Systemic Allergic Dermatitis

Systemic allergic (contact) dermatitis (SAD or SCD) is a type IV or delayed cell-mediated hypersensitivity in the skin.⁸¹ It is caused by systemic exposure to a specific allergen to which the patient has a prior exposure and preexisting sensitization. This systemic reexposure may come from ongoing contact with an internal substance, either consumed, such as foodstuffs or medications, or implanted, such as prostheses. Although traditionally described following topical exposure, in the case of medications the primary exposure may also be systemic, such as prior exposure to the same drug or a cross-reacting medication.¹⁰² With implants, although it is theoretically possible for the primary exposure to be from a systemic exposure, it is believed that usually these sensitizing exposures occur first in the skin from environmental exposure to components of the implant or a cross-allergen.³⁶ In such cases, the patient may report a prior history of dermatitis occurring in localized areas from contact with external allergens, a syndrome termed allergic contact dermatitis (ACD). SAD is also known by other names, including mercury exanthema, internal-external contact-type hypersensitivity, systemically induced allergic contact dermatitis, baboon syndrome, paraptic eczema, nonpigmented fixed drug eruption, symmetrical psychotropic and nonpigmented drug eruption, intertriginous drug eruption, drug-induced intertrigo, and flexural eruptions,⁵⁰ although some of these syndromes may have been erroneously classified as SAD.

In orthopedic patients, this syndrome is important because it not only may lead to a chronic and debilitating skin condition, but also may be associated with failure of joint prosthesis. The latter remains somewhat controversial but is supported by observations of a high rate of metal sensitivities in patients with prosthetic loosening,⁴⁵ shorter joint life span in patients with positive patch tests,⁴¹ and common finding of hypersensitivity-like reactions on histopathologic testing of the tissue surrounding loosened joints.¹¹²

Pathophysiology

The pathophysiology of a dermatitis becoming systemic is poorly understood. However, in SAD, the initial sensitization is traditionally described as coming from external exposure and thus the pathomechanism of this stage is identical to that of ACD. On primary exposure to a chemical or contact allergen responsible for SAD or ACD, haptens from these allergens bind to proteins found on epidermal Langerhans cells.⁷⁴ This is termed the afferent stage of sensitization. Subsequently, these Langerhans cells migrate to the lymph nodes, where they present haptenated peptides on major histocompatibility complex class I and II molecules, resulting in the induction of hapten-specific CD8⁺ and CD4⁺ T cells, respectively.^14,⁵⁷ This initial sensitization stage takes approximately 10 to 14 days. However, on reexposure, the response is much quicker, typically taking between 12 and 48 hours. During this latter efferent, stage, cloned memory Th1 cells are activated, releasing a cascade of inflammatory cytokines, promoting spongiosis and dermal edema. In ACD, the reexposure is from direct skin contact, whereas in SAD the hapten must be distributed hematogenously from its site of origin to the skin site to elicit a cutaneous response. Conversely, response at the site of implant may cause inflammation, leading to implant complications (see later).

Histopathologic immunophenotyping of cutaneous SAD to systemic nickel reveals CD4⁺ and CD8⁺ T lymphocytes in the epidermis and dermis,³⁷ but decreased CD4⁺ and CD8⁺ T cells, along with decreased CD3⁺, CD45RO⁺, and CD19⁺ T and B cells in peripheral blood.¹² Similarly, in the gastrointestinal mucosa of nickel-sensitive patients, CD4⁺, CD45RO⁺, and CD8⁺ lymphocytes are increased when orally challenged.²⁷ Nickel-sensitive patients have also been shown to have a higher fraction of skin-homing CLA⁺ (cutaneous lymphocyte antigen) CD3⁺ CD45RO, CD4⁺ CD45RO, and CD8⁺ CD45RO T cells when compared with healthy controls, but a decrease in blood CLA⁺ CD8⁺ CD45RO memory T cells after nickel provocation. This suggests that these cells may have migrated to the skin and is consistent with a delayed cell-mediated hypersensitivity.⁵⁴

Cytokine dysregulation is also consistent with a Th1–driven delayed-type hypersensitivity. Tumor necrosis factor-alpha (TNF-α), soluble TNF receptor type 1 (sTNF-R1), interleukin-1 (IL-1 receptor antagonist, and neutrophil gelatinase-associated lipocalin (NGAL) have been shown to be upregulated in gold-sensitive patients when challenged.⁷⁰ In nickel-sensitive patients, similar challenge only provoked increases in sTNF-R1,⁷¹ although other studies using high-dose nickel also noted upregulated IL-2, IL-5,¹² IL-6, and IL-10.⁵⁴ In a zinc-sensitive patient, both TNF-α and migration inhibitory factor (MIF), which upregulates TNF-α, were found to be increased.¹¹³ In addition, upregulation of IL-1β, TNF-α, IL-6, and prostaglandin E2 (PGE2) have been associated with proinflammatory cytokine–induced bone resorption via activation of osteoclasts and suppression of osteoblasts,⁷⁹ providing a mechanism for the observed aseptic loosening.

Other Types of Reaction to Implants

This type of reaction should be distinguished from other forms of inflammation that may exist, at times ranging from during surgery to months or years afterward. One such reaction is IgE-mediated immediate hypersensitivity response, which presents with a variety of clinical signs including contact urticaria, angioedema, asthma, and anaphylaxis within minutes of exposure and is most commonly associated in the clinical setting with natural rubber latex allergy. A second type is a granulomatous reaction, which often occurs in response to a foreign body such as plastic particulate matter from a worn prosthesis or talc from surgical gloves. This type of reaction is also associated with joint loosening.⁷ Although starting immediately, this reaction may not show clinical significance for weeks to months after particle deposition. The normal healing and repair response also invokes significant stages of inflammation and repair in the weeks to months after surgery. Finally, an autoimmune reaction, in which the body produces antibodies against itself, is theoretically also possible to trigger with prosthesis placement. A humoral type 3 immune reaction mediated by circulating antigen-antibody complexes that cause inflammation on tissue deposition has been postulated; it is supported by the identification of antibodies against hapten-albumin complexes in the blood.^80,¹⁰⁷

Epidemiology

More than 3000 environmental chemicals have been identified as causing SAD or ACD. However, although one contact dermatitis is a risk factor for developing another, it is important to assess the clinical relevance of allergies identified because only some of these are relevant to orthopedic implants.

Allergenic Substances Used in Orthopedics

There are a variety of designs of knee prosthesis used throughout the world. However, as their components generally do not vary greatly, we will review here standard constituents and not specific designs. In orthopedic prostheses, the metals used are alloys, or combinations of metals, very rarely pure metals. It is often the minor impurities that are the common allergens.

Vitallium, a commonly used alloy, is composed of 70% cobalt, 25% to 30% chromium, and 6% to 7% molybdenum, with trace amounts of nickel. Austenitic stainless steel, which is also commonly used, occasionally contains up to 35% nickel, but generally contains 8.5% to 14% nickel, 17% to 20% chromium, 2% to 3% molybdenum, and less than 1% carbon, nitrogen, manganese, silicon, sulfur, phosphorus, and niobium. Alloys of cobalt-chromium-tungsten-nickel with 9% to 11% nickel and cobalt-chromium-molybdenum with 2% nickel are also sometimes used. Titanium is used in its pure form and alloyed with 6% aluminum or 4% vanadium for improved tensile strength.³⁶

In addition to the metals in prosthesis, bone cement is sometimes used. The most common bone cement is polymethylmethacrylate (PMMA)-based, but this and others may have allergenic additives, the most common of which are gentamicin and benzoyl peroxide.^94,⁹⁵ These allergens are likely less significant than the metal sensitivities, given the lower prevalence of these allergies in the general population and the lower amounts of these substances in prostheses. The allergenicity of all these standard components is reviewed below.

Genetics of Contact Sensitization

In general, there is no known genetic predilection for the development of contact sensitization to prosthesis components. The exception to this is in nickel sensitivity. The increased prevalence of nickel allergy in monozygotic over dizygotic twins in epidemiologic studies suggests a possible genetic component to nickel allergy,⁶⁷ although these results have not been consistently reproduced.¹⁸ Null mutations in filaggrin have been found to be associated with nickel allergy.⁷⁷ Filaggrin is a highly phosphorylated, histidine-rich polypeptide important in keratin filament aggregation and formation of the skin barrier.²³ This may be particularly important in nickel-sensitive patients because histidine-rich polypeptides are strong nickel-chelating agents and thus may also cause the accumulation of nickel in the stratum corneum.⁹⁷ The discrepancy in epidemiologic study findings noted earlier may therefore partly be explained by the fact that the study showing no correlation included a significant number of patients exposed via ear piercing, thus circumventing the need for a genetic basis for barrier disruption, whereas the earlier study looked primarily at patients with topical clothing-based exposures.

Prevalence of Contact Sensitization

Risk factors for the development of SAD are directly linked to the risk factors for developing an initial cutaneous sensitization. These will therefore be reviewed to facilitate patient risk stratification for SAD. Because exposure risk is controlled by the environment, prevalence data and risk factors vary by geographic location and therefore need to be assessed based on the patient population. For example, nickel allergy rates are lower in Denmark and Germany than in the United States, presumably because nickel content for clothing (buttons, fasteners) and piercings is more stringently regulated in these countries.^68,^88,^98,⁹⁹ The effects of such local restrictions on prevalence can be seen by the decline in nickel allergies after these nickel content regulations were introduced in these countries.

Nickel allergy is among the most common contact allergies, with an estimated prevalence based on positive patch testing of 16.7%⁶¹ to 19%¹¹⁶ in the United States. In Thailand, this figure is reported to be as high as 33.8%,¹¹¹ whereas in Europe a similar incidence has been reported but with a clear disparity between women (17%) and men (3%).^29,¹⁰¹ This gender discrepancy may be explained by a discrepancy in the rates of ear piercings,⁷⁵ with 80% of women and 10% of men estimated to have piercings in this population. As noted, however, the prevalence of nickel allergy in Denmark is reported to have dropped to 6.9% (P = .004) in piercings in women since the introduction in 1990 of more stringent nickel content restrictions.^98,⁹⁹ Cobalt and chromium sensitivity are estimated to be approximately 1% to 3% range,^68,⁸⁸ although chromium sensitivity is believed to be increasing in Denmark, Singapore,³⁹ and the United States.⁷³

When the results of 10 European patch-testing centers were pooled, cobalt sensitivity was seen to have an age-dependent prevalence of 6.2% to 8.8% and chromium of 2.4% to 5.9%. Gold ^11,⁸² and palladium¹ allergies are seen in about 10% of dermatitis patients, although gold sensitivity increases to 30% in patient with gold dental² and cardiac³¹ implants. In contrast, aluminum sensitivity is rarely reported.⁵⁶

Titanium hypersensitivity is regarded as extremely rare.⁶⁰ However, it has been reported in hip replacement⁵⁹ and with a static titanium implant in which dermatitis was observed overlying the site and a positive lymphocyte transformation test result was obtained.⁹³

Coreactivity to metals is common, with one study showing nickel reactivity in 79% of cobalt-sensitive patients, 39% of chromium-sensitive patients and 95% of palladium-sensitive patients.⁵⁶ This high rate of coreactivity between nickel and palladium, and the low rate of palladium exposure outside the electronics and chemicals industry, has led many to question whether this is in fact simply a cross reaction to a nickel allergy.¹⁰⁵ In contrast, the high rate of concordance of cobalt and nickel sensitization is believed to be caused by concomitant sensitization rather than cross sensitization because of the prevalence of cobalt in consumer products, and is thus clinically relevant.

Risk Factors

For Nickel Sensitivity

Nickel allergy has been identified in a variety of occupational exposures. These include plating industry workers,¹⁰³ retail clerks, hairdressers, domestic cleaners, metal workers, caterers,⁹⁰ locksmiths, and carpenters.⁶² Nickel dermatitis has also been reported as being caused by clothing, ranging from suspenders²² to jeans’ buttons and zippers.¹⁶ Other sources include headsets and mobile phones.¹⁰⁰ Jewelry and body piercings are a common cause in women^13,³² and in men the number of body piercings has been shown to correlate positively with the risk of nickel allergy.³⁰ Significant long-time exposure to metals that release nickel are also likely to be a risk factor. These include white gold, gold plating, German silver, Monel solder, nickel plating, and stainless steel. Finally, some reports have implicated nickel-containing cosmetics and devices such as eye shadow,³⁸ mascara,¹⁰⁶ eyeliner pencils,¹¹⁵ and eyelash curlers.¹⁵

For Chromium Sensitivity

Occupational exposure to chromium is possible in locksmiths and carpenters ⁶² and those working with cement,¹¹⁴ dyeing agents, metal alloys, pottery, colorant, and antirust agents in coolants, such as mechanics.⁴ Although cement workers have historically been the most important of these, addition of iron sulfate to reduce the amount of water-soluble hexavalent chromium reduced chromium sensitivity from 12.7% in 1989 to 1994, prior to its addition, to 3.0% in 1995 to 2007. In contrast, in the same time period, chromium sensitivity from consumer exposure to leather has increased from 24.1% to 45.5%.^34,^89,⁹⁶

For Cobalt Sensitivity

Occupational exposures to cobalt include hard metal workers, painters in the glass and pottery industry,^33,⁸⁶ locksmiths, carpenters, cashiers, and secretaries.⁶² In consumer exposure many of the same risk factors are present as in nickel as chromium has historically often been mixed with nickel. These include jewelry and piercings.^65,⁷⁵

Sensitivity to Implants

Systemic Allergic Dermatitis to Implants

Metal implants have been used for the repair of fractures since the 1950s and in joint replacements since 1962, with the first prosthetic hip. All metal implants, even static ones such as those used to repair fractures or in pacemaker devices, are inevitably in contact with body fluids. Therefore, they will corrode and release metal ions, which have the potential to bind proteins and activate T cells ^46,⁶⁹ and macrophages.²¹ As noted, T-cell activation can lead to an allergic contact dermatitis in the overlying skin in pacemakers¹⁹ and joints,⁸⁵ or to a more extensive SAD. In contrast, macrophage activation has been associated with device failure.

The degree of allergenicity in stainless steel implants has been shown to be directly related to the sulfur content, which reflects the alloy’s ability to liberate nickel ions. High-sulfur (0.3%) stainless steel AISI 303 can release up to 1.5 µg/cm²/week, sufficient to induce dermatitis in nickel-sensitive patients.^48,⁶⁶ In contrast, stainless steel with less than 0.03% sulfur release only 0.03 µg/cm²/week and does not result in nickel dermatitis. Despite multiple reports of localized dermatitis over a static implant^64,⁸³ in a prospective study of 48 subjects receiving static stainless steel orthopedic implants, none developed dermatitis, even the three subjects shown by patch testing prior to implantation to have a nickel allergy.³⁶ However, restenosis of stainless steel stents, which contain nickel, cobalt, and molybdenum, have been associated with the presence of a nickel allergy,^52,^55,⁸⁷ although this association continues to be debated.^51,⁹¹ In contrast, coronary artery stent restenosis is strongly associated with gold allergy in gold-plated stents.³¹

In nonstatic implants, there is an even greater theoretical potential for metal ion exposure because of the mechanical wear inherent in the device’s function. Today, the most commonly reported systemic allergic dermatitis associated with an implanted orthopedic joint prosthesis is that following hip replacement. This may be caused by combination of factors, including the fact that hip replacement is the most common type of implanted joint prosthesis performed, the high load and frictional forces of this joint contributing to high wear and particulate production, and the materials and design used, most especially in early hip prostheses, which were more prone to allergenic particulate production. Early prosthetic hips consisted of metal on metal components resulting in much higher frictional wear, subsequent release of particulate metals and metallic ions, and eventual loosening in up to one quarter of cases.³⁶

Subsequent to this, hips with metal femoral but plastic acetabular components were introduced. Currently, the former is most commonly austenitic stainless steel and the latter is composed of high or ultra–high-molecular-weight polyethylene (UHMWPE), ceramic, or carbon fiber. In younger patients, porous-coated implants are sometimes used that require no cement but in older, and other higher risk patients, PMMA bone cement is used. Because acrylic bone cements are not easily biodegraded, inflammatory reactions to PMMA and other bone cements have been reported.⁴² In addition, additives such as benzoyl peroxide¹⁰⁴ and gentamicin^43,⁶³ can be allergenic. In one study, 28 of 113 patients with cemented prostheses had a sensitivity to bone cement components⁹⁴; 16.8% were sensitive to gentamicin and 8.0% to benzoyl peroxide, although N,N-dimethyl-p-toluidine and hydroquinone sensitivities were also identified in a minority of patients. With the metal on plastic hips, many reports exist of patch test–confirmed metal-sensitive patients receiving implants with no development of cutaneous problems or loosening.^6,²⁰ The downside of these metal on plastic hips is that although they release less metal, they do produce greater overall wear, losing 0.2 mm/year from the polyethylene surface,¹⁰⁹ in comparison to 0.1 10 µm/year for the ball and 0.2 to 6 µm/year for the cup in all-metal prosthesis. The polyethylene particles also tend to be larger, inducing greater tissue reaction and thus more osteolysis. For this reason, metal on metal hips consisting of a cobalt alloy femoral stem and a titanium acetabular cup were reintroduced in the 1980s. Three years after prosthesis insertion, serum levels of titanium were found to be threefold higher and that of chromium fivefold higher,^17,⁵³ leading to the potential for development of SAD long after implantation.

The knee, after the hip, is the next most common prosthetic joint replacement. Although many of the same concerns that exist for hip replacement could be extended to knee prosthesis, the design and biomechanics of the two joints are different. In particular, in knee prostheses, the articulation in more commonly metal on plastic, leading to decreased metal debris. Despite this, SAD cases have been reported in nickel- and cobalt-sensitive patients even after preoperative patch testing.⁸ Localized dermatitis over an artificial knee joint has also been reported occurring 2 months after the use of a condylar knee joint replacement.⁴⁷ This prosthesis contained a Cu²⁺-Cr³⁺ alloy femoral component. The patient did not go on to SAD and serum levels of copper, nickel, and chromium were shown to be normal, but the patient had a positive patch test result to copper sulfate and cobalt chloride, but not to nickel. An 80-year-old Japanese patient also developed knee dermatitis after total knee arthroplasty (TKA) with a Co-Cr alloy knee.⁷⁸ The patient patch-tested positive to Co, Ni, Cr, Mn, Pt, Ir, In, Hg, Sn, and Zn; the dermatitis was resolved when the prosthesis was replaced with a titanium-ceramic device.

In one case series, 30 patients were observed to develop localized dermatitis around the knee 1 to 3 months after joint revision with a total condylar knee prosthesis (DePuy Orthopaedics, Warsaw, Ind).¹⁰⁸ This prosthesis contains a femoral component consisting of a cobalt-chrome alloy made up of 27% to 30% chromium, 5% to 7% molybdenum, 0.7% nickel, 59% to 64% cobalt, and 4% other elements. The tibial component was a titanium alloy containing 5.5% to 6.5% aluminum, 3.5% to 4.5% vanadium, and 88% to 91% titanium. Of the 30 original patients, 15 consented to patch testing. Testing was performed to nickel sulfate (5% in pet), cobalt chloride (1% pet), and potassium dichromate (0.5% pet). At 3 days, 7 of 15 patients showed a positive metal sensitivity, 4 to nickel, 2 to chromium, and 1 to cobalt.

In a different case series, four German female patients were reported to have persistent dermatitis after a Co-Cr alloy TKA.²⁸ Each patient underwent a lymphocyte transformation test (LTT) to Ni, Cr, Co, Mo, Mn, and Ti and patch testing to a standard series containing Ni, Cr, and Co, an expanded metal series, including Mn, Mo, V, and Ti, and a bone cement series, comprised of 2-hydroxyethylmethacrylate, PMMA, copper sulfate pentahydrate, benzoyl peroxide, gentamicin sulfate, hydroquinone and N,N-dimethyl-p-toluidine. The first patient in the series had a positive nickel sensitivity. The second patient in the series was sensitive to both cobalt and nickel. In the third patient, patch testing was positive to cobalt and the LTT showed elevated sensitivity to both nickel and cobalt. The final patient was sensitive to nickel and cobalt but was also sensitive to ethylene glycol dimethacrylate, 2-hydroxyethylmethacrylate, and 2-hydroxypropylmethacrylate. Resolution of all symptoms in all four patients, including joint effusion and dermatitis, was achieved after switching to a titanium-plated prosthesis and, in the case of the final patient, removal of residual cement at the time of titanium prosthesis implantation.

In a prospective study of 92 patients undergoing TKA between 2000 and 2002, preoperative modified lymphocyte stimulation tests (mLSTs) to Ni, Co, Cr, and Fe were performed.⁷⁶ Of these, 26% showed positive sensitivity to at least one of these metals. Five of the patients with preoperative metal sensitivity went on to develop implant-related dermatitis, although the only association reaching statistical significance was with Cr (P < .05). Two of the metal-sensitive patients had TKA revision, with resolution of the dermatitis.

In a different study, 94 subjects were recruited, 20 prior to TKA, 27 with a well-functioning TKA, and 47 with loosening of the joint after revision.⁴⁰ Patch testing for 5% nickel sulfate, 1% cobalt chloride, 2% chromium trichloride, 0.5% potassium dichromate, 2% ferric chloride, 2% molybdenum chloride, 1% niobium chloride, 2% titanium dioxide, 5% PMMA, 2% butyl methacrylate, 2% triethylene glycol dimethacrylate, 2% ethylene glycol dimethacrylate, 2% N,N-dimethyl-p-toluidine, 5% hydroxylethylmethacrylate, 2% benzoyl peroxide, and 1% hydroquinone monobenzyl ether was performed. In preimplant patients, a positive patch test result was seen in 20% of patients. In postimplant patients, positive patch tests were shown to be higher in both groups tested after TKA, with a slight increase in patch test positives in patients with TKA loosening (59.6%) over those with stable TKA (48.1%). The most important factor identified in this study in predicting joint loosening was a prior history of contact allergy to metals. This single item in the history increased the risk of failure fourfold.

Finally, one case has been reported in which TKA failure was attributed to a preexisting contact sensitivity to PMMA.⁴⁹ The patient had previously had a periungual dermatitis believed to be from acrylic nail use, which resolved with the avoidance of acrylic nails and glues. She subsequently underwent TKA with a PMMA-containing bone cement. The patient went on to experience significant early joint loosening and was patch-tested to both the bone cement and metals. PMMA was positive but metal patch testing was negative. The patient subsequently underwent joint revision with a cementless prosthesis, with no recurrence of the problem.

Only gold members can continue reading. Log In or Register to continue