressive loads in the lumbar vertebral column during normal level walking. J Orthop Res 1984;1:292–301. [4] Cromwell R, Schultz AB, Beck R, Warwick D. Loads on the lumbar trunk during level walking. J Orthop Res 1989;7:371–7. [5] McGill SM, Marshall L, Andersen J. Low back loads while walking and carrying: comparing the load carried in one hand or in both hands. Ergonomics 2013;56:293–302. http://dx.doi.org/10.1080/00140139.2012.752528. [6] Callaghan JP, Patla AE, McGill SM. Low back three-dimensional joint forces, kinematics, and kinetics during walking. Clin Biomech (Bristol Avon) 1999;14:203–16. [7] Nachemson A, Elfstrom G. Intravital dynamic pressure measurements in lumbar discs. A study of common movements, maneuvers and exercises. Scand J Rehabil Med Suppl 1970;1:1–40. [8] Rohlmann A, Claes L, Bergmann G, Graichen F, Neef P, Wilke H-J. Comparison of intradiscal pressures and spinal fixator loads for different body positions and exercises. Ergonomics 2001;44:781–94. [9] Dreischarf M, Bergmann G, Wilke HJ, Rohlmann A. Different arm positions and the shape of the thoracic spine can explain contradictory results in the literature about spinal loads for sitting and standing. Spine (Phila Pa 1976) 2010;35:2015–21. http://dx.doi.org/10.1097/BRS.0b013e3181d55d52. [10] Rohlmann A, Hinz B, Bluthner R, Graichen F, Bergmann G. Loads on a spinal implant measured in vivo during whole-body vibration. Eur Spine J 2010;19:1129–35. http://dx.doi.org/10.1007/s00586-010-1346-5. [11] Rohlmann A, Zander T, Graichen F, Bergmann G. Lifting up and laying down a weight causes high spinal loads. J Biomech 2013;46:511–4. http://dx.doi.org/ 10.1016/j.jbiomech.2012.10.022. [12] Rohlmann A, Zander T, Graichen F, Dreischarf M, Bergmann G. Measured loads on a vertebral body replacement during sitting. Spine J 2011;11:870–5. http:// dx.doi.org/10.1016/j.spinee.2011.06.017. [13] Graichen F, Arnold R, Rohlmann A, Bergmann G. Implantable 9-channel telemetry system for in vivo load measurements with orthopedic implants. IEEE Trans Biomed Eng 2007;54:253–61. http://dx.doi.org/10.1109/ TBME.2006.886857. [14] Rohlmann A, Gabel U, Graichen F, Bender A, Bergmann G. An instrumented implant for vertebral body replacement that measures loads in the anterior spinal column. Med Eng Phys 2007;29:580–5. http://dx.doi.org/10.1016/ j.medengphy.2006.06.012.
A. Rohlmann et al. / Gait & Posture 39 (2014) 750–755 [15] Magerl F, Aebi M, Gertzbein SD, Harms J, Nazarian S. A comprehensive classification of thoracic and lumbar injuries. Eur Spine J 1994;3:184–201. [16] Rohlmann A, Graichen F, Bergmann G. Influence of load carrying on loads in internal spinal fixators. J Biomech 2000;33:1099–104. [17] Rohlmann A, Graichen F, Kayser R, Bender A, Bergmann G. Loads on a telemeterized vertebral body replacement measured in two patients. Spine 2008;33:1170–9. [18] Zander T, Bergmann G, Rohlmann A. Large sizes of vertebral body replacement do not reduce the contact pressure on adjacent vertebral bodies per se. Med
755
Eng Phys 2009;31:1307–12. http://dx.doi.org/10.1016/j.medengphy.2009.08. 013. [19] Schwab F, Lafage V, Boyce R, Skalli W, Farcy JP. Gravity line analysis in adult volunteers: age-related correlation with spinal parameters, pelvic parameters, and foot position. Spine 2006;31:E959–67. [20] Frigo C, Carabalona R, Dalla Mura M, Negrini S. The upper body segmental movements during walking by young females. Clin Biomech 2003;18: 419–25.
Contents lists available at ScienceDirect
Gait & Posture journal homepage: www.elsevier.com/locate/gaitpost
Loads on a vertebral body replacement during locomotion measured in vivo A. Rohlmann *, M. Dreischarf, T. Zander, F. Graichen, G. Bergmann Julius Wolff Institute, Charite´ – Universita¨tsmedizin Berlin, Augustenburger Platz 1, 13353 Berlin, Germany
A R T I C L E I N F O
A B S T R A C T
Article history: Received 27 May 2013 Received in revised form 10 October 2013 Accepted 12 October 2013
Walking is one of the most important activities in daily life, and walking exposes the spine to a high number of loading cycles. Little is known about the spinal loads during walking. Telemeterized spinal implants can provide data about their loading during different activities. The aim of this study was to measure the loads on a vertebral body replacement (VBR) during level and staircase walking and to determine the effects of walking speed and using walking aids. Telemeterized VBRs were implanted in five patients suffering from compression fractures of the L1 or L3 lumbar vertebral body. The implant allows measurements of three force and three moment components. The resultant force on the VBR was measured during level and staircase walking, when walking on a treadmill at different speeds, and when using a wheeled invalid walker or crutches. On average, the resultant force on the VBR for level walking was 171% of the value for standing. This force value increased to 265% of the standing force when ascending stairs and to 225% when descending stairs. Walking speed had a strong effect on the implant force. Using a walker during ambulation on level ground reduced the force on the implant to 62% of standing forces, whereas using two crutches had only a minor effect. Walking causes much higher forces on the VBR than standing. A strong force reduction can be achieved by using a walker. ß 2013 Elsevier B.V. All rights reserved.
Keywords: Walking Load measurement Vertebral body replacement Telemetry Spinal loads
1. Introduction Walking is one of the most important activities of daily living. During each step, the discs and vertebrae of the lumbar spine are loaded with a large force [1,2]. However, few in vivo data exist on the spinal loads experienced during level and staircase walking. Furthermore, almost no in vivo measured data exist on the impact of walking speed and walking aids on the loads. For preclinical implant tests, however, and for finite element studies, realistic loads are mandatory to obtain relevant results. In addition, knowledge about spinal loads may help to advice patients regarding using of walking aids. Several groups used electromyography, force plate and body kinematic data to calculate spine loading during walking [3–6]. For the validation of their models in vivo measured data are required. The intradiscal pressure during level walking has been measured in a few subjects [2,7]. Wilke et al. measured a value in one subject that was 30% higher than the value measured for standing [2,8]. For slow walking, Nachemson and Elfstro¨m [7]
* Corresponding author. Tel.: +49 30 2093 46128; fax: +49 30 2093 46001. E-mail addresses: [email protected], [email protected] (A. Rohlmann). 0966-6362/$ – see front matter ß 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gaitpost.2013.10.010
measured an increase of intradiscal pressure of 15% in four patients compared with standing. These measurements were performed only during one measuring session per subject. The loads in internal spinal fixation devices during level walking were measured in 10 patients by Rohlmann et al. [1,8]. On average, they determined a maximum implant load during walking that was 28% higher than that for standing. For two patients [1], they reported only a small influence of walking speed on implant load. The normal use of crutches led to a small reduction of the implant loads, while a wheeled invalid walker reduced the loads on the internal spinal fixators by approximately 25%. Staircase walking caused higher intradiscal pressure values than level walking did [2,8]. Ascending stairs led to a maximum pressure in the disc that was 40% higher than for walking, while the value for descending stairs was only 20% higher [2]. Similar load values were measured in spinal fixators [1], with the loads usually being higher for ascending than descending stairs. However, no in vivo measured data exists about the loads on an anterior spinal implant during level and staircase walking. The telemeterized vertebral body replacement (VBR) allows the in vivo measurement of the loads acting in the anterior column of the lumbar spine. Such an implant has already been used to quantify the loads for several activities, including sitting, standing, lifting up and laying down a weight, and whole body vibration
A. Rohlmann et al. / Gait & Posture 39 (2014) 750–755
[9–12]. The force values varied strongly from patient to patient, while the measured moments were usually small. The aims of the present study were (1) to measure in vivo the resultant force on a vertebral body replacement during level walking; (2) to determine the influence of walking speed on implant loads; (3) to measure the resultant implant force during ascending and descending stairs; and (4) to determine the effects of walking aids such as a wheeled invalid walker and crutches on implant loads.
2. Materials and methods 2.1. Telemeterized vertebral body replacement A modified VBR (Synex, Synthes Inc., Bettlach, Switzerland) was used to measure the loading in vivo. Six semiconductor strain gauges (Type KSP 1-350-E4, Kyowa, Tokyo, Japan) were glued to the inner wall of the hollow implant, and they served as load sensors. A nine-channel telemetry unit and a coil for the inductive power supply of the electronic circuits were also integrated in the hermetically sealed implant. The instrumented implant allows the in vivo measurement of the three force and three moment components at a frequency of approximately 125 Hz. The resultant force is the geometric sum of the three force components. Prior to implantation, the VBRs were calibrated in the laboratory. The average measuring errors were within 2% for the force and 5% for the moment components as related to the maximum calibration values of 3000 N and 20 Nm, respectively. The sensitivity of the measuring implant is smaller than 1 N and 0.01 Nm. The telemeterized implant and the measuring system have been described in detail elsewhere [13,14]. 2.2. Patients Telemeterized VBRs were implanted into five patients who had suffered from an A3 type compression fracture [15] of a lumbar vertebral body. Four patients (WP1–WP4) had a fracture of the L1 vertebral body, and one patient (WP5) had a fracture of the L3 vertebral body. The fractures were first stabilized with pediclescrew-based internal fixation devices, implanted from the posterior. In a second surgery, parts of the fractured vertebral body and the adjacent discs were removed, and the instrumented VBR was inserted into the corpectomy defect. To enhance fusion of the adjacent segments, autologous bone material was added to the VBR. Table 1 provides data on patients and surgical procedures. The Ethics Committee of our hospital approved implantation of the modified implant in patients (Registry number 213-01/22520). The procedure was explained to the patients before surgery, and they gave their written consent for implantation of the telemeterized VBR, participation in measurements, and publication of their images.
751
2.3. Measurements Measurements were taken with an inductive power coil placed around the patient’s trunk at the level of the VBR and a small loop antenna fixed to the patient’s back with a harness. The patients were videotaped during the measurements, and the load-dependent telemetric signals were recorded on the same videotape. This allowed a detailed analysis of implant loads and activity at a later state without the patient having to be present. Parallel to this, the implant forces and moments were calculated from these signals and displayed online on a monitor [13]. Measurements were taken during the following activities: Level walking at a self-chosen walking speed (approximately 3– 4 km/h) in five patients during a total of 66 measuring sessions (9–17 per patient) up to 65 months postoperatively; Walking on a treadmill at different speeds in three sessions with patient WP1 and in one session each with patients WP2, WP4, and WP5. The speed of the treadmill was varied between 1 and 5 km/h in increments of 1 km/h. However, not all patients agreed to have measurements taken at all speeds. The patients did not use the handrail during the measurements; Walking with a wheeled invalid walker with arm support in four patients (WP1–WP4) during at least one session and walking with two crutches in all five patients in up to four sessions. The crutches were loaded alternatingly at the contralateral side of the supporting leg; Ascending and descending stairs in all five patients in up to six measuring sessions. The standard stairs had six treads. During some trials the patients used the handrail for safety; A total of about 600 trials, each consisting of approximately 5 steps, were evaluated for these four basic activities. These trials were performed during 66 measuring sessions. The patients reported no pain during the measuring sessions. 2.4. Evaluation During walking, the measured load components changed periodically. The peak values of a step were evaluated in this study. For each activity and session the median value of the peak resultant forces on the implant was determined. From these medians, the median value was determined for each activity and patient to obtain a representative value. The ranges of the peak resultant force values from several measuring sessions were also calculated for each patient. In order to determine the load ranges during walking the differences between the maximum and minimum values of the resultant force (max–min difference of each measurement) were calculated as well. The implant loads for standing relaxed were measured on the average nine times during each session. The median resultant force
Table 1 Data on patients and surgical procedures. Parameter
Gender Age at the time of surgery (y) Height (cm) Body mass (kg) Fractured vertebra Level of internal fixation device Bone material added Implantation date (mo/y)
Patient WP1
WP2
WP3
WP4
WP5
Male 62 168 66 L1 T12-L2 Yes 09/2006
Male 71 169 74 L1 T12-L2 Yes 11/2006
Female 69 168 64 L1 T11-L3 Yes 03/2007
Male 63 170 60 L1 T11-L3 Yes 01/2008
Male 66 180 63 L3 L2–L4 Yes 07/2008
752
A. Rohlmann et al. / Gait & Posture 39 (2014) 750–755
Fig. 1. Load components and force vectors during walking on a treadmill at 4 km/h. The results from five double steps are shown. The patient is shown at the instant of force maximum (vertical black line in the diagram). The nonsymmetric load in the frontal plane is mainly caused by nonsymmetric implantation of the anterior and posterior implants.
during standing was determined for each session and set to 100%. The force values for the other activities were related to this force value during standing, which was measured on the same day. Only descriptive statistics could be applied because no more than five patients had a telemeterized VBR and not all of them agreed to perform all activities. 3. Results 3.1. Level walking The peak load during each step usually occurred during the stance phase of the leg, directly after toe-off of the contralateral leg (Fig. 1). The peak force acted in a nearly axial direction of the VBR. The peak forces and moments acted simultaneously. The direction of the force vectors on the VBR varied only slightly during walking (Fig. 1).
The median peak resultant force on the VBR for 9–17 measuring sessions per patient was between 120% and 183% the values for standing (%STG) (Fig. 2). The median value for the five patients was 171 %STG. The average change of the resultant force during walking for the five patients was 105 %STG (range 47–118 %STG) (Fig. 2). There was a large variation in the median peak forces and the ranges of the resultant force for the different measuring sessions. Compared with the peak force value during walking, the median max-min difference was 56% (range 30–63%). 3.2. Effect of walking speed The peak resultant force on the implant generally increased with the walking speed when walking on a treadmill (Fig. 3). For patient WP1, the forces differed strongly for the three different measuring sessions. After linear regression, the force on the implant increases by approximately 22 N per 1 km/h speed increase and has a value of approximately 150 N for a speed of zero. Normal level walking, measured on the same day, corresponded to a walking speed on a treadmill of approximately 1 km/h, except for patient WP5, for whom it corresponded to a speed of approximately 3 km/h. 3.3. Ascending and descending stairs Ascending stairs generally caused higher forces than descending stairs (Fig. 4). The values for level walking, measured on the same day, were similar to or lower than those for descending stairs. The force value for going up the staircase varied strongly between the five patients. The force varied between 150 and 430 %STG with a median value of 265 %STG.
Fig. 2. Median values and ranges of peak resultant forces on the vertebral body replacement and difference between the maximum and minimum resultant force measured during level walking in 9–17 measuring sessions per patient. The forces are relative to the value for standing (=100%) measured on the same day. %STG = percentage of value for standing.
3.4. Effect of walking aids Using a wheeled invalid walker strongly reduced the forces on the VBR. Walking without a walking aid caused a median peak
A. Rohlmann et al. / Gait & Posture 39 (2014) 750–755
753
Fig. 3. Effect of walking speed on the median peak resultant force measured during a single session. The forces are relative to the value for standing measured on the same day.
Fig. 4. Comparison of peak resultant forces while ascending stairs, descending stairs and level walking. Median values and ranges of the forces from several measuring sessions relative to the value for standing.
force for the five patients which was approximately 170 %STG. Walking with a walker decreased the implant force to 62 %STG while walking with two crutches and loading each crutch contralaterally to the supporting leg had only a minor influence on the resultant force on the VBR. 3.5. Measured moments in the VBR The measured bending moments during level walking were usually less than 2 Nm, and the torsional moments were less than 1.5 Nm. High lateral bending moments (up to 4.6 Nm) were measured for stair climbing. Flexion bending moments of up to 4 Nm were measured when climbing stairs. This corresponds to a shift of the point of load application of less than 6 mm in the anterior direction. High bending moments were accompanied by high axial and resultant forces. The highest torsional moments (up to 2.1 Nm) were measured again for staircase walking. 4. Discussion The loads acting on a VBR during level and staircase walking were measured in five patients. The effects of walking speed and walking aids were investigated. The resultant force on the implant
was on the average approximately 170 %STG during level walking and more than 250 %STG for stair climbing. The walking speed had a strong influence on the resultant force. The force on the VBR was strongly reduced when a wheeled invalid walker was used, but not when using crutches. Level walking caused implant loads that were approximately 170 %STG (Fig. 2). For most patients, even the lowest relative median value from any of the measuring sessions was higher than the relative median values described in the literature for the intradiscal pressure [2,7] and for the loads in the internal fixators [1,8]. In our patients, the spinal load is shared by the VBR, the internal fixators, and the bone around the implant. Due to the long, pliable pedicle screws, the axial stiffness of the posterior internal spinal fixation device is much smaller than that of the anterior column at the implant level. The load changes in the fixators are therefore small [16,17]. Thus, the main portion of the spinal load is transferred by the VBR, and this has been confirmed by a finite element study [18]. Possible reasons for the differences compared with previous studies are the higher age of the patients in the present study and therefore the associated differences in shape and position of the spine and pelvis [19]. The walking speed had a considerable influence on the resultant implant force (Fig. 3). In previous studies [1,2,8], the walking speed
754
A. Rohlmann et al. / Gait & Posture 39 (2014) 750–755
had only a marginal effect on the measured loads. Again, in contrast to the subjects in previous studies, the patients of the present study were older, and most of them were not used to walking on a treadmill. In order not to fall backward while walking on a treadmill, the patients bent their upper body forward. We were able to measure the flexion angle of the upper body in the videos and found that it increased with increasing walking speed. This is in agreement with the findings of Frigo et al. [20] who also measured a forward inclination of the trunk during walking. Flexion of the upper body is accompanied by shifting the center of gravity anteriorly. This requires higher back muscle forces to balance the upper body, which in turn leads to higher spinal forces [11,12]. During level walking, the risk of falling backward is much lower than during treadmill walking. Thus, the forward inclination angle of the upper body is lower during level walking, and the measured loads are therefore lower than at the same speed on the treadmill. Cappozzo [3] used a biomechanical model of the trunk and a photogrammetric technique to obtain three-dimensional kinematic data and predicted an increase of spinal loads by the factor of 2.5 when doubling the speed from 4 to 8 km/h. This is much higher than what we measured here. However, his subjects were young and healthy, and the speeds were higher than in the present study. Ascending stairs caused higher forces on the VBR than descending stairs or level walking. This is in agreement with previous measurements by Wilke et al. [2] and Rohlmann et al. [1,8]. However, the load increase was generally much higher in the present study. This too may be due to the increased age of our patients. The upper body is usually more bend forward during ascending stairs than during descending stairs, especially in elder patients. The corresponding shift of the center of gravity explains the differences in the implant forces [11,12]. Using a wheeled invalid walker strongly reduced the force on the VBR. Placing one’s arms on the walker causes most of the upper body weight to be supported by the walker, and this reduces the force on the spine. The small load differences between the four patients are surprising and show the effectiveness of this walking aid. In three patients, the measurements with the walker were performed within three weeks after surgery. The load reduction on spinal fixators by using a walker without arm support, measured in two patients, was only 25% [1]. This is less than that observed in the present study. Walking with two crutches and loading them alternatingly did not reduce the force on the VBR. Previous measurements with telemeterized spinal fixation devices also showed only a small reduction from the normal use of the two crutches [1]. The main benefit of crutches is not a spinal load reduction, but an increase in walking safety and a reduction in the risk of stumbling and falling. During walking, changes of the bending moments in the sagittal and the frontal plane are most likely caused mainly by inclination changes of the upper body in the different planes [20]. Bending moments in the VBR are due to a shift of the point of load application on the VBR. This shift also depends on the axial force component. The maximum shift was less than 6 mm because high bending moments were accompanied by high axial forces. Due to the high stiffness of the VBR, a high bending moment has only a minor effect on the intersegmental rotation at the implant level. A high bending moment will, however, strongly increase the stresses in the VBR. The variation of the torsional moments measured during walking depends mainly on the rotation of the spine. Frigo et al. [20] found a total shoulder–pelvis rotation of 138 in the horizontal plane for a double step. This study has some limitations: Unique telemeterized VBRs were implanted in a small cohort of only five patients, and thus, only descriptive statistics could be applied. All patients were older than 60 years at the time of surgery. Not all patients agreed to
perform all activities described in this paper. The number of measurements was in some cases small, e.g., when walking with a wheeled invalid walker or with crutches. The number of measurements of an activity often differed between the various patients. The postoperative time of the measurements varied, and the resultant implant force for standing differed for the various measuring sessions. All results were therefore related to the values during standing, which were measured several times in each measuring session. In conclusion, the resultant forces on the VBR during walking were considerably higher than during relaxed standing. The large force variations between patients and from one measuring session to the next should be noted. The implant force can be lowered by reducing the walking speed and by using a wheeled invalid walker, but not by using crutches. Staircase walking, especially climbing upward, caused higher forces on the VBR than level walking did. Patients trying to prevent high loads should avoid or reduce this activity. Acknowledgments The authors greatly appreciate the friendly cooperation of their patients. They thank Dr. A. Bender, J. Dymke, J. Schleusener, and H. Srbinoska for technical assistance and Dr. U. Weber, Dr. C. Heyde, and Dr. R. Kayser for their medical support. Conflict of interest statement: The authors do not have any conflict of interest. Funding: This study was supported financially by the Deutsche Forschungsgemeinschaft, Bonn, Germany, (Ro 581/18-1) and by the Deutsche Arthrose-Hilfe, Frankfurt, Germany. Sponsors role: None. References [1] Rohlmann A, Bergmann G, Graichen F. Loads on an internal spinal fixation device during walking. J Biomech 1997;30:41–7. [2] Wilke H-J, Neef P, Caimi M, Hoogland T, Claes LE. New in vivo measurements of pressures in the intervertebral disc in daily life. Spine 1999;24:755–62. [3] Cappozzo A. Compressive loads in the lumbar vertebral column during normal level walking. J Orthop Res 1984;1:292–301. [4] Cromwell R, Schultz AB, Beck R, Warwick D. Loads on the lumbar trunk during level walking. J Orthop Res 1989;7:371–7. [5] McGill SM, Marshall L, Andersen J. Low back loads while walking and carrying: comparing the load carried in one hand or in both hands. Ergonomics 2013;56:293–302. http://dx.doi.org/10.1080/00140139.2012.752528. [6] Callaghan JP, Patla AE, McGill SM. Low back three-dimensional joint forces, kinematics, and kinetics during walking. Clin Biomech (Bristol Avon) 1999;14:203–16. [7] Nachemson A, Elfstrom G. Intravital dynamic pressure measurements in lumbar discs. A study of common movements, maneuvers and exercises. Scand J Rehabil Med Suppl 1970;1:1–40. [8] Rohlmann A, Claes L, Bergmann G, Graichen F, Neef P, Wilke H-J. Comparison of intradiscal pressures and spinal fixator loads for different body positions and exercises. Ergonomics 2001;44:781–94. [9] Dreischarf M, Bergmann G, Wilke HJ, Rohlmann A. Different arm positions and the shape of the thoracic spine can explain contradictory results in the literature about spinal loads for sitting and standing. Spine (Phila Pa 1976) 2010;35:2015–21. http://dx.doi.org/10.1097/BRS.0b013e3181d55d52. [10] Rohlmann A, Hinz B, Bluthner R, Graichen F, Bergmann G. Loads on a spinal implant measured in vivo during whole-body vibration. Eur Spine J 2010;19:1129–35. http://dx.doi.org/10.1007/s00586-010-1346-5. [11] Rohlmann A, Zander T, Graichen F, Bergmann G. Lifting up and laying down a weight causes high spinal loads. J Biomech 2013;46:511–4. http://dx.doi.org/ 10.1016/j.jbiomech.2012.10.022. [12] Rohlmann A, Zander T, Graichen F, Dreischarf M, Bergmann G. Measured loads on a vertebral body replacement during sitting. Spine J 2011;11:870–5. http:// dx.doi.org/10.1016/j.spinee.2011.06.017. [13] Graichen F, Arnold R, Rohlmann A, Bergmann G. Implantable 9-channel telemetry system for in vivo load measurements with orthopedic implants. IEEE Trans Biomed Eng 2007;54:253–61. http://dx.doi.org/10.1109/ TBME.2006.886857. [14] Rohlmann A, Gabel U, Graichen F, Bender A, Bergmann G. An instrumented implant for vertebral body replacement that measures loads in the anterior spinal column. Med Eng Phys 2007;29:580–5. http://dx.doi.org/10.1016/ j.medengphy.2006.06.012.
A. Rohlmann et al. / Gait & Posture 39 (2014) 750–755 [15] Magerl F, Aebi M, Gertzbein SD, Harms J, Nazarian S. A comprehensive classification of thoracic and lumbar injuries. Eur Spine J 1994;3:184–201. [16] Rohlmann A, Graichen F, Bergmann G. Influence of load carrying on loads in internal spinal fixators. J Biomech 2000;33:1099–104. [17] Rohlmann A, Graichen F, Kayser R, Bender A, Bergmann G. Loads on a telemeterized vertebral body replacement measured in two patients. Spine 2008;33:1170–9. [18] Zander T, Bergmann G, Rohlmann A. Large sizes of vertebral body replacement do not reduce the contact pressure on adjacent vertebral bodies per se. Med
755
Eng Phys 2009;31:1307–12. http://dx.doi.org/10.1016/j.medengphy.2009.08. 013. [19] Schwab F, Lafage V, Boyce R, Skalli W, Farcy JP. Gravity line analysis in adult volunteers: age-related correlation with spinal parameters, pelvic parameters, and foot position. Spine 2006;31:E959–67. [20] Frigo C, Carabalona R, Dalla Mura M, Negrini S. The upper body segmental movements during walking by young females. Clin Biomech 2003;18: 419–25.
