One aspect of bath photochemotherapy (PUVA) that requires clarification is
the duration of psoralen-induced cutaneous photosensitisation under conditi
ons simulating clinical use. Using a half back comparison study technique,
we investigated the persistence of trimethylpsoralen (TMP)-induced photosen
sitivity in skin irradiated to simulate a first PUVA exposure compared with
un-irradiated skin. Baseline UVA minimal erythema dose and minimal phototo
xic dose (MPD) were determined in 13 healthy volunteers. After readings at
72 h, subjects were bathed in TMP bath water for 15 min and one half of the
back was immediately exposed to 40% of the MPD. Test sites (1.5 cm(2)) on
both halves of the back were then irradiated with a UVA dose series at 15 m
in, 5, 10, 24, 34, 48 and 72 h after the bath. MPD readings were recorded v
isually at 72 h after each UVA exposure. The UVA MED was >25 J/cm(2) in all
the subjects. At each time point, a phototoxic index (PI) was calculated a
s UVA MED/MPD, In un-irradiated skin, photosensitivity returned to normal (
PI=1) within 24 h after the TMP bath, In contrast, skin pre-irradiated to s
imulate the first PUVA treatment was still significantly photosensitive (PI
=2.3; P=0.002) at 48 h. Contrary to previous recommendations, these data su
ggest that patients should be advised to avoid ambient or artificial source
s of UVA throughout their course of TMP bath PUVA to reduce the risk of pho
totoxic erythema.