The Developing Spine

The story of every human spine begins not with bone but with a delicate rod of cells called the notochord. During the third week after conception, in the process known as gastrulation, the flat embryonic disc reorganises itself into three layers, and along its midline a column of cells condenses into the notochord. This structure is, in evolutionary terms, the defining feature of the entire chordate lineage; in the human embryo it serves as the central axis around which the spine will be built and as a signalling beacon that instructs the tissues around it.

On either side of the notochord, the middle layer of the embryo, the paraxial mesoderm, begins to divide into paired blocks of cells called somites. The first pair appears at around day twenty, and new pairs are added in strict rhythm towards the tail, on average one pair roughly every ninety minutes, until something close to forty-four pairs have formed. The somites are the segmental units from which the vertebrae, the ribs, the deep back muscles and the overlying dermis will all be carved. Their orderly, repeating arrangement is the reason the adult spine is itself segmental, a stacked series of similar units rather than a single fused rod.

Each somite then splits its allegiance. Its ventral and medial portion, the sclerotome, gives rise to the vertebral column and ribs; its dorsal portion, the dermomyotome, gives rise to the trunk muscles and skin. The sclerotome cells undergo a transformation from a tidy epithelium into migratory mesenchyme and stream inwards to surround the notochord and the developing neural tube. There they condense, and from these condensations the bodies, arches and processes of the vertebrae take shape, while the notochord that they engulf is largely squeezed out of existence within the vertebral bodies and persists only between them, contributing to the centre of each intervertebral disc, the nucleus pulposus.

One of the most elegant and historically contentious ideas in spinal embryology is resegmentation. If each somite simply hardened into one vertebra, the muscles spanning the spine and the nerves emerging from it would be poorly arranged for movement. Instead, each sclerotome divides into a looser rostral half and a denser caudal half, and a vertebral body forms from the caudal half of one sclerotome fused with the rostral half of the sclerotome below. The result is that each vertebra is built from two adjacent half-somites, and the segmental muscles end up bridging the joints between vertebrae rather than lying over the middle of each bone, which is precisely the arrangement needed to bend the column. The spinal nerves, which form in relation to the original somite boundaries, come to exit between the newly assembled vertebrae. This shift of one half-segment is sometimes called the resegmentation-shift, and the notochord and the surrounding signals are essential to making the pattern repeat cleanly along the whole length of the embryo.

While the bony spine is being assembled, the spinal cord is being formed within it. In the third and fourth weeks the flat neural plate folds along its midline and zips shut into the neural tube, a process called primary neurulation that produces the brain and the great majority of the spinal cord. The lowest, most caudal part of the cord, including the tapering tip called the conus medullaris and the cauda equina, is made differently, by secondary neurulation: a solid cell mass at the tail end of the embryo hollows out and canalises to join the tube already formed above. Failures at these early steps underlie the various forms of spinal dysraphism, including the open neural tube defect of spina bifida.

A striking consequence of later growth is the so-called ascent of the conus. Early in development the spinal cord fills the entire length of the vertebral canal, with its tip reaching the lowest sacral or coccygeal levels. But the bony spine then grows faster and longer than the cord within it, so that the tip of the cord appears to climb. By around twenty-four weeks the cord ends near the first sacral level, in a newborn it lies near the second or third lumbar vertebra, and by adulthood it ends at roughly the first lumbar vertebra. This differential growth is why, in adults, the lower vertebral canal contains only the loose strands of nerve roots of the cauda equina rather than the cord itself.

The somites are at first remarkably similar to one another, yet they go on to form cervical, thoracic, lumbar, sacral and coccygeal vertebrae of strikingly different shape. What tells a somite which kind of vertebra to become is a code of regulatory genes, and the discovery of that code is one of the triumphs of twentieth-century biology. The HOX genes, a set of master regulators arranged along the chromosomes in the same order in which they act along the body, assign positional identity to each level of the axial skeleton. Their combined activity establishes where the rib-bearing thorax begins and ends, why the lumbar region is rib-free, and where the transitions between regions fall. Experimental inactivation of particular HOX groups in animals transforms one region into another, for example converting rib-free lumbar vertebrae into rib-bearing thoracic ones, demonstrating that these genes truly specify regional identity.

Working alongside the HOX genes are the PAX genes, particularly PAX1 and PAX9, which act within the sclerotome to drive the formation of the vertebral bodies and discs. The two families cooperate, with shared regulatory elements responding to both HOX and PAX proteins, and upstream regulators such as the PBX genes coordinating the whole programme. The picture that has emerged is of a layered genetic hierarchy that translates the segmental rhythm of somite formation into the precise, region-specific architecture of the finished spine.

For a long time the regularity of somite formation was a puzzle. Why should new segments appear at such even intervals, and how is the spacing controlled? A conceptual answer was proposed in 1976 by Jonathan Cooke and Christopher Zeeman, who suggested a "clock and wavefront" mechanism: a biochemical oscillator, or clock, ticking within the cells of the unsegmented presomitic mesoderm, combined with a maturation front, or wavefront, sweeping down the embryo, so that cells are partitioned into a somite each time the clock and the front coincide. For two decades this remained an elegant theory in search of a molecule. Then, from the late 1990s, the clock was made visible: genes of the Notch, Wnt and FGF signalling pathways were shown to switch on and off in rhythmic waves sweeping through the presomitic mesoderm, with each cycle corresponding to one new pair of somites. This vindication of the clock-and-wavefront idea, and the dissection of the molecular machinery behind it, transformed developmental biology and explained, at last, the orderly construction of the backbone.

The vertebrae are first laid down as cartilage and only later converted to bone. Each vertebra develops three primary ossification centres, one in the body and one in each half of the posterior arch, which appear during fetal life. At birth a vertebra is still part bone and part cartilage, with cartilaginous bridges, the neurocentral synchondroses, joining the body to the arch; these fuse during early childhood. Around and after puberty a second wave of bone formation appears at the secondary ossification centres, including the ring apophyses, the rims of bone around the top and bottom of each vertebral body to which the intervertebral disc attaches, as well as centres at the tips of the spinous and transverse processes. The ring apophyses ossify through the early-to-mid teenage years and fuse to the vertebral body towards the end of the second decade, an event that marks the near-completion of vertebral growth. Because their maturation closely tracks overall skeletal maturity, the ring apophyses have themselves become a useful marker of where a young person stands in the growth process.

The spine roughly triples in length between birth and adulthood, but it does not grow at an even pace. Growth is rapid in the first years of life, slows during middle childhood, and accelerates again sharply at the pubertal growth spurt. During puberty the trunk grows proportionately more than the limbs, so that much of the adolescent height gain is in sitting height, which is to say in the spine itself. This uneven tempo matters greatly for understanding childhood spinal deformity, because a curve in the spine tends to worsen fastest during the periods of fastest growth, and a child with a great deal of growth remaining is at far higher risk of progression than one whose growth is nearly finished.

Judging how much growth remains became a central practical question, and the answer came from an unexpected place. In 1958 Joseph Risser described that the bony cap of the pelvis, the iliac apophysis, ossifies progressively from front to back and then fuses, and that this excursion runs in parallel with the maturation of the spinal growth plates. The "Risser sign", graded from zero to five, allowed the maturity of the spine to be estimated from a single pelvic X-ray, and it became, for decades, the standard way of gauging how much spinal growth and therefore how much risk of curve progression lay ahead. More recent schemes have refined this estimate by reading the growth plates of the hand and elbow, but the underlying principle that treatment of the growing spine must be timed against the trajectory of growth has remained constant.

Before the twentieth century, knowledge of the developing spine rested almost entirely on dissection of embryos and stillbirths. The advent of imaging changed this at every stage. The discovery of X-rays at the end of the nineteenth century allowed the bony spine to be seen in the living child for the first time, made the measurement of curves and the grading of skeletal maturity possible, and underpinned the whole discipline of paediatric spinal deformity. Computed tomography, from the 1970s, rendered the complex three-dimensional architecture of congenital vertebral anomalies visible in a way that plain films never could, clarifying exactly which vertebrae had failed to form or to segment. Magnetic resonance imaging, arriving in clinical use in the 1980s, did something different again: without any radiation, it revealed the spinal cord itself, the discs and the soft tissues, allowing tethered cords, syrinxes and cord malformations hidden within an apparently bony problem to be recognised.

Most remarkable of all has been the development of prenatal imaging. Obstetric ultrasound, pioneered for medical use in the 1950s, eventually became sensitive enough to detect open spina bifida and other spinal anomalies before birth, often through indirect signs in the skull and brain. Fetal MRI then added detailed soft-tissue views of the fetal spine and cord. Together these techniques created something entirely new in the history of the spine: the ability to diagnose, characterise and even plan the treatment of a spinal condition in a child who has not yet been born.

The scientific study of the spine has a long arc. The Renaissance anatomist Andreas Vesalius, whose great illustrated work *De humani corporis fabrica* appeared in 1543, established anatomy as an evidence-based discipline grounded in direct dissection, and depicted the adult vertebral column with unprecedented accuracy, correcting many inherited errors. But Vesalius and his successors described the finished structure; how it came to be remained unknown.

The understanding of development came later and in stages. In the eighteenth century Caspar Friedrich Wolff argued, against the prevailing belief that organisms were preformed in miniature, that structures arise gradually from undifferentiated material, an idea too far ahead of its time to be accepted. It was vindicated in the nineteenth century by a generation of embryologists. Karl Ernst von Baer, who set out his laws of embryology in works of 1828 and 1837, transformed the comparative study of development. Robert Remak, working on the chick embryo around 1850 to 1855, identified the segmental tissue that gives rise to the vertebral column and advanced the much-debated theory that vertebrae are formed by resegmentation of the somites, an idea that, in refined form, remains central today. Across the late nineteenth and twentieth centuries this descriptive embryology was steadily deepened, until in the molecular era the genes and signals behind the old observations were finally identified, joining the careful morphology of the Victorians to the gene expression studies of the present.

When vertebrae fail to form or to separate properly in the embryo, the result can be a congenital spinal deformity. Congenital scoliosis arises from vertebral anomalies that cause one side of the spine to grow faster than the other. Early in the twentieth century these anomalies were already being classified by their embryological origin. In 1910 Vincenzo Putti distinguished types of hemivertebra by how completely they were separated from their neighbours by discs. The framework that endured, associated with MacEwen and adopted by the Scoliosis Research Society, divides the anomalies into failures of formation, such as the hemivertebra, failures of segmentation, such as the unsegmented bar or block vertebra, and mixed forms. This was not mere taxonomy; understanding the embryological basis allowed surgeons to predict which anomalies would drive the worst progressive curves, and therefore which children might benefit from early intervention. In recent years the molecular roots of some of these anomalies have been identified, with genes of the segmentation machinery, such as DLL3, MESP2, LFNG and HES7 in spondylocostal dysostosis, and TBX6 in a substantial fraction of congenital scoliosis, shown to disrupt the very clock-and-wavefront process that builds the segmented spine.

Perhaps nowhere has growing knowledge of the developing spine reshaped treatment as dramatically as in childhood scoliosis. The earliest approaches were external and non-operative. Through the nineteenth and early twentieth centuries deformities were managed with casts and braces, and a turning point came in 1946 when Walter Blount and Albert Schmidt devised the Milwaukee brace, the first widely used spinal orthosis. Lower-profile designs such as the Boston brace followed from the 1960s and became the mainstay of conservative management. For many years the evidence that bracing actually altered the natural history of a curve was debated; that question was settled only in 2013, when the BrAIST randomised trial demonstrated that bracing significantly reduced the chance of a curve progressing to the threshold for surgery.

Surgery for scoliosis also has deep roots. Russell Hibbs performed early posterior spinal fusions from around 1911, and by the 1930s fusion combined with cast immobilisation was established practice. The great surgical leap came in the 1950s and early 1960s, when Paul Harrington, working in Houston, developed an internal metal rod with hooks that could be distracted to straighten the spine from within. The Harrington rod transformed scoliosis surgery and was implanted in very large numbers over the following decades. But it embodied an assumption that would later be questioned: that the right response to a curve was to correct it and fuse it solid.

As understanding of the developing spine matured, that assumption proved inadequate for the youngest patients. Fusing the spine of a very young child arrests its growth, and a short, fused thorax cannot allow the lungs to develop, producing the grave problem of thoracic insufficiency. This recognition drove the central modern idea in early-onset scoliosis: that growth must be preserved, not stopped. The result was a family of growth-friendly techniques. Traditional growing rods, anchored above and below the curve and surgically lengthened every few months, allowed the spine to keep growing while the deformity was controlled. The vertical expandable prosthetic titanium rib, or VEPTR, addressed the chest wall and thorax directly. Then, from the 2010s, magnetically controlled growing rods such as the MAGEC system allowed lengthening to be performed non-invasively, by an external magnet in the clinic, sparing children the repeated operations that traditional rods demanded.

The most recent shift goes further still, harnessing growth rather than merely accommodating it. Growth modulation rests on the Hueter-Volkmann principle, the long-known observation that compression slows growth at a growth plate while reduced loading allows it to proceed. Vertebral body tethering applies this idea directly: a flexible cord is anchored along the convex, faster-growing side of a curve, gently restraining its growth so that the slower concave side can catch up and the curve straightens as the child grows. Approved for use in suitable skeletally immature patients from 2019, tethering represents a conceptual inversion of the Harrington era, treating remaining growth not as an obstacle to be eliminated by fusion but as the very force that corrects the deformity.

The treatment of spina bifida has been transformed by a parallel insight: that some of the damage in an open neural tube defect is not fixed at the moment the defect forms but accumulates as the exposed cord is bathed in amniotic fluid and dragged through the womb over the months of pregnancy. If that ongoing injury could be stopped earlier, outcomes might improve. This reasoning led to the audacious step of operating on the fetus in the womb to close the defect before birth. The hypothesis was tested in the landmark Management of Myelomeningocele Study (MOMS), a randomised trial whose results were published in the New England Journal of Medicine in 2011. The trial was stopped early for benefit: prenatal repair reduced the need for shunting of fluid on the brain, lessened the hindbrain herniation of the associated Chiari malformation, and improved motor function and the likelihood of independent walking compared with the traditional repair after birth. MOMS established fetal surgery for spina bifida as a genuine option and stands as one of the clearest examples of how an understanding of spinal development, in this case the timing of the injury during gestation, can redirect treatment to an earlier and more effective moment.

The recognition of the tethered cord is another instance of developmental understanding reshaping care. The normal ascent of the conus during growth means that a cord abnormally fixed at its lower end, by a thickened filum terminale, a lipoma or scar, cannot rise as it should and is placed under increasing tension as the child grows. The condition was crystallised in 1976, when the Canadian neurosurgeon Harold Hoffman and colleagues described a series of patients with an elongated, low-lying cord whose symptoms improved after the filum terminale was divided, and coined the term "tethered cord". The understanding that the problem is mechanical and developmental, a failure of the normal relative ascent, explained why symptoms could emerge or worsen during growth spurts and informed the rationale for surgically releasing the cord. It also illustrates a recurring theme: that the timing of treatment for many conditions of the immature spine is dictated by the trajectory of growth itself.

The contemporary picture of the developing spine is one in which classical embryology, molecular genetics and clinical innovation have converged. The genes that build the segmented spine are increasingly known, and sequencing studies have begun to explain why particular children develop congenital vertebral anomalies, with TBX6 and the genes of the segmentation clock now firmly implicated. Growth modulation has moved from principle to practice, and the idea of guiding rather than arresting growth continues to expand. Fetal surgery, once unthinkable, is an established field, and less invasive fetoscopic techniques for closing spinal defects before birth are under active study. Regenerative ideas, including the use of stem cells and engineered biomaterials to improve fetal repair, are being explored at the research frontier.

Open questions remain in abundance. The precise genetic architecture of common deformities such as adolescent idiopathic scoliosis is still incompletely understood. The long-term outcomes of the newest growth-friendly and growth-modulating devices continue to be gathered. And the fundamental biology of how the human notochord, somite and segmentation clock differ from those of the laboratory animals in which so much has been learned is only beginning to be studied directly in human tissue. What is clear is that the developing spine, far from being a settled chapter, remains one of the most active meeting points of developmental biology and surgical innovation, and that each advance in understanding how the backbone is built has, time and again, changed how its disorders are treated.

• —Pourquié O, et al. Somite Development: Constructing the Vertebrate Body. *Cell.* 1998
• —Hubaud A, Pourquié O. Signalling dynamics in vertebrate segmentation. *Nat Rev Mol Cell Biol.* (Somitogenesis review.) Embryology overview: UNSW Embryology, Somitogenesis
• —Gilbert SF. Paraxial Mesoderm: The Somites and Their Derivatives. *Developmental Biology.* NCBI Bookshelf NBK10085
• —Ward L, Pang ASW, Evans SE, Stern CD. The role of the notochord in amniote vertebral column segmentation. *Dev Biol.* 2018. PMC5971204
• —A resegmentation-shift model for vertebral patterning. *J Anat.* 2017. PMC5244455
• —Christ B, Wilting J. From somites to vertebral column. *Ann Anat.* 1992. PMID: 1605355
• —Criswell PJ, et al. Resegmentation is an ancestral feature of the gnathostome vertebral skeleton. *eLife.* 2019
• —Mallo M. Building the backbone: the development and evolution of vertebral patterning. *Development.* 2015;142(10):1733
• —Wellik DM. Hox patterning of the vertebrate axial skeleton. *Dev Dyn.* 2007;236(9):2454–2463
• —Mallo M, et al. Hox genes and vertebrate axial pattern. *Curr Top Dev Biol.* 2009. PMID: 19651308
• —Casaca A, Santos AC, Mallo M. Controlling Hox gene expression and activity to build the vertebrate axial skeleton. *Dev Dyn.* 2014
• —Capellini TD, et al. Pbx1/Pbx2 govern axial skeletal development by controlling Polycomb and Hox in mesoderm and Pax1/Pax9 in sclerotome. *Dev Biol.* PMC5918304
• —Guerreiro I, et al. Role of a polymorphism in a Hox/Pax-responsive enhancer in the evolution of the vertebrate spine. *PNAS.* 2013;110(26)
• —Cooke J, Zeeman EC. A clock and wavefront model for control of the number of repeated structures during animal morphogenesis. *J Theor Biol.* 1976. Clock and wavefront model overview
• —Wahl MB, et al. FGF signaling acts upstream of the NOTCH and WNT signaling pathways to control segmentation clock oscillations in mouse somitogenesis. *Development.* 2007;134(22):4033. PMID: 17965051
• —Modelling coupled oscillations in the Notch, Wnt, and FGF signaling pathways during somitogenesis. *PLoS One.* PMC4381657
• —Neuroanatomy, Neural Tube Development and Stages. *StatPearls.* NCBI Bookshelf NBK557414
• —Neuroanatomy, Conus Medullaris. *StatPearls.* NCBI Bookshelf NBK545227
• —Ascent of the Conus Medullaris. ISPN Guide (embryology of the central nervous system)
• —Early spinal cord development: from neural tube formation to neurogenesis. *Nat Rev Neurosci.* 2025
• —Spinal dysraphism illustrated; embryology revisited. *Indian J Radiol Imaging.* PMC5761168
• —Ossification and Fusion of the Vertebral Ring Apophysis as an Important Part of Spinal Maturation. *J Clin Med.* 2021;10(15):3217. PMC8347734
• —Maturation of the vertebral ring apophysis is delayed in girls with adolescent idiopathic scoliosis compared to the normal population. *Spine Deform.* 2024. PMC11499316
• —Dimeglio A, Canavese F. The immature spine: growth and idiopathic scoliosis. *Ann Transl Med.* 2020
• —Normal growth of the spine and skeletal maturation. *EMC Rhumatologie-Orthopédie / Orthop Traumatol.*
• —Hacquebord JH, Leopold SS. In Brief: The Risser Classification: A Classic Tool for the Clinician Treating Adolescent Idiopathic Scoliosis. *Clin Orthop Relat Res.* 2012. PMC3392381
• —Risser sign. Wikipedia (history of the Risser sign, described 1958)
• —Splinter R, et al. De Humani Corporis Fabrica surgical revolution (Vesalius, 1543). *PubMed.* PMID: 19160130
• —De Humani Corporis Fabrica Libri Septem. Wikipedia (Vesalius, 1543)
• —Karl Ernst von Baer's Laws of Embryology. Embryo Project Encyclopedia, Arizona State University
• —The somites of the chick (Remak's resegmentation theory, historical). UNSW Embryology
• —Burnei G, et al. Congenital scoliosis: an up-to-date (embryology and classification). *J Med Life.* PMC2856388
• —Burnei-Gavriliu classification of congenital scoliosis. *J Med Life.* 2015. PMC4392099
• —Congenital Spinal Deformity: Background, Embryology of Spine, Pathophysiology. Medscape eMedicine
• —Molecular landscape of congenital vertebral malformations: recent discoveries and future directions. *Bone Res / review.* 2024. PMC10829358
• —Wu N, et al. TBX6 Null Variants and a Common Hypomorphic Allele in Congenital Scoliosis. *N Engl J Med.* 2015;372:341–350
• —Progress and perspective of TBX6 gene in congenital vertebral malformations. *Bone Res.* 2016. PMC5302999
• —Clinical genetics of spondylocostal dysostosis: a mini review (DLL3, MESP2, LFNG, HES7). *Front Genet.* 2022
• —A History of Bracing for Idiopathic Scoliosis in North America. *Clin Orthop Relat Res.* 2009. PMC2816759
• —Historical Timeline. Scoliosis Research Society (Hibbs 1911 fusion; bracing history)
• —Weinstein SL, Dolan LA, Wright JG, Dobbs MB. Effects of bracing in adolescent idiopathic scoliosis (BrAIST). *N Engl J Med.* 2013. Design paper: PMC4536980
• —The Legacy of Harrington's Rod and the Evolution of Long-Segment Constructs in Spine Surgery. *Cureus / review.* 2024. PMC11432481
• —Harrington rod. Wikipedia (Paul Harrington, history of internal correction)
• —Magnetic Controlled Growth Rods in the Treatment of Scoliosis: Safety, Efficacy and Patient Selection. *Med Devices (Auckl).* PMC7085947
• —Growing rods in Early Onset Scoliosis: the current scenario. *J Clin Orthop Trauma.* PMC10372050
• —Vertebral Body Tethering (VBT) in Pediatric and Adult Spinal Deformity. Scoliosis Research Society Informational Statement (FDA HDE approval 2019; Hueter-Volkmann)
• —Growth Modulation for Childhood Scoliosis: From Where Have We Come, and Where are We Going? *Semin Spine Surg / ScienceDirect.* 2024
• —Adams MA, et al. Vertebral Body Tethering: Indications, Surgical Technique, and a Systematic Review of Published Results. *J Clin Med / review.* 2022. PMC9099651
• —Adzick NS, Thom EA, Spong CY, et al. A Randomized Trial of Prenatal versus Postnatal Repair of Myelomeningocele (MOMS). *N Engl J Med.* 2011;364(11):993–1004. PMC3770179
• —MOMS Trial. Wikipedia (Management of Myelomeningocele Study, 2011)
• —Myelomeningocele Surgery over the 10 Years Following the MOMS Trial: A Systematic Review. *J Clin Med.* 2021. PMC8307221
• —Advances in Fetal Repair of Spina Bifida Integrating Prenatal Surgery, Stem Cells, and Biomaterials. *Review.* 2024. PMC12838692
• —Hoffman HJ, Hendrick EB, Humphreys RP. The tethered spinal cord: its protean manifestations, diagnosis and surgical correction. *Childs Brain.* 1976. Reviewed in: Tethered cord syndrome: a review of the literature from embryology to adult presentation. *Neurosurg Focus.* 2010;29(1)
• —Tethered Cord Syndrome (TCS). *StatPearls.* NCBI Bookshelf NBK585121
• —Fetal magnetic resonance imaging in the diagnosis of spinal cord neural tube defects: a prospective study. *Front Neurol.* 2022. PMC9393549
• —Ultrasound and magnetic resonance imaging in the prenatal diagnosis of open spina bifida. *Ultrasound Obstet Gynecol / review.* 2018. PMID: 29730690

This information is for general patient education and is not a substitute for an individual consultation. Always seek personalised advice from a qualified spinal specialist.