The first recorded successful attempt at spinal instrumentation was performed in Ottawa, Kansas, in 1888 by W. F. Wilkins. Wilkins performed surgery on a newborn infant who had sustained a vertebral dislocation with traumatic rupture of a lumbar intervertebral disk. After reducing the disk herniation, Wilkins stabilized the spine by wiring the pedicles together with carbolized silver wire, and the infant recovered uneventfully [4].

A similar technique was subsequently employed by Berthold Ernest Hadra in Texas in 1891, though he was not aware of Wilkins’ procedure at the time. Hadra was a Prussian-born surgeon who had immigrated to the United States after serving as an assistant surgeon in the Prussian army during the Austro-Prussian war [5]. He used silver wire to join the spinous processes of the sixth and seventh cervical vertebrae in a 30-year-old patient with an unstable cervical fracture-dislocation and neurological deterioration. The procedure was initially a success, although the surgery was repeated 3 weeks later when the wires loosened. Undaunted, Hadra posited that his interspinous wiring method was a success and could be adapted for use in the treatment of “any deviation of a vertebra” [6].

As it happened, the closing of the nineteenth century found physicians facing widespread outbreaks of tuberculosis worldwide. Progressive spinal deformity and neurological compromise as a result of extrapulmonary tuberculosis, first described by Percivall Pott in 1779, had long been treated with externally applied braces and devices, but remained a debilitating malady that surgeons of the day struggled to address. Pott himself had advocated surgical drainage of paraspinal tuberculosis abscesses as a solution, although this procedure did nothing to correct deformity and was associated with the development of secondary infections and draining sinus tracts [7].

Inspired by his relative success with interspinous wiring, Hadra suggested that his procedure might have applicability in the treatment of Pott’s disease, although he had not attempted this himself [8]. Two years later, French neurosurgeon Antoine Chipault performed the first internal fixation for this very purpose. In 1898, Lovett published a report of his experience in five patients with tuberculosis spondylitis, describing a procedure in which he denuded adjacent spinous processes and then wired them together with silver wire [9]. However, despite achieving stability initially, this process resulted in short-lived success and patients frequently required a second surgery [10].

Inspired by the work of Hadra and Chipault, Fritz Lange of Munich was himself beginning to consider the problem of internal splinting for tuberculous spines. He envisioned an “artificial spinal column of steel,” composed of bilateral tin-plated steel rods wired to the spinous processes. After his first attempt in 1902 resulted in postoperative complications from the sharp ends of the silver wire used to secure the rods, he set to work trying to characterize the optimum materials for use in his procedure. After several years of animal experiments, he settled on 5-mm-thick, 10-cm-long, tin-plated steel rods secured to the spinous processes with silk thread [11].

In 1911, Fred Albee and Robert Hibbs, both orthopedic surgeons in New York, independently struck upon the importance of arthrodesis in halting the progressive deformity of Pott’s disease when they each developed procedures that used autologous bone grafts to achieve bony fusion of the spinous processes, thereby obviating the need for wires. In both procedures, this process corrected deformity and reinforced the dorsal spinal elements. The Albee procedure used tibial grafts as struts between the spinous processes of adjacent vertebrae [19], while Hibbs used bone from the spinous processes themselves to bridge vertebrae [20, 21]. The theoretical advantages of these procedures were immediately obvious, and their potential for halting spinal deformity from other causes was quickly recognized, such that these techniques were widely adopted soon after their invention. In 1914, Herbert Galloway became the first surgeon to use fusion techniques to correct paralytic scoliosis [22]. Nevertheless, these initial efforts at fusion resulted in unacceptably high rates of pseudarthrosis, and as a result, these procedures were often modified by others to include silver wiring to minimize graft movement and maximize arthrodesis [23].

The 1930s saw the development of more elegant techniques for spinal fusion, including interbody fusion of the lumbar spine. In 1933, Burns published his description of anterior lumbar interbody fusion for an L5–S1 spondylolisthesis [24]. Beginning in 1939, Cloward would go on to develop the posterior lumbar interbody fusion approach [25]. Such procedures were developed and refined over the course of the next 30 years. While all were promising methods for halting the progression of spinal deformity, all were also initially associated with fairly high rates of pseudarthrosis.

Importantly, the foregoing sections illustrate the point that spinal instrumentation and fusion were developed more or less independently. These earliest methods for addressing deformities of the spine were limited in two ways: (1) They generally halted the progression of deformity, rather than correcting it, and (2) they failed to marry the techniques of spinal fusion, which provided the durable stabilization sought by these early pioneers, with spinal instrumentation, which helped to maintain the alignment and rigidity necessary to promote arthrodesis. It was not until the 1960s that the interdependence of these two parallel advances would become fully and widely recognized, and the modern era of spinal stabilization would begin. In the immediate aftermath of World War II, American physicians faced a recurrence of widespread poliomyelitis epidemics. In the late 1940s and early 1950s, polio crippled approximately 35,000 individuals each year, with the development of an effective vaccine still several years away. Aside from the immediate complications of polio, many sufferers subsequently developed progressive and debilitating thoracic scoliosis, often with attendant cardiorespiratory compromise that made treatment with braces and casts impossible. The treatment of these patients occupied spinal surgeons of the day.

One such individual was Paul Randall Harrington, an American orthopedic surgeon in Houston, Texas. After finishing his residency in orthopedic surgery in 1942, he joined the US Army, serving as the chief of orthopedic surgery for the 77th Evacuation Hospital. At the war’s end in 1945, Harrington settled in Houston, where he worked at the Jefferson Davis City-County Hospital. City-County saw many polio patients and by 1953 had become the second Respiratory Center in the nation.

Recognizing that a large number of the polio patients seen at City-County often developed scoliosis, Harrington began in 1947 to develop a method to stop the progression of these patients’ deformity and improve their cardiopulmonary function. Harrington’s first attempts used screws to fix the facet joints in a corrected position, a technique introduced by Toumey [26], and King [27], as a means of providing rigid internal fixation to aid in arthrodesis. While Harrington’s facet-screw procedure appeared promising at first, ultimately this method proved a failure, and he was forced to seek an alternative means for correcting the deformity [28]. The next iteration of Harrington’s approach involved hooks connected to a threaded rod, by which he was able to apply distracting corrective forces. The hooks were placed at the spinous processes of the vertebrae at the superior and inferior extremes of a curve, with the rod spanning the concavity of the scoliotic curve.

Harrington continued to refine his surgical technique and instrumentation over the next 15 years, making the early instrumentation by hand. When his early handmade constructs failed, he partnered with engineers to develop sturdier instruments that could withstand the repetitive stress to which they were subjected. Ultimately, Harrington’s work led to the creation of a system of stainless steel rods and distraction hooks. As he continued to develop an instrumentation system that could address the spinal deformity from polio, Harrington also began to apply it to cases of idiopathic scoliosis [29].

Despite the initially satisfactory correction of the deformity, Harrington was cognizant of the fact that hardware failure was inevitable, often as early as 6 months postprocedure. To achieve lasting deformity correction, he recognized that his instrumentation would have to incorporate a fusion within its extent. Rather than using hooks and rods as a dynamic means of establishing correction, he determined that instrumentation should serve as a means of permitting arthrodesis [28].

By demonstrating that instrumentation was merely a temporary means to the end of spinal fusion, rather than an end in itself, Harrington laid the foundation for future developments in spinal instrumentation. His experience revealed that instrumentation failure is inevitable, but that permitting arthrodesis can achieve long-term stability. By acknowledging the “race between instrumentation failure and the acquisition of spinal fusion,” Harrington united the parallel advances in internal fixation and spinal fusion, thereby ushering in the modern era of spinal stabilization [28].

Almost immediately, Harrington’s rod system became the state-of-the-art spinal instrumentation through the 1970s. Indeed, use of Harrington instrumentation continues to the present day. The utility of the procedure was easily recognized, and the indications were soon extended beyond polio and idiopathic scoliosis to include trauma [30, 31], degenerative processes, and malignancy [32, 33]. With its widespread adoption, however, came the recognition of its limitations. Most notably, the effectiveness of distraction rods in correcting coronal curvature of the spine came at the expense of the natural sagittal curvature of the spine, resulting in loss of lumbar lordosis, the so-called flatback syndrome [34, 35]. Additionally, the procedure required that patients spend several months in plaster braces postoperatively. Lastly, repeated stresses occasionally resulted in hook dislodgement or rod breakage [36].

If Harrington provided the bedrock on which the modern era of spinal instrumentation was built, much of the foundation was laid by Eduardo Luque in the mid-1970s. Luque, an orthopedic surgeon in Mexico City, sought to address some of the limitations of the Harrington rod system. His work ultimately led to the introduction of segmental spinal fixation. Whereas Harrington used straight rods with hooks at either end of the deformity, Luque used a contoured steel rod, attached to the vertebrae at several points with sublaminar wiring. By doing so, Luque was able to more evenly distribute the forces across the construct, which not only increased the rigidity of the construct and reduced the potential for hardware failure but also offered the potential for greater correction and reduced the need for postoperative bracing. Most importantly, however, segmental spinal fixation permitted deformity correction without sacrificing sagittal curvature [37].

While spinal surgeons recognized the promise of Luque’s segmental fixation technique, their enthusiasm was tempered by concern for potential neurological injury caused by the sublaminar wiring used to secure the rods. Indeed, reports of complications such as epidural hematoma and direct trauma caused by passing the sublaminar wire left surgeons searching for alternative means of achieving segmental fixation [38, 39]. To address these concerns, Drummond et al. published a description in 1984 of a technique using a button-wire implant threaded through a hole drilled at the base of the spinous process [40]. Although the construct was not as strong as Luque’s sublaminar wiring construct, it had the advantage of reducing the potential for neurological injury, thereby making it a more attractive alternative to spinal surgeons of the time.

A major advance during the refinement of segmental fixation techniques was the introduction of pedicle screws as a means to secure the constructs to the vertebrae. As mentioned previously, the first accounts of the use of screws for internal fixation of the spine as an adjunct to fusion were published in the early 1940s by Toumey [26] and King [27]. Both used short bone screws that traversed the facet joints of the segment to be fused bilaterally. King reported a pseudarthrosis rate of only about 10 %, although Thompson and Ralston subsequently reported much higher rates of pseudarthrosis using a similar technique [41]. Thus, the technique did not gain widespread acceptance, and Bosworth suggested in 1957 that the benefits of screw fixation did not justify the difficulty in placing the screws [42].

H. H. Boucher, a Canadian spinal surgeon, believed that the concept of screw fixation was nevertheless sound. He developed and implemented a method of screw fixation using longer screws to traverse the facet joint and enter the pedicle and vertebral body below. In 1959, he published a paper describing his experience with this technique over a 12-year period for single- and multiple-level fusions and to treat spondylolisthesis [43]. Although he reported excellent results, he conceded that screws were a temporary solution and were likely to loosen over time, but that well-placed screws could offer superior fixation to minimize pseudarthrosis.

In Paris, Raymond Roy-Camille, under the guidance of Robert Judet, used pedicle screws as fixation points for a multiple-segment posterior plate system. Although he performed his first surgery in 1963, he did not publish the results until 1970 [44]. His system of plates and pedicle screws was applied to a variety of spinal disorders, and its success confirmed the superiority of pedicle screw segmental fixation [45]. He is largely credited with pioneering the use of pedicle screws in segmental instrumentation.

The advantages of pedicle screws over facet screws, hooks, and wire fixation are manifold. Biomechanically, pedicle screws gain better purchase and are capable of withstanding a wider range of forces across multiple vectors [46, 47]. Additionally, unlike hooks and wire fixation, which require posterior vertebral structures, they may be placed even after a laminectomy, a benefit that permitted the extension of instrumentation techniques to address various degenerative spine conditions.

In the United States, the first surgeon to popularize the use of pedicle screws for segmental fixation was Arthur Steffee, an orthopedic surgeon in Cleveland, Ohio. In the early 1980s, Steffee et al. noted that the pedicle acted as a “force nucleus” for the vertebra and was the conduit by which the forces acting upon the posterior vertebral elements could be transmitted anteriorly [48]. Accordingly, he posited that the pedicle was an ideal structure for placing a screw in a segmental fixation construct. Steffee placed headless screws in the pedicles and modified standard long bone fracture plates with slots that could slide onto the screws. He secured the plates in place with nuts, thereby pioneering “variable screw placement.”

In Paris during roughly the same time, Cotrel and Dubousset were developing a segmental fixation system that used rods, as opposed to the plates that their contemporaries were using. Initially, their system used hooks at multiple segments, secured with bolts [49], and while this offered satisfactory correction, the bolt-locked hook system proved difficult to remove. Eventually, they adopted monoaxial pedicle screws as fixation points for their rods [50], allowing for easier adjustment or removal of the construct. Thus, Cotrel and Dubousset married the advantages of segmental rod fixation of the spine with those of pedicle screws. While plate constructs were preferred for their strength in spinal instrumentation, rods offered much greater flexibility and a lower profile, thus allowing surgeons to more precisely correct deformities in three dimensions and leaving more room for bone grafting.