Christopher J. Colloca, DC, Robert Gunzburg, MD, PhD, Brian J. Freeman, MD,
Marek Szpalski, MD, Mostafa Afifi, PhD, and Robert J. Moore, PhD
Graduate Student, PhD Kinesiology Program,
School of Nutrition and Health Promotion,
Arizona State University, Tempe, AZ.
OBJECTIVE: The purposes of this study were to quantify the biomechanical and pathologic consequences of surgically induced spinal lesions and to determine their response to spinal manipulation (SMT) in an in vivo ovine model.
METHODS: Of 24 Merino sheep, 6 received L5 spondylolytic defects, 6 received L1 annular lesions, and 12 served as respective controls. Dorsoventral (DV) stiffness was assessed using oscillatory loads (2-12 Hz). Two SMT force-time profiles were administered in each of the groups using a randomized and repeated-measures design. Stiffness and the effect of SMT on the DV motions and multifidus needle electromyographic responses were assessed using a repeated-measures analysis of variance (a = .05). Postmortem histologic analysis and computed tomography validated the presence of lesions.
RESULTS: L5 DV stiffness was significantly increased (40.2%) in the spondylolysis (6.28 N/mm) compared with the L5 control group (4.48 N/mm) (P < 03). Spinal manipulations delivered to the spondylolysis group resulted in less DV vertebral displacement (P < .01) compared with controls. Dorsoventral stiffness of the disc degeneration group was 5.66 N/mm, 94.5% greater than in the L1 control group (2.91 N/mm) (P < .01). One hundred-millisecond SMTs resulted in significantly reduced DV displacements in the disc degeneration group compared with the L1 control group (P < .01). Animals in the disc degeneration group showed a consistent 25% to 30% reduction in needle electromyographic responses to all SMTs.
CONCLUSIONS: Quantifiable objective evidence of spinal lesions and their response to SMT were confirmed in this study. Neuromechanical alterations provide novel insights into quantifying manipulable spinal lesions and a means to biomechanically assess SMT outcomes.
From the Full-Text Article:
Spinal disorders represent a broad range of conditions for which patients with musculoskeletal pain seek treatment at an enormous cost to society and governments alike. [1, 2] Abnormal mechanics of the spine have been hypothesized as a precursor to musculoskeletal pain via nociceptive afferents in the innervated spinal tissues.  The association between abnormal spinal mechanics and nociceptive stimulation includes considerations of inflammation,  biomechanical and nutritional challenges,  immunologic factors,  changes in the structure and material of the vertebral end plates, and discs such as neoneuralization. [7, 8]
Abnormal spinal mechanics may be due to degenerative changes in the spinal column [9, 10] and/or injury of the ligaments.  Most likely, the initiating event is some kind of trauma involving the spine.  It may be a single trauma due to an accident or a microtrauma caused by repetitive motion over a long time.  It is also possible that spinal muscles will fire in an uncoordinated way in response to sudden fear of injury, such as when one misjudges the depth of a step.  Any or all these events may cause spinal ligament injury. Adverse psychosocial factors may also play an important role in transforming the musculoskeletal pain into disability. 
The intervertebral disc (IVD) is a known pain generator among low back pain patients, and the IVD is therefore a primary target of intervention for clinicians applying spinal manipulation (SM).  Progressive degenerative changes of the IVD are associated with increased age, trauma, and abnormal postural loading.  Indeed, a large proportion of the population undergoing SM has some degree of disc disease.  To influence the peripheral pain generator, patients with discogenic disease commonly undergo spinal manipulative therapy (SMT) with the primary goal of normalizing loads and improving spinal mobility.  Although not as common a spinal entity undergoing SM treatment, spondylolysis is another common spinal disorder that results from pars defects acquired from trauma, fatigue failure, or nonunion. 
In the treatment of spinal disorders, many conservative treatment modalities have been proposed, yet quantification of lesions to better mold the therapeutic response remains elusive. Recent reviews of the limited basic science research in chiropractic noted that nearly all of the theories of the effects and mechanisms of action of SM still lack adequate research and that no definitive anatomic or biomechanical studies have yet identified the lesion manipulated. [20, 21]
In vivo animal studies require the validation of adequate models. For disc degeneration, Osti et al  developed an ovine model that has been universally accepted and used. A stab wound in the disc, reaching into the nucleus, is followed by accelerated disc degeneration as early as 3 months postsurgery.  Only recently has this ovine model been used to study spinal manipulaton. [24, 25] For pars defect or spondylolysis, no such model has been validated. The search for the quantifiable detection of spinal lesions has been identified as a major need in spinal research aimed toward defining their biomechanical consequences and clinical relevance. The purpose of this experimental study was to quantify the biomechanical and prospective pathologic consequences of surgically induced spinal lesions of pars fracture (spondylolysis) and IVD degeneration and determine their response to SM in an in vivo ovine model.
The primary aim of this study was to determine if spinal lesions could progress to a clinical state that could be objectively quantified biomechanically. Using a previously validated ovine lumbar degenerative disc lesion model  and a novel ovine spondylolysis lesion model, we were able to prospectively track the progression of surgically induced spinal lesions histologically and radiologically through a 6-month follow-up. Objective evidence of an increase in dynamic spinal stiffness as well as reductions in vertebral displacements occurring in response to SM was observed in the spondylolysis and disc degeneration groups compared with their age-matched and exposure level controls. Electromyographic evidence of multifidus muscle inhibition was found in the disc degeneration group compared with the L1 controls.
Analogous to disc degeneration in humans,  the healing response of the L1-L2 ovine disc was associated with histologically verified annular disruption, nuclear migration, and granulation tissue formation in the outer annular region of the ovine disc. The disc lesion resulted in substantial loss of L1 disc height and a breakdown of disc matrix with radial and circumferential extension of the initial annular lesion with secondary displacement of the nucleus toward the anterior aspect. This, in turn, results in prominent inversion of the posterior annular fibers from their usual concave orientation as observed in our results. Such lesions have been described previously.  No such changes were visualized in the L1 control group that underwent the same surgical exposure but without stab incision to their IVD. It is thus likely that the progressive degenerative changes histologically observed in the Degenerative Disc group were due to the stab incision itself and not the surgical exposure.
Previous investigations have demonstrated anatomical, biomechanical, and biochemical similarities between sheep and human IVDs. [26, 27, 37] Thus, the ovine degenerative disc model is deemed to be a valid model to investigate biomechanical responses to dorsoventral mechanical excitation for dynamic spinal stiffness assessment and responses to SM. Indeed, this is the first study to use the ovine degenerative disc model for in vivo biomechanical comparisons of dynamic spinal stiffness and SM outcomes.
Conversely, while there have been a number of studies of spondylolisthesis using animal models both in vitro and in vivo, [38-41] to our knowledge, our study is the first animal model designed specifically to investigate both the in vivo mechanical and pathologic consequences of surgically created spondylolytic defects. Ardern et al  reported the in vitro RSA results of this group of animals as part of another report. In this work, the instability from the surgically induced spondylolytic defects observed initially progressively stabilized over the 6-month study period to become less mobile than controls in all axes of movement. The stabilization of the spinal motion segments observed in the RSA portion of this research is consistent with our in vivo biomechanical findings of increased PA spinal stiffness identified at 2 Hz and 12 Hz and decreased spinal displacements that were observed for the 80-N, 100-millisecond spinal manipulative thrusts.
Irrespective of bony or fibrous ankylosis observed histologically, the 6-month postinjury period provided ample opportunity for healing that was characterized not by a chronic instability but rather by an increase in stiffening properties of the spine to PA forces. The fact that L5 control animals that received the same posterior surgical exposure did not develop this progressive spinal stiffening suggests that it was not caused by surgical scarring of the exposure alone. In this spondylolysis model, the creation of the spondylolytic defects leads to a stiffening response that was both progressive and excessive. An important implication for practitioners of SM is the fact that our results found that spondylolytic lesions exhibited greater spinal stiffness and less vertebral motion when exposed to spinal manipulative thrusts. The notion that spondylolytic lesions may be contraindicated for SM treatment because of risk of instability is therefore not supported by our results. Rather, analogous to welded steel being harder than the parent metal, 6 months after induced lumbar spondylolytic fracture, the ovine spine healed to a state of increased stiffness over normal ovine spines.
We did not find any evidence of disc or growth plate injury or degeneration in any of the animals in the spondylolysis arm of the study. It is possible that the progressive stiffening of the spinal motion segments protected against this. It is also possible, however, that such change would have been observed if the study had progressed longer than 6 months given that the defects had not healed with bone in two-thirds of the animals. It should be remembered that spondylolysis does not lead to slip (spondylolisthesis) in most of the cases in humans either and that when spondylolisthesis does occur, it develops over a prolonged period. 
Electromyographic findings also did not demonstrate any increases in paraspinal muscle activity responsible for the increased stiffening of the spine. It is thus likely that the increases in spinal stiffness observed in the spondylolysis group were largely influenced by the fibrous and bony ankylosis response to healing in the passive structures of the spine. It should be noted that, in some cases, this increase in stiffness is partly due to facet joint ankylosis at a distance to the original fracture. Because we only analyzed one spinal level in our study, the extent to which this may be an influencing factor cannot be verified by our study. Future work should aim to determine the exact anatomical and physiologic contributors to the biomechanical increases in spinal stiffness observed.
Animals with degenerated discs showed significantly increased L1 PA stiffness at 2 Hz and 12 Hz and similarly showed decreased vertebral displacements for spinal manipulative thrusts of 100-millisecond duration. This corroborates the findings of others who have demonstrated increased stiffness among dehydrated or degenerated discs in situ.  However, statistically significant disc degeneration–related changes in segmental and intersegmental kinematics were not observed for the shorter-duration (10 milliseconds) SM thrusts, which in this study seems to reflect the fact that impulsive loading produces a more variable kinematic response. Our finding of decreased nEMG responses among animals with degenerated discs likely reflects a disturbance in the neuromuscular reflex control of the stabilizing trunk muscles characteristic of the IVD lesion. This result was not obtained in the spondylolysis group or in the L1 control group, whose discs were found to remain intact. Further research should aim to target the relationships between degenerated discs and neuromuscular responses.
Biomechanically, when examining dynamic PA spinal stiffness at 6 Hz were any differences observed neither between the disc degeneration group and its L1 control nor in comparing the spondylolysis group with its L5 control. The resonant frequency of the ovine spine has been previously reported to be approximately 4 to 6 Hz,  and thus, at the 6-Hz mechanical excitation frequency, the spine would be expected to move the most with the least amount of force. This increase in mobility from exciting the spine close to its resonant frequency may be responsible for similar stiffness results between the active and control groups that did not yield a significant difference at 6 Hz. Other research has demonstrated increases in spinal stiffness observed at certain mechanical excitation frequencies and not at others when comparing independent variables. 
This is the first study demonstrating differences in vertebral kinematics for specimens with degenerated discs and spondylolytic lesions, an important finding for clinicians. Clinicians practicing SMT cognitively and kinesthetically gauge the amount of force they deem appropriate for a particular patient or condition based on biomechanical (ie, anatomical) and clinical (ie, pain tolerance) variables alike. Knowledge that degenerated or ankylosed functional spinal units will undergo substantially less PA motion for a given PA force as demonstrated in the current study provides clinicians with important biomechanical information that can be considered in clinical practice.
Measurement of vertebral movement using intraosseous pins equipped with accelerometers [44-46] and other invasive motion measurement devices [47, 48] has been previously shown to be a precise measure of spine segmental and intersegmental motion, but invasive procedures currently have limited clinical utility. Noteworthy, however, was our finding that increases in dynamic PA spinal stiffness and decreases in PA displacement in the animals with spinal lesions were able to be observed biomechanically using a noninvasive measurement system. The ability to noninvasively detect biomechanical changes in degenerated discs in vivo using an indenter over the spinous processes may have implications for the development of quantitative biomechanical spinal assessment strategies. 
It is important to point out that the animals examined in this study were anesthetized, which may have altered the mechanical responses (force displacement) slightly. Most likely, any effects of muscle tone on the force-displacement response of the spine were minimal, however, because the ligamentous spine is a highly damped structure.  Recent work  indicates that although sustained supramaximal muscle stimulation increases the PA stiffness (force/displacement) of the ovine spine up to 2-fold, the effects of low amplitude muscle stimulation were much less dramatic (less than 5% increase in PA stiffness). Therefore, the absence of muscle tone should have a minimal effect on the mechanical responses reported in this study.
The creation of spinal lesions in animal models consistent with clinically relevant spinal disorders is an important first step in understanding our ability to identify clinical correlates of the lesion and/or targets for SM or other types of treatments. Further work examining the dynamic mechanical response of the normal and degenerated spine as well models of other spinal disorders will assist in understanding both the etiology of spinal disorders and putative effects of spinal manipulative therapy among different patient populations.
Using a previously validated ovine lumbar degenerative disc lesion model and a novel ovine spondylolysis lesion model, surgically induced spinal lesions were prospectively tracked histologically and radiologically through a 6-month follow-up. Objective evidence of an increase in dynamic spinal stiffness, as well as reductions in vertebral displacements occurring in response to SM, was observed in the spondylolysis and disc degeneration groups compared with their age-matched and exposure level controls. Histologic evidence of pathology consistent with an alteration of spinal stiffness accompanied by alterations in the neuromuscular system provide novel insights into quantifying manipulable spinal lesions as well as a means to biomechanically assess SMT outcomes.
Using a previously validated ovine lumbar degenerative disc lesion model and a novel ovine spondylolysis lesion model, surgically induced spinal lesions were prospectively tracked histologically and radiologically through a 6-month follow-up.
Objective evidence of an increase in dynamic spinal stiffness, as well as reductions in vertebral displacements occurring in response to SM, were observed in the spondylolysis and disc degeneration groups compared with their age-matched and exposure level controls.
Histologic evidence of pathology consistent with an alteration of spinal stiffness accompanied by alterations in the neuromuscular system provides novel insights into quantifying manipulable spinal lesions as well as a means to biomechanically assess SMT outcomes.