➢ There is anecdotal evidence that deep-seated metal implants may induce cutaneous sensitivity reactions.
➢ There is controversial evidence with regard to the role of metal hypersensitivity as a cause of ongoing pain or aseptic loosening in patients with conventional orthopaedic implants.
➢ Patients should be counseled, as part of the consent process, with regard to the small risk of potential reactions to metal and the limitations in our understanding of the mechanisms accounting for these reactions.
➢ For patients reporting only mild localized cutaneous reactions, the use of conventional cobalt-chromium and stainless steel implants seems justified without any additional investigations.
➢ For patients who report substantial localized reactions (blistering, hives, extensive rash) or systemic cutaneous reactions to metal, patch testing is recommended, which could then guide the choice of metal implant.
Metal implants are widely used in orthopaedic surgery for elective joint arthroplasty or fracture fixation. Metal alloys also are widely used in other medical specialties for devices such as dental implants, cardiovascular stents, and gastrointestinal wire mesh stents. Metal hypersensitivity, as detected with skin patch testing, is common, with a prevalence of between 10% to 17% in the general population1-5. This raises the concern that hypersensitivity reactions may occur in patients with metallic orthopaedic implants and the possibility that metal hypersensitivity reactions may account for ongoing pain or aseptic loosening following joint arthroplasty6,7.
The purpose of the present review is to examine the current understanding of metal hypersensitivity and its potential clinical importance in orthopaedic surgery. The methods of testing for metal hypersensitivity and the limitations of such tests are also discussed. In this article, we consider conventional arthroplasty implants (metal-polyethylene and metal-ceramic articulations) and fracture-fixation devices. Coverage of the localized reactions reported in association with metal-on-metal arthroplasty is beyond the scope of this article as the local and systemic levels of metal ions achieved following those arthroplasties are substantially higher than those following procedures performed with conventional implants8.
When considering metal hypersensitivity, it is important to distinguish between cutaneous contact sensitivity and sensitivity to deep implanted devices.
Cutaneous Metal Hypersensitivity
Metal hypersensitivity due to cutaneous contact is manifested in the form of local or systemic reactions. Local reactions at the area of contact include rash, urticaria, and swelling. Systemic reactions in response to cutaneous contact are unusual and may present as generalized eczematous rash, swelling, or pruritus9.
Cutaneous hypersensitivity reactions to metal are mediated by activation of the immune system. Four classic types of hypersensitivity reactions have been described. Types I, II, and III are antibody-mediated, and Type IV is cell-mediated. Type-I (anaphylactic) reactions to metal are extremely rare and therefore not discussed in the current report. Cutaneous hypersensitivity to metal is typically a Type-IV reaction. It is a delayed response mediated by the activation of T-lymphocytes and the recruitment of macrophages. Skin contact with metal allows metal particles to cross the stratum corneum and to combine with epidermal proteins, forming metal particle-protein conjugates that act as antigens. These larger conjugate antigens are recognized as a foreign body and are taken up by Langerhans cells (antigen-presenting dendritic skin cells). These cells present the antigen to T-lymphocytes, setting off an inflammatory cascade that involves macrophage activation. Macrophages and inflammatory cytokines are responsible for local or systemic cutaneous hypersensitivity reactions2.
Cutaneous metal hypersensitivity usually becomes apparent following skin contact with metallic objects such as household appliances, tools, or, more commonly, jewelry. Metal jewelry can vary from precious metals such as platinum, gold, and silver to other metals such as nickel, cobalt, chromium, copper, steel, bronze, brass, aluminum, and palladium (Fig. 1)2,10.
Nickel is used in jewelry because it is cheap and durable. The reported amount of nickel in jewelry items varies. White gold consists of 10% nickel and 90% gold, whereas nickel silver (also known as German silver) consists of 20% nickel, 60% copper, and 20% zinc11,12. Nickel is used with white gold and other metal alloys to produce its characteristic white color. It gives a bright finish to costume jewelry items such as earrings and clothing apparel that has metal buttons and zippers. Denim wear and ear-piercings became popular in the 1970s and 1980s, which could be an explanation for the increase of nickel contact hypersensitivity in those decades13-15.
Cobalt and chromium are used in rings, hair clasps, and work tools such as metal cutting tools16-18. Cobalt is also used in jet and gas turbine engines because of its heat-resistant and corrosion-resistant properties19. Occupational contact dermatitis resulting from contact with cobalt and chromium is of concern in certain occupations such as construction workers, hairdressers, barbers, and machine operators20. Aluminum is commonly used in antiperspirants, dental cements, and metal jewelry such as bracelets and earrings. However, aluminum contact hypersensitivity is not as common as nickel contact hypersensitivity21. Palladium is a silvery-white metal used in white gold and in dental crowns. Palladium and nickel tend to cross-react with each other as they have identical coordination geometry in their protein-bound state22. Hence, palladium contact hypersensitivity is commonly seen together with nickel hypersensitivity23. Palladium is now becoming more popular in jewelry because of the increasing price of gold and platinum. Titanium is usually mixed with other metals such as vanadium and aluminum to create durable and lightweight alloys. Titanium alloys are used in jewelry, spectacle frames, medical implants, and sporting goods. A titanium spectacle frame may contain palladium or nickel, which can cause cutaneous reactions24,25. Hypersensitivity to pure titanium is rare, but cases of pure titanium allergy have been reported26,27. Zirconium oxide has similar properties to ceramic. It is biocompatible, corrosion-free, and strong. It is used widely in dental implants. Rare cases of hypersensitivity reactions such as cutaneous reactions, granuloma formation, and pneumonitis have been reported28,29.
Even though genetic predisposition to cutaneous metal hypersensitivity may exist, prolonged cutaneous contact with metal may sensitize an individual, who may then exhibit hypersensitivity reactions in response to shorter periods of exposure. To limit such sensitization, European Union legislation was introduced in 2000 to restrict the nickel composition of products that come into direct and prolonged contact with skin, such as jewelry30. The restriction only involves limiting the amount of nickel released from human body piercings (0.2 µg/cm2/wk) and other materials that come into contact with the skin (0.5 µg/cm2/wk). This legislation led to a reduction in the prevalence of nickel hypersensitivity in European countries, including Denmark, which had implemented its own nickel regulation in 1990, prior to the European directive30,31. The prevalence of nickel hypersensitivity in women who had ear piercings following the implementation of this nickel regulation was significantly lower than that in women who had ear piercings prior to it (6.9% compared with 15.6%; p = 0.004)32. There is no such legislation in the United States, which may explain the increasing prevalence of nickel allergy5. The North American Contact Dermatitis Group (NACDG) reported an increase in the prevalence of nickel allergy in their patients from 14.5% in 1992 to 1995 to 18.8% in 2003 to 200433. In the United Kingdom, the nickel content was initially restricted to <0.05% in products that are in direct and prolonged contact with the skin, such as body piercings and earrings34. This restriction has been replaced with a new regulation complying with the European Union legislation, which puts more emphasis on limiting the rate of nickel release from body piercings (0.2 µg/cm2/wk) and other nickel-containing products in contact with the skin (0.5 µg/cm2/wk)35.
Identifying Cutaneous Metal Hypersensitivity
Patients may report previous reactions to metal such as jewelry. However, self-reporting relies on recall, is subjective, and is based on the previous presence or absence of exposure. Skin irritation resulting from other causes may be confused with symptoms of hypersensitivity. It has been reported that women with a history of childhood eczema perceived themselves to have a higher prevalence of nickel hypersensitivity due to an increase in skin sensitivity resulting from eczema36,37. Thus, self-reported nickel hypersensitivity may overestimate the true prevalence of nickel hypersensitivity.
Different types of metals can be implanted into a patch that is then placed on the skin to test for a hypersensitivity reaction. The patch usually is placed on the patient’s back, and several patches may be used, depending on the number of metals tested. The British Contact Dermatitis Society (BCDS) has recommended a standard series for patch testing that includes soluble metal salts such as potassium dichromate, cobalt chloride, and nickel sulphate38. Other soluble metal salts such as titanium chloride and zirconium chloride also can be patch tested even though they are not part of the standard series recommended by the BCDS. Ready-to-use patch tests, including the Thin-layer Rapid Use Epicutaneous Test (T.R.U.E. Test; Mekos Laboratories, Hillerød, Denmark)39 and the Epiquick (Hermal, Reinbek, Germany)38, are available. The metal allergens are already mixed with a specific material such as methylcellulose or hydroxypropyl cellulose to produce the allergen films that coat the patches.
As metal hypersensitivity is a delayed-type hypersensitivity reaction, it takes several days to complete the test. The results of patch testing may be influenced by the timing of patch application and inspection. Although the usual time to inspect the skin is between Day 2 and Day 4, inspection on Day 6 or 7 may demonstrate an additional 10% increase in the rate of positive results. The British Association of Dermatologists therefore recommends skin inspection on Day 6 or 738. The severity of reactions can range from an area of redness to blisters. The International Contact Dermatitis Research Group (ICDRG) has produced a standard for clinical scoring of allergic patch test reactions40,41 (Table I).
Dry Metal Taping
Testing by taping a dry metal such as a disc or coin onto a patient’s skin has been described42,43. However, this practice generally has been abandoned because the metal is not in the form of a water-soluble ion, which is necessary to elicit a hypersensitivity reaction. In addition, the metal disc itself may cause nonspecific irritation because of its physical pressure on the skin44.
Subcutaneous Metal Implantation
Implanting a piece of metal in the subcutaneous tissue leads to an increased concentration of metal ions in the tissue, which triggers a deep-tissue immune response with proliferation of monocytes and lymphocytes45,46. This method is not routinely used in practice, and only one case of using subcutaneous metal implantation to identify metal hypersensitivity reaction has been reported in the literature so far3.
Lymphocyte Transformation Test
The lymphocyte transformation test aims at directly examining the cellular response to an allergen in terms of lymphocyte proliferation. Lymphocytes are isolated from a blood sample taken from the patient and are cultivated with the metal ion solution for six days. Radioactive [H3] thymidine is mixed with the lymphocytes and is incorporated into cellular DNA during lymphocyte division. This allows measurement of the proliferative response of lymphocytes against an unexposed lymphocyte sample (control), known as the stimulation index (SI)47. The SI is calculated in measured radiation counts per minute (cpm) according to the formula SI = (mean cpm with treatment)/(mean cpm without treatment)6,7. The result is graded with use of standard criteria (with a twofold to fourfold response graded as mild reactivity, a fivefold to eightfold response graded as moderate reactivity, and a more than eightfold response graded as high reactivity)48.
There is still no strong evidence about whether the lymphocyte transformation test is more accurate than patch testing for the detection of hypersensitivity reactions. However, the lymphocyte transformation test is useful in cases in which the result of the patch test is equivocal49.
Leukocyte Migration Inhibition Test
The leukocyte migration inhibition test involves exposing leukocytes to a specific antigen and measuring the leukocyte migration activity. Leukocytes are obtained from a blood sample from the patient. They are then cultured in a medium, and the antigen in question is introduced. Another leukocyte sample is used as a control. If the leukocytes are sensitized toward the antigen, they migrate slower than the control leukocytes, thus exhibiting migration inhibition. Several techniques have been used to measure the migration rate, including the capillary tube, Boyden chamber, leukocyte migration agarose test, and collagen gel techniques50.
Hypersensitivity to Deeply Implanted Metals
Orthopaedic implants usually are made of alloys (mixtures of several metals). Such alloys include cobalt-chromium, stainless steel, titanium, and zirconium alloys. These alloys contain traces of other metals such as nickel, aluminum, and molybdenum (Table II)51,52. Stainless steel and cobalt-chromium alloys are commonly used as they provide high tensile strength, hardness, and resistance to corrosion. They are also inexpensive in comparison with titanium and zirconium alloys. However, both stainless steel and cobalt-chromium implants may contain nickel, which has been associated with cutaneous hypersensitivity reactions50. Titanium and zirconium alloys are strong, lightweight, and highly biocompatible and may be associated with fewer hypersensitivity reactions53,54. However, there have been reports of titanium-alloy implants that have contained impurities of metals such as nickel and palladium, suggesting that they are not proof-free of these potential sensitizers55,56.
Deeply implanted metallic materials corrode. Chemically or mechanically assisted corrosion releases metal debris and ions, which may combine with native proteins to form larger complexes that are then taken up and presented by antigen-presenting cells. Langerhans cells are found in skin and mucosal surfaces, but fibroblasts, endothelial cells, and macrophages have been shown to act as antigen-presenting cells in deep tissues57,58. Certain areas of deep-seated implants (for example, the screw-plate junction) show encapsulation by a fibrous membrane containing macrophages, foreign-body giant cells, and lymphocytes59. It is believed that large metal-protein complexes are taken up by deep-tissue antigen-presenting cells and that these complexes are presented to major histocompatibility complex molecules. T-helper cells may then be activated, causing the release of inflammatory cytokines and the recruitment of more macrophages, monocytes, neutrophils, and other pro-inflammatory type cells. Various inflammatory cytokines are released, such as tumor necrosis factor-alpha (TNF-α), interferon-gamma (IFN-γ), interleukins (IL-β, IL-1, IL-6, IL-8, IL-10, IL-12, IL-15, IL-18), prostaglandin E2 (PGE2), and receptor activator of nuclear factor kappa-B ligand (RANKL). RANKL causes differentiation of osteoclast precursors into mature osteoclasts, which could lead to an increase in bone resorption around metal implants. Metal wear debris has been shown to impair functional osteoblast formation by inhibiting mesenchymal stem-cell differentiation. Theoretically, these inflammatory cytokines may cause local or systemic cellular responses leading to implant loosening, osteolysis, or dermatitis (Fig. 2)6,7.
The clinical support for the potential development of metal hypersensitivity in response to deep-seated metal implants comes from several reports, both from orthopaedics and from other disciplines, in which local or generalized cutaneous reactions developed following the use of a metal implant and resolved following implant removal (Table III)56,60-67. These reports suggest that deep-body contact rather than simply cutaneous contact can elicit cutaneous reactions. Nevertheless, these are rare reactions, with few cases having been reported in literature.
Additional clinical support for the development of metal hypersensitivity in response to deep-seated metal implants comes from reports on patients who have developed metal hypersensitivity (as determined with patch testing and the lymphocyte transformation test) only after deep metal implantation. Frigerio et al. reported on seventy-two patients undergoing hip or knee arthroplasty who underwent patch testing preoperatively and postoperatively68. Five patients who had a negative result preoperatively had a positive result one year postoperatively. Four of these patients developed sensitivity to nickel, and one developed sensitivity to cobalt. However, none of these five patients developed cutaneous hypersensitivity reactions, unexplained joint pain, or aseptic loosening. Merritt and Rodrigo reported similar results in a report on seven patients who had a negative preoperative lymphocyte transformation test but tested positive for nickel, cobalt, or chromium hypersensitivity following joint arthroplasty69. Notably, the contrary has also been reported, with patients having a positive result on preoperative patch testing but a negative result after joint arthroplasty70.
It may be postulated that orthopaedic implants can lead to locally mediated hypersensitivity reactions rather than just to cutaneous reactions. The activation of the hypersensitivity cascade and the resultant inflammation might account for aseptic loosening or unexplained ongoing pain following joint arthroplasty. Whether such loosening or pain occurs as part of true hypersensitivity reaction is currently controversial. There have been clinical reports of patients who tested positive for cutaneous metal hypersensitivity and developed implant loosening or otherwise unexplained implant pain that resolved after implant revision. However, these reports are anecdotal, and a causal relationship cannot be established. Bergschmidt et al.71 reported on a patient who had total knee arthroplasty with cobalt-chromium femoral and titanium-alloy tibial implants and continued to experience pain and loss of motion. Biopsy of intra-articular adhesions showed lymphoplasmacellular fibrinous tissue suggestive of a Type-IV hypersensitivity reaction. The patient was sensitive to nickel and palladium on patch testing. The femoral component was revised to a ceramic implant, after which the patient had improvement in terms of both pain and range of motion. Kosukegawa et al.72 reported on a patient who underwent total hip arthroplasty with a titanium acetabular cup, polyethylene liner, and cobalt-chromium femoral implant and continued to experience substantial hip pain. At the time of the revision operation, a large joint effusion with extensive tissue necrosis was found containing extensive infiltration of lymphocytes, plasma cells, and giant cells. Patch testing was positive for cobalt and chromium. The pain resolved following revision of the femoral component to a titanium stem and ceramic head.
In contrast, there have been several studies of patients with positive cutaneous hypersensitivity as determined with patch testing who did not develop reactions in response to deeply implanted metals following orthopaedic (Table IV) and non-orthopaedic procedures (Table V)4,70,73-79. Thyssen et al., in a study from Gentofte Hospital, Denmark, compared the prevalence of metal hypersensitivity in patch-tested patients with and without total hip arthroplasty73. The study group comprised 356 patients who were registered in both the patch test database at Gentofte Hospital and the Danish Hip Arthroplasty Registry (DHAR). The control group comprised 712 patients who were registered in the Gentofte patch test database but not the hip arthroplasty registry. The reason for patch testing of these patients was unknown but was assumed to be largely related to contact dermatitis. There was no significant difference in the prevalence of metal sensitivity between the study and control groups. The prevalence of revision hip arthroplasty in the study group was also similar to that in the overall DHAR database. Carlsson and Möller reported on eighteen patients who were exposed to an orthopaedic implant containing a metal component to which they were allergic (as determined with patch testing)74. Eight patients had joint arthroplasty implants, and ten had fracture fixation implants. None of the patients developed complications attributed to metal allergy. Rooker and Wilkinson patch tested sixty-nine patients before hip arthroplasty and found that six were sensitive to chromium, nickel, or cobalt70. One patient had a titanium femoral prosthesis, and the remaining sixty-six who underwent the procedure had a Charnley prosthesis. None of the patients developed metal hypersensitivity reactions postoperatively. Webley et al. patch tested fifty patients with knee arthroplasty and thirty-three controls75. Sixteen patients in the study group and three patients in the control group had a positive test result for nickel, cobalt, chromium, or manganese. Seven patients developed implant loosening, but only one of the seven had a positive patch test result. Despite the high prevalence of metal hypersensitivity in these patients, there was no correlation with implant loosening. Carlsson et al. studied two groups of patients who were managed with hip arthroplasty76. Of the 134 patients who were studied retrospectively, only thirteen had a positive patch test for nickel, cobalt, or chromium. Of the 112 patients who were studied prospectively, nine had a positive patch test for nickel before and after the operation. Three developed sensitivity against nickel or cobalt only after surgery. There was no evidence of cutaneous or deep hypersensitivity reactions induced by hip arthroplasty. Granchi et al. examined the frequency of metal sensitization in patients who had undergone knee arthroplasty and its relationship with clinical outcomes77. Three groups of patients were studied, including twenty patients who were awaiting primary knee arthroplasty, twenty-seven who had a stable knee implant, and forty-seven with a loose knee implant. The frequency of a positive patch test result was higher in patients with a knee implant in situ, but there was no significant difference between the stable and loose implant groups.
A recent systematic review and meta-analysis80 examining metal hypersensitivity testing in patients undergoing joint arthroplasty demonstrated that the probability of developing metal hypersensitivity was higher postoperatively and that the risk was greater in patients with failed implants as compared with those with stable implants. However, that systematic review included several studies evaluating metal-on-metal hip implants, which potentially could have influenced the reported outcomes.
There is anecdotal evidence supporting the existence of metal hypersensitivity to deep-seated implants manifesting as cutaneous reactions, but this is a rare occurrence, with few cases having been reported.
Currently, there does not seem to be strong evidence supporting or disputing the role of metal hypersensitivity in the development of aseptic loosening, deep local reactions, or ongoing pain in patients with deep-seated implants. The levels of metal ions that are released vary between articulating and non-articulating implants, and there is a paucity of literature in this regard. Malfunctioning articulating implants potentially can release high levels of metal ions. Fracture fixation devices are less likely to generate the same amount of metal ions as conventional arthroplasty implants. Even though a cellular response is seen around conventional joint arthroplasties or fracture fixation devices, there does not seem to be a strong relationship between loosening and ongoing pain and the presence of cutaneous hypersensitivity. This lack of strong evidence to show a relationship between loosening and ongoing pain and the presence of cutaneous hypersensitivity may be either because the cellular mechanisms around loose or symptomatic implants are not true hypersensitivity reactions or because hypersensitivity to cutaneous contact is not a surrogate of hypersensitivity to deeply seated implants.
These considerations must be taken into account when one is faced with a patient who is about to undergo orthopaedic surgery involving the implantation of a metallic implant. Several approaches can be adopted, ranging from universal screening by means of clinical questioning and/or patch testing, to clinical questioning and selective patch testing, to no questioning or patch testing at all.
Universal screening with patch testing is costly. In the United States, the patch test kit costs an average of $80 USD. In 2010, >300,000 knee arthroplasties and >700,000 hip arthroplasties were performed in the United States according to the National Center for Health Statistics (NCHS). Patch testing all of these patients would cost about $80 million USD without taking into account the clinical appointment costs. In the United Kingdom, the patch testing process involves at least one patch test kit with three outpatient appointments, with an estimated total cost of $300 USD per patient. In 2010, a total of 68,907 primary hip arthroplasties and 76,870 primary knee arthroplasties were performed in England and Wales81. If all of those patients were patch tested, the cost would amount to nearly $45 million USD. There is no strong evidence to show that universal screening would be beneficial for these patients. Therefore, routine patch testing of all patients prior to joint arthroplasty is not recommended.
When using orthopaedic metal implants, the authors favor counseling patients, as part of the consent process, with regard to the small risk of potential reactions to metal, the risk of ongoing symptoms and aseptic loosening, and the limitations in our understanding of the mechanisms accounting for metal hypersensitivity reactions. For patients with a history of metal hypersensitivity, the specifics of any previous reactions should be sought. For patients reporting only mild localized cutaneous reactions, the use of conventional implants seems justified without any further investigations. In the rare case in which a patient reports substantial localized (blistering, hives, extensive rash) or systemic cutaneous reactions to metal, patch testing is recommended. Such patch testing should include the components of potential implants (stainless steel, cobalt-chromium, or titanium-zirconium), and the results can then be used to guide the choice of metal implant. Often in these cases, titanium or zirconium implants are appropriate, provided that cutaneous sensitivity to these metals is not demonstrable. However, this approach is more caution-based than evidence-based. Indeed, in a consensus study of arthroplasty surgeons in the United Kingdom (unpublished data), most participants agreed on the use of conventional implants even for patients with a positive result on patch testing.
A randomized controlled trial comparing stainless steel or cobalt-chromium implants with identical implants made of titanium or zirconium is needed to provide high-quality evidence regarding the effects of implant material on clinical outcomes. Further eliciting the cellular processes occurring in aseptic loosening or ongoing symptomatic implants may help us to determine the extent to which hypersensitivity has a role to play in these issues. The development of tests that will reliably identify individuals who are prone to such a response is also awaited.
Source of Funding: No external funds were received.
Investigation performed at the Department of Orthopaedic Surgery, Blackpool Victoria Hospital, Blackpool, United Kingdom
Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any aspect of this work. None of the authors, or their institution(s), have had any financial relationship, in the thirty-six months prior to submission of this work, with any entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. Also, no author has had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.
- Copyright © 2014 by The Journal of Bone and Joint Surgery, Incorporated