The currently used 8% fragrance mix (FM I) does not identify all patients with a positive history of adverse reactions to fragrances. A new FM II with 6 frequently used chemicals was evaluated in 1701 consecutive patients patch tested in 6 dermatological centres in Europe. FM II was tested in 3 concentrations - 28% FM II contained 5% hydroxyisohexyl 3-cyclohexene carboxaldehyde (Lyral), 2% citral, 5% farnesol, 5% coumarin, 1% citronellol and 10%alpha-hexyl-cinnamic aldehyde; in 14% FM II, the single constituents' concentration was lowered to 50% and in 2.8% FM II to 10%. Each patient was classified regarding a history of adverse reactions to fragrances: certain, probable, questionable, none. Positive reactions to FM I occurred in 6.5% of the patients. Positive reactions to FM II were dose-dependent and increased from 1.3% (2.8% FM II), through 2.9% (14% FM II) to 4.1% (28% FM II). Reactions classified as doubtful or irritant varied considerably between the 6 centres, with a mean value of 7.2% for FM I and means ranging from 1.8% to 10.6% for FM II. 8.7% of the tested patients had a certain fragrance history. Of these, 25.2% were positive to FM I; reactivity to FM II was again dose-dependent and ranged from 8.1% to 17.6% in this subgroup. Comparing 2 groups of history - certain and none - values for sensitivity and specificity were calculated: sensitivity: FM I, 25.2%; 2.8% FM II, 8.1%; 14% FM II, 13.5%; 28% FM II, 17.6%; specificity: FM I, 96.5%; 2.8% FM II, 99.5%; 14% FM II, 98.8%; 28% FM II, 98.1%. 31/70 patients (44.3%) positive to 28% FM II were negative to FM I, with 14% FM II this proportion being 16/50 (32%). In the group of patients with a certain history, a total of 7 patients were found reacting to FM II only. Conversely, in the group of patients without any fragrance history, there were significantly more positive reactions to FM I than to any concentration of FM II. In conclusion, the new FM II detects additional patients sensitive to fragrances missed by FM I; the number of false-positive reactions is lower with FM II than with FM I. Considering sensitivity, specificity and the frequency of doubtful reactions, the medium concentration, 14% FM II, seems to be the most appropriate diagnostic screening tool.

In order to find sensitizers additional to the current fragrance mix (FM) a series of fragrance materials (series II) was evaluated in 6 dermatological centres in Europe. 11 of the test materials were essential oils, the remaining 7 being either mixtures of isomers or simple chemicals of frequent usage in the perfume industry. 1606 patients were consecutively tested with series II and 8% FM. Each patient was classified regarding a history of adverse reactions to scented products: certain, probable, questionable, none. Reactions to FM occurred most frequently in 11.4% of the subjects. The 6 materials with the highest reactivity after the FM were ylang-ylang oil (YY) I (2.6%), YY II (2.5%), lemongrass oil (1.6%), narcissus absolute (1.3%), jasmine absolute (1.2%) and sandalwood oil (0.9%). 48 (3.0%) of the patients reacted only to materials of series II and not to FM. 6.0% of 1606 patients gave a history of adverse reactions to fragrances which was classified as certain. This group reacted to FM only in 22.9%, to series II and FM in 15.6% and to series II only in 5.2%. 63.5% of the patients reacting to both FM and 1 of the materials of series II had some type of positive fragrance history, which was higher in comparison to those with isolated reactions to FM (46.2% of 121) or to series II, respectively, (45.8% of 48). However, this difference was not statistically significant. In conclusion, the materials of series II identified a further subset of patients with a fragrance problem, which would have been missed by the current FM as the single screening tool for patch testing.

Periorbital dermatitis is common and can be due to the external use of ophthalmic drugs. We evaluated patch test results of the Information Network of the Departments of Dermatology. During a 5-year period (1995-99), of a total 49,256 patch-tested patients, 1053 (2.1%) were eventually diagnosed as allergic periorbital contact dermatitis (APD) and 588 (1.2%) as non-allergic periorbital dermatitis (NAPD). Patient characteristics between APD, NAPD and other cases (OCs) differed with respect to sex (19.7% male in both periorbital groups versus 36.3% in OCs), atopic dermatitis (10.4% in APD versus 60.2% in NAPD versus 16.9% in OCs) and age, APD being substantially more often (68.2%) aged 40 and above than NAPD (52.6%). Several of the top allergens in OCs [such as fragrance mix, Myroxylon pereirae resin (balsam of Peru), lanolin alcohol and potassium dichromate] caused significantly fewer positive test reactions in both periorbital groups. In contrast, thimerosal, phenylmercuric acetate, sodium disulfite, gentamicin sulfate, phenylephrine hydrochloride and benzalkonium chloride tested positively significantly more often in APD but not in NAPD, verifying them as true ophthalmic allergens. Finally, in 42 cases (4%) of APD patients, additional allergens were identified by testing of the patients' own substances (mostly beta-blockers, oxybuprocaine and dexpanthenol), supporting the necessity of testing with ophthalmic drugs as is where individual substances are not readily available.

50 patients (2.9%) showed a positive reaction to 14% FM II and 70 patients (4.1%) to 28% FM II. 24/50 (48%) produced a positive reaction to 1 or more of the individual constituents of 14% FM II and 38/70 (54.3%) to 28% FM II, respectively. If doubtful reactions to individual constituents are included, the break-down testing was positive in 74% and 70%, respectively. Patients with a positive reaction to 14% FM II showed a higher rate of reactions to the individual constituent of the 28% FM II: 36/50 (72%). Positive reactions to individual constituents in patients negative to FM II were exceedingly rare. If doubtful reactions are regarded as negative, the sensitivity, specificity, positive predictive value and negative predictive value for the medium concentration of FM II towards at least 1 individual constituent was 92.3% (exact 95% confidence interval 74.9-99.1%), 98.4% (97.7-99.0%), 48% (33.7-62.6%) and 99.9% (99.6-"100.0%), respectively. For the high concentration, the figures were very similar. The frequency of positive reactions to the individual constituents in descending order was the same for both FM II concentrations: hydroxyisohexyl 3-cyclohexene carboxaldehyde (Lyral) > citral > farnesol > citronellol > alpha-hexyl-cinnamic aldehyde (AHCA). No unequivocally positive reaction to coumarin was observed. Lyral) was the dominant individual constituent, with positive reactions in 36% of patients reacting to 14% FM II and 37.1% to 28% FM II. 5/11 patients developed a positive ROAT after a median of 7 days (range 2-10). The 5 patients with a doubtful or negative reaction to 28% FM II were all ROAT negative except 1. There were 7 patients with a certain fragrance history and a positive reaction to either 28% or 14% FM II but a negative reaction to FM I. Analysis with GC-MS in a total of 24 products obtained from 12 patients showed at least 1-5 individual constituents per product: Lyral (79.2%), citronellol (87.5%), AHCA (58.3%), citral (50%) and coumarin (50%). The patients were patch test positive to Lyral, citral and AHCA. In conclusion, patients with a certain fragrance history and a negative reaction to FM I can be identified by FM II. Testing with individual constituents is positive in about 50% of cases reacting to either 14% or 28% FM II.

Contact allergy to fragrances is a common problem world-wide. The currently used fragrance mix (FM) for patch testing has only eight constituents and does not identify all fragrance-allergic patients. As perfumes may contain 100 or more substances, the search for markers for allergy continues. The synthetic fragrance 4-(4-hydroxy-4-methylpentyl)-3-cyclohexene carboxaldehyde (Lyral) was tested together with the FM and 11 other fragrance substances on consecutive patients in six European departments of dermatology. All patients were carefully questioned regarding a history of reactions to scented products in the past and were grouped into four categories: 'certain', 'probable', 'questionable' and 'none'. Lyral (5% in petrolatum) gave a positive reaction in 2.7% of 1855 patients (range 1.2-17%) and ranked next to 11.3% with FM allergy. Twenty-four patients reacted to both Lyral and FM, but 21 (1.1%) reacted positively only to Lyral. Of 124 patients with a 'certain' history, 53.2% reacted to the FM and a further 7.2% to Lyral only. If any kind of history of fragrance intolerance was given, 80% (40 of 50) of Lyral positive patients had a 'positive' history while only 58.6% (123 of 210) of FM positive patients had such a history; this difference was significant at P < 0.01. Lyral was identified by gas chromatography-mass spectrometry in some products which had caused an allergic contact dermatitis in four typical patients who showed a patch test positive to Lyral and negative or doubtful to FM. In conclusion, we recommend the testing of 5% Lyral (in petrolatum) in patients suspected of contact dermatitis.