1 Prevalence of Allergy and Epidemiology

Chromium (Cr) has traditionally been the third most common metal allergen (after nickel and cobalt) [1, 2], but ranges from the first to the fifth most common metal allergen in different occupations and countries (Fig. 27.1). Its prevalence varies largely among different groups and countries and over time, closely related to occupations, exposure, different regulations, and work hygiene. Figure 27.1 shows the large variation in the percentage of positive patch test reactions among adults to potassium dichromate (0.25% or 0.5% in petrolatum) found in different studies after the year 2000. Studies on the prevalence of allergic Cr contact dermatitis before the year 2000 have been summarized in [3] and in [4].

Fig. 27.1
figure 1

Prevalence of Cr sensitization obtained through patch testing with 0.25% or 0.5% potassium dichromate (in petrolatum) in occupational (blue-collar workers), clinical (e.g., patients suspected to have contact dermatitis), and general population groups in different countries. The study country/region and number of tested persons in each study are noted above the bars, and the years during which the patch testing was conducted (two last numbers of the year only) can be found below the bars and reference letters—A [14], B [2], C [15], D [16], E [17], F [18], G [19], H [20], I [21], J [22], K [23], L [24], M [25], N [26], O [27], P [1], Q [28], R [29], S [30], T [31], U [32], V [33], W [34], X [35], Y [36]. Note that there are differences in patch testing, readings, statistical analysis, etc., among the studies. The reader is referred to the respective study for exact numbers and study details

Full size image

Generally, prevalence of allergic Cr contact dermatitis is lower in North America and Western Europe compared with Eastern Europe and Asian countries (Fig. 27.1). This corresponds with a long-term decrease in allergic Cr contact dermatitis in Western Europe and North America, and a stagnant or increasing trend in Asia [4], and is believed mainly to be related to differences in exposure sources, occupations, and regulatory measures. Differences in referral patterns and false positive or false negative readings could be other possible explanations.

Generally, prevalence of allergic Cr contact dermatitis increases with age [4]. Although contact dermatitis to Cr is usually less common in children compared to adults, it remains an important contact allergen [5,6,7,8] and, in some studies and countries (e.g., Italy, Switzerland, India), represents one of the most common contact allergens for children [5].

Traditionally, allergic Cr contact dermatitis has been mostly found in men due to the high prevalence in construction workers [4]. This has, however, recently changed in many countries due to regulatory measures to limit the soluble hexavalent Cr (CrVI) content in cement in some countries, first implemented in Denmark in 1981 [4, 9, 10]. Other important regulations are the recently implemented restriction of CrVI released from leather [11] and the general attempt by industry and regulators to replace CrVI in processes and products, e.g., for electronic equipment [12] and chemicals [13].

2 Relevant Chromium Chemistry

Trivalent Cr (CrIII) is the hapten that is bound to a carrier (a protein), forming a conjugate (antigen) which causes Cr sensitization and elicitation [37, 38]. Both trivalent and hexavalent Cr species can be stable in aqueous solutions depending on the pH and solution redox potential [39, 40]. Figure 27.2 gives some examples of important trivalent and hexavalent Cr species in aqueous solutions such as patch test solutions. Generally, CrVI species cannot form cations and do not easily bind to organic species and proteins (unless they are first reduced to CrIII) [41, 42]. This, in addition to their negative charge (at neutral and alkaline pH), has been suggested to be the reason for their relatively high skin penetration [42,43,44]. CrIII, in contrast, forms a vast number of different cationic, anionic, and neutral charged species in aqueous solutions and binds easily to abundant blood and skin proteins [42, 45,46,47], but has a low skin penetration and is reported to be rejected by the skin [44].

Fig. 27.2
figure 2

Some examples of relevant cationic, anionic, and neutral hexavalent (top) and trivalent (bottom) Cr species in aqueous solutions of relevance for patch test solutions and clinical and laboratory studies. Species were analyzed for several patch test solutions (13% CrCl3 and 0.5% potassium dichromate in water) and for the most common leather tanning agent CrSO4OH in water, using the Medusa software [50], and drawn by the ChemSketch software (ACD/Labs Freeware 2012)

Full size image

Metallic Cr, Cr0, is neither stable in air nor water due to its high oxygen affinity, and it rapidly forms a trivalent Cr oxide. This surface oxide or passive film is important for all corrosion-resistant Cr alloys such as stainless steel, Inconel (a common Ni-Cr alloy), and cobalt-chromium alloys (e.g., used for dental implants and artificial joint prostheses). This surface oxide of noncorroding Cr alloys and metal is very stable. It cannot be seen by the naked eye, because it is very thin (e.g., 2–5 nm in water or air) [48, 49].

3 Skin Deposition

The skin deposition of Cr from different items is not necessarily similar to the amount and chemical form that is released from these items in different simulants, such as artificial sweat. There are two major reasons: (1) most skin contact includes wear processes which are often not considered in laboratory testing, such as in artificial sweat according to the EN 1811 standard [51], and (2) the chemical reactions in skin contact are thin-film reactions and not necessarily comparable to bulk solution conditions [49,50,51,52]. It has been shown that the skin deposition on the index finger after touching different metal and alloy surfaces is higher compared with the amount released into artificial sweat [53]. Skin deposition of Cr from different items and in certain occupations has been investigated using acid wipe sampling [54,55,56] and wipe sampling [57]. The recovery of Cr was between 90% and 102% [58] in acid wipe sampling, however, not tested for CrVI, which is not possible to distinguish from CrIII using this technique. The skin deposition of Cr has been found to be larger for Cr-tanned leather compared with Cr-containing metal discs [54], which is in accordance with release data (see Use and Sources of Exposure below). It has been highlighted in these studies that even brief skin contact results in significant amounts of deposited metals. This is also true for release processes in bulk solutions, which mostly take place in the first seconds to minutes for passive metals [49, 59] and Cr-tanned leather [60]. With both metals [53] and Cr-tanned leather [61], it was found that surfaces previously stored in air can release more metals compared with previously touched [53] or immersed [61] surfaces. Unpublished results of Cr skin deposition from leather bracelets suggest that Cr is deposited from Cr-tanned leather bracelets (Y. Hedberg, B. Erfani, M. Matura, C. Lidén, unpublished data).

4 Skin Penetration

Penetration of Cr through human skin is governed by (1) pH, (2) Cr concentration, and (3) chemical speciation of Cr, as well as biological factors such as the skin barrier (Table 27.1). Since the Cr chemical speciation (Fig. 27.2) is governed by the solution pH, solution composition, and Cr concentration, systematic studies comparing the skin penetration of different Cr compounds are difficult to conduct and interpret. Table 27.1 summarizes several skin penetration studies, with a focus on the comparison of different Cr compounds and solution pH values. Since CrVI species are generally more soluble at alkaline pH, and CrIII species at acidic pH, they cannot be compared directly without affecting their solubility and ionic charge. It is, however, clear from Table 27.1 and [62] that anionic species of both CrVI and CrIII penetrate the skin to a larger extent compared with cationic or neutral species. The solution pH affects not only the Cr speciation (which Cr species exist) but also the skin permeability [43] and skin charge [42]. At alkaline pH, both the skin membrane and CrVI species are negatively charged (CrIII species are not soluble at this pH), which results in electrostatic repulsion and hence no binding. The consequence of the equal skin membrane and CrVI species charge, as well as the higher skin permeability to water, is that CrVI species at alkaline pH show the highest Cr skin penetration compared with CrIII species and acidic pH.

Table 27.1 Relative order of skin penetration rates of different Cr solutions of varying pH values
Full size table

5 Immunology

In order to be a sensitizer, Cr must bind to a protein (a carrier) to form an antigen [66, 67]. It is now generally accepted that the antigen is a CrIII-protein conjugate and that any CrVI first needs to be reduced before it binds to a protein [37, 38, 41]. The fact that CrVI is a stronger sensitizer compared to CrIII (Tables 27.2 and 27.3), even if skin penetration is excluded (e.g., by intra-/subdermal injection or in in vitro assays), may be explained by (1) the strong binding of CrIII to proteins compared with CrVI and (2) the ability of CrVI to penetrate the cell membrane and bind to proteins after reduction inside the cell [37, 38, 68]. It has been suggested that Cr (and metal cations in general) changes protein structure upon binding and that this structurally modified protein is processed and presented to T cells by cutaneous dendritic cells (Langerhans cells) [67] (sensitization step) or recognized by circulating hapten-specific T cells (elicitation step) [66, 67, 69, 70]. Activation of human monocyte-derived dendritic cells through direct ligation with human Toll-like receptor (TLR)-4, shown to be important for nickel, palladium, and cobalt, has not been found to be important for Cr [71]. It is well known that CrIII can modify protein structure, or even strongly aggregate proteins, such as normally pure serum albumin [45,46,47, 72]. Whether this structural modification or any other reaction is responsible for the antigenic properties is unclear. Several CrIII-protein conjugates have been identified that are able to be recognized as antigen by Cr-sensitized persons (investigated by different methods), in decreasing order: human serum albumin [73,74,75,76], heparin [75, 76], hyaluronic acid from human umbilical cord [76], undefined skin proteins [73], and γ-globulin [74, 75].

Table 27.2 Relative order of magnitude of positive skin reactions in Cr-sensitized persons to Cr solutions of varying pH
Full size table
Table 27.3 Minimum elicitation threshold (MET) values of CrIII or CrVI solutions of different studies, all through occluded patch testing on intact skin for 48 h, on Cr-sensitized persons
Full size table

Generally, due to the strong effect of skin penetration (for all Cr species) and protein binding (in the case of CrIII only), allergic responses are greater for sub- or intracutaneous testing relative to epicutaneous testing, as well as irritated or stripped skin relative to intact skin (Table 27.2). Furthermore, skin penetration largely determines the wide variation among the different Cr solutions summarized in Table 27.2. However, for anionic CrIII solutions, such as CrIII-oxalate, the difference from CrVI solutions is smaller (Table 27.2). Several minimum elicitation threshold values for CrIII and CrVI are summarized in Table 27.3, and other studies are summarized in [77]. These were obtained via occluded patch testing, and it has been argued that more realistic methods such as repeated open application tests (ROAT), where the test solution is applied for brief discontinuous periods on non-occluded skin, are needed. ROAT tests have been conducted for aqueous potassium dichromate solutions, where comparable threshold values were obtained as in occluded patch tests [78, 79].

Concomitant reactivity to other haptens may occur, possibly due to cross-reactivity or co-sensitization. For example, concomitant reactivity to Cr and Co has been observed [80]. A recent study on 656 consecutive dermatitis patients, of which 200 patients reacted positively to either Co, Cr, or Ni, demonstrated that reactivity to each of these metals can either exist independently or for one or several more metals, suggesting co-sensitization rather than cross-reactivity to be the reason behind concomitant reactivity [81].

6 Use and Sources of Exposure

Cr is used or present in a large variety of articles, products, and alloys. In contrast to other metal allergens, the most important sources of skin exposure, both occupationally and environmentally, are nonmetals: Cr-containing cement, Cr-tanned leather [86, 91], and different fluids and chemicals, such as detergents and bleaching chemicals [92]. Table 27.4 summarizes amounts of released CrIII and CrVI from different sources into select environments. Generally, the release of Cr from noncorroding metals and alloys is significantly lower compared with Cr-tanned leather (Table 27.4). The release of Cr from noncorroding metals is not proportional to their Cr content, which is also illustrated in Table 27.4. Corroding metals and alloys can, however, release a significant amount of CrIII. Corroding metals and alloys without any chromate-containing coatings can only release significant amounts of CrVI if a high voltage is applied (Table 27.4), which does not occur in ambient environments, skin contact, or the human body, but might be important in certain manufacturing processes, electrical applications, or implants that make use of high pulsed voltage. Chromate coatings have been used as anti-fingerprint coating on metal surfaces during transport and storage, such as roof sheets, screws, and other metal products [93]. These coating types can result in CrVI release upon skin contact when a product is new or has been stored at dry conditions, and they are increasingly replaced by manufacturers with alternative coatings [93]. Cr-releasing particles include welding fume from Cr-containing alloys and cement particles (Table 27.4). Since ultrafine particles can be inhaled and reach the alveolar region in the lung [94], special protection is required. Certain combinations of exposure factors and source chemistry should also be avoided. These include for Cr-tanned leather (Table 27.4): (1) alkaline fluids such as wet cement contact or detergents, (2) dry storage in low humidity followed by wet skin exposure, and (3) frequent skin contact with Cr-tanned leather that was not treated with antioxidants.

Table 27.4 Reported release of CrIII and CrVI (μg) per surface area of different sources (cm2)
Full size table

Cr can also be released from Cr-containing alloys (nearly all implant materials that require wear and corrosion resistance) inside the human body, such as from different stainless steel and cobalt-chromium-alloy dental or artificial joint prostheses. The released form is ionic CrIII or wear particles including Cr0 or CrIII in the form of Cr2O3 [95, 96].

Other sources vary largely among countries and occupations and depend on prevailing safety procedures, technologies, and regulations. Industrial/professional use includes chromate containing anticorrosion inhibitors or coatings, catalysts (CrIII based and converted to other oxidation states during the process), electroplating (CrO3) and anodization agents, pigments (green Cr2O3, decreasingly yellow/orange chromates) for the production of glass or ceramics, chromates in paints, mordant dyes of textiles, CrVI-containing wood preservatives that are converted to CrIII during the process, CrVI-containing oxidative bleaching chemicals, and wet cement [4, 97, 98].

Green pigments in cosmetics and tattoo ink (Cr2O3); Cr-containing ash, e.g., due to combustion of Cr-tanned leather and preserved wood; and dry/wet cement are examples of other sources that are not necessarily an occupational exposure [4, 70, 92, 97, 98]. A more detailed summary of different sources, including older technology and processes, can be found in [4].

7 Clinical Manifestations

7.1 Contact Dermatitis

Contact dermatitis to Cr is often located on the hands and feet [86, 113]. Cr dermatitis is associated with greater severity of hand eczema [114], a lower quality of life [115], and higher prevalence of sick leave [115], compared to other dermatitis patients. Cr contact dermatitis is very persistent [116, 117] and has a poor prognosis [113,114,115, 118,119,120]. It has been suggested that this might be due to the multitude of different Cr sources and/or an ability of Cr to remain absorbed in the skin [116, 117]. However, it has been shown that strict allergen avoidance, mainly by a change of workplace or early retirement, resulted in the improvement of 72% of Cr contact dermatitis patients’ symptoms within a few years [121]. Systemic contact dermatitis is not very common and of minor importance, except in the setting of a high oral intake of CrVI [97] and CrIII food supplements [122].

7.2 Hypersensitivity to Implant Materials and Their Wear Debris

In regard to the general prevalence of metal contact dermatitis, nickel and cobalt are more commonly reported to cause hypersensitivity reactions compared with Cr [123]. However, Cr is an alloying element in many biomedical metallic implant materials and might be released in nano-sized wear particles [95] or as ionic species (CrIII) [96]. Cr ions or Cr-containing wear debris released from implant materials can cause cutaneous allergic reactions [123,124,125], peri-implant inflammation [123, 126], and other reactions [127]. Large aggregates of Langerhans cells have been found in the lymph nodes of a patient with high cobalt and Cr serum ion levels [124], suggesting a type IV (cell-mediated) reaction. Complications for patients with articulating implants, especially knee arthroplasty, are associated with a higher rate of metal sensitization [123]. It is, however, unclear whether the metal sensitization caused complications or vice versa [123]. Symptomatic relief and/or disappearance of associated eczema has been reported after replacement of the implant, but the role of metal allergy is still not clear for several other complicating conditions (e.g., persistent pain, aseptic loosening, pseudotumors) [123]. Two studies concluded that the overall risk of knee arthroplasty failure is not increased due to metal hypersensitivity [128, 129]. It was also shown (in 52 patients) that, despite higher serum ion levels as compared to metal-on-polymer implants, having cobalt- and Cr-releasing metal-on-metal hip implants did not increase the prevalence of Cr or cobalt contact dermatitis after a 5-year follow-up [130]. This was confirmed in a Danish registry-based study, where it was, however, found that the prevalence of cobalt and Cr contact dermatitis was increased for patients having two or more episodes of revision surgery [131].

7.3 Allergic Asthma

Asthma is a common disease worldwide, with a physician-diagnosed prevalence rate of 4.3%, ranging from 0.19% (China) to 21% (Australia) [132]. Its prevalence is not declining [133]. The incidence of allergic asthma is approximately equal to that of non-allergic asthma, but dependent upon age [134]. Cr-induced allergic asthma has been reported relatively scarcely [135] for some occupations, such as welding, electroplating/metal plating, and cement work [136,137,138,139]. It has been suggested that the metal sensitivity involved in allergic asthma may be IgE mediated without a clear association with allergic contact dermatitis to the same metal [135, 138, 140] or may be IgE mediated in some cases, but not in others [136, 139]. Thus, there is some evidence for both delayed and immediate types of allergic asthma to Cr.

8 Patch Testing, Spot Testing, and Other Testings

For patch testing, potassium dichromate 0.5% in petrolatum (e.g., baseline series for Europe) or 0.25% in petrolatum (e.g., North America) is most often used [97]. These CrVI patch test allergens can cause irritant reactions that are sometimes difficult to distinguish from allergic reactions without retesting. However, lower concentrations of CrVI, shorter patch test durations than 48 h, or the use of CrIII salts may instead result in a large percentage of false negatives (Tables 27.2 and 27.3) [90]. A recent study found that Cr was associated with a significantly higher percentage of potentially irritant reactions for at least one patch test reading (day 3 or days 6–7 after the start of patch test placement for 48 h), in comparison to cobalt and nickel [81]. It seems, therefore, particularly important that two readings be conducted when patch testing to Cr [81]. Of note, the number of irritant reactions for Cr patch testing increased with lower temperature and lower humidity [141], probably related to an impaired skin barrier under these conditions. For children, the same patch test concentration as for adults (0.5% potassium dichromate in petrolatum) was recommended, if there was a history of reactions to shoe allergens [142]. In addition to patch testing, less commonly used tests are immersion tests [143] and repeated open application or prolonged tests [79, 91].

In order to identify potential sources of Cr in a simple and inexpensive way, CrVI spot testing using diphenylcarbazide (DPC) has been suggested and conducted [4, 144, 145]. The test is, however, more challenging to interpret and conduct correctly, as compared to other spot tests such as for nickel and cobalt. This is due to its specificity to CrVI, relatively high detection limit of 0.5 mg/L CrVI [145], and the potential loss of specificity to CrVI upon oxidation in air, facilitated by illumination [146]. CrVI is easily reduced to CrIII on many surfaces, including leather and metal surfaces, at a sufficiently high relative humidity in air [61, 93, 109]. The DPC spot test can hence miss CrVI that has been reduced to CrIII, but may be oxidized again on leather surfaces when the relative humidity is low [61, 109]. The DPC spot test may furthermore miss lower CrVI concentrations and all CrIII, which are also able to cause positive reactions (Tables 27.2 and 27.3). Despite these limitations, the DPC spot test might be a reasonable initial rapid screening tool, and contact with items that test positive should be avoided for a Cr-sensitive person.

Currently, other tests cannot be performed as easily and inexpensively. X-ray fluorescence can detect the total Cr content of a product, but does not give information on its potential Cr release, which often differs (Table 27.4). Release tests can be very relevant and should ideally be able to distinguish between CrIII and CrVI (speciation testing). Candidates for speciation testing are the DPC test by means of spectrophotometry (UV-vis), testing by means of stripping voltammetry (polarography), and distinction by chromatography techniques, before/in conjunction with analytical techniques for total Cr such as atomic absorption spectroscopy and inductively coupled plasma spectroscopy [147]. The DPC spectrophotometry test is used both in standard CrVI testing in cement [148] and in leather [149]. More details on the testing of CrVI in leather can be found in Chap. 4.