Heritability

Although scoliosis has been recognized for centuries, its familial nature was not noted until the late nineteenth century.1 The first credible evidence supporting a familial bias of scoliosis was reported in the 1930s.2– 4 Although not definitive, this early work was important because it initiated systematic investigation of the hereditary nature of AIS and established three general categories of its study that continue today: (1) studies of single, multigenerational families; (2) studies of relatives of index patients or probands; and (3) twin studies.

Perhaps the most significant early study of a single, multigenerational family with idiopathic scoliosis (?not due to neurological or congenital causes?) was reported by Garland in 1934.3 The mode of transmission of the disease in this five-generation family was described as father to son over the first three generations and then through both sexes over the last two generations. With approximately half of the offspring in the latter two generations being affected with scoliosis, a dominant Mendelian mode of inheritance was proposed. Faber, in 1936, published the first population study of scoliosis, involving 660 patients with what was initially thought to be rachitic scoliosis.2 The patients included in this study, in the author?s conclusion, probably had AIS. More than 200 of these patients were found to have family members with scoliosis, with 7% of siblings and 14% of parents affected. Faber also proposed a dominant mode of inheritance for scoliosis In the first twin study of the disease, described in 1933, Nitsche and Armknecht reported on several twin pairs, some of whom shared the diagnosis of scoliosis whereas others did not.4

Over the remainder of the twentieth century, the accumulation of data in each of the three study categories named above has led to general acceptance of the conept that hereditary factors contribute to the transmission of AIS. Family and population studies by Harrington,5 Wynne-Davies, 6 Riseborough and Wynne-Davies, 7 Cowell et al8 Filho and Thompson, 9 and others10 support a greater incidence of idiopathic scoliosis in families than in the general population. These studies found that up to a third of the members of families in which the disease recurs are affected with scoliosis. These studies documented an increased incidence of scoliosis in first-degree relatives of index patients (up to 11%) as compared with second- and third-degree relatives (up to 3.7% and 1.6%, respectively), as is characteristic of polygenic traits. Although all of these studies support a genetic basis for idiopathic scoliosis, they are fraught with potential limitations including the size of the study population, family size, genealogical resources, false paternity, clinical and genetic heterogeneity, and diagnostic accuracy.

Our group sought to overcome many of these limitations by initially studying the heritability of AIS with a unique population database for Utah and the Intermountain West region of the United States. Known as GenDB, this database derives its power from multiple factors, including a massive size (21 million birth, death, and marriage records), a population that has been geographically stable for over 150 years, extensive genealogical resources, the largest families in the developed world over multiple generations, and low false-paternity rates. Furthermore, the database does not represent a population isolate but rather a group that is outbred and representative of the general population of the United States. Using GenDB, our group demonstrated that AIS had familiar origins for 97% of 145 patients or probands. Kinship coefficients for the scoliosis patients had standard deviations threefold greater than those of controls, and a substantial founder effect was demonstrated. More than 50% of probands or individual families were connected by a common ancestor in sixteenth-century England.11 One group of 14 previously unconnected families created an extended pedigree with a common ancestor in Essex, England, ca. 1520 AD, and another 17-family group had a common ancestor in Kent, England, ca. 1560 AD ( Fig. 28.1 ). This statistically well-powered study further supports the familial nature of AIS.

Twin studies have consistently supported the familial nature of AIS, 10,12,13 demonstrating higher concordance in monozygotic (up to 73%) than in dizygotic (up to 36%) twins. However, the usefulness of twin studies in distinguishing between environmental and heritable effects in complex disorders has recently been questioned. It is now apparent that many monozygotic twins are not identical, eitherphe-notypically or genotypically, and have major differences often evident in birth weight, the presence of genetic disease, and congenital anomalies. Although certain postzy-gotic events that lead to discordance in monozygotic twins are well understood (e.g., mosaicism, X-chromosome inactivation), other factors, resulting in discordance for lateral asymmetries (e.g., handedness, situs inversus), major malformations (e.g., vertebral anomalies alone or in combination with the VATER association of v ertebral abnormalities, an imperforate a nus, [c ardiac anomalies], t rans e sophageal fistula, r enal anomalies, and l imb anomalies), and fetal growth (e.g., pla-cental vascular anatomy) are poorly understood.14,15 Thus, although twin studies are somewhat helpful in supporting the familial nature of AIS, the complexities of twinning make these types of studies less than ideal for defining the genetics of a structural asymmetry such as that in AIS.

Multiple modes of inheritance for AIS have been proposed over the years.6– 8,11,16– 21 For example, in some families AIS follows simple mendelian inheritance patterns, whereas in others it folows more complex patterns. Still, most data suggest that the inheritance of AIS is polygenic and multifactorial.

Establishing the heritability of AIS has been an arduous process involving contributions from multiple investigators over the past century. Although the clinical impact of this body of work has been small, perhaps serving only to increase vigilance in screening patients? family members for scoliosis, the overall impact of this research has been substantial in that it has focused efforts at centers around the world on defining the genetic etiology of this condition.

Gene Search

The search for specific genes associated with AIS has been greatly aided by recent advances in genetic technology, statistics, and computers, as well as by completion of the mapping of the human genome.22,23 Whereas previous investigations of AIS had relied primarily on a candidate-gene approach alone, current investigations, using newer technologies, have used a genome-wide assessment based on linkage analysis, genetic association studies, or both. Multiple approaches are often combined in a complimentary fashion to improve the chance of success and validate the results with particular genetic techniques.

The standard candidate-gene approach is one way of identifying genes involved in a disease process. This type of analysis depends on genes with known protein products that appear relevant to the physiological basis of the disease and that may be investigated individually in patients or the families of patients with the disease. Through a process of trial and error, genes known to be associated with a specific disease process in animals or thought to be related to the underlying biology or physiology of the disease process can be individually evaluated to determine their association with the condition. Unfortunately, this approach has found limited success in defining genes for AIS17,18,20,21,24,25 probably because of poor understanding of the biology and physiology of this particular idiopathic source of spinal asymmetry and the lack of an animal model for it. However, a candidate-gene approach can be used in combination with other approaches that identify particular areas on the human genome that are associated with AIS.

The mapping of the human genome, as well as advances in genetic technology, now allow screening of the entire genome for one or more genes. In contrast to the more limited standard candidate-gene approach, both linkage analysis and genome-wide association studies (GWAS) allow screening of the full genome with known genetic markers. A genetic linkage occurs when a particular gene is inherited jointly. Specifically, genetic loci on the same chromosome are physically connected and tend to stay together during meiosis, and are thus genetically linked, as shown in Fig. 28.2 . Within a linkage analysis, benign variations in the deoxyribonucleic acid (DNA) among different individuals are used as markers to identify regions or loci containing genes that predispose to a condition in related individuals. The algorithm used to study the linkage may be designated as parametric or nonparametric. A parametric linkage analysis is used to study major gene disorders, and assumes that the inheritance of such a disorder follows a specific and known genetic model, such as mendelian inheritance. A nonparametric linkage analysis is used for more complex diseases and when the inheritance of a disease does not follow a specific genetic model. The results of a simple linkage analysis are expressed as a logarithm of the odds (LOD) score, which compares the probability or odds of obtaining a particular test result when two genes are linked to one another with the probability or odds of obtaining the same score when the same two genes are not linked to one another.

The LOD score is calculated as follows:

[Equation 28.1]

where NR = the number of offspring with a nonrecombinant gene pattern or offspring that have a genotype identical to the parent?s genotype

R = the number of offspring with a recombinant gene pattern or offspring that have a genotype that differs from the parent?s genotype

θ or the recombinant fraction = R / (NR + R).

An LOD score >3.0 is considered evidence for linkage

The techniques currently used for linkage analysis have become very robust; however, they must be applied to large families with reliable clinical- and family-history data if they are to yield significant findings. Furthermore, in complex genetic diseases such as AIS, linkage analysis is usually most efficient at discovering rare genes with very large effects.

The large-scale screening of the genome of an individual with a particular condition may be done through the use of known genetic markers or polymorphisms. Genetic polymorphisms are benign variations in the structure of DNA that may vary among individuals. In GWAS, single nucleotide polymorphisms, or SNPs, are used as the markers to detect an association with a disease or condition in the population as a whole. These studies are similar to epidemiological studies that often employ case-control or cohort designs to assess multiple risk factors for disease (e.g., smoking as one of many risk factors for lung cancer). In GWAS, SNP markers are analyzed as risk factors. Typical microarrays that are gene ?chips? with a million or more genetic probes attached to them are used to rapidly and systematically analyze a patient?s genetic makeup ( Fig. 28.3 ). A GWAS is done over short region of the genome and may help target discrete regions of the genome for fine mapping, defining them as regions of interest (ROIs).

Several recent studies have used linkage analysis in an effort to identify specific regions of human chromosomes that may be associated with AIS. Wise et al20 identified potential linkage regions on chromosomes 3, 6, 10, 12, and 18 in a single family with scoliosis, with the 18q and 6p regions being the most significant. Chan et al25 subsequently evaluated the regions that Wise et al identified as most significant (6p, 10q, and 18q) but found no evidence of linkage in a single family with scoliosis. In the same family, Chan et al went on to identify linkage regions 19p13.3 and 2q as potentially important. In another single family with scoliosis, Salehi et al24 identified a region on 17p11 as potentially important. Miller, reporting on a large group of 202 families with scoliosis, noted multiple areas of linkage in two different subgroups. In a subgroup that demonstrated an X-linked dominant mode of inheritance, linkage region Xq 23–26 was found to be important. However, in the subgroup, which demonstrated an autosomal dominant mode of inheritance, five primary (6p, 6q, 9, 16, and 17) and eight secondary (1, 3, 5, 7, 8, 11, 12, and 19) regions of possible linkage were identified.

Although it has been suggested that the multiple locations identified in these sophisticated genetic studies of AIS represent conflicting results, 17,18,20,24–26 it has also been suggested that this great variability merely reflects the poly-genic nature of the heritability of scoliosis. It is likely that multiple factors contributed to the wide variability of results in these studies. Certainly the small size of many of the studies was a factor in this variability. But perhaps the most significant issue relates to the quality of the populations studied. The quality of any population used in a genetic study is directly proportional to the genetic informative-ness of the families selected for study. Informativeness can be adversely affected by small family size, poor genealogical resources, high false-paternity rates, significant clinical and genetic heterogeneity, and low diagnostic accuracy. Linkage-analysis studies that are substantially compromised in one or more of these areas will yield inconsistent results of questionable significance.

In a preliminary investigation, our group reported on 500 probands with AIS and identified regions on chromosomes 3 and 7 that were more statistically significant in terms of linkage to the condition than were regions found in any previous study (with LOD scores of 7.0 and 7.3, respectively).27 The significance of these markers was ~10,000 greater than that of regions found in the next largest study. We then conducted an additional case-control genetic association analysis to confirm the significance of these loci with regard to AIS. The high level of significance in our study was probably related to the informativeness of the families studied. Use of the GenDB database minimized or eliminated many of the limitations evident in previous studies. Work is underway to further validate these two linkage regions and identify possible additional markers associated with AIS.

Once areas or regions on the human genome are identified as being highly linked or associated with AIS, these loci can be further investigated with a positional-candidate approach. The ROI in a positional-candidate approach is relatively small as compared with that in candidate-gene analysis, and the association with AIS is clearer ( Fig. 28.4 ). Several studies have employed linkage analysis in an attempt to target specific regions of the human genome for subsequent positional-candidate-gene analysis. Although several investigators have targeted linkage regions containing the genes for melatonin, aggrecan, and other genes related to the structure of the extracellular fluid matrix, they have yet to reveal any specific association with idiopathic scoliosis.24,28–30

The search for specific genes associated with AIS represents an ongoing effort at multiple centers throughout the world. Although great progress has been made, the ultimate goal of fully understanding the molecular basis of this unique deformity has yet to be achieved. Once the genes responsible for AIS are mapped, substantial opportunities will become available for improving the management of this disease. These include the potential for early diagnosis through broad population screening and novel treatment options using cellular, molecular, or pharmacological strategies alone or in combination. Two-thirds of disease-gene discoveries made today are exploitable, allowing the targeting of related proteins or biochemical pathways for diagnosis or treatment. AIS may eventually evolve from a condition treated mechanically (either with a brace or surgery) to one treated medically. At present, the greatest clinical effect of the search for genes involved in AIS has not come from the mapping of genes themselves but through the related discovery of prognostic genetic markers.

Prognostics

Genome-wide linkage analysis and genetic association studies commonly use genetic markers merely as means to the ends of identifying genes associated with a particular condition. Yet genetic markers can also be used as ends in themselves. As an interim result in our quest to fully understand the molecular basis of AIS and fine map the genes responsible for this condition, we have identified genetic markers associated with severe and progressive scoliosis. An initial panel of markers associated with surgical-grade scoliosis (defined as spinal deformity progressing to 40 degrees or more before skeletal maturity) has proven highly accurate in identifying such deformity at an early stage. In an early study, involving 675 skeletally mature Caucasian subjects with AIS at centers across the United States, our group identified 12 SNP markers as having prognostic utility. Using an additive model involving both genetic and clinical risk factors (which had weights of ~90% and 10%, respectively), we obtained a summed risk score that correctly identified 97% of patients at low risk and 92% of those at high risk for progressive AIS, for an overall sensitivity of 93% and specificity of 90% (P <2.2 × 10”’ 16).26

The inclusion in this early work of >5000 patients from across the United States has allowed additional refinement of the initial marker panel. With the goal of improving the accuracy of genetic analysis in the prognosis of AIS, and including multiple races and ethnicities in such analysis, we have identified a panel of 53 genetic markers that now allows the prediction of progression of scoliosis to severe curvature with a range of accuracy of 90%. Although a highly accurate assessment of such risk can be achieved with these markers alone, the addition of clinical parameters, such as curve magnitude and skeletal maturity, further improves the sensitivity and specificity of this assessment, minimizing the frequency of false-positive and false-negative results.

Because most AIS is diagnosed at an early stage, either through school screening efforts or awareness among pediatricians, methods for predicting its progression have been sought for some time. However, accurate determination of the prognosis in individual cases of AIS has been elusive, perhaps because of a reliance on clinical variables that change with time. Lonstein and Carlson described the most widely accepted current method for clinically assessing the risk of progression of AIS in 1984.31 With this method, the magnitude of a patient?s curve and the patient?s skeletal maturity (Risser grade) are used to estimate the risk of progression. Yet these radiographic parameters are accurate only in cases of extreme curve magnitude and skeletal maturity, in which there is little question about the decision to treat. For example, a large curve in a skeletally immature patient (e.g., 39 degrees and Risser grade 0) is at high risk for continued progression and is likely to require treatment, whereas a small curve in a skeletally mature patient (e.g., 11 degrees and Risser grade 4) is unlikely to progress or ever require treatment.

For the majority of skeletally immature patients who present with AIS in the mild to moderate range, the current clinical methods of diagnosis of AIS provide significantly less ability to predict outcome than do prognostic genetic markers. To demonstrate the substantial differences in sensitivity and specificity in assessing the risk of progression of AIS with these two methods, we compared the Lonstein and Carlson method with the use of prognostic genetic markers in a cohort of our patients. Using radiographs from a patient?s initial presentation to determine the risk of progression with the Lonstein and Carlson method, we found that it had 60% sensitivity and 55% specificity in predicting outcome at skeletal maturity, whereas the prognostic genetic markers had a 95% and 93% sensitivity and specificity, respectively.

The uncertainty created by the radiographic methods for assessing the risk of progression of AIS has led to great inefficiencies in management of this disease. Patients with a mild scoliosis, at least 90% of whom are unlikely to experience progression of their curvature to a grade of surgical severity, are observed for years with serial X-rays at 4- to 6-month intervals until skeletal maturity. For patients who present with a moderate curve or whose curves progress into a moderate range, bracing is often recommended. Yet bracing is probably unnecessary in some such patients and ineffective in others, even with good compliance. For those whose curves progress beyond 40 degrees during skeletal immaturity, fusion surgery is often considered. However, fusion surgery essentially represents a nonphysiological salvage procedure that eliminates growth, motion, and function of the spine, and is justifiable only for curves of large magnitude. Whenever possible, avoidance of such a procedure would be ideal.

The discovery of prognostic genetic markers will profoundly affect the management of AIS by providing powerful information that accurately defines the risk of its progression at the earliest possible stage ( Fig. 28.5 ). For patients with a low risk of curve progression, the use of such markers may permit their reassurance and the reassurance of their families at a single encounter, eliminating years of uncertainty and serial radiographic evaluation. This will result in great individual and aggregate savings. For patients at high risk of curve progression, early intervention with standard bracing treatment or more novel procedures of fusionless scoliosis surgery will be possible. For patients thought to be reasonable candidates for bracing, it could be offered at an earlier stage and undertaken with conviction. For patients for whom bracing is not optimal (because of anticipated or evident compliance issues or psychological factors or both), or is contraindicated (thoracic lordosis, pulmonary issues), unwanted, or likely to fail (high-risk prognostic markers), novel, more physiological treatment options may be possible. These include fusionless scoliosis surgery on the convexity or concavity of a curve, anteriorly or posteriorly, with the goal of guiding growth for the correction of deformity while preserving motion and function of the spine.

Related posts:

7 The Case for Bracing 6 The Importance of the Sagittal Plane: Spinopelvic Considerations 15 The Use of Traction in Treating Large Scoliotic Curves in Idiopathic Scoliosis 23 Spinopelvic Fixation in Idiopathic Scoliosis 17 Surgical Treatment of the Right Thoracic Curve Pattern 10 Biomechanics and Reduction of Scoliosis

Stay updated, free articles. Join our Telegram channel

Join