The double rib contour sign (DRCS) in lateral spinal radiographs. Aetiologic

implications for scoliosis Theodoros B GRIVAS, Spyros DANGAS, Basil D POLYZOIS, Panagiotis SAMELIS

Orthopaedic Department, “Thriasio” General Hospital G. Genimata Avenue, Magula, 19600 Greece

E-mail: tgri69@otenet.gr

Abstract: All lateral spinal radiographs in idiopathic scoliosis show a DRC sign of the thoracic cage, a radiographic expression of the rib hump. The outline of the convex overlies the contour of the concave ribs. The aim of this study is to assess this DRC sign in children with and without late onset idiopathic scoliosis (LOIS) with 10 -20 Cobb angle, and to examine whether in scoliosis the deformity of the thorax or that of the spine develops first. Methods and material. The radiographs of 133 children referred to hospital in a school screening study were examined. There were 47 boys and 86 girls, 13.28 and 13.39 years old respectively. The Cobb angle was measured and the radiological lateral spinal profile (LSP) was appraised from an angle made by a line drawn down the posterior surface of each vertebral body (T1-L5) and by the vertical. The children, boys and girls, were divided in 5 groups, namely: 1) with straight spines, 2) with spinal curvature having a Cobb angle <10, 3) with thoracic, 4) thoracolumbar and 4) lumbar curves 10-20. For quantification of the DRC sign, the “rib index” was defined as d1/d2 ratio, where d1 expresses the distance from the most extended point of the most projecting rib contour (RC) to the posterior margin of the corresponding to that point vertebra and d2 the distance from the posterior margin of the same vertebra to the most protruding point of the least projecting RC. In a symmetric and non-deformed thorax these two RC lines are superimposed and the “rib index” is 1. Results: The statistical descriptives of d1, d2 are presented for boys and girls together because they are not statistically different. There are no sex differences of the “rib index” which is 1.45, 1.51, 1.56, 1.59, 1.47 for the 5 aforementioned groups respectively. According to statistical analysis there is no correlation of the Cobb angle to the “rib index” of thoracic, thoracolumbar and lumbar scoliosis groups. The DRC sign is present in all referrals and scoliotics. The data show a correlation of the “rib index” to each of T2, T3, T4, T5, T6 and T7 LSP in girls with lumbar curvatures. Discussion: The DRCS results primarily due to rib deformation and secondarily to vertebral rotation, because DRCS could be present in straight spines with no vertebral rotation. In all our school-screening referrals, (having ATI 7°), the thorax deformity, in terms of the DRC sign, has already been developed. 70% of these children were scoliotics. The rest had a curvature of less than 9° of Cobb angle (10%) or they were children with straight spines (20%) who were followed due to the existing rib hump. The non-scoliotics were 1,5-2 years younger than the ones who had already developed scoliosis, and they had both approximately a “rib index” of 1,5. The DRC sign is present in all referrals; in contrary there is no scoliotic spine without it, as the DRC sign is always present in scoliotic lateral spinal radiographs with no exception. This observation supports our hypothesis that in idiopathic scoliosis the deformity of the thorax develops first and the deformity of the spine succeeds.

1. Introduction

The scoliotic thoracic cage, in lateral spinal radiographs at the standing position presents the double rib contour sign (DRCS). This is the consequence of the following: The rib outline of the convex hemithorax protrudes and delineates separately from the outline of the concave hemithorax, which is overlain by the former, (figure 1). The DRCS is present not only in the spinal radiographs of scoliotic children but in children referred from school-screening with straight spines as well. The DRC sign is present in all our referrals, in contrary there is no scoliotic spine without it, as the DRC sign is always present in scoliotic lateral spinal radiographs with no exception.

In a normal and symmetric thorax, these two outlines almost coincide. This observation caused us to hypothesize the following: in the patho-biomechanics of the

spine the deformation of the thoracic cage is observed first and the deformation of the central axis, namely the spine, succeeds.

The aim of this report therefore is the study of this double contour sign of the hemithoraces in the lateral spinal radiographs in children with and without late onset idiopathic scoliosis 10°-20° of Cobb angle, who were referred from the school screening program. The referred children had an ATI of 7°. To our knowledge there is no similar study in the literature dealing with the DRCS and its relevant aetiological implications. 2. Material and Method

2.1 The examined children. The spinal radiographs of the referred children from the school-screening program were studied. All the referred children had a scoliomerer reading of ATI 7°. They were 47 (35.3%) boys, and 86 (64.7%) girls, a total of 133 children. The boys had a mean age of 13.28 (range 9-18) years and the girls 13.39 (7-18) respectively. The statistical analysis between the ages of the boys and the girls showed that they were not statistically significantly different, so the sample of the study comprises as far as age is concerned a homogenous population (independent Samples T-test). The children were divided in five groups. Group 1. 27 children, (12 boys [44.4%] and 15 girls [55.6]), without scoliosis, with straight spines. Group 2. 13 children, (6 boys [46.2%] and 7 girls [53.8]), without scoliosis but with curves of 1-9 Cobb angle. Group 3. 47 children, (17 boys [36,2%] and 30 girls [63.8%]) with a thoracic scoliosis of 10°-20° Cobb angle. Group 4. 14 children (4 boys [28.6%] and 10 girls [71.4%]) with thoracic lumbar scoliosis of 10°-20° Cobb angle. Group 5. 28 children (7 boys [25%] and 21 girls [75%]), with lumbar curves 10°-20°.

2.2 The measurements. With the aid of a goniometer the lateral spine profile was appraised, using the spinal radiographs in standing position, from an angle made by a line drawn down the posterior surface of each vertebral body (T1-L5) and by the vertical.

For quantification of the deformation of the thoracic cage the “rib index” was calculated. This index was defined as a ratio of two distances, that is the d1/d2 ratio. d1 expresses the distance between the most extended point of the most projecting rib contour (RC) and the posterior margin of the corresponding to that point vertebra. d2 expresses the distance between the posterior margin of the same vertebra and the most protruding point of the least projecting RC, (figure 2). In a symmetric and not deformed thorax these two RC lines are superimposed and the “rib index” is 1.

2.3 The statistical analysis was performed using various statistical tests with the aid of SPSS-PC package. The statistical techniques used were frequencies, descriptives including mean, range, minimum, maximum, standard deviation, the Kruskal-Wallis and Mann-Whitney tests and correlations using the Pearson Correlation coefficient.

3. Results The spinal profiles of the 5 groups of the children were presented elsewhere, [2]. In the tables 1-5 the descriptives of d1, d2 and of the “rib index” of the examined children are presented. Moreover, in Table 6 the correlation of the spinal profile to the “rib index” for the thoracic (1), thoracolumbar (2), and lumbar (3) curves, scoliosis 10°-20° (Pearson Correlation), is presented, for boys and girls respectively.

figure 1 figure 2

d2

d1

Table 1. Descriptives of d1, d2 and of the “rib index” of 27 children, 12 boys [44.4%] and 15 girls [55.6%] without scoliosis, with straight spine. Variable N Minimum Maximum Mean Std. Deviation age 20 7,00 18,00 12,0500 2,9465 rib index 26 1,00 2,30 1,4517 , 2499 d1 26 18,00 60,00 41,2308 10,6294 d2 26 15,00 47,00 28,6538 7,2219 Table 2. Descriptives of d1, d2 and of the “rib index” of 13 children, 6 boys [46.2%] and 7 girls [53.8%], without scoliosis, but with 1 - 9 Cobb angle. Variable N Minimum Maximum Mean Std. Deviation age 13 7,00 18,00 12,4615 2,9330 rib index 13 1,26 1,90 1,5149 , 1957 d1 13 25,00 58,00 41,3846 9,5003 d2 13 15,00 38,00 27,5385 6,2131 Table 3. Descriptives of d1, d2 and of the “rib index” of 47 children, 17 boys [36,2%] and 30 girls [63.8%] with thoracic scoliosis of 10°-20° Cobb angle. Variable N Minimum Maximum Mean Std. Deviation age 42 8,00 17,00 14,0000 2,5567 rib index 46 1,11 2,20 1,5606 , 2994 d1 46 28,00 66,00 47,1087 9,8007 d2 46 15,00 47,00 30,9783 7,3560 CobbThoracic 47 10,00 22,00 13,0851 3,0348 Table 4.Descriptives of d1, d2 and of the “rib index” of 14 children, 4 boys [28.6%] and 10 girls [71.4%] with thoracolumbar scoliosis of 10°-20° Cobb angle. Variable N Minimum Maximum Mean Std. Deviation age 13 9,00 18,00 13,8462 2,4781 rib index 14 1,19 2,16 1,5913 , 3181

d1 14 40,00 80,00 51,3571 10,7102 d2 14 22,00 42,00 32,9286 6,4267 CobbThoraco- 14 10,00 19,00 11,0000 6,1644 lumbar Table 5. Descriptives of d1, d2 and of the “rib index” of 28 children, 7 boys [25%] and 21 girls [75%], with lumbar curves of 10°-20° Cobb angle. Variable N Minimum Maximum Mean Std. Deviation age 28 8,00 18,00 13,6429 2,1979 rib index 28 1,13 2,42 1,4774 , 3004 d1 28 35,00 63,00 47,4286 6,9251 d2 28 23,00 46,00 32,9286 6,3882 CobbLumbar 27 10,00 24,00 14,3704 3,9336 Table 6. Correlation of the spinal profile to the rib index for the thoracic (1), thoracolumbar (2), and lumbar (3) curves, scoliosis of 10°-20° Cobb angle (Pearson Correlation), for boys/girls respectively, (e.g. NSS/NSS means NSS for boys and NSS for girls). The figures present r while the figure in the parenthesis the statistical significance p (* p< 0.05, **p< 0.001). NSS = not statistically significant. Vertebra (1) ♂/♀ (2) ♂/♀ (3) ♂/♀ T1 NSS/NSS NSS/NSS NSS/NSS T2 NSS/NSS NSS/NSS NSS/0.492(0.045) T3 NSS/NSS NSS/NSS NSS/0.520(0.027) T4 NSS/NSS NSS/NSS NSS/0.513(0.025) T5 NSS/NSS NSS/NSS NSS/0.556(0.013) T6 NSS/NSS NSS/NSS NSS/0.582(0.007) T7 NSS/NSS NSS/-.654(0.04) NSS/0.464(0.03) T8-12 NSS/NSS NSS/NSS NSS/NSS L1-5 NSS/NSS NSS/NSS NSS/NSS

No statistical difference for the rib index was observed among the 5 studied groups of children, (Kruskal-Wallis test). Similarly no statistical difference was observed for the rib index between sex, (Mann-Whitney test), among the 5 studied groups of children. No statistical significant correlation (Pearson Correlation coefficient) of the Cobb angle with the rib index for the thoracic, thoracolumbar and lumbar curves, (scoliosis 10°-20°) in boys and girls was observed either. 4. Discussion

The surface back contour asymmetry is considered very important for the selection of children at risk to develop scoliosis. Traditionally the existence of a hump constitutes the main indicator for referral during School-screening for scoliosis, because the rib hump reflects the rib-cage deformation.

The role of rib asymmetry, [1,6,9] and the importance of rib cage in the aetiology of idiopathic scoliosis, [3,4,5,7,8], has previously been examined.

The DRCS is mainly produced due to the deformation of the ribs (asymmetric growth), and to the subsequent deformation of the vertebrae. It is very interesting to note that this sign is present in thoraces with straight spines and in our scoliotics with not only thoracic or thoracolumbar but with lumbar curves as well. The predominant role of rib deformation is also confirmed in the transverse sections of CT scans of cases with small scoliotic curves. In these cases the deformation of the

thoracic cage expressed as asymmetrical and deformed ribs of both hemithoraces is present, but it is not accompanied by similar severity to ribs vertebral deformation.

Our results show that the mean rib index for the 5 groups studied is 1,45, 1,51, 1,56, 1,59 and 1,47 respectively. This means that the protruding hemithorax in its most projecting spot maintains the ratio to the overridden one almost constant. Therefore it is obvious that in all referred children for scoliosis in OPD with ΑΤΙ7 the deformation of the thoracic cage has started. It is noticeable that from a total of 133 children referred for radiological examination, almost 70% had scoliosis. The rest of the children (10 %) had curves with Cobb angles 1-9 or a straight spine (20%), and are followed up due to their significant hump. These latter children were 1,5-2 years younger than the children who had already developed scoliosis and had, like the scoliotics, rib index of approximately 1,5. More specifically there were a) 27 straight spine children (20,3%) with a mean age of 12 years, b) 13 minor curve 1- 9 children (9,77%), with a mean age of 12, 46 years, c) 47 thoracic curve 10- 20, children (35, 3%) with a mean age of 14 years, d) 14 thoracolumbar curve 10-20 children (10, 52%) with a mean age of 14 years and e) 28 lumbar curve 10-20, children (21%) with a mean age of 13, 4 years. The lack of difference that was observed between rib index and both the sex of the studied groups and the Cobb angle, shows that this DRC sign is a common characteristic of all the referred children, while not all of them have scoliosis. In contrary there is no scoliotic spine without it, as the DRC sign is always present in scoliotic lateral spinal radiographs with no exception. In a normal and symmetric thorax these two hemithoraces rib contour outlines coincide.

Conclusively this study supports our hypothesis, that in scoliosis, the thoracic cage deformation precedes and then the deformation of the central axis, namely the spine, follows.

The correlation pattern of the rib index in the levels T2, T3, T4, T5, T6 and T7 with the lateral spinal profile measured in the lateral radiographs of girls with lumbar curves is also interesting. This shows that there is a special mechanism, which correlates the deformation of the thoracic cage to the spinal profile in these curves. This correlation is very strong for T6, namely in the middle thoracic spine, while scoliosis is in the lumbar. This finding requires more research. References [1]Agadir, M., Sevastik, B., Sevastik, J.A., Persson, A. and Isberg, B. (1988). Induction of scoliosis in the growing

rabbit by unilateral rib-growth stimulation. Spine 13, 1065-1069. [2]Grivas Th B, Dangas S, Polyzois DG. The radiological lateral spinal profile of referred from school - screening

children without and with late onset scoliosis 10°-20°. 25th “N. Giannestras – P. Smyrnis” Spinal Meeting, Hotel Porto Rio, Patras , Hellas, 21-23 May 1999. Abstract book p 38.

[3]Grivas TB, Burwell RG, Purdue M, Webb JK and Moulton A (1990-1992): The rib-cage deformity in infantile idiopathic scoliosis-the funnel-shaped upper chest in relation to specific rotation as a prognostic factor. An evaluation of thoracic shape in progressive scoliosis and control children during growth. In Proceedings of VIth International Symposium on Surface Topography and Spinal Deformity. Hotel Estoril Eden, 19-20 September, ed. A. Alberti. Lisbon: Faculdade Ciencias Medicas 1990. In: Surface Topography and Spinal deformity VI, Alberty, Drerup, Hierholzer (ed), Gustav Fischer Verlag, Stuttgart, Jena, New York, 1992, pp. 93-109.

[4]Grivas TB, Burwell RG, Purdue M, Webb JK (1991): A segmental analysis of thoracic shape in chest radiographs of children. Changes related to spinal level age sex side and significance for scoliosis. J Anat, 178, 21-38.

[5]Grivas TB, Burwell RG, Purdue M, Webb JK and Moulton A (1992): Segmental patterns of rib-vertebra angles in chest radiographs of children. Changes related to rib level, age, sex, side and significance for scoliosis. Clin Anat, 5(4): 272-288.

[6 ]Pal GP (1991) Mechanism of production of scoliosis: a hypothesis, Spine 16, 288-292. [7] Sevastik, J.A., Aradir M, Sevastik, B (1990). Effects on the rib elongation on the spine. II Correction of scoliosis in the rabit. Spine 15, 826-829.

[8] Sevastik, J.A., Aaro, S., Lindholm, S.T. and Dalhborn, M. (1987). Experimental scoliosis in growing rabbits by operations on the rib cage. Clinical Orthopaedics 136, 282-286.

[9]Stokes, I.A.F., Dansereau, J. and Moreland, M.S. (1989). Rib cage asymmetry in idiopathic scoliosis. In: Combined Meeting of Scoliosis Research Society and European Spinal Deformities Society, Amsterdam, 17-22 September.