Polymorphic Light Eruption Reassessed

Polymorphic light eruption (PLE) is possibly the most common chronic skin disorder in people living in temperate climes. Self-report questionnairesurveys suggest a prevalence of about 15% in the northern hemisphere.¹ It is therefore appropriate that the current research programs of some specialist centers include a reassessment of this disease,including its cause and diagnosis.

In this issue of the ARCHIVES, 2 articles address different aspects of PLE. A team of Dutch investigators² haveexamined the mechanism of the apparent underlying immune defect, whereas a British group³ reassessed the methodology ofprovocation of the disorder.

In an earlier article, Kölgen et al⁴ reported that UV-induced Langerhans cell (LC) migration from the skin was impairedin PLE patients. It was therefore postulated that the pathologic defect underlying PLE might be a failure of normal photoimmunosuppression. Thus the balanceof UV-induced suppression and UV-induced provocation would be altered, allowing sunlight exposure to provoke the PLE eruption. In their article in this issue,Kölgen et al² further examine this hypothesis. They assess whether there are abnormalities of UV-induced secretion of tumornecrosis factor (TNF) α and interleukin (IL) 1β, cytokines known to be important in effecting LC migration. Second, they examine the effectsof UV on secretion of T helper cell type 2 (T_H2) cytokines IL-4 and IL-10, which mediate immunosuppression. These cytokines were assessedby protein immunostaining in sections of skin biopsy specimens taken after application of 6 minimum erythemal doses (MEDs) of UV-B to buttock skin.

Interestingly, the authors² found a reduction in UV-B induction of a number of the cytokines under study. The UV inductionof TNF-α was significantly lower than in healthy subjects, although IL-1β was unchanged. Additionally, UV induction of the T_H2cytokine IL-4 was significantly reduced in PLE patients compared with healthy subjects. A similar finding was noticed for IL-10, although the study methodused did not allow for assessment of significance. A reduced induction of TNF-α, which affects LC migration, and of the immunosuppressive cytokineIL-4 and probably also IL-10, provides confirmation of a defect of UV-inducedimmunosuppression in PLE and goes a long way to provide an explanation ofthe underlying mechanism. A further notable finding was that virtually all (>98%) of the observed TNF-α and IL-4 expression, and to a lesser degreethe IL-10 expression, was seen in neutrophils rather than in keratinocytes. Fewer neutrophils were apparent in the UV-irradiated skin of PLE patientsthan in healthy subjects, and therefore it seems that failure of a normal degree of neutrophilic infiltration might explain the reduced cytokine inductionand consequently the lesser degree of photoimmunosuppression in PLE.

Further exploration is required of this fascinating area. As discussed by the authors,² it would be helpful if cytokinemessenger RNA and protein could be quantified to fully assess the significance of the findings by using reverse transcriptase polymerase chain reaction andby measuring suction blister fluid cytokines. The expression of other neutrophil-derived mediators that might be important in the mechanism of PLE such as the cytolyticmoiety nitric oxide could also be examined. It would be wise to repeat the work with age-matched subjects; the PLE subjects in this report were of widelyvarying ages, while all the healthy subjects were young adults. The 6-MED UV-B dose represents a severe challenge, and therefore the effect of morephysiologic doses of UV could be assessed. To validate that differences in LC migration may have a bearing on disease initiation, the site of early provokedlesions might also be examined in addition to the sunburn response. As illustrated in the other PLE article in this issue,³ thesunburn response and rash provocation appear to be quite separate phenomena.

van de Pas et al³ review the literature on PLE photoprovocation published since 1980, discuss the desirability ofidentifying an optimal method for this procedure, and make useful progress toward describing this method. Over the last 23 years, several published studiesof PLE provocation using a range of irradiation sources and protocols have reported positive provocation rates ranging from 0% to 100%. These includestudies by Hölzle et al,⁵ who reported a UV-A method successful in 90% of patients; Neumann et al,⁶ whofound that repeating exposure over 2 to 4 days increased the chance of success; and McFadden and Larsen,⁷ who discovered thatthe dose for positive provocation testing may vary by a factor of 4 between patients. Approaches have varied between the administration of high-dose UV(several multiples of the MED) and low-dose (suberythemal) challenges.

van de Pas et al³ in this issue of theARCHIVES examine several factors that may affect response with the use ofa solar-simulated radiation (SSR) source. These include the effect of skin site (previously affected vs unaffected), the influence of multiple challenges,and the relationship between rash provocation and erythemal responses. In 25 PLE patients, van de Pas et al³ used a 6-pointSSR dose range on 4 × 4-cm areas of exposed (arm or upper back) and unexposed (buttock) skin. They performed repeated daily challenges to thesesites, to a maximum of 3 challenges in most subjects. Including all sites, the investigators reported that 68% of subjects had a positive provocationafter 2 or 3 exposures and that the probability of a positive response increased with number of exposures. They also found there was an equal likelihood ofthe rash being provoked on exposed and unexposed skin, that repeated suberythemal challenges were more successful than a single higher-dose challenge, and thatprovocation response did not correlate with the slope of UV-erythemal response.

It should be noted that the provocation doses applied were multiples of the MED as derived on buttock skin. As the authors discuss,³ thisindicates that equivalent MED-related doses were in fact not given to exposed and unexposed sites. Hence, the exposed sites received a lesser inflammatorydose of UV, and it might be speculated that equivalent MED-related doses might have resulted in a higher provocation rate. It was interesting that with repeateddoses, PLE could be relatively easily provoked at several skin sites. The previous literature on this topic is conflicting, with some reports of 0%provocation rate on unexposed skin. Hence, it has now been clarified that many patients will respond on unexposed skin as long as their threshold forprovocation is met, both in terms of individual doses delivered and in repetition of doses. Since large subgroups of patients responded only on either exposedor unexposed skin, it appears that with the method described, it is necessary to challenge both sites to reach a reasonable response rate. Further studiesmight be considered to formally compare differences in provocation yield for skin test areas of different sizes.

Another recently devised successful method of PLE provocation supports the use of multiple lower-dose challenges. An irradiation "arm-box" comprisinga cylindrical arrangement of fluorescent tubes, originally devised for psoralen–UV-A studies,⁸ permits challenge to a large area,including the whole of the forearm, with a low (suberythemal) UV dose. With this equipment, up to 3 challenges to the whole forearm and hand at 20 J/cm² of broadband UV-A over consecutive days resulted in positive provocationin 59 (88%) of 67 PLE patients referred for photoinvestigation.⁹ Das et al¹⁰ used the arm-box equipped with eitherUV-A or narrowband UV-B lamps to challenge a 10 × 10-cm area of each arm up to 3 times in 68 PLE patients. While 38 patients (56%) showed a positiveresponse to UV-A, 34 (50%) responded to narrowband UV-B, and the probability of a positive response following irradiation with both lamps was 81%.

The detailed testing system described by van de Pas et al³ could be useful for defining a patient's provocation threshold. This could be appliedto objectively evaluate whether PLE provocation threshold correlates with clinical severity of the disorder. Although more exacting than a single-dosechallenge, it would certainly be feasible to perform multiple, medium-sized area provocation challenges within the confines of a research study, and thiscould be particularly beneficial in the assessment of potential photoprotective agents. Using a less detailed (3-dose) UV-A challenge administered by thearm-box system, researchers have reported a dose-related PLE provocation response, which was previously used to test the photoprotective properties of an oralagent.¹¹

The differences in PLE provocation rates between reported studies cannot be fully accounted for by differences in challenge methods, since similarlamps and protocols have sometimes been used. Since PLE clinical severity varies widely, it is conceivable that patient groups have not been exactlycomparable. In particular, those patients who have been referred to a tertiary investigation center tend to have clinically severe or atypical clinical patterns.The complexity of the underlying pathogenesis is also likely to contribute to the variable provocation rate, with the current work of Kölgen etal² providing further evidence that UV-immunosuppression pathways and UV-provocation pathways may be involved. Approaches from thedirections illustrated in the 2 articles in this issue, which facilitate a better understanding of the underlying immune etiology and clarify optimalprovocation methods, should ultimately help identify the chromophore(s) involved in the common but deceptively complex disorder recognized clinically as PLE.

This work was supported by the European Union Framework V Programme, project No. QLK4-CT01-0015.