The hardware placed in a fusion takes three forms. Pedicle screws anchor into the vertebrae, one on each side of the bone. Rods connect those screws and hold the segment in position while it heals. Between the vertebrae, where the disc used to be, sits an interbody cage — a spacer that restores height and gives bone a scaffold to grow across. Each of these components can be made from more than one material, and the choice among them turns on one property: stiffness, relative to bone.
The short list: four materials
The first is titanium — strong, corrosion-resistant, and the material of most modern screws, rods, and cages. The second is cobalt-chromium, a metal alloy used when the stiffest rod is required. The third is PEEK, short for polyetheretherketone — a medical-grade polymer, not a metal — used most often for cages. The fourth is carbon fiber–reinforced PEEK, the same polymer reinforced with carbon fibers. It is radiolucent and produces no artifact on CT or MRI, allowing better radiographic assessment on follow-up imaging.
Stainless steel, standard in earlier decades, has largely given way to these four in modern practice. All four are strong enough for the loads involved. What separates them is stiffness.
Why stiffness is the property that matters
Stiffness is measured as elastic modulus. Bone has a modulus of its own, and it comes in two forms: cortical bone, the dense outer shell, and cancellous bone, the lighter, spongier bone inside.
Bone responds to load: load carried through bone maintains it, and load carried around it does not. An implant far stiffer than the bone it supports can take forces that would otherwise pass through the bone and the graft — a pattern called stress shielding. Stiffer implants are more likely to stress shield the bone graft, "and potentially result in a pseudarthrosis" — a fusion that does not heal (Heary et al., Journal of Spine Surgery, 2017). A material closer to bone in stiffness shares the load with it. It is called load sharing.
PEEK measured 3.84 GPa, nearly identical to cancellous bone at 3.78 GPa. The carbon-fiber composite measured 17.94 GPa, close to cortical bone at 14.64 GPa. The metals measured higher: titanium at 50.20 GPa, stainless steel at 51.07 GPa, and cobalt-chromium at 53.15 GPa. The authors note that those metal values were extrapolated from their tests and may underestimate the true modulus (Heary et al., Journal of Spine Surgery, 2017). Review tables using textbook values place the metals higher still: solid titanium at approximately 110 GPa and cobalt-chromium at approximately 210 GPa (Garg et al., Journal of Functional Biomaterials, 2026). The two sets come from different methods and are not interchangeable. In both, the polymers sit near bone and the metals far above it.
A finite-element model of a single lumbar segment found the same pattern: titanium fixation produced higher stress shielding than PEEK, whose stiffness let load pass to the bone (Ceddia et al., Biomimetics, 2025).
Titanium
Titanium is strong at a relatively light weight, resists corrosion inside the body, and is well tolerated by bone over decades. In a spinal fusion it serves as screws, as rods, or as the cage itself, and most of the hardware placed in modern spine surgery is titanium.
It has two limitations, both related to stiffness. Solid titanium is far stiffer than bone, so a solid titanium cage carries load the graft could have shared. And while titanium disturbs MRI less than the other metals — in standardized testing of implant constructs, titanium alloys produced the smallest image artifacts of the metals tested (Heinrich et al., European Radiology, 2021) — it still produces artifact immediately around the hardware.
The answer to the stiffness limitation is structural rather than a different metal: titanium printed as a porous lattice. A 3D-printed cage is titanium at its surface, where bone meets it, and open space through its body. The open structure brings the effective modulus down to approximately 2–6 GPa across designs, with porosity of 60%–80% and pore sizes of 300–700 μm (Yang et al., Frontiers in Bioengineering and Biotechnology, 2026). That range is a span across designs, not a single product's value. At its open end, one printed design — 61.3% porous, with pores of 500–700 μm — measured 1.7 GPa (Li et al., Scientific Reports, 2016). Against cancellous bone at 3.78 GPa and PEEK at 3.84 GPa, printed porous titanium can be engineered to mimic the modulus of cancellous bone more closely than solid titanium — and, at the open end of the range, more closely than PEEK.
The pores also give bone a structure to grow into: osseointegration, bone bonding to and through the implant rather than resting against it. In an animal comparison, bone volume within the graft window was significantly greater for 3D-printed porous titanium cages than for PEEK cages at both time points measured (McGilvray et al., The Spine Journal, 2018). In patients, a network meta-analysis pooling 47 studies and 4,881 patients found 3D-printed titanium cages had higher odds of radiographic fusion than PEEK (odds ratio 2.78) and lower odds of subsidence, the cage sinking into the endplate (odds ratio 0.45) (Perez Rodriguez Garcia et al., European Spine Journal, 2026). A multicenter randomized trial found that a fully porous titanium cage with no graft window fused as well as the windowed version (Ham et al., Journal of Bone and Joint Surgery, 2024). One Mayo Clinic cohort ran the other way on subsidence, finding more of it with porous titanium on first analysis, even as reoperation for a failed fusion ran lower — 2.0% with titanium versus 11.4% with PEEK (Levy et al., Global Spine Journal, 2026). Size, shape, and surface all ride along with the material choice, and longer-term data are still accumulating.
Cobalt-chromium
Cobalt-chromium is chosen when rigidity is the surgical goal — most often as rods in larger reconstructions and deformity corrections, where holding a corrected position against strong forces matters more than flexing with the bone. It is the stiffest material on the list by either set of measurements above.
That stiffness carries an imaging cost. In standardized construct testing, cobalt-chromium produced the largest MRI artifacts of the materials studied (Heinrich et al., European Radiology, 2021). The artifact around the hardware is wider, which matters when follow-up imaging must show the nerve or the cord next to the metal.
PEEK
PEEK is a high-performance polymer, machined into cages — including the cages placed in the neck during an ACDF. Its measured stiffness of 3.84 GPa sits almost exactly on cancellous bone's 3.78 GPa (Heary et al., Journal of Spine Surgery, 2017). A PEEK cage therefore loads the graft much the way bone itself would.
Its limitation is biological rather than mechanical. PEEK is inert: bone does not bond to it, and fusion across a PEEK cage happens through the graft inside it rather than into the cage wall.
Modern cages address that inertness at the surface. A plasma-sprayed titanium coating gives bone a surface it can bond to directly: in an animal study, bone grew directly onto titanium-coated PEEK, and the bond strengthened with time (Walsh et al., The Spine Journal, 2015). In patients, a multicenter randomized trial found titanium-coated PEEK fused earlier, with significantly higher fusion at six months (Hasegawa et al., Spine, 2020). A second randomized trial found no difference in final fusion rates between coated and uncoated PEEK through 24 months (Schnake et al., European Spine Journal, 2021). Hydroxyapatite — the mineral bone is made of — can instead be blended through the PEEK itself; in an animal study, it produced more direct bone apposition than PEEK alone (Walsh et al., Clinical Orthopaedics and Related Research, 2016). The supported claim for these surfaces is direct bone bonding and possibly earlier fusion — not a higher final fusion rate.
Carbon fiber
Carbon fiber–reinforced PEEK is radiolucent: it produces no artifact on CT or MRI, allowing better radiographic assessment on follow-up imaging. In most spine surgery that is a convenience. In tumor surgery it can be the deciding property, because follow-up imaging watches for recurrence at the operated level and radiation treatment is planned from those same images.
In a study of 35 spinal-tumor patients instrumented with carbon fiber–reinforced PEEK screws, the screws reduced image artifacts on CT and MRI and were a valuable option where postoperative imaging and radiation planning are necessary (Ringel et al., World Neurosurgery, 2017). A later review of the same material in tumor fixation found a statistically significant improvement in CT imaging quality compared with titanium, whose hardware complicates both the images and the radiation dose planning (Ho et al., Journal of Spine Surgery, 2024). Its stiffness can also be tuned by how the fibers are laid: the composite measured 17.94 GPa, close to cortical bone at 14.64 GPa (Heary et al., Journal of Spine Surgery, 2017).
Will my hardware set off the airport metal detector?
It depends on the machine. Airports use three different tests, and the findings differ by machine type.
The walk-through arch is the least sensitive of the three. In a study that tested modern spinal implants both on the bench and in 40 patients who had them, the standard arch detector did not detect any implant — not in patients, and not on the bench up to a titanium mass of 215 g (Chinwalla and Grevitt, Spine, 2012). In a children's scoliosis cohort, the only patients who triggered a walk-through detector — 5.6% of the cohort — had stainless steel instrumentation from before 2008; no patient with titanium instrumentation triggered any detector in any venue (Fabricant et al., Journal of Child Orthopaedics, 2013). Size matters as well: in a general orthopedic study, patients with four or fewer screws and nothing else were never detected by the arch (0 of 7), while joint replacement prostheses — a much larger mass of metal — were detected every time (Obremskey et al., Journal of Orthopaedic Trauma, 2007).
The handheld wand is a different instrument. Wanded directly over the body, it detected all posterior spinal implants in the same bench-and-patient study, at a distance of 5 cm (Chinwalla and Grevitt, Spine, 2012). Posterior hardware can register under a wand at close range.
The full-body scanner is the third test. In a survey of 182 spinal-implant patients, 40% reported a detector alert at an airport requiring additional screening, and 86% of those alerts came from full-body scanners, which were far more likely to flag implants than traditional metal detectors (odds ratio 5.1) (Arhewoh et al., The Spine Journal, 2022). Material played a smaller role than machine type in that survey. In a children's scoliosis cohort, cobalt-chrome constructs triggered additional screening in 24% of patients (8 of 33), compared with none of 17 with stainless-steel rods (Woon et al., Spine Deformity, 2015).
The practical guidance is TSA's. Tell the officer you have an implant; you may show a notification card or other medical documentation; and if you alarm, expect a pat-down. A card documents the hardware, but it does not waive screening: in the children's cohort study, implant cards did not prevent further screening once a detector had triggered (Fabricant et al., Journal of Child Orthopaedics, 2013). Carry the card if your surgical team provided one, and allow a few extra minutes.
Can I still have an MRI?
Yes. Spinal implants are classified as MR Conditional: MRI is performed with them routinely, under the conditions the implant system was tested for. "MRI-safe" is a different regulatory designation, and it does not apply to these implants.
The testing behind that classification is standardized. Pedicle-screw-and-rod constructs from 130 to 430 mm long, in titanium, titanium alloy, and cobalt-chromium, were tested at both common scanner strengths (1.5 T and 3 T) with the patient lying on their back: the maximum heating measured in the magnetic field was 1.3 K, and the magnetically induced torque and displacement force were negligible (Heinrich et al., European Radiology, 2021). MRI after this surgery is performed under defined, tested conditions. The practical effect is image artifact around the hardware — widest with cobalt-chromium, narrowest with titanium — and scan settings are adjusted to read around it.
Summary
Four materials cover nearly all spinal hardware, and the choice follows the problem the operation has to solve: titanium for most screws, rods, and cages, now printed porous where matching bone's stiffness and giving bone a surface to grow into matter most; cobalt-chromium where a correction must be held rigidly; PEEK where a cage should share load with bone, with coated or integrated surfaces where direct bone bonding matters; and carbon fiber where follow-up imaging must remain unobstructed.
Frequently asked questions
What are spinal implants made of, in plain words?
Four materials account for nearly all modern spinal hardware. Titanium makes most screws, rods, and cages. Cobalt-chromium makes the stiffest rods, used when a correction must be held firmly. PEEK — polyetheretherketone, a medical polymer rather than a metal — makes many cages, especially in the neck and lower back. Carbon fiber–reinforced PEEK appears mainly in tumor surgery, where its radiolucency is the point: it produces no artifact on CT or MRI. Older stainless steel still exists in patients operated on years ago but has largely been replaced in current practice.
Why does it matter how stiff an implant is?
Because bone stays strong by carrying load, and a much stiffer implant can carry the load instead — stress shielding. Stiffer implants are more likely to stress shield the bone graft, and potentially result in a pseudarthrosis, a fusion that does not heal (Heary et al., Journal of Spine Surgery, 2017). Stiffness is measured as elastic modulus. Cancellous bone measured 3.78 GPa and the denser cortical shell 14.64 GPa. Materials near that range share the work with bone; materials far above it, like the metals, take more of the work onto themselves.
What is PEEK, and why is it used for cages?
PEEK is short for polyetheretherketone, a high-performance medical polymer — not a metal. It is used for interbody cages, the spacers placed between vertebrae, because its stiffness closely matches the spongier inner bone it sits between: 3.84 GPa for PEEK against 3.78 GPa for cancellous bone (Heary et al., Journal of Spine Surgery, 2017). That match means the cage shares load with the graft in a way a solid metal cage does not. Its trade-off is that bone does not bond to PEEK itself; fusion happens through the graft the cage holds. Modern variants address that with titanium-coated or hydroxyapatite-integrated surfaces, which bone can bond to directly.
Show all questionsShow fewer questions7 more ▾
What is carbon-fiber hardware for?
Carbon fiber–reinforced PEEK is radiolucent: it produces no artifact on CT or MRI, allowing better radiographic assessment on follow-up imaging. That property matters most in tumor surgery, where follow-up imaging watches for recurrence at the operated level and radiation treatment is planned from those images. In a series of 35 spinal-tumor patients, carbon-fiber screws reduced artifacts on both CT and MRI (Ringel et al., World Neurosurgery, 2017), and a later review found CT image quality significantly improved compared with titanium fixation (Ho et al., Journal of Spine Surgery, 2024). Outside tumor cases, it is used far less often.
What is 3D-printed porous titanium — and how is it different from solid titanium?
It is the same titanium, printed as an open lattice instead of machined as a solid. The pores change two things. First, stiffness: a 2026 review of printed porous titanium cages reports an effective modulus of approximately 2–6 GPa across designs — porosity of 60%–80%, pore sizes of 300–700 μm — down from the 50–110 GPa of solid titanium, and into the range of PEEK and cancellous bone (Yang et al., Frontiers in Bioengineering and Biotechnology, 2026). Second, biology: the pores give bone a structure to grow into, which solid titanium and PEEK do not offer in the same way.
Will my hardware set off the airport metal detector?
It depends on the machine. The walk-through arch is the least sensitive test: in a study of 40 patients with modern spinal implants, the arch detected none of them, and bench testing found no detection up to 215 g of titanium (Chinwalla and Grevitt, Spine, 2012). Full-body scanners flag implants more readily — in a survey of 182 implant patients, 40% reported an alert requiring extra screening, and 86% of those alerts came from full-body scanners (Arhewoh et al., The Spine Journal, 2022). If you are flagged, expect a pat-down or a wand check; tell the officer about your hardware and allow a few extra minutes.
Should I carry an implant card when I fly?
It can help you explain, and TSA's guidance is to inform the officer about an internal implant and to offer a notification card or other medical documentation if you have it. What a card does not do is waive screening: in a study of children with spinal instrumentation, implant cards did not prevent further screening once a detector had triggered (Fabricant et al., Journal of Child Orthopaedics, 2013). Carry the card if your surgical team gave you one, mention the implant before you are screened, and allow a few extra minutes for screening.
Can I still have an MRI with spinal hardware?
Yes. Spinal implants are MR Conditional — MRI is performed with them under the conditions the implant was tested for, which is the accurate term; implants are not described as "MRI-safe." In standardized testing of screw-and-rod constructs at both 1.5 T and 3 T, the maximum heating measured was 1.3 K, and the magnetic torque and pulling force on the constructs were negligible (Heinrich et al., European Radiology, 2021). Tell the imaging team about your hardware beforehand, as you would for any implant — the scan then proceeds under its tested conditions.
Will my implants block the view on follow-up scans?
Metal distorts the image immediately around it; the degree varies by material. In standardized construct testing, titanium alloys produced the smallest artifacts of the metals studied and cobalt-chromium the largest (Heinrich et al., European Radiology, 2021). Carbon fiber–reinforced PEEK produces no artifact on CT or MRI, so follow-up images remain unobstructed where the hardware sits — which is why it is chosen for tumor patients whose surveillance depends on those images (Ringel et al., World Neurosurgery, 2017). Scan settings are adjusted to read around ordinary metal artifact.
Does the choice of material change my chance of a successful fusion?
It is one factor among several, and the newest evidence concerns cage design. A 2026 network meta-analysis of 47 studies and 4,881 patients found 3D-printed titanium cages carried higher odds of radiographic fusion than PEEK cages (odds ratio 2.78) and lower odds of the cage subsiding into the bone (odds ratio 0.45) (Perez Rodriguez Garcia et al., European Spine Journal, 2026). The findings are not one-directional in every study — one cohort found more subsidence with porous titanium even as reoperation for failed fusion ran lower (Levy et al., Global Spine Journal, 2026). Fusion also depends on bone quality, the graft, and the stability of the construct.