Metal allergy



You’ve referred your patient whom you’ve been following for a number of years for knee replacement surgery and the operation was a success. Clinical examination was excellent with a range of motion of 0 to120 degrees, the x-rays look great and the patient has been given a yearly review or even a discharge from follow-up. Yet the patient feels dissatisfied – the knee is persistently swollen, mildly warm and the skin over it is a little dry and scaly. The patient does not have fever. Alarmed that this may be an infected prosthesis you refer the case back to the orthopedic surgeon who does an aspiration – it is aseptic. The patient then undergoes a full course of antibiotic treatment but the pain persists and the swelling continues.

If this is not a scenario that you have faced rest assured that it is only a matter of time. As with many in the orthopedic community, the subject of metal hypersensitivity in joint replacement surgery is not commonly discussed. Rather than suspect that this is some industry-wide cover-up to rival covert arms deals and fund mismanagement schemes the answer to this is much simpler. The evidence it seems is not completely clear.


The metals used in a standard implant system has to balance a few seemingly conflicting requirements. First you need a highly biocompatible metal at the bone-metal interface. This has reliably been served with Titanium. Pure titanium however is too soft for implant manufacture and therefore needs to be produced as an alloy with aluminium and vanadium. Similarly on the articulating surface, a harder surface which is polished needs to be provided and this comes in the form cobalt chrome which also has a bit of nickel and molybdenum in the mix.

A review of 3047 patients by ATJ Goon et al has shown that in Singapore, the frequency of positive patch tests to the following metals were nickel 19.9%, chromate 5.6%, cobalt 8.2% and gold 8.3%. That should mean that one in five patients (with Nickel allergies) should develop allergies to joint replacements right? Wrong.

The effects of metal ions on the body are delayed type IV T-cell hypersensitivity, dose-dependent cell necrosis, and mutagenic changes amounting to carcinogenesis. These effects are predictably more common in implant designs that result in metal on metal articulation. Thankfully, and perhaps more by chance than by forethought, the standard articulation couple is cobalt chrome on polyethylene or ceramic on ceramic. Unfortunately, there do exist metal on metal implants in the form of metal on metal hip replacements and hinge knee prostheses (Figure 1) and in this group of patients the incidence of metal hypersensitivity can be as high as 30 percent which proves it right? Wrong.

Figure 1. Twenty-four year-old female with an osteosarcoma in the right distal femur that had been operated at a different hospital twice before. It was decided to perform an en bloc resection of all contaminated tissue and insert a hinged knee prosthesis (b). The hinge mechanism (arrowed) allows for knee articulation in a situation without any ligamentous support. Among other disadvantages the hinge results in high metal wear and liberation. This may predispose to metal hypersensitivity reactions. The surgery was a success (c) and the patient was able to return to full weight bearing with full range of motion despite being virtually ankylosed before this last operation. The patient however had prolonged healing of the wound with an eczematous wound. Although this eventually healed, the patient continues to have dry eczematous skin over the incision even 2 years later. She has never been shown to be infected despite multiple attempts at aspirations and culture and sensitivity tests. She is not on prolonged courses of antibiotics and continues to be on close follow-up.

Historically, testing for delayed-type hypersensitivity has been conducted in vivo by skin testing (that is, so-called patch testing or intradermal testing) and in vitro by lymphocyte transformation testing (LTT) and leukocyte migration inhibition testing (termed LIF or MIF testing). While there are general patch-testing protocols and commercial kits for a variety of commonly antigenic substances, there is continuing concern about the applicability of skin testing to the study of immune responses to implants; in particular, there is a lack of knowledge about, and availability of, appropriate metal challenge agents. Unlike periprosthetic exposure, patch testing involves incorporating an antigen (for example, 1% aqueous nickel sulfate) in a carrier, such as petrolatum, and exposing this to dermal tissue by means of an affixed bandage. After exposure for approximately forty-eight to ninety-six hours, reactions are graded on a scale of 1 (mild or absent response) to 4 (severe red rash with small and possibly encrusted weeping blisters). This is quite different from the weeks to months of constant exposure prior to typical reports of eczemic reactions to orthopaedic implants. Additionally, the haptenic potential of metals on open-testing dermal contact (in which dermal Langerhans cells are the primary hypersensitivity effector cells) is likely quite different from that in a closed periprosthetic in vivo environment. Other concerns are that the diagnostic utility of patch testing possibly could be affected by immunological tolerance (that is, suppression of dermal response to implants) or by impaired host immune response and that the testing possibly could induce hypersensitivity in a previously insensitive patient. Moreover, even if patch testing were a biologically reliable means of assessing metal sensitivity, no suitable standardized battery of tests of relevant metals currently exists. In vitro proliferation testing (also known as lymphocyte transformation testing or LTT) involves measuring the proliferative response of lymphocytes following activation. A radioactive marker is added to lymphocytes along with the desired challenge agent. The incorporation of radioactive [H3]-thymidine marker into cellular DNA upon division facilitates the quantification of a proliferation response through the measurement of incorporated radioactivity after a set time-period. On the sixth day, [3H]-thymidine uptake is measured with use of liquid scintillation. The proliferation factor, or stimulation index, is calculated with use of measured radiation counts per minute (cpm): proliferation factor = (mean cpm with treatment)/ (mean cpm without treatment). Although the use of proliferation testing in the assessment of metal sensitivity is less popular than patch testing, it has been well established as a method for testing metal sensitivity in a variety of clinical settings. The use of lymphocyte transformation testing for implant-related metal sensitivity has been limited, and therefore few conclusions can be drawn. These investigations indicate that metal sensitivity can be more readily detected by lymphocyte transformation testing than by dermal patch testing. Such reports seem to indicate that, compared with dermal patch testing, lymphocyte transformation testing may be equally or better suited for the testing of implant-related sensitivity. So unfortunately there does not exist a reliable pre-emptive protocol to predict the patient who is going to develop metal allergy to an implant- doing it is not cost-effective, does not provide conclusive evidence and worse may sensitize the patient to developing it in the first place.

Approach to the patient with metal allergy

In the author’s practice as a joint replacement surgeon in a tertiary joint replacement center he faces patients with this scenario in two settings. The patient who has had joint replacement surgery and now presents with metal allergy and the patient with known metal allergy who needs surgery.

The patient with metal allergy after joint replacement surgery

The opening scenario of this article pretty much sums the presentation of patients with metal allergy with implants in situ – they essentially present as infections but without fever or warmth. Variations on the theme include patients with persistent sterile drainage and eczema in the lower limb over the incision. The author’s additional clinical practice as a musculoskeletal oncologist involving tumor resections and reconstructions also allows the observation that these occurrences occurs much more commonly in older patients and patients who have not received chemotherapy- possibly because the immunosuppression involved in the latter allows the patient ironically to tolerate the metal better akin to immunosuppression in organ transplant.

Such a patient should receive a full clinical evaluation. He should have an aspiration done preferably by an orthopedist. There should on no account be an attempt at ‘reducing the inflammation’ by injecting the knee with a steroid. This is an unproven measure that may inadvertently introduce an infection and convert a difficult situation into a disaster. Fluid should be cultured preferably in medium. Infection must be excluded.

Assuming there is no infection one of 2 scenarios exists. In the more straightforward situation, the symptoms are severe enough to warrant a revision. The implants are removed, the joint thoroughly washed and an alternative articulating system re-implanted. Hence, if the previous implant was a cobalt chrome implant then a non-cobalt chrome hypoallergenic implant may be inserted. This approach has had some success in scattered reports recently in the literature in the knee. In the hip the revision to a non metal-based articulation (eg. ceramic on ceramic) may be attempted, however, the evidence to support this is limited.

If the patient has only mild symptoms the author’s preference is to manage the patient conservatively and follow the patient closely (Figure 1). When loosening develops there will be no choice but to revise the implants. If there is good bone stock at that point, ceramic type prosthesis may be used. However if there is poor bone stock one may be forced to use a non-ceramic implant – at present no revision system exists to allow non-metal articulation. The results are ultimately likely to be dissatisfying- the patient is likely to develop metal allergy problems again (Figure 2).

Figure 2. Forty-one year-old male had chronic left knee pain which was found to be due to an osteosarcoma in the medial femoral condyle (a). A biopsy was performed and metal rods inserted to prevent collapse of the condyles while he underwent chemotherapy (b). He subsequently had the tumor resected and the defect replaced with a hinged mega-endoprosthesis (c). He developed an eczematoid eruption over the left lower limb about 3 months later (d). There was no fever. The presumptive diagnosis was metal allergy and the patient observed. Unfortunately he subsequently presented to a different specialist who put in a needle to aspirate the joint in good intention. This necessarily had to be done through unhealthy eczematoid skin. The culture was negative but the patient developed an infection from the needle insertion. He underwent an exchange arthroplasty and is currently without infection but with a relapse of eczema.

The patient with known metal allergy who needs surgery

In this patient there exists a ready non-metal articulation initiative in the form of a ceramic on ceramic bearing in the hip (Figure 3). There is still the risk of metal allergy to the stem and acetabular shell but the evidence to suggest this is elusive – it appears to be safe. Most implant companies produce some form of ceramic on ceramic implant. The ceramic on polyethylene option has also been popular but the rationale other than cost considerations is unclear.

Figure 3. Allergies in hip surgery. It is key to avoid metal articulations in this case. The options are therefore to use ceramic on ceramic or ceramic on polyethylene articulations (a). Theoretically there may be ion liberation from the head stem junction or the liner shell junction but this is negligible. There is no ceramic coated implant in the hip available presently. Hip revisions allow similar articulation options. In this patient with multiple allergies a ceramic on ceramic option was elected. Three years after surgery there has been no signs of allergy (c).

In the knee, where there appears to be more evidence, one may take the approach that metal skin allergies do not equate to metal allergies in prosthesis. This would probably be a defensible stand medico-legally as there is no good evidence either way. There are multiple studies that suggest an association between metal allergy on the skin and that to implants. The subcutaneous placement of prosthesis in the knee as opposed to deep placement in the hip may have some bearing on the occurrence. In the past the orthopedic surgeon in this situation would either have to recommend that the patient not have surgery or with full disclosure have the surgery expecting that there may be complications.

Thankfully implant companies have now considered that the situation is sufficiently prevalent to warrant a non-allergenic knee replacement solution. This generally fall under 2 categories: transformation of standard metal alloys (eg. Zimmer and Stryker) or having an inert substance for articulation (BBraun and Smith and Nephew). As this is a rare situation one can only make recommendations based on one’s own experience and in the author’s practice he uses the line of AS (allergy solution) implants from BBraun (Figure 4). These are Zirconium coated implants and result in no metal ion elution. In theory, the coating could wear but it is known that ceramic wear characteristics are excellent. In the author’s limited experience it has reasonable ease of use and implantation. It is based on the Colombus knee design with asymmetric femoral radii and all polyethylene tibial tray inserts and a zirconium coated tibial component. Knee design is in line with present long lasting knee designs.

Figure 4. Allergy protection knees typically are based on standard proven designs that are given a hypoallergenic surface treatment, in this case Zirconium. This imparts a remarkable color signature – in this a gold knee (a). This patient had advanced osteoarthritis in the medial and patello-femoral joints radiologically. She also had lateral compartment pain clinically. The author performed a total knee arthroplasty using standard techniques (c) and the patient was able to walk the next day (d).There has been no inflammation to date despite being a known case of metal allergy.


The likelihood of increased awareness of metal allergy in joint replacement surgery will lead to increased use of allergy resistant knee designs. This may be a patient derived demand at which time this may become a dominant solution in knee arthroplasty. The extremely low frequency of occurrence however suggests that this will remain a niche product put in by a select group of surgeons with a special interest in the field. Nevertheless, because of the relatively high incidence of metal allergy on patch testing this may well result in higher demand of such solutions. The patient should understand that as an emerging field our experience in this area is limited. Nevertheless, by following the arguments presented in this article it appears that there is some justification for their use in metal sensitive patients.