The polymorphous light eruption–severity assessment score does not reliably predict the results of phototesting

Abstract

Polymorphous light eruption (PLE) is a very common photodermatosis in which patient history is highly specific. Phototesting is used to confirm the diagnosis and to determine the action spectrum and the severity of this disease. In daily practice and in research studies, it would be convenient to assess disease severity by patient history only. This study aims to assess PLE disease severity via patient history and compares this with severity assessment via phototesting. Sixty-one patients with PLE were asked 10 standard questions and all were phototested. The answers to the standard questions were coded with linear scores ranging from 0 to 10. The score of each question was plotted as independent variable in a multiple linear regression model against the score of the phototest (minimal number of irradiations necessary to elicit a positive skin lesion, with a maximum of 6 irradiations) as dependent variable using an enter approach. Furthermore, the scores of the separate questions were added to form a total score, the PLE-severity assessment score (PLE-SAS). The medians of these PLE-SASs were compared with the result scores obtained by phototesting. Phototesting was done with ultraviolet A and ultraviolet B irradiation. Fifty-seven of the 61 patients had a positive test result (93%). Using the multiple linear regression model, the severity assessment by patient history (PLE-SAS) compared with the result of phototesting showed two significant contributing questions (adjusted PLE-SAS) (P < 0.05) but with a regression coefficient of 0.2. A significant difference in median scores with the severity assessment (PLE-SAS and adjusted PLE-SAS) between patients testing positive after 1-3 irradiations compared with those testing positive after 4-6 irradiations was present (P < 0.05). However, the overlap quartile range between both groups was such that the PLE-SAS and the adjusted PLE-SAS have little predictive value in individual patients. We showed that in PLE, disease severity as determined using the PLE-SAS or adjusted PLE-SAS did not reliably predict severity as assessed by phototesting. Two significant contributing questions were not discriminating enough to be used as predicting questions to assess severity. Accurate patient history proved to be a reliable method to diagnose PLE. Phototesting is useful to determine the responsible ultraviolet action spectrum and to exclude differential diagnoses like photosensitive eczema, lupus erythematosus or chronic actinic dermatitis. PLE-SAS cannot replace phototesting for determining the severity of PLE.