We investigated 21 pairs of twins for zygosity and

Idiopathic scoliosis in twins studied by DNA fingerprinting THE INCIDENCE AND TYPE OF SCOLIOSIS Masatoshi Inoue, Shohei Minami, Hiroshi Kitahara, Yoshinori Otsuka, Yoshinori Nakata, Masashi Takaso, Hideshige Moriya From Chiba University and Chiba Higashi Hospital, Chiba, Japan We investigated 21 pairs of twins for zygosity and idiopathic scoliosis. DNA fingerprinting confirmed that 13 pairs were monozygotic and eight were dizygotic. There was concordance for idiopathic scoliosis in 92.3% of monozygotic and 62.5% of dizygotic twins. Of the 12 pairs of monozygotic twins concordant for idiopathic scoliosis, six showed discordant curve patterns but eight had differences in Cobb angle of less than 10. Seven of the ten pairs of monozygotic twins had similar back shapes. Our findings suggest that there is a genetic factor in the aetiology of idiopathic scoliosis; they also indicate that there is a genetic factor in both the severity of the curve and the general shape of the back. J Bone Joint Surg [Br] 1998;80-B:212-7. Received 2 January 1997; Accepted after revision 11 September 1997 Idiopathic scoliosis often develops in several members of large families, suggesting that its incidence is influenced by genetic factors. 1,2 Twins with idiopathic scoliosis have been reported, but few series have accurately separated these into monozygotic and dizygotic pairs. We used DNA fingerprinting to determine whether each of 21 pairs of twins was monozygotic or dizygotic. M. Inoue, MD S. Minami, MD H. Moriya, MD, Professor Department of Orthopaedic Surgery H. Kitahara, MD Department of Clinical Radiology School of Medicine, Chiba University 1-8-1, Inohana, Chuo-ku, Chiba 260, Japan. Y. Otsuka Y. Nakata M. Takaso Department of Orthopaedics, National Sanatorium, Chiba Higashi Hospital, Chiba, Japan. Correspondence should be sent to Dr M. Inoue. 1998 British Editorial Society of Bone and Joint Surgery 0301-620X/98/27544 $2.00 Patients and Methods In the 21 pairs of twins, there were 39 females and three males. Their mean age was 15.2 years (9 to 23). Idiopathic scoliosis was assessed by radiological and clinical examination including the forward-bending test and the measurement of leg length. A structural lateral curvature of over 10 by the Cobb method was regarded as diagnostic for this study. DNA from peripheral blood samples was digested with the restriction enzyme Hinf I at 37 C for three hours, and DNA fragments were separated by electrophoresis through a 0.7% agarose gel, with transfer to nylon membranes by standard procedures. 3 A 32 P-labelled oligonucleotide (CAC) 5 (Fresenius AG, Bad. Hamburg, Germany) was used as the fingerprinting probe, and prehybridisation, hybridisation, and washing were performed according to the instructions provided by the manufacturer. 4,5 The membrane was placed in contact with an X-ray film at 80 C overnight. We examined the proportion of concordance for idiopathic scoliosis in both monozygotic and dizygotic twins. In concordant twins, we also evaluated the similarity in the pattern and severity of the curve and the shape of the back. Curve patterns were classified according to the SRS classification. 6 The angle of kyphosis was measured by the Cobb method from T4 to T12 7 since T1 to T3 vertebrae were seldom clearly shown on radiographs. We used the Mann- Whitney test to compare quantitative data from monozygotic and dizygotic pairs of twins. Results There were 13 pairs of monozygotic and eight of dizygotic twins. Concordance for the presence of idiopathic scoliosis was found in 12 of the 13 (92.3%) pairs of monozygotic twins, and in five of the eight (62.5%) pairs of dizygotic twins (Table I). The mean Cobb angle was 31.6 ± 15.6 (SD). Twelve of the 25 monozygotic twins with idiopathic scoliosis (48.0%) had a curvature of more than 30, as did six of 13 dizygotic twins (46.2%). There were 26 patients with a right thoracic curve, eight with a double curve, and four with a thoracolumbar curve. Of the 12 pairs of monozygotic twins who were concordant for idiopathic scoliosis, 212 THE JOURNAL OF BONE AND JOINT SURGERY

IDIOPATHIC SCOLIOSIS IN TWINS 213 Table I. Clinical characteristics of twins with idiopathic scoliosis. DNA fingerprinting confirmed that 12 pairs were monozygotic and five were dizygotic: 92.3% of monozygotic and 62.5% of dizygotic twins were concordant for idiopathic scoliosis Age Cobb angle (degrees) studied Curve Pairs Gender (yr) Concordance* pattern Lateral Kyphotic (T4 to 12) Monozygotic 1 F 22 C Double 51 43-8 R thoracic 68-7 2 F 15 C R thoracic 32 53 Double 31 32 36 3 F 18 C R thoracic 64 19 R thoracic 53 20 4 F 19 C R thoracic 32 38 R thoracic 25 32 5 F 14 C Double 30 20 32 R thoracic 21 32 6 F 13 D Straight 0 23 R thoracic 11 23 7 F 18 C R thoracic 18 26 Double 25 22 28 8 F 14 C R thoracic 25 - R thoracic 27-9 F 12 C R thoracic 80 26 R thoracic 40 15 10 F 11 C R thoracic 42 - Double 31 25-11 F 16 C R thoracic 18 - R thoracic 25-12 F 13 C R thoracic 25 14 R thoracic 28 13 13 F 15 C Double 18 10 28 R thoracic 15 25 Dizygotic 1 F 17 C R thoracic 36 23 F R thoracic 27 29 2 M 17 D Double 32 22 29 F Straight 0 16 3 F 23 C R thoracic 47 - F R thoracic 16-4 F 13 C Double 24 34-2 F R thoracic 48-3 5 M 9 D R thoracic 17 21 M Straight 0 16 6 F 12 D R thoracic 20 28 F Straight 0 15 7 F 16 C R thoracolumbar 20 - F R thoracolumbar 20-8 F 15 C R thoracolumbar 30 17 F R thoracolumbar 16 38 * C, concordant; D, discordant one twin more mature, see text Table II. Severity of lateral curve and back shape of 21 pairs of twins with idiopathic scoliosis. The entries are frequencies of individuals not pairs of twins Lateral Cobb angle (degrees) Kyphotic Cobb angle (degrees) Differences between Differences between Severity pairs of twins Severity pairs of twins <20 20 to 29 30 to 39 40 3 4 to 9 10 <10 10 to 19 20 to 29 30 3 4 to 9 10 Monozygotic 5 8 5 7 4 4 4 2 6 6 6 7 1 2 Dizygotic 3 4 4 2 1 1 3 2 4 5 1 1 2 3 nine were also concordant for treatment. Of these, two had surgery, three received brace treatment, and four were not treated. The other three pairs had received different treatments. In each pair of twins, four of 12 (33.3%) concordant monozygotic twins had differences of less than 3, as did one of five (20.0%) concordant dizygotic twins (Table II). In one hospital, radiographs were used as little as possible and therefore lateral views were seldom available unless the patient had surgical treatment. Of the 21 pairs of twins, 16 pairs were measured for kyphosis. The mean VOL. 80-B, NO. 2, MARCH 1998

214 M. INOUE, S. MINAMI, H. KITAHARA, ET AL kyphotic angle of these patients was 21.7 ± 13.3. Eighteen of the 32 patients (56.3%) with idiopathic scoliosis had kyphosis of 20 to 40. If it is considered that two kyphotic curves differing by less than 3 are concordant, seven of ten pairs (70.0%) of monozygotic twins were concordant while only one of six pairs (16.7%) of dizygotic twins was concordant. This shows that in most monozygotic twins the shape of the back was relatively similar (Table II). In addition, in all three pairs of dizygotic twins who were discordant for idiopathic scoliosis, the twin without scoliosis had less kyphotic curvature than the one with scoliosis. Case Reports Pair 1. A pair of 14-year-old monozygotic female twins was concordant for idiopathic scoliosis (Fig. 1). The Cobb angle was 84 in one and 40 in the other. One twin had Harrington instrumentation and the other was treated in a Milwaukee brace. Pair 2. A pair of 13-year-old monozygotic female twins was discordant for idiopathic scoliosis (Fig. 2). One had a right thoracic curvature of 11 ; the other had a straight spine. The girl with scoliosis was taller than the other, and had reached the menarche while the other had not. It seems possible that the second twin may be affected with scoliosis when she is more mature. Pair 3. A pair of 16-year-old female dizygotic twins was concordant for idiopathic scoliosis (Fig. 3). Their mother considered them to be monozygotic twins, but they were shown to be dizygotic by their patterns of DNA fingerprinting. Despite this they were similar in regard to curve pattern and severity. Discussion Studies on twins are of considerable value in the investigation of genetic and environmental aetiological factors. In entirely genetic disorders, monozygotic twins should be affected similarly since they have all genes in common. In fact, most inheritable disorders are not solely genetic, but also have environmental origins. Monozygotic twins are more often concordant than dizygotic twins, but some will show discordance in some respects. Verschuer 8 first report- Fig. 1a Fig. 1b Radiographs of 14-year-old female monozygotic twins concordant for idiopathic scoliosis. The Cobb angle was 84 in one (a) and 40 in the other (b). One had Harrington instrumentation and the other was treated in a Milwaukee brace. THE JOURNAL OF BONE AND JOINT SURGERY

IDIOPATHIC SCOLIOSIS IN TWINS 215 Radiographs of 13-year-old female monozygotic twins, discordant for idiopathic scoliosis. One had a right thoracic curvature of 11 (a) and the other was straight (b). The girl with no scoliosis was shorter and had not reached the menarche. Fig. 2a Fig. 2b Radiographs of 16-year-old female dizygotic twins concordant for idiopathic scoliosis. One had a right thoracic curvature of 29 (a) and the other of 26 (b). They were similar with regard to curve pattern and severity. Fig. 3a Fig. 3b VOL. 80-B, NO. 2, MARCH 1998

216 M. INOUE, S. MINAMI, H. KITAHARA, ET AL Fig. 4 Reported differences in curve severity in monozygotic and dizygotic twins. Our series and other reported cases showed the significant difference in the severity of the curve between monozygotic and dizygotic twins. The bars indicate the 10th and 90th percentiles, and the box shows the 25th, 50th, and 75th percentiles, respectively. The circles show individuals outliers of less than 10 percentiles in each direction. ed a twin study in 1927, but few reports have accurately separated monozygotic and dizygotic sets of twins. DNA fingerprinting provides accurate identification using parentage testing and twin zygosity assessment, and we have been able to confirm that nearly all monozygotic twins were concordant for idiopathic scoliosis. There are two other reports which have distinguished accurately between monozygotic and dizygotic twins by genotype based on blood type. Fisher and De George 9 reported that all of six pairs of monozygotic twins were concordant for idiopathic scoliosis. Only six of eight pairs of dizygotic twins were concordant, but if a curvature of more than 10 was considered, only three of the eight pairs were concordant. Oguma, Kaneda and Kikoshi 10 found that eight of nine pairs of monozygotic twins were concordant for idiopathic scoliosis as against none of three pairs of dizygotic twins. Oguma et al 10 examined 13 blood types, considering that the identity of all 13 in both twins gave an accuracy of 99.6% when using (CAC) 5 as a probe, confirming that all band patterns were present in both twins. If these two studies are considered together with our series, there were 26 concordant pairs (92.9%) in 28 pairs of monozygotic twins with idiopathic scoliosis, but only eight of 19 pairs (42.1%) of dizygotic twins were concordant. In the single discordant pair of monozygotic twins (our pair 3), the twin without scoliosis was less mature than the one with scoliosis and we considered that they would probably become concordant for scoliosis in the future. In addition, the other reported pair of monozygotic twins without concordance was 11 years old and could have become concordant after bone maturity. The general findings indicate a strong genetic aetiology for idiopathic scoliosis. Other researchers have suggested a genetic component in idiopathic scoliosis because of the similar curve pattern in twins concordant for scoliosis. Cowell, Hall and Mac- Ewen 11 reported that the convexity is right-sided in almost 90% of patients with a thoracic curvature: 29 of the 38 individuals had a right thoracic curvature, and the curve patterns were clearly discordant not only between the twins in a set, but also in most of the other patients. Cowell et al 11 also reported that the average difference between the severity of the curve in monozygotic twins was less than that between dizygotic twins. Our study and the literature review indicate that there are marked differences in curve severity between monozygotic and dizygotic twins based on non-parametric statistical analysis (Mann-Whitney test, p < 0.05) (Fig. 4). This provides evidence for the genetic basis of curve severity in scoliosis. Riseborough and Wynne-Davies 2 considered that curve severity depended on a number of factors, such as the age at which it first develops, the speed of development and the efficacy of treatment. Our findings confirm that some individuals have a genetic predisposition to curve development. Our observation that dizygotic twins concordant for idiopathic scoliosis tended to be dissimilar in curve severity, indicates that the disease gene causing idiopathic scoliosis and the candidate gene for the development of scoliosis are different. Archer and Dickson 12 and Dickson et al 13 suggested that the familial trend in idiopathic scoliosis was explained by the genetically determined shape of the spine in the sagittal plane. Carr, Jefferson and Turner-Smith 14 reported that flat backs were associated with a marked genetic susceptibility to adolescent idiopathic scoliosis. Inoue 7 studied 40 normal individuals and found that the mean kyphotic angle from T4 to T12 was 28.2 ± 6.7. In our series, we were unable to find differences in flat back between affected and normal twins. Furthermore, our observation that individual discordant twins without scoliosis had less kyphotic curvature than those with scoliosis conflicts with the assumption that a flat back confers a marked genetic susceptibility to idiopathic scoliosis. Despite this, the similar back shapes in monozygotic twins seem to indicate that there is a genetic factor in the determination of the shape of the back and further investigations are needed. We thank Makoto Matsubara, Keiichi Funazu, and Keijiro Isobe for referring some of the twins to us. No benefits in any form have been received, or will be received, from a commercial party related directly or indirectly to the subject of this article. THE JOURNAL OF BONE AND JOINT SURGERY

IDIOPATHIC SCOLIOSIS IN TWINS 217 References 1. Wynne-Davis R. Familial (idiopathic) scoliosis: a family survey. J Bone Joint Surg [Br] 1968;50-B:24-30. 2. Riseborough EJ, Wynne-Davies R. A genetic survey of idiopathic scoliosis in Boston, Massachusetts. J Bone Joint Surg [Am] 1973; 55-A:974-82. 3. Sambrook J, Fritsch EF, Maniatis T. Molecular cloning. 2nd ed. New York: Cold Spring Harbor Laboratory Press, 1989. 4. Jeffreys AJ, Wilson V, Thein SL. Hypervariable minisatellite regions in human DNA. Nature 1985;314:67-73. 5. Schafer R, Zischler H, Epplen JT. (CAC) 5, a very informative oligonucleotide probe for DNA fingerprinting. Nucleic Acids Res 1988; 16:5196. 6. Lonstein JE, Bradford DS, Winter RB, Ogilvie JW. Moe s textbook of scoliosis and other spinal deformities. 3rd ed. Philadelphia: W. B. Saunders Company, 1996. 7. Inoue K. Radiographic study of sagittal plane alignment of scoliosis. J Jpn Orthop Ass 1985;59:505-16. 8. Verschuer O von. II. Die vererbungsbiologische Zwillingsforschung. Ihre Biologischen Grundlagen: studien an 102 eineiigen und 45 gleichgeschlechtlichen zweieiingen Zwillings - und an 2 Drillingspaaren. Ergebn inn Med Kinderheilk 1927;31:35. 9. Fisher RL, De George FV. A twin study of idiopathic scoliosis. Clin Orthop 1967;55:117-26. 10. Oguma T, Kaneda K, Kikoshi T. A twin study of idiopathic scoliosis. Clin Orthop Surg (Japanese) 1983;18:228-35. 11. Cowell HR, Hall JN, MacEwen GD. Genetic aspects of idiopathic scoliosis. Clin Orthop 1972;86:121-31. 12. Archer IA, Dickson RA. Stature and idiopathic scoliosis. J Bone Joint Surg [Br] 1985;67-B:185-8. 13. Dickson RA, Lawton JO, Archer IA, Butt WP. The pathogenesis of idiopathic scoliosis: biplanar spinal asymmetry. J Bone Joint Surg [Br] 1984;66-B:8-15. 14. Carr AJ, Jefferson RJ, Turner-Smith AR. Familial back shape in adolescent scoliosis: a photogrammetric population study. Acta Orthop Scand 1991;62:131-5. VOL. 80-B, NO. 2, MARCH 1998