Additive Manufacturing for Medical Devices: Material Selection, Biocompatibility, and Regulatory Pathway

AM has genuine structural advantages for medical implants: patient-specific geometry from a CT scan, bone-ingrowth lattice structures that would be impossible to machine, and low-volume production economics that make custom implants financially viable. This article covers the material options, porosity and surface finish targets, post-processing requirements, and the regulatory framework that applies to implantable AM components.

Why AM suits medical

Three properties of AM make it particularly relevant for implants and surgical devices.

Patient-specific geometry. A CT scan or MRI can be converted to a STL file within hours. LPBF or EBM can print the exact anatomy of a patient's acetabulum, cranial plate, or mandible reconstruction at no geometric surcharge. Conventional machining of patient-specific parts is extremely expensive; AM makes it routine.

Osseointegration lattices. Bone grows into porous scaffolds with interconnected pore networks in the 300–600 µm range. These structures — trabecular-like lattice infills on implant surfaces — dramatically improve long-term fixation and reduce stress shielding. They cannot be manufactured by any subtractive method. EBM and LPBF produce them routinely.

Low-volume production economics. Spinal implant product families may run 5–50 parts per size variant per year. AM eliminates tooling, making low volumes economically viable. The entire business model of the modern orthopaedic implant sector has shifted toward AM for this reason.

Material selection framework

Material selection for implantable AM is constrained by biocompatibility, mechanical properties, and applicable standards. The four dominant choices are:

Ti-6Al-4V ELI (Grade 23)

The workhorse material for structural load-bearing implants — spinal interbody cages, acetabular cups, tibial trays, cranial plates.

Standards: ISO 5832-3, ASTM F136 (wrought), ASTM F3001 (AM-specific), ASTM F2924 (LPBF).

Why ELI over standard Grade 5 (GR5): ELI (Extra Low Interstitial) grade has tighter limits on oxygen (≤0.13 wt% vs ≤0.20 wt%), nitrogen, carbon, and iron. This reduces the density of hard brittle interstitial phases, improving ductility and fatigue crack propagation resistance. For a structural implant that must survive 10+ years of cyclic loading, the fatigue performance improvement is significant. Typical ELI yield strength is 795–875 MPa depending on build direction and heat treatment — sufficient for most skeletal load cases.

Biocompatibility: Ti-6Al-4V has an extensive clinical record and is well-characterised under ISO 10993 (biological evaluation of medical devices). The native TiO₂ oxide layer is stable and non-reactive in vivo.

Typical AM process: LPBF and EBM both qualify. EBM is preferred for complex porous acetabular cups because the hot powder bed reduces residual stress in intricate lattice structures.

CoCrMo (ASTM F75 / ISO 5832-4)

Preferred for bearing surfaces in articulating implants — femoral heads, knee condyles — where tribological performance is the primary requirement.

Standards: ISO 5832-4, ASTM F75 (cast), ASTM F1537 (wrought), ASTM F2979 (AM powder).

Why CoCrMo for bearings: CoCrMo has significantly better wear resistance than titanium alloys. Titanium is a poor bearing material — it galls and generates abrasive debris. CoCrMo articulating against UHMWPE (ultra-high molecular weight polyethylene) or CoCrMo-on-CoCrMo are the standard bearing couples in hip and knee replacements.

Strength: LPBF CoCrMo can achieve UTS >1000 MPa, yield >700 MPa. Hot isostatic pressing is typically required to achieve acceptable fatigue performance.

Limitations: Concerns exist about ion release (Cr, Co, Mo) in metal-on-metal bearing applications. Regulatory and clinical focus has shifted away from large-diameter metal-on-metal bearings; CoCrMo remains standard for femoral head components against polyethylene liners.

Commercially Pure Titanium Grade 2 (CP-Ti GR2)

Used for dental implants, maxillofacial implants, and applications where corrosion resistance is prioritised over maximum strength.

Standards: ISO 5832-2, ASTM F67.

Properties: Lower strength than Ti-6Al-4V (UTS 400–540 MPa, yield 275–410 MPa), but superior corrosion resistance and superior osseointegration characteristics. The absence of alloying elements (aluminium and vanadium) eliminates any concern about ion release from those elements.

AM considerations: CP-Ti is more susceptible to oxygen pickup during processing than Ti-6Al-4V, requiring tighter atmosphere control. LPBF with well-maintained inert atmosphere is appropriate. EBM is also viable.

PEEK (Polyether Ether Ketone)

The primary polymer material for load-bearing implants, most commonly spinal interbody fusion cages.

Why PEEK for spinal cages: PEEK's elastic modulus (~3.5–4.5 GPa) closely matches that of cortical bone (~15–25 GPa; trabecular bone is 0.1–5 GPa). Metal cages with much higher modulus create stress shielding at the vertebral endplate, impeding fusion. PEEK eliminates this problem.

Radiolucency: PEEK is radiolucent — it does not appear on X-ray or CT. This allows post-operative imaging to assess fusion progress without the metal artefact that obscures bone grown around titanium cages. Some PEEK cages include titanium markers for radiographic positioning.

AM process: High-temperature FDM (PEEK requires print temperatures of 380–420°C and bed temperatures of 120–200°C) or PEKK variants. SLS of PEEK is at pre-commercial stage for medical. AM PEEK implants are used but the dominant manufacturing route remains machining from PEEK rod stock; AM is growing for custom geometries.

Surface limitations: AM PEEK surfaces have lower osseointegration than titanium lattices. Hybrid implants — PEEK body with titanium coating or integrated titanium lattice end-plates — are increasingly common.

Porosity requirements for implants

Porosity requirements split into two entirely different regimes depending on function:

Structural porosity (must be minimised)

For the load-bearing matrix of an implant, porosity must be below 0.5% by volume, and below 0.1% for fatigue-critical applications. This is the same requirement as aerospace fracture-critical parts.

• Residual pores act as fatigue crack initiation sites
• Regulatory qualification (FDA, CE) for implants requires documented evidence of density
• Accepted methods: Archimedes density, X-ray CT (pores >50 µm), metallographic cross-section
• LPBF optimised parameters for Ti-6Al-4V ELI routinely achieve >99.9% density (< 0.1% porosity)

HIP is the most reliable route to achieving <0.1% porosity even if the as-built density is marginal. ASTM F3001 and F2924 recommend HIP as standard for fatigue-critical implants.

Scaffolded open porosity (must be maximised in the right range)

Bone ingrowth into a porous surface requires interconnected pores of 300–600 µm. This is intentional, designed porosity — entirely separate from the structural residual porosity discussed above.

Key requirements for effective osseointegration scaffolds:

• Pore size: 300–600 µm optimal. Below 100 µm: fibrous tissue ingrowth only. Above 800 µm: reduced surface area, poor initial fixation.
• Interconnectivity: All pores must connect to adjacent pores. Dead-end pores do not support bone ingrowth.
• Porosity level: 60–80% for the lattice layer; the solid substrate beneath must retain full structural density
• Surface area: Higher surface area per unit volume improves cell attachment. Diamond, gyroid (TPMS), and octet-truss lattices each have different surface area characteristics.

The critical point is that a single implant typically contains both regimes: a fully dense load-bearing core and an intentionally porous surface layer. Design and qualification must address both.

Surface finish requirements

Surface finish requirements for medical AM vary sharply by the function of the surface:

Bearing surfaces require Ra < 0.8 µm. LPBF as-built surfaces are Ra 8–20 µm. Bearing surfaces must be machined and polished. The final finish step for metal-on-metal bearings is typically superfinishing to Ra < 0.05 µm.

Osseointegration surfaces should be Ra 2–4 µm. This roughness range has been shown in clinical and in vitro studies to maximise osteoblast spreading and mineralisation. The ideal surface is not smooth. Bead blasting or acid etching after AM processing is commonly used to achieve this range. The AM lattice structure itself contributes geometric roughness at a larger scale, but the micro-roughness of the strut surfaces also matters.

Ra > 4 µm increases bacterial adhesion risk. Rough surfaces trap bacteria and make sterilisation less effective. For implant surfaces in contact with soft tissue (not bone ingrowth lattices), keeping Ra below 4 µm is a clinical requirement, not merely cosmetic.

Post-processing: what is mandatory

HIP

For any load-bearing implant that must pass fatigue qualification, HIP is effectively mandatory. ASTM F3001 (Ti-6Al-4V ELI AM) and F2924 (Ti-6Al-4V LPBF) both include HIP as the standard route to achieve the fatigue performance data in the standard. Building to a lower porosity level using process optimisation alone and arguing that HIP is unnecessary requires substantial fatigue data and regulatory justification.

Typical HIP cycle for Ti-6Al-4V: 900–950°C / 100 MPa / 2 h in argon. This closes residual pores, partially relieves residual stress, and coarsens the alpha-beta microstructure. A subsequent vacuum anneal at 700°C/2h is typically applied after HIP to restore some fine-scale microstructure.

Use the HIP cycle designer to specify cycle parameters by material.

Electropolishing

Used to achieve bearing surface finish requirements and to remove the thin oxide layer and alpha case on titanium, which are artefacts of the AM process. For dental and maxillofacial titanium implants, electropolishing is commonly the final step before packaging.

Cleaning and sterilisation

AM implants must undergo validated cleaning protocols to remove metallic debris, powder particles, and process residues. ISO 17664 governs reprocessing instructions. Implants are typically sterilised by gamma irradiation or ethylene oxide. The cleaning and sterilisation validation is part of the design verification package.

Regulatory pathway

ISO 13485 — Quality management system

The foundation of any medical device regulatory programme. ISO 13485 is the QMS standard for medical device manufacturers, analogous to ISO 9001 but with additional requirements for traceability, risk management, and post-market surveillance. It is required by CE marking (MDR 2017/745 in Europe) and FDA expects its equivalent for 510(k) and PMA submissions.

For AM specifically, ISO 13485 requires that the manufacturing process — including machine qualification, parameter sets, build layout, and powder lot — be treated as controlled processes with documented procedures.

FDA 510(k) vs. PMA

510(k) (Premarket Notification): Demonstrates substantial equivalence to a legally marketed predicate device. Most orthopaedic implants (hip, knee, spine) have predicate devices and follow the 510(k) pathway. The FDA's guidance document "Technical Considerations for Additive Manufactured Medical Devices" (2017, updated 2024) is the key reference.

PMA (Premarket Approval): For Class III devices with no predicate — higher-risk, novel implants. Requires clinical data. More expensive and longer than 510(k). Relevant for truly novel AM implant architectures where no predicate exists.

FDA AM-specific guidance areas: Build and design documentation, material characterisation (including powder lot testing), process validation, post-processing validation, device testing (mechanical, biocompatibility, sterility), and labelling. The FDA expects the following specific to AM: the 3D model file management, slicing and build preparation records, machine qualification documentation, and in-process monitoring records.

Design History File (DHF)

The DHF is the complete record of the design and development process, required under 21 CFR Part 820 (FDA QSR) and ISO 13485. For an AM implant it must include:

• Design inputs (anatomical requirements, mechanical requirements, biocompatibility requirements)
• Design outputs (CAD files, build parameter sets, post-processing procedures)
• Design verification testing results (mechanical tests, dimensional, surface finish, biocompatibility)
• Design validation (clinical or simulated use data demonstrating the device meets user needs)
• Design changes and their justification

The DHF is audited during FDA inspections and CE technical file review. Gaps in the DHF are the most common finding in 510(k) deficiency letters.

Device Master Record (DMR)

The DMR is the manufacturing recipe — the complete set of documents needed to produce the device in production. For AM implants it includes:

• CAD model version control records
• Build parameter set (locked and version-controlled)
• Machine qualification records
• Material specification (powder grade, certificate of conformance requirements)
• Post-processing procedures (HIP cycle, heat treatment, cleaning, electropolishing, packaging, sterilisation)
• Inspection and test procedures

Risk management: ISO 14971

ISO 14971 requires systematic identification, analysis, evaluation, and control of all risks associated with the device throughout its life. For AM implants this means explicitly addressing risks that are unique to AM:

• Parameter drift leading to porosity above specification
• Powder contamination between builds affecting biocompatibility
• Surface roughness variation affecting bacterial adhesion
• Dimensional variation from build-to-build affecting fit and fixation
• Fatigue life variation between builds

Each risk requires a documented control measure and residual risk evaluation.

Traceability requirements unique to medical AM

Medical AM has traceability requirements beyond what is expected in industrial AM:

Powder lot traceability: Every implant must be traceable to the specific powder lot used. Powder lot certificates of conformance (PSD, chemistry, oxygen content, Hall flow) must be retained. This drives the decision about powder reuse — reused powder must be retested and recertified; many medical AM manufacturers use virgin-only powder for implants.

Parameter qualification: The build parameter set must be validated (mechanical testing, density testing) and locked. Any parameter change triggers a revalidation. The qualification data must be retained in the DMR.

Build record per part: Each individual implant must have a build record: which machine, which build job, position in the build (XY location, Z height), powder lot, build date, operator, and machine maintenance status. This is the device history record (DHR) and is required per 21 CFR 820.184.

Post-processing traceability: HIP cycle records (temperature, pressure, time, atmosphere), heat treatment furnace records, and electropolishing bath chemistry and time must be retained per part.

EBM vs. LPBF for medical

Both processes are used for medical implants. Key differences:

EBM's near-zero residual stress makes it attractive for the complex lattice acetabular cups used in hip revision surgery — intricate lattice structures that would crack or distort under LPBF residual stress. The vacuum atmosphere also eliminates oxygen pickup risk for titanium.

LPBF's better resolution is critical for dental applications where fine features at sub-millimetre scale are required. For complex spinal cage geometry with thin walls, LPBF's resolution advantage justifies the additional post-processing steps to manage residual stress.

Related tools

• HIP cycle designer — specify HIP cycle parameters for Ti-6Al-4V ELI, CoCrMo, and other medical alloys
• Fatigue life estimator — estimate fatigue performance before and after HIP and surface treatment
• Surface roughness reference — Ra by process, orientation, and surface condition
• NDT selector — choose the right inspection method for implant qualification
• Porosity predictor — estimate residual porosity from LPBF process parameters