Comparison of Demographic and Photobiological Features of Chronic Actinic Dermatitis in Patients With Lighter vs Darker Skin Types

Key Points

Question  Do demographic and photobiological features differ between patients with chronic actinic dermatitis and darker vs lighter skin types?

Findings  This retrospective evaluation of 70 patients with chronic actinic dermatitis found that patients with darker skin types present at a younger age and are more frequently female compared with patients with a lighter skin type, although phototest reactions are equally severe. Reactions to photopatch testing are also common in patients with chronic actinic dermatitis.

Meaning  In contrast to classic chronic actinic dermatitis, patients with darker skin types are more frequently female and may present at a much earlier age, but patients with all skin types can safely undergo photopatch testing.

Importance  Chronic actinic dermatitis (CAD) is classically described in older, white men, although increasing reports describe younger patients with darker skin types, particularly South Asians. Photocontact allergy occurs in CAD but is less studied than contact allergy in this exquisitely photosensitive condition.

Objective  To evaluate for differences in demographic and photobiological features between persons with darker and lighter skin types who have CAD.

Design, Setting, and Participants  This retrospective review included 70 consecutive adult patients (≥18 years) undergoing investigation for photosensitivity who were diagnosed with CAD from November 1, 2000, through August 31, 2015, at the specialist Photobiology Unit of a tertiary academic referral center.

Main Outcomes and Measures  Patient age, sex, ethnicity, clinical features, and phototesting outcomes.

Results  A total of 70 patients (37 men [53%] and 33 women [47%]; mean [SD] age, 50.9 [2.3] years) were diagnosed with CAD. Of these, 36 were non-Hispanic and non-Latino white, 31 were Asian (including 24 South Asian, 4 East Asian, and 3 Middle Eastern), and 3 were black. Patients were aged 9 to 83 years at diagnosis, with a mean (SD) age at onset of 42.6 (2.4) years and duration of disease of 8.8 (1.3) years. Forty-one had lighter skin types (Fitzpatrick skin types I-IV), and 29 had darker skin types (Fitzpatrick skin types V and VI). Patients with darker skin types and CAD were younger at diagnosis (mean [SD] age, 40.7 [3.5] vs 58.1 [2.5] years; P < . 001) and had earlier onset of photosensitivity (mean [SD] age, 35.5 [3.9] vs 47.5 [2.9] years; P = .01) compared with patients with lighter skin types. Of note, the male to female ratio in the lighter skin group was 2:1 compared with 1:2 in the darker skin group. Phototest reactions were equally severe in Fitzpatrick skin types V to VI and I to IV, with minimal erythemal doses to monochromatic UV-B, UV-A, and visible radiation and broadband provocation testing showing similar results. Photoallergic contact reactions to UV filters, personal sunscreen products, and nonsteroidal anti-inflammatory drugs were seen in both groups; 14 of 61 patients (23%) undergoing photopatch testing showed positive photopatch reactions.

Conclusions and Relevance  Chronic actinic dermatitis presents with an earlier age at onset and an inverted male to female ratio in patients with darker compared with lighter skin types. Clinicians should thus be cognizant of CAD in younger women with darker skin types. Photopatch testing should be considered in patients with CAD, with coexistent photocontact allergy occurring in a substantial proportion.

Chronic actinic dermatitis (CAD) is a photosensitivity disorder classically described in older, white men.¹ Although the exact cause of CAD remains elusive, increased susceptibility to development of delayed-type allergic responses to endogenous photoallergens and exogenous allergens has been proposed.² The condition is clinically distinct, defined by a persistent dermatitis and/or pseudolymphomatous eruption affecting predominantly photoexposed sites and by results of monochromator phototesting, which typically show severely reduced minimal erythemal doses (MEDs), especially in the UV-B and shorter UV-A wavelengths.^3,4 Chronic actinic dermatitis is associated with multiple contact allergies, including to sesquiterpene lactones, fragrance, colophony, rubber, and sunscreens; however, the association with photocontact allergy is less well known because photopatch testing is less commonly performed. Younger patients,^5,6 especially those with darker skin types,⁷ appear to be increasingly developing CAD, although the prevalence is unknown. We evaluated patients diagnosed with CAD during a 15-year period in a specialized photoinvestigation unit and investigated differences in demographic and photobiological features between patients with lighter and darker skin types.

We performed a retrospective review of the records of patients with CAD diagnosed in the Photobiology Unit, Dermatology Centre, Salford Royal NHS (National Health Service) Foundation Trust, Manchester, England, from November 1, 2000, through August 31, 2015. All cases were diagnosed by a photodermatologist (T.C.L. and L.E.R.) through clinical and photobiological assessment. The latter consisted of monochromator phototesting to narrowband UV-B, UV-A, and visible radiation; provocation testing to broadband UV-A and solar simulated radiation; and photopatch testing to sunscreen filters, sun-protection products, and nonsteroidal anti-inflammatory drugs. Blood and urine sampling was performed as described below. Ethical approval and informed consent were waived for this retrospective review of medical records by the Salford Royal Hospital Department of Research and Development.

Patients referred for photosensitivity assessment attended a standardized 4-day photoinvestigation program. A detailed history was obtained, including age at onset; distribution and natural history of the skin condition; seasonal variation; whether photosensitivity improved or worsened with use of sunscreens; detailed history of therapeutic drug use at the onset of photosensitivity and at the time of assessment; excessive ingestion of foods or drinks with phototoxic constituents, namely, quinine and psoralens (eg, tonic water, parsley, and parsnip); personal and family history of atopy or photosensitivity; Fitzpatrick sun-reactive skin type⁸; ethnicity; occupation; and recreational activities. Morphologic features and distribution of skin lesions were recorded.

Patients were exposed to narrowband UV and visible radiation from 300 to 600 nm (±half-maximum bandwidth) using the following geometric series of doses at each wave band: 0.0018 to 0.08 J/cm² (300 ± 5 nm), 0.13 to 4.00 J/cm² (320 ± 10 nm), 0.44 to 14.00 J/cm² (330 ± 10 nm), 0.90 to 40.00 J/cm² (350 ± 20 nm), 1.80 to 57.00 J/cm² (370 ± 20 nm), 3.50 to 113.00 J/cm² (400 ± 20 nm), and 50.00 J/cm² (500 ± 20 and 600 ± 20 nm) (1-kW xenon arc lamp coupled to a 0.25-m grating monochromator; Newport Spectra-Physics Ltd). Reference ranges were originally established at another center in Northern England.⁹ Irradiance was measured using a calibrated thermopile (Medical Physics Department, Dryburn Hospital, Durham, England) and digital voltmeter (Medical Physics Department, Royal Liverpool University Hospital, Liverpool, England).

Provocation testing was performed on 5 × 5-cm areas of ventral forearm for as many as 3 consecutive days to 15 J/cm² of broadband UV-A, using a custom-built circumferential arm exposure unit incorporating 310- to 400-nm fluorescent bulbs (Cleo Performance, Phillips Healthcare UK Ltd) and, separately, to 10 J/cm² of solar simulated radiation (290-400 nm, 1-kW xenon arc plus atmospheric attenuation filter; Newport Spectra-Physics Ltd).

Photopatch testing with control patches was performed to a series of 25 agents,¹⁰ including 19 UV filters, 4 nonsteroidal anti-inflammatory drugs (Chemotechnique Diagnostics), and 2 prescription sunscreen products, from 2009 to 2015. Before 2009, the photopatch series consisted of 10 agents,¹¹ including 9 organic UV filters and 1 sunscreen product. Patients’ own sunscreen products were also applied. Duplicate patches were applied (day 1) to the skin of the midback for 24 hours, after which 1 set was irradiated (day 2) with 0.50 to 5.00 J/cm² of broadband UV-A (310-400 nm) (UVAL 801; Herbert Waldmann GmbH & Co KG). The UV-A was dosed according to the depth of erythemal response on day 2 at the UV-A provocation site (ie, 5.00 J/cm² was used if mild erythema was seen; 2.50 J/cm², if moderate erythema was seen; and 0.50-1.00 J/cm², if severe erythema was seen). Readings were made at 24 and 48 hours after UV administration (days 3 and 4) to examine for a crescendo response, using the International Contact Dermatitis Research Group (ICDRG) grading (range, nil to +++, indicative of more intense contact reaction).^11,12

Routine assessment included a plasma and urine porphyrin scan, serum autoantibody screen, and tests for levels of IgE and 25-hydoxyvitamin D. Skin biopsy was rarely indicated.

We used the Dermatology Life Quality Index (DLQI)¹³ questionnaire to assess the effect of skin disease on quality of life. Because clinical photosensitivity can fluctuate depending on the season and ambient UV or visible radiation, questionnaires focusing on events in the past week may underestimate the effect. Thus, questionnaires separately assessed the effect during the past week and the past year.¹⁴ The maximum potential score on the DLQI is 30, with higher scores equating to greater impairment of quality of life.¹³

The data were analyzed using analysis of variance (version 2.7.9; StatsDirect, Ltd). Statistical significance was accepted at the P < .05 level. Unless otherwise indicated, data are presented as mean (SD).

A total of 2025 patients underwent photoinvestigation from 2000 to 2015. Characteristics of the 70 patients (37 men [53%] and 33 women [47%]; mean [SD] age, 50.9 [2.3] years) diagnosed with CAD are shown in Table 1. Patients were aged 9 to 83 years at diagnosis, with a mean (SD) age of onset of 42.6 (2.4) years and a mean (SD) duration of disease of 8.8 (1.3) years. Five patients (7%) were 21 years or younger; 19 (27%), 22 to 40 years; 18 (26%), 41 to 60 years; and 28 (40%), older than 60 years. Forty-one patients had lighter skin types (Fitzpatrick I-IV), and 29 had darker skin types (Fitzpatrick V and VI). The population included 36 non-Hispanic and non-Latino white patients (25 male; mean [SD] age, 61.1 [14.1] years; mean [SD] age at onset, 49.6 17.9] years); 31 Asian patients, including 24 South Asian patients (9 male; mean [SD] age 41.9 [19.9] years; mean [SD] age at onset, 36.4 [21.7] years), 4 East Asian patients (3 Chinese or white Chinese and 1 Laotian; 3 male; mean [SD] age, 45.8 [25.7] years; mean [SD] age at onset, 36.3 [21.7] years), and 3 Middle Eastern patients (2 Saudi Arabian and 1 Kuwaiti; 3 female; mean [SD] age, 37.0 [11.5]years; mean [SD] age at onset, 32.0 [14.2] years); and 3 black patients (1 Libyan, 1 Somalian, and 1 Afro-Caribbean; 3 female, mean [SD] age, 29.7 [13.6] years; mean [SD] age at onset, 25.3 [14.0] years).

Twenty-six patients (37%) had a history of atopic eczema, and an additional 10 patients (14%) without eczema had other features of atopy, including asthma and allergic rhinitis. In addition, of the 44 patients without atopic eczema, 4 had a history of contact allergic dermatitis, 3 had a history of hand eczema, and 2 had unspecified eczema. Patients with CAD and darker skin types were younger at diagnosis (mean [SD] age, 40.7 [3.5] vs 58.1 [2.5] years; P < .001) and had an earlier age at onset of photosensitivity (mean [SD], 35.5 [3.9] vs 47.5 [2.9] years; P = .01) compared with patients with lighter skin types. A detailed history of therapeutic drug use was taken; 8 patients were using potentially photosensitizing drugs, but only in 3 did the medication use predate photosensitivity, with a latent period of at least 5 years.

Patients showed characteristic clinical and photobiological features of CAD (Figure 1). Chronic actinic dermatitis typically manifested as a photodistribution of eczematous and sometimes lichenified lesions, including the face, V area of the chest, nape of the neck, dorsal forearms, and hands, with sharp demarcation between affected and sun-protected areas. In patients with darker skin, responses on monochromator phototesting were more evident from the raised and palpable nature of responses than the erythema (Figure 1F).

For each wavelength tested, a mean MED for the patients was calculated using the lowest dose point at which a response was seen for that wavelength; this value may not represent the absolute threshold in cases in which patients responded to the lowest dose tested. Severely reduced MEDs at 24 hours were seen in all patients, with the action spectrum predominantly in the UV-B and shorter UV-A range and often spreading to longer-wavelength UV-A and including the 400-nm border of visible light, but infrequently beyond this (Table 1 and Figure 2). Four patients had normal UV-B thresholds but were classified as having CAD owing to their severely low UV-A thresholds and consistent clinical findings. Mean MEDs for the different wave bands were similar for darker and lighter skin types, except at 400 nm, where the mean MED was lower in the group with lighter skin types.

All patients had positive results of provocation to broadband UV-A and solar simulated radiation (Figure 1C and D). Most patients developed an erythemal response after the first test, often followed by development of eczematous features (particularly scaling) if the test was repeated.

Detailed photopatch testing was performed concurrently with phototesting. Despite the challenge posed by UV-A irradiation in such severely photosensitive patients, in total 61 underwent photopatch testing with control patch testing, and 5 underwent the control patch testing component alone. The irradiation dose used ranged from 0.50 to 5.00 J/cm² (1 had 0.50 J/cm²; 34, 1.00 J/cm²; 12, 2.50 J/cm²; and 14, 5.00 J/cm²). Overall, 14 of 61 patients (23%) had positive photopatch reactions, including 11 of 37 with lighter skin types (28%; types I-IV) and 3 of 24 with darker skin types (12%; types V and VI). Most patients had 1 reaction on photopatch testing (8 of 14) (Figure 1G), whereas 3 patients had 2 reactions and 3 had 3 or more reactions. Details of the positive photopatch test results are shown in Table 2. Among the 14 patients with a positive photopatch test result, 9 were given only 1.00 J/cm² of UV-A; 2, 2.50 J/cm²; and 3, 5.00 J/cm².

We additionally found that 15 of 66 patients (23%) showed contact reactions to our photopatch test panel, constituting 12 of 40 patients (30%) with skin types I to IV and 3 of 26 patients (12%) with skin types V and VI. Standard patch testing for contact allergy was also performed in 25 patients by their dermatologists (21 patients with skin types I-IV and 4 with skin types V and VI); the most frequent positive contact reactions were to fragrance, nickel, balsam of Peru, and thiazolinone (eTable 1 in the Supplement).

Results of the plasma and urine porphyrin scans were negative in all patients. Test results for antinuclear antibody were positive in 17 of 70 patients (11 with skin types I-IV and 6 with skin types V and VI). Fourteen of these patients had a low titer of 1:100, and 3 had a titer greater than 1:1000; these findings did not appear to be clinically relevant, and test results for DNA and extractable nuclear antigen antibodies were negative. Serum IgE levels were elevated in 35 of 49 patients (71%), with a similar proportion in patients with light (20 of 29 [69%]) and dark (15 of 20 [75%]) skin types (Table 1). Vitamin D status (ie, 25-hydroxyvitamin D level) is shown in Table 1. Skin biopsy was rarely performed because CAD was diagnosed using clinical and phototest findings; 1 patient underwent biopsy of his naturally occurring condition, revealing histologic features of chronic dermatitis.

A subset of patients completed the DLQI questionnaire (33 for the past week and 31 for the past year; Table 3) because this measure was routinely introduced in 2011. Comparison of the week and year scores revealed no significant difference (P = .86). The effect on quality of life was designated very large (DLQI score, 11-20) to extremely large (DLQI score, 21-30), with 15 of 33 patients (45%) having a DLQI score of greater than 10 for the past week and 24 of 31 (77%) having a DLQI score of greater than 10 for the past year. These findings were similar in patients with darker and lighter skin types.

The patient demographic characteristics in our study highlight differences in sex distribution and age at onset of photosensitivity in patients with lighter (I-IV) and darker (V and VI) skin types who presented with CAD. The former are predominantly older male individuals, consistent with the earlier CAD literature,^1-4 whereas the latter are more often younger female individuals. We found a female to male ratio of 2:1 among those with darker skin types (ie, a reversal of the female to male ratio of 1:2 seen in patients with lighter skin types). This result is consistent with the Michigan study of African American patients with photodermatoses by Kerr and Lim,¹⁵ which found the female to male ratio in 15 Afro-Caribbean patients with CAD was 2:1. This pattern may be more widespread in patients with darker skin types and across continents. Wadhwani et al¹⁶ reported the female to male ratio of 1:3.2 among 50 patients in New Delhi, India, although the diagnosis of CAD was made without monochromator phototesting.

Five patients with CAD were 21 years or younger when they presented to our unit. The youngest was a 9-year-old girl of mixed white and Chinese heritage with skin type IV who developed photosensitive features 1 year earlier. The other younger patients were of South Asian (n = 3) and Afro-Caribbean (n = 1) descent who developed similar features 4 to 7 years earlier.

The photobiological characteristics in our patients with CAD were typical of the literature, with classic markedly reduced MEDs to UV-B and with UV-A involvement (Table 1 and eTable 2 in the Supplement). Predominantly UV-A involvement was reported in 25 of 507 patients (4.9%) in a large study of this rare disorder from Dundee, Scotland.² This finding was consistent with our finding of severe UV-A sensitivity alone in 4 of 70 patients (6%) (2 with lighter and 2 with darker skin types); of note, none had a history of photosensitizing drug use. In our study, lighter and darker skin types were as severely affected on results of monochromator phototesting, with markedly reduced erythemal thresholds, especially in the 300-nm wave band, and with phototesting frequently producing palpable lesions.

Photopatch testing with control patch testing is a routine part of photosensitivity investigation in our unit owing to frequent positive findings in the patients with photosensitivity. We use the European Academy of Dermatology and Venereology standardized battery of sunscreen filters and nonsteroidal anti-inflammatory drugs for testing.^10,11 Positive photocontact reactions were seen in 23% of all patients undergoing testing, whereas contact reactions to patch controls alone were seen in another 23% of patients. Benzophenone-3 was the most common sunscreen filter that caused photocontact and contact reactions in people with lighter and darker skin types (Table 2). Barber and Cronin¹⁷ found positive photopatch reactions in 5 of 47 patients with CAD, with musk ambrette (4 of 5 patients) as the main photoallergen, although only 6 potential photocontact allergens were included in their study from 1980 through 1981. Menagé et al¹ reported 12% positive photopatch reactions in 89 patients with CAD who underwent testing (to musk ambrette, oxybenzone, and para-aminobenzoic acid) from 1987 through 1992. Our data, produced from photopatch testing of virtually all patients presenting with CAD and using a wide, standardized contemporary battery, provide a salient addition to the literature.

Most of our patients with CAD and a positive photopatch test reaction underwent irradiation with only 1.00 J/cm² of UV-A radiation, indicating that this low dose is sufficient to elicit a positive response. This information is of practical significance in these patients with severe photosensitivity for whom lesion provocation could complicate the procedure. We cannot rule out that the low UV-A radiation dose might have been insufficient to activate some photoallergens; thus, these patients might have an even higher rate of associated photocontact allergy. Although reluctance by some departments to subject patients with severe photosensitivity to UV-A irradiation for photopatch testing is understandable, we show a substantial positive response of 23% of patients with CAD, indicating that pursuit of this investigation is informative, with complications mitigated by using a small dose of UV-A radiation.

Conversely, coexistent contact allergy is well reported in CAD.¹ Results of contact testing to a standard patch test battery were available in 25 of our 70 patients, provided by the referring dermatologists. The more common contact allergic reactions were to fragrance, balsam of Peru, thiazolinone, sesquiterpene lactones, colophony, nickel, and cobalt; the latter 2 might reflect the background atopy of many of the patients.^18-20 Contact sensitivity to sesquiterpene lactones, known to be associated with CAD but thought to be decreasing in prevalence in patients with CAD,²¹ is still relatively frequent in our study. Positive patch test reactions to paraphenylenediamine, possessing cross-reactivity with sesquiterpene lactones,²² are also seen in CAD,⁷ although a role in pathogenesis remains unproven. Parthenium dermatitis, a common cause of plant (Parthenium hysterophorus) dermatitis in India, is classically an airborne contact dermatitis²³ but is reported to develop into a photodermatitis resembling CAD.²⁴ Such plants are native to tropical America, India, and Australia, whereas the patients in our review lived mostly in the United Kingdom.

Among the 2025 patients undergoing photoinvestigation during the 15 years of our study period, an additional 378 (18.7%) were diagnosed with photoaggravated eczema. These patients had reduced erythemal thresholds predominantly in the UV-A rather than the UV-B range and of a less severe degree. However, this finding questions the relationship between the more severe photoaggravated eczema and the small percentage of patients with CAD with severely low UV-A rather than UV-B thresholds, that is, whether they represent a continuum rather than completely distinct disorders. Serum IgE levels were elevated in 71% of patients with CAD undergoing this assessment, with similar proportions seen in patients with lighter and darker skin types (Table 1), consistent with the reported association of CAD with atopy.^5,6

Assessment of vitamin D status, measured as circulating levels of 25-hydroxyvitamin D in 30 patients (from 2011 onward), showed that 16 (53%) were in the vitamin D deficiency range (<10.0 ng/mL [to convert to nanomoles per liter, multiply by 2.5]), at which the bone disorders rickets (in children) and osteomalacia (in adults and children) most often occur. Patients with lighter and darker skin types were similarly affected, whereas nonphotosensitive persons with darker skin typically have lower vitamin D status than those with lighter skin, including in Greater Manchester (53.5° N).²⁵ This finding illustrates the vigilant sun avoidance and photoprotection practices that these patients with severe photosensitivity adopt.²⁶ Low vitamin D status is well documented in photosensitivity; in a mixed-diagnosis group, insufficient 25-hydroxyvitamin D levels (≤20.0 ng/mL) were found in 47% of patients in summer, increasing to 73% in winter, whereas deficient levels were seen in 9% summer and 32% in winter.²⁶ Assessment of vitamin D status is recommended in patients with photosensitivity, and supplementation should be instituted for patients with insufficient or deficient levels.^27,28

Dermatology Life Quality Index scores for the past week and year were obtained for 33 and 31 patients, respectively (Table 3). The largest category of patients (15 of 33) had DLQI scores greater than 10, indicating a very large to extremely large impairment of quality of life,²⁹ similar in patients with lighter and darker skin types. This finding is consistent with that of a multicenter study of 790 patients with photodermatoses in the United Kingdom,¹³ among whom 39% (31) of 127 patients with CAD had DLQI scores of greater than 10.

The main limitation of our review was its retrospective nature, although all patients underwent assessment according to standardized clinical and phototest proforma. Interpretation of MED in patients with darker skin types is sometimes visually challenging, whereas the palpable “doughy” texture assists detection of responses. This interpretation may be assisted in the future by devices that objectively determine variations in skin perfusion.³⁰ Photopatch test batteries will continue to evolve to reflect current photoallergen prevalences.

Our review found that a substantial proportion of CAD now presents in younger female patients with darker skin; similar photosensitivity and effect on quality of life are seen as for the typically presenting older white male patients. What is to our knowledge the largest reported series of photopatch testing for this photosensitive disorder shows that the investigation can be performed successfully, with a high yield of photopatch positivity.

Corresponding Author: Lesley E. Rhodes, FRCP, Photobiology Unit, Dermatology Centre, University of Manchester, Salford Royal NHS Foundation Trust, Salford, Manchester M6 8HD, England (Lesley.e.rhodes@manchester.ac.uk).

Accepted for Publication: December 7, 2016.

Published Online: March 22, 2017. doi:10.1001/jamadermatol.2016.5861

Author Contributions: Drs Tan and Rhodes had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Haylett, Ling, Rhodes.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Tan, Haylett, Rhodes.

Critical revision of the manuscript for important intellectual content: Ling, Rhodes.

Statistical analysis: Haylett.

Administrative, technical, or material support: Tan, Haylett.

Study supervision: Ling, Rhodes.

Conflict of Interest Disclosures: None reported.

Additional Contributions: Donald Allan, PhD, Salford Royal Hospital, Manchester, England, oversaw correct use of light sources. Vivien Robinson, University of Manchester, Manchester, England, provided administrative assistance. Neither was compensated for this work. We thank the patients for granting permission to publish this information.