Review of Scoliosis Research at the University of Michigan

Joel D. Mack, M.D. and Robert A. Greenberg, M.D.

Scoliosis is a spine curvature that can considerably disable humans. About 75-80% of scoliosis cases are idiopathic; the rest are caused by musculoskeletal abnormalities, neuromuscular diseases, trauma, nerve root compression, spinal cord tumors, and other conditions.¹ The etiology of idiopathic scoliosis is unknown; a better understanding of the causes of scoliosis may improve scoliosis management. However, it is difficult, if not impossible, for animal models to provide clinically relevant insights into human idiopathic scoliosis: artificial induction of scoliosis in animals does not parallel spontaneous human scoliosis, and the erect human has fundamentally different forces on the spine than four-legged animals. Having reviewed a research grant proposal submitted by J.A. Miller² to study spinal muscle functions in a canine model of scoliosis, we conclude that there are severe methodological flaws.

Ashton Miller claims three general points of significance for this study. First, he writes, "This proposal is aimed at improving our understanding of the mechanisms of curve progression in scoliosis."² While this issue has theraputic and prognostic importance, the value of these dog studies is dubious. Ashton Miller plans to induce experimental scoliosis in dogs by external forces. It is unlikely that this artificial protocol reproduces the situation in human scoliosis. Also, the rate of scoliosis formation in the animal model differs dramatically from that in humans. Ashton Miller notes:

In man, the average curve progression rate seldom exceeds 10 degrees per year. Of course, it would be best to study similar rates in dog [sic]. However, firstly we are not assured that the trends in the muscular response to such slow rates can be easily detected given the noise in myoelectric estimates of muscle contraction intensity and secondly, the study of an adequate number of animals to achieve statistical significance over 12 months or longer is prohibitively expensive.²

Range of curve progression will vary between 10 degrees per hour to 5 degrees per week, all of which greatly exceed normal curve progression in humans. This raises the possibiliy that he will be studying artifactual consequences of rapid scoliosis induction, rather than the real progression of scoliosis.

A second area of significance, according to Ashton Miller, is that 'There are no longitudinal studies of paravertebral muscle response to the presence of scoliosis."² He continues, 'Usually, attention has been focused exclusively on the convex and concave sides of the curve apex, rather than at multiple levels as in the present proposal."² Myoelectric swdies have been done on human scoliotic patients, and it is possible that Ashton Miller's electromyographic data could be gathered from human clinical subjects. Indeed, Ashton Miller proposes to perform simultaneous human clinical studies in an effort to validate the animal model. If it were possible to obtain the desired information safely and ethically with human patients, then the animal studies would be superfluous. Ashton Miller, recognizing the limitations of animal models, writes, 'Extrapolation of the results from dogs to the human case is fraught with uncertainties."² Consequently, an animal model appears to be inappropriate.

Ashton Miller concludes:

The significance of the animal experiments is that we will, for the first time, study the spatial and temporal response of the intact musculoskeletal system to ramp and step changes in the magnitude of an experimentally-induced scoliosis. This may be done in animal experiments at a level of detail not normally possible in the clinical setting.. . . The significance of this proposal is the recognition of the importance of understanding the muscular response to a scoliosis.²

Although Ashton Miller contends that he will obtain useful data, the relevance of this animal model is doubtful. Ashton Miller appears to misunderstand the nature of scoliosis. He writes, "Scoliosis, or a lateral curvature of the spine, occurs in about 3% of adolescents."² This is inaccurate. In fact, scoliosis generally has a rotational component,^3,4 which may be the more important component of the deformity. Roaf has summarized:

There are two types of scoliosis, which occur independently of etiology. The first is an uncommon type in which lateral flexion is the dominant element. The curve usually remains relatively mobile, is fairly easily corrected, has less tendency to relapse after fusion, and presents minimal rib deformity; basically it is an exaggeration of a normal movement

The other type of curve, in which rotation is the main feature, tends to be pmgressive; it soon becomes rigid and resistant to correction, and tends to relapse after fusion for two reasons, Firstly, there is an internal torsion and asymmetry of the vertebrae themselves which are largely confined to the neural arch.. . . Secondly, in consequence of the rotation, six secondary deforming factors come into action which inevitably tend to increase the deformity.⁴

Thus, lateral deformity of the spine rarely causes severe clinical problems, but the more common variety of scoliosis, which has a rotation component, often causes disabilities. Ashton Miller's model, at best, addresses the less common type of scoliosis.

A further problem with this animal model is that the dog's quadruped structure results in markedly different forces on the spine than occurs in the biped human. Ashton Miller writes:

Although dog [sic] as a quadruped is a less-than-ideal model for studying the musculoskeletal response of the bipedal spine, there are some redeeming features that make it reasonable to start with. In upright man the primthy loading of the spine is axial compression arising from trunk extensor response to the forward flexion moment due to trunk weight. Any lateral movements arising from the body weight, inertia or muscle action must be equilibrated by muscles lateral to the spine. In dog, the primary loading is compression due to trunk flexor response to the extension moment created by bodyweight. Lateral movements arising from inertial effects or muscle action must be equilibrated by muscles lateral to the spine. So, although the basic spine compression loading may arise for different reasons in the two situations, the lateral movements which can arise in scoliosis must be equilibrated in a similar manner.¹

Although Ashton Miller dismisses the musculoskeletal differences between dogs and humans, erect posture has important implications for human scoliosis. Erect posture puts gravitational force on the spine, which increases the curvature. X-ray evaluation of scoliosis patients always includes erect views, in order to assess the contribution of the gravitational forces to the scoliosis.³

In any case, experimental artifact may invalidate the research results. The "spine compression devices" used to induce scoliosis will, like all foreign bodies, induce scarring of the surrounding tissues, including the muscles. This will likely affect those muscles' functioning. This renders the research results uninterpretable.

Ashton Miller finishes his "Significance" section with the observation, "Finally, but not least, the success of this proposal would launch the investigator on an independent research career in the area of spine biomechanics." Career advancement hardly seems a proper reason for either conducting such research or receiving funds to do so. The history of artificially induced scoliosis in a dog differs markedly from the natural history of human scoliosis. Any data obtained by Ashton Miller will be of questionable relevance to humans.

This project, funded by the Public Health Service, represents an inappropriate use of research dollars.

References

1. Edmonson AS: Scoliosis, in Crenshaw AH (ed): Campbell's Operative Orthopedics, Seventh Edition. St Louis, CV Mosby, 1987.

2. Ashton Miller JA: Muscle activity in congenital and experimental scoliosis. PHS Grant application 85-716-J1, submitted 10/22/84 for funding from 7/1/85 to 6/30/88.

3. Winter RB: The spine, in Lovell WW, Winter RB (eds): Pediatric Orthopedics. Philadelphia, SB Uppincott, 1978, pp 573-584.

4. Roaf R: Rotation movements of the spine with special reference to scoliosis. J Bone Joint Surg 1958;4OB:312-332.

Contents