Contact allergy to >4000 substances has been described (17
), and the allergy has been considered to persist throughout life (18
). However, there are only a few modern studies addressing this issue. Lee and Maibach in 2001 (19
) published a review article on patch test follow-ups years after the initial test, under the heading: ‘Is contact allergy in man life long?’ They found four articles on different separate substances, colophonium (20
), nickel (21
), and cobalt (23
). The number of retested persons varied from 50 to 104 in the different studies. The proportion who had become negative in the test ranged from 4% to 23%. Josefson et al. (24
) retested 369 women in Sweden more than 20 years after the initial study. Of 38 schoolgirls positive for nickel in the early 1980s, 28 were verified to be positive at retesting, indicating that 30% had become negative. In a study from 1954 (25
), dermatology patients in Denmark were patch tested 2–19 years after the first occasion, and it was found that 111 of 188 positive reactions in the standardized test at that time had ‘disappeared’ (59%).
Even if the contact allergy remains, the connected contact dermatitis can heal if the individual is not exposed to the allergen. Some of the most important contact allergens are metals, for example nickel, cobalt, and chromium. Aluminium can be added to this list, although clinically relevant contact dermatitis is very uncommon (26
). The most important routes of aluminium exposure are vaccination with aluminium-containing vaccines, hyposensitization with aluminium-adsorbed extracts, and contact with aluminium in antiperspirants and other skin care products.
Ideally, the present study should have had a control group consisting of the children who had vaccine-related itchy nodules with negative patch test results at the initial testing when they were tested with ongoing or recent pruritus for the first time ≥ 5 years ago (7
). Informal contacts between one of the authors (B.T.) and members of the regional ethics committee indicated, however, that it would be difficult to obtain permission to retest children who had never had a positive test result. We agreed with that evaluation, and therefore did not offer retesting of initially negative children.
The main finding in the present study was that aluminium allergy as shown by a positive patch test result and related to vaccination with aluminium-containing vaccines had decreased significantly. With the assumptions made above, it had probably ‘disappeared’ in 77% of children with a previous positive test result. In follow-up patch test studies, such a large proportion turning negative in the second test has never been shown before; in contrast, contact allergies are mostly described as being chronic (18
). In light of the common opinion among dermatologists and in the major textbooks in the field (27
) that aluminium is a very uncommon allergen among unselected children, the rate of positive test results among the children retested here still is very high. The consequence of this is that a large number of children may bring their allergy with them into adolescence and adult life, leading to problems when they need other aluminium-adsorbed vaccines, and when they need, but cannot use, antiperspirants.
We are aware of only one previous study in which children with demonstrated aluminium allergy had been retested (10
). In that study from Denmark, 4 patients with a positive patch test result following vaccination with aluminium hydroxide-adsorbed DTaP were retested (interval not stated). Two of them had become negative. The same study showed, in agreement with the present study, that the subcutaneous nodules and the local itching could resolve with time. Of 21 children followed for 1–8 years, 5 healed and 11 improved clinically.
Disappearance of aluminium allergy as measured with the patch test was significantly correlated with disappearance of local itching. The chance of having a negative second test was also higher with increasing age and a longer time interval between the first vaccination and the second test. The numerical difference for the time elapsed between the first vaccine dose and the date of the test was, however, rather small; only 0.5 years between those with a negative test result and a those with a positive test result. In this study, this time interval seems to be of little clinical relevance.
The results must be regarded with some caution, as it has been suggested that a concentration of aluminium chloride hexahydrate of >2% is needed to detect all cases of aluminium allergy (28
). In one study of 37 children who had undergone hyposensitization therapy with aluminium-containing extracts, aluminium allergy was shown in 8 patients with 10% aluminium chloride hexahydrate, whereas only 3 were clearly positive when the concentration was 2% (26
). There are many publications on aluminium in the literature concerning contact dermatitis, but we have not found a consensus about which aluminium compound, vehicle and concentration could be the gold standard for patch testing. In the handbook commonly used in patch test clinics (17
), more than a handful of different aluminium compounds in different vehicles and concentrations are listed. The reason why 2% was used in the previous study (7
) and present study was that this concentration is used in the baseline series in patch testing and that it is the only one commercially available. Furthermore, a large number of children reacted very strongly to the 2% solution in the previous study (8
), making us hesitant to apply a higher concentration. Besides, there is no consensus regarding what amount of pet. preparation to apply while patch testing. In the present study, one of the authors (U.B.) performed the application of aluminium in pet. and the Finn Chamber®. In the previous study, these applications were performed by a physician (E.B.) or two nurses, who had all been instructed by U.B. on how to perform the applications. Swedish investigators have shown that there is both intraindividual and interindividual variation in the amount of pet. applied, but the individual technician can keep the variation within a limited range (29
Thus, even though we cannot state that the children/young adults in this study had become free from their aluminium allergy, it is clear that the magnitude of the allergy must have decreased.
The interaction between itching nodules and aluminium allergy is unclear, and cannot be clarified by this study. As vaccinations start at an early age in healthy children, it is hardly conceivable that the children were aluminium-allergic prior to the vaccinations. It is more likely that the sensitization reaction to aluminium has led to the granuloma formation. Sarcoidal-type allergic contact granuloma was first described in 1983 from gold earrings (30
) Palladium-induced contact granuloma has also been reported (31
). Case reports of injected aluminium in a tattoo or cosmetic eyebrows have been described (32
). Histopathological examination of the nodules has shown granuloma formations in which aluminium can be shown by staining and atomic absorption spectrometry (5
In conclusion, the present study showed that previously diagnosed aluminium allergy in 241 children following vaccination with aluminium-adsorbed vaccines could be shown in only 55 (22.8%) when ≥5 years had elapsed after the initial testing, with the same procedure and a concentration of 2% aluminium chloride hexahydrate in the patch test. Disappearance of aluminium allergy as measured by the patch test was significantly correlated with disappearance of local itching. The chance of having a negative second test was also higher with increasing age and a longer time interval between the first vaccination and the second test. The numerical difference for the time elapsed between the first vaccine dose and the date of the test was, however, rather small, at only 0.5 years between those with a negative patch test result and those with a positive patch test result. In this study, this time interval seems to be of little clinical relevance. Furthermore, the likelihood of having a negative patch test result at the second test was significantly higher if the intensity of the positive patch test result in the first study was low (+) than if it was high (++ or +++).