Progressive Congenital Spinal Deformity

K.S. is a 3–year, 10–month–old girl with multiple congenital anomalies of her thoracolumbar spine. She was first evaluated at age 1 year, 3 months when her left shoulder was noted to be lower than the right. Her birth history was notable only in that she was born by Cesarean section due to extended labor. There is no family history of spinal congenital anomalies. At the initial evaluation, radiographs revealed a mid–thoracic right hemivertebra with probable fusion of ribs on the left side in the same region, creating a 21° right thoracic curve. A second hemivertebra on the left at the thoracolumbar junction and a third hemivertebra between L3 and L4 were also noted. Initial management consisted of observation with interval radiographic studies. At 2 years, 5 months of age, her right shoulder was 2 cm higher than the left and radiographs demonstrated a 38° right thoracic curve. At age 3 years, 8 months the right upper thoracic curve had increased to 43° (Fig 1).

To better delineate the complex anatomy of her congenital anomalies, a CT scan of the thoracolumbar spine with 3–D reconstruction was obtained (Fig 2a, b). This demonstrated a hemivertebra at approximately the T7 level, and another hemivertebra at the T11 level. A third unincarcerated hemivertebra was noted at the T12–L1 junction on the left (opposite side from hemivertebra noted above) (Fig 3). Congenital fusion of ribs at the T6 through T8 level on the left was clearly identified. Also, an isolated hemivertebra was noted at the L3 level. Management options include:
1. Continued observation.
2. Application of a thoracolumbar spinal orthosis.
3. In–situ fusion of the levels of progressive deformity to prevent curve progression.
4. Hemi–epiphysiodesis of the convexity of the deforming regions both anteriorly and posteriorly, to stabilize the curve and perhaps achieve partial correction with growth.

Fig. 1	Fig. 2a	Fig. 2b	Fig. 3
Click on thumbnail for full view.

Progressive Congenital Spinal Deformity: Discussion

Keith H. Bridwell, MD
Washington University School of Medicine St. Louis, MO

The authors have documented progression from a 21° curve to a 43° curve. The pathology of the congenital curve with an apex at T7 appears to be the following: On the left side there is an unsegmented bar with three ribs fused together. On the right side there appears to be a large congenital block. At the top of that block is a hemivertebra which appears to be fused into two segments below. This block is taller on the right side than the left.

The most malignant combination of congenital deformities is that of an unsegmented bar on the concave side and unincarcerated hemivertebra on the convex side.3,5 Because the hemivertebra seems to be fused into the two segments below in this case, we are not dealing with such a malignant combination. No sagittal x–rays have been provided. If the spine is in kyphosis, this suggests that the hemivertebra is the principal deforming force. If the spine is in lordosis, this usually suggests that the unsegmented bar is the principal deforming force.

There are three hemivertebra below the area of principal deformity. Two are at the thoracolumbar junction and the third one is in the mid–lumbar spine. Two of those hemivertebrae are incarcerated and the left hemivertebra at T12/L1 is unincarcerated. The hemivertebra at T12/L1 probably has the greatest potential for progression in the future. But to date, the three hemivertebrae below seem to balance each other out. Continued observation is therefore appropriate for them.

Clearly for the major deformity centered on T7, continued observation and application of orthosis will not be helpful. The spine is growing faster on the right side than the left, and the deformity is starting to wedge the growth plates just above and below.

There appear to be open disc spaces and, therefore, open growth plates at the few segments above and below the principal congenital deformity. The best treatment would be a hemi–epiphysiodesis of the convexity, both anteriorly and posteriorly. By including one or two segments below and two or three segments above the area of principle deformity, we will hopefully slow down the convexity and allow the concavity to catch up. The main potential for "catch–up" will be the included fusion segments immediately above and below the congenital block. The greatest potential for correction with growth exists if the procedure is performed in a patient under the age of five years.1,4

Hopefully, with this approach, we would not see too much "adding on" to the deformity. With a posterior fusion alone, we would not have any hope of seeing correction of the deformity and it would be more likely that we would see "adding on" from continued wedging of the segments above and below. With an unsegmented bar on one side and essentially a block vertebra on the other side, the potential for crankshaft should not be as great as in cases that are purely caused by failure of formation. However, in recently reviewing our patients2 who experienced crankshaft phenomenon, we did find that there was a significant incidence of crankshaft in congenitals with the same combination that this patient has.

In conclusion, based on the available information, I would recommend a convex anterior and posterior hemi–epiphysiodesis. T6 to T8 should be included, and I would probably include three or four segments above and one segment below that block.

References
1. Andrew T, Piggott H. Growth arrest for progressive scoliosis. Combined anterior and posterior fusion of the convexity. J Bone Joint Surg. 1985;67B(2):193–197.

2. Hamill CL, Bridwell KH, Lenke, LG, et al. Posterior arthrodesis in the skeletally immature patient: Assessing the risk for crankshaft. Is an open triradiate cartilage the answer? Abstract submitted March 1995 to the Scoliosis Research Society.

3. McMaster MJ, Ohtsuka K. The natural history of congenital scoliosis. A study of two hundred and fifty–one patients. J Bone Joint Surg. 1982;64A(8):1128–1147.

4. Winter RB, Lonstein JE, Denis F, Sta–Ana de la Rosa H. Convex growth arrest for progressive congenital scoliosis due to hemivertebrae. J Pediatr Orthop. 1988;8:633–638.

5. Winter RB, Moe JH. The results of spinal arthrodesis for congenital spinal deformity in patients younger than five years old. J Bone Joint Surg. 1982;64A(3):419–432.

Progressive Congenital Spinal Deformity: Discussion

Maurice Bergoin, MD
Department of Pediatric Orthopaedics
CHU Nord, Marseille, France

This complex congenital spine deformity must be divided into three problems.

I. The Upper Right Thoracic Curve
In this area we observe one hemivertebra on T7 on the right side, as well as rib fusion on the left side from T6 through T8 and from T3 to T5. This deformity looks like a unilateral unsegmented bar with a controlateral hemivertebra as described by Nasca in 1975. The unsegmented vertebrae are suspected on the 3–D reconstruction.

The natural history of the deformity usually indicates a progression of 8š to 10š each year. In this case the progression between 1 year, 3 months and 3 years, 8 months is 22°. Waiting with continued observation is contraindicated because this deformity is one of the more progressive.

Rib resections and osteotomies of the bar itself are hazardous and are not able to induce growth on the concave side. The safer solution is to reduce the growth on the convex side by a hemi–epiphysiodesis of the convexity, including four vertebrae above and three vertebrae under the T7 right hemivertebra. The hemi–epiphysiodesis should be performed by anterior and posterior approaches (in one or two stages). A short–compression instrumentation on the convex side during the posterior approach should be considered.

A post–operative cast is necessary for six months. We recommend a partial wearing of a Milwaukee™ brace.

II. The Thoracolumbar Area
There are two alternate hemivertebrae: one on T11 on the right side and one between T12 and L1 on the left side. At present the two hemivertebrae appear balanced but this will not continue.

The hemivertebra on T11 is semi–segmented and the progression on this level is usually very small. The hemivertebra between T12 and L1 is fully segmented in this area (thoraco&endash;lumbar junction). The progression is usually important.

When hemiepiphysiodesis is obtained on the upper thoracic level (one year later) it will be necessary by a simultaneous approach to fuse in situ the left thoracolumbar hemivertebra anteriorly and posteriorly. The lower rib can be resected and used as a bone graft. This must be followed by cast immobilization for six months.

The left thoracolumbar hemivertebra is surgically altered to an unsegmented hemivertebra, to induce a balanced growth with the right T11 hemivertebra. No surgery is required on the T11 hemivertebra.

III. Hemivertebra Between
L3 and L4 on Left Side

On the x–ray we are not able to determine if it is a semi–segmented or a full–segmented hemivertebra. Currently there is no deformity and no need for early surgery. Observation is appropriate, and perhaps in the future a hemivertebral in situ fusion will be required.

Conclusion
All management options are valuable on different levels:

Hemiepiphysiodesis on the right upper thoracic curve;

In situ fusion on left thoracolumbar hemivertebra;

Spinal orthosis (Milwaukee brace™) after surgery during the growth; and

Continued observation of the lumbar left hemivertebra.

References
1. Bergoin M, Bollini G. Spinal surgery before age 5 years. Editorial Round Table. J Pediatr Orthop. 1993(pt B):1:93–94.

2. Bergoin M, Bollini G, Gennari JM, One stage hemivertebral excision and arthrodesis on congenital oblique take off in children aged less than five years. J Pediatr Orthop. 1993(pt B);1:108–112.

3. Bollini G, Bergoin M, and Coll., Hemivertebrae excision and fusion in children aged less than five years. J Pediatr Orthop. 1993(pt B);1:95–101

4. Dubousset J, Katti E, Serine R. Epiphyseodesis of the spine in young children for congenital spinal deformations. J Pediatr Orthop. 1993(pt B);1:292;123–130.

5. McMaster M, and Ohtsuka K. The natural history of congenital scoliosis. J Bone and Joint Surg. 1982;64A:1128–1147.

6. Nasca RJ, Stelling FH, Steek HH. Progression of congenital scoliosis due to hemivertebrae and hemivertebrae with bars. J Bone and Joint Surg. 1975;57A:45–46.

7. Winter RB, Moe, JH. The results of spinal arthrodesis for congenital spinal deformity in patient younger than five years old. J Bone and Joint Surg. 1982;64A:419–432.