Porous 3D Printed PEEK: How a Porous Spinal Implant Is Made

• 1
  What the First Porous 3D Printed PEEK Spinal Implant Proved
• 2
  Why Pores Change Load Transfer and Bone Ingrowth
• 3
  How Porous PEEK Is Printed, and Where It Fails
• 4
  Where Machined PEEK Still Beats a Printed Lattice
• 5
  What to Put on the Drawing and What to Ask a Supplier
• 6
  Porous 3D Printed PEEK vs Machined PEEK: Feature by Feature

In April 2023 a US surgical team implanted a spinal device built from porous PEEK, printed rather than machined. The material was Evonik VESTAKEEP, a medical-grade polyetheretherketone. The claim behind the surgery was not that PEEK is new. PEEK has sat inside interbody cages for years. The claim was that a printed PEEK body can carry controlled porosity and still hold its mechanical job.

That distinction matters for anyone who writes implant drawings. A machined PEEK cage is dense, and its stiffness sits close to cortical bone. A porous 3D printed PEEK cage is mostly air by volume, so the same outer envelope weighs less and behaves like a lattice. Bone can grow into the pores instead of only fusing around the surface.

The trade is not free. Porosity lowers stiffness, ultimate load and fatigue life at the same time it improves bone ingrowth. Designers who want the ingrowth have to give up some of the safety margin that a solid cage gives them for nothing. That is the whole engineering story in one sentence.

For a shop like ours the interesting part is the handoff. Printing sets the pore network. Machining still sets the seating face, the screw threads and the lordotic angle. Both operations usually land on the same drawing revision.

Bone adapts to the load it sees. A stiff metal cage shields the graft from stress, and the graft resorbs. Titanium cages carry this problem because their modulus runs well above bone. PEEK sits much closer, which is why dense PEEK cages have been used for fusion for years.

Pores add a second effect. An open network with pore sizes in the 300–600 μm range gives cells a path into the body of the implant. When bone forms inside those pores, the interface is no longer a flat contact. Load transfers through a graded structure rather than a single plane.

Fully interconnected pores matter more than total porosity. A closed pore is dead volume. It adds weight and cost without helping fixation. This is where printing earns its place: the pore network is designed, not incidental to a foaming process.

The boundary condition is pore size against fatigue. Open pores act as stress concentrators. Push pore size up for better ingrowth and fatigue strength falls. Most published design windows settle near 60–70% porosity with 300–500 μm pores for spinal fusion devices.

Medical-grade PEEK prints mainly by high-temperature material extrusion, with nozzle and chamber temperatures well above the range used for PLA or ABS. Layer heights typically sit between 0.1 mm and 0.2 mm. Below that, print time climbs fast and the pore network closes up.

The hard part is heat history. Every thermal cycle changes crystallinity, and crystallinity drives stiffness and toughness. A printed lattice with uneven cooling has uneven mechanical properties, part to part and layer to layer. Validation has to follow the process, not just the geometry.

Sterilization is the next gate. Steam autoclave cycles above the glass transition of PEEK will relax residual stress and can distort a thin lattice. Gamma and ethylene oxide are the usual routes for printed PEEK devices. Confirm the cycle before you lock the design.

Where printing fails: thin solid walls, sharp internal corners, and any feature that needs a sealing surface. Those are the places where a printed part leaks, warps or cannot hold a tolerance. Machine them instead, or design them out.

Machined PEEK holds ±0.005 mm on a seating face. A printed lattice does not, and no amount of process control changes that. If your drawing carries a tight flatness callout or a threaded hole for a set screw, that feature belongs on a mill.

PEEK also machines cleanly at the right parameters. Sharp tooling, high surface speed and generous coolant or air blast keep heat out of the cut. Heat is the enemy: it smears the surface, raises crystallinity at the skin, and leaves a finish that looks fine and measures wrong.

Solid machined PEEK is the right answer for trial implants, instrument trays, drill guides and anything that has to be sterilized repeatedly and stay dimensionally stable. It is also the faster path to a first article, because no print validation is needed to cut one.

A practical hybrid works well: print the porous fusion body, machine the mating plate and screw holes, then assemble or press-fit. Each process does what it is good at. Neither one has to pretend it can do the other's job.

Separate the critical-to-function dimensions from the cosmetic ones. Pore size, porosity and interconnectivity are process specifications, not tolerances. State them as ranges and say which one wins if they conflict. Machining tolerances go on the solid features only.

Ask for the material certificate and the lot traceability on the PEEK stock. Medical-grade resin is not the same as industrial-grade rod, and the difference is not visible in a photo. For machined parts, ask whether the shop runs ISO 13485:2016 and can supply inspection reports.

Ask how the supplier validates the pore network. Optical measurement on a cut section is common. Micro-CT gives more data and costs more. Neither is useful if the sampling plan is one part from one build.

Finally, ask about the handoff between printing and machining. If two vendors share one drawing, someone has to own the datum scheme. That is usually where a hybrid implant program loses a week.