Updated Analysis of Standardized Photoprovocation in Patients With Cutaneous Lupus Erythematosus

Authors


Abstract

Objective

To determine the frequency and reproducibility of standardized photoprovocation in patients with cutaneous lupus erythematosus (CLE) and report our long-term experience.

Methods

Photoprovocation using a standardized protocol was evaluated retrospectively in 566 patients. A diagnosis of CLE was clinically and/or histologically confirmed in 431 patients, and 315 patients with polymorphic light eruption (PLE) were additionally included as controls. Data were statistically analyzed using an SPSS database.

Results

A total of 61.7% of the 431 CLE patients exhibited a positive photoprovocation, with a significantly longer latency period for the development of skin lesions after ultraviolet (UV) A and/or UVB irradiation than PLE patients (P < 0.001). The frequency of positive photoprovocation varied among the CLE subtypes, and intermittent CLE was the most photosensitive disease entity (74.8%). Subsequent photoprovocation in 35 patients demonstrated that CLE patients with an initial positive result exhibited a significantly higher frequency of a positive photoprovocation at a later time point (P = 0.013). However, an initial positive photoprovocation did not definitively predict a positive reaction at a later time point. Moreover, patient history of photosensitivity was not a predictor for the photoprovocation outcome.

Conclusion

Standardized photoprovocation is a useful tool to reproducibly induce skin lesions and objectively evaluate photosensitivity in patients with CLE. These data further suggest that the reaction to UV light may change during the course of this heterogeneous disease and that photosensitivity should not be excluded in patients with a negative history of photosensitivity.

INTRODUCTION

The heterogeneous autoimmune disease lupus erythematosus (LE) is characterized by a broad range of cutaneous and systemic manifestations (1). Cutaneous involvement occurs in over 75% of patients during the course of systemic LE (SLE), and skin lesions may precede the onset of systemic organ manifestations by several weeks or months (2). In a high number of patients, the skin can be the only affected organ, which resulted in the definition of cutaneous LE (CLE) (3). Skin lesions of CLE mainly occur in sun-exposed areas. Moreover, it is commonly accepted that sun exposure can induce and exacerbate skin lesions in patients with all subtypes of CLE, which supports the role of ultraviolet (UV) light in the pathogenesis of cutaneous involvement (4). Sun exposure can also induce significant organ involvement in SLE, including lupus nephritis (5).

In 1851, Cazénave initially suggested a relationship between the induction of LE and environmental factors, such as sun exposure (6). Several decades later, Freund (7) characterized 507 patients with LE and described a significant increase in disease frequency in the spring and summer. In the following years, numerous reports linked sun exposure to disease induction and exacerbation (8–10). Accordingly, photosensitivity is included in the American College of Rheumatology (ACR) criteria for the classification of SLE, along with malar rash, discoid rash, and oral ulcers (11, 12). However, the ACR criteria poorly define photosensitivity as a “skin rash as a result of unusual reaction to sunlight, by patient history or physician observation.” In addition, photosensitivity is not specific to the disease, and it is observed in other conditions, such as polymorphous light eruption (PLE), which might have important therapeutic and prognostic implications (13). Therefore, the ACR criteria often overestimate the occurrence of SLE in patients with primarily cutaneous disease manifestations. Accordingly, it would be important to develop a more detailed definition of photosensitivity in the international literature.

In 1965, Epstein and coworkers (14) conducted the first experimental photoprovocation in 25 patients with LE, inventing a repeated exposure technique to objectively determine photosensitivity. In further applications, other investigators used monochromatic light to determine the wavelength by applying the repeated UV exposure technique (15, 16). A standardized photoprovocation protocol based on the discovery that the action spectrum of CLE reaches into the long-wave UVA region was developed by Lehmann and coworkers (17) in 1986. A subsequent study demonstrated that CLE can be induced by experimental UVA or UVB irradiation in 43% of patients (18). We have optimized this standardized photoprovocation protocol by accounting for multiple factors, such as light source, test area of irradiated skin, dose of UV exposure, and frequency of irradiation. In 2001, we previously presented an analysis of 323 CLE patients and reported on our 15-year experience with photoprovocation (19).

Meanwhile, photosensitivity is a well-established factor in the manifestation of CLE, and it is an important tool for the investigation of disease pathogenesis in clinical and experimental settings (4). We previously demonstrated an abnormally high number of apoptotic keratinocytes in skin biopsy specimens of CLE patients after UV irradiation prior to lesion formation (20). In addition, a recent multicenter European study applied the standardized photoprovocation protocol and demonstrated that it is a safe and reproducible technique to induce skin lesions in different CLE subtypes (21). Moreover, we also showed in a randomized controlled trial that the application of a broad-spectrum sunscreen prevents UV-induced skin lesions under standardized conditions (22). In the present monocentric study, we retrospectively analyzed photoprovocation in 566 CLE patients using the standardized protocol and correlated the results with the patients' history of photosensitivity and clinical and laboratory features. These data suggest that standardized photoprovocation is a useful tool to reproducibly induce skin lesions in CLE, and that photosensitivity should not be excluded in patients with a negative history of photosensitivity.

Significance & Innovations

  • To our knowledge, this analysis of photoprovocation using a standardized protocol includes the highest number of patients with cutaneous lupus erythematosus (CLE) to date.

  • In CLE patients, the reaction to ultraviolet light may change during the course of this heterogeneous disease.

  • These data suggest that photoprovocation is a useful tool to reproducibly induce skin lesions in CLE, and that photosensitivity should not be excluded solely on the basis of patient history.

PATIENTS AND METHODS

Patients.

Photoprovocation was performed in 881 patients (626 women [71.1%] and 255 men [28.9%]) with photosensitive disorders at the Department of Dermatology, University of Duesseldorf, Duesseldorf, Germany, between 1990 and 2005. Data sets of 135 patients were excluded from the final statistical analysis because the diagnosis of CLE applying clinical and/or histologic features of primary or UV-induced lesions had not been confirmed. Consequently, data from 431 patients with different subtypes of CLE (278 women [64.5%] and 153 men [35.5%]) and from a control group of 315 patients with PLE (253 women [80%] and 62 men [20%]) were included in the final statistical analysis. In the present analysis, 323 patients with CLE and all patients with PLE were included from a previous report (19). The final data set included 25 acute CLE (ACLE) patients, 77 subacute CLE (SCLE) patients, 102 chronic CLE (CCLE) patients, 123 intermittent CLE (ICLE) patients, and 104 patients with a CLE subtype that could not be further specified (i.e., undifferentiated CLE) (Table 1). The various CLE subtypes were diagnosed using clinical features, histologic changes, and laboratory findings according to the Duesseldorf Classification 2004 (23). In patients with >1 CLE subtype, the respective subtype that was initially diagnosed was defined as the main diagnosis in the statistical analysis. If subtypes occurred simultaneously, the form with the more frequent risk of developing systemic organ manifestation was declared as the main diagnosis (ACLE > SCLE > CCLE > ICLE). The mean ± SD age of CLE patients was 43.2 ± 13.7 years at the time of photoprovocation; the mean ± SD age of PLE patients was 37.2 ± 13.8 years.

Table 1. Demographic data of CLE and PLE patients included in the study*
DiseaseNo. of patientsFemale/maleMean ± SD age, years
  • *

    CLE = cutaneous lupus erythematosus; PLE = polymorphic light eruption; ACLE = acute CLE; SCLE = subacute CLE; CCLE = chronic CLE; ICLE = intermittent CLE.

CLE   
 ACLE2520/535.7 ± 11.4
 SCLE7759/1845.3 ± 15.5
 CCLE10269/3344.1 ± 13.3
 ICLE12364/5941.7 ± 11.9
 Undifferentiated CLE10466/3844.3 ± 14.5
 Suspected CLE13595/4043.6 ± 16.8
 Total566373/19343.0 ± 14.5
PLE315253/6237.2 ± 13.8

All information, including clinical, histologic, laboratory, and photoprovocation data, was collected retrospectively from patient files. Written informed consent was obtained from all patients, and the Local Ethical Committee of the University of Duesseldorf, Duesseldorf, Germany, approved the clinical investigations.

Photoprovocation procedure.

The UVA light source was a high-pressure metal-halide lamp (UV wavelength of 323–436 nm; Sellamed 3000 [Sellas, Medizinische Geraete]). The lamp for UVB irradiation was a UV-801 unit lamp with fluorescent bulbs (UV wavelength of 285–350 nm; Philips TL 20 W/12 [Waldmann]). The minimal erythema dose (MED), threshold dose for immediate pigment darkening (IPD), and minimal tanning dose (MTD) were determined using a standard procedure (17, 24). Areas (4 × 5 cm) of uninvolved skin on the upper back, extensor aspects of the arms, shoulders, thighs, or buttocks were irradiated with single doses of UVA (60–100 J/cm2), UVB (1.5 MED), and UVA and UVB daily for 3 consecutive days for photoprovocation (Figures 1A and B). In the present study, 429 CLE patients were irradiated with single doses ofUVA, 427 were irradiated with single doses of UVB, and 124 were irradiated with combined UVA/UVB irradiation, while 122 patients received all 3 types of UV irradiation (UVA, UVB, and UVA/UVB). In all patients, test reactions were evaluated at 24, 48, and 72 hours up to 3 weeks after irradiation. The criteria for positive photoprovocation required that 1) the induced lesions clinically resembled CLE, 2) the histopathologic findings were compatible with CLE, and 3) the skin lesions developed slowly and persisted for several days or weeks (4, 19). In most of the 431 patients with CLE, photoprovocation was conducted in the context of a routine diagnostic procedure; only data from 25 patients were obtained from photoprovocation conducted in the context of a clinical study (22). In 35 CLE patients, photoprovocation was performed a second time, and 4 of these patients received a third photoprovocation during the course of the disease.

Figure 1.

Positive photoprovocation after ultraviolet (UV) A and UVB irradiation in a patient with discoid lupus erythematosus (DLE). A, Overview of photoprovocation on the back of a patient with UVA (left), UVA/UVB (middle), and UVB light (right). B, Higher magnification of DLE-specific skin lesions after UVA/UVB irradiation. C, Histopathology after UVA/UVB irradiation: hyperplastic and orthohyperkeratotic epidermis with signs of interface dermatitis, superficial perivascular lymphocytic infiltration (original magnification × 100).

Histologic analysis of skin biopsy specimens.

Skin biopsy specimens were obtained from the primary lesions of patients with untreated CLE and/or from fully developed UV-induced lesions after photoprovocation (Figure 1C). Tissues were fixed, paraffin embedded, and stained with hematoxylin and eosin using standardized procedures for routine histologic evaluation.

Laboratory features.

Antinuclear antibodies (ANAs) were assayed using standardized indirect immunofluorescence with commercially available HEp-2 cells (Bio-Rad). Serum ANA titers ≥1:160 were considered abnormal. Anti-Ro/SSA and anti-La/SSB antibodies were determined using standardized enzyme-linked immunosorbent assays (DPC Biermann).

Statistical analysis.

Data collection and analysis were performed using PASW Statistics 18.0 (IBM SPSS). An SPSS database enabled a consistent, detailed statistical analysis of the obtained information. The association between qualitative parameters such as photoprovocation results and ANAs within the different subgroups was examined using the chi-square test and Fisher's exact test, respectively, to adjust for small case numbers. Quantitative data, including age, UVA/UVB irradiation doses, and reaction time, were analyzed using the Kruskal-Wallis and Mann-Whitney U tests for nonsymmetric values. Analyses of variance and t-tests were applied for Gaussian distributions. McNemar's test was used to compare the first photoprovocation to the second or third photoprovocation. P values less than 0.05 were considered statistically significant. Means are shown with the SD. Each analysis included the total number of CLE patients unless otherwise indicated. In cases of multiple photoprovocations, the results of the initial procedure in each patient were included in the statistical analysis. The data of multiple photoprovocations were only used for the evaluation of significant differences between the multiple photoprovocation results in the 35 patients who received multiple photoprovocations.

RESULTS

UV irradiation dose.

In 881 patients with photosensitive disorders, threshold testing for MED, MTD, and IPD as well as standardized photoprovocation were performed using UVA and/or UVB irradiation. Prior to irradiation, 367 patients (41.7%) presented with a confirmed diagnosis of CLE; however, 199 patients (22.6%) exhibited suspected CLE because the skin lesions had not been previously confirmed by clinical and/or histologic criteria or were not visible on the date of presentation to our clinic. Photoprovocation and subsequent clinical and/or histologic analyses confirmed the diagnosis of CLE in 64 (32.2%) of these 199 patients. Therefore, in addition to the control group of 315 PLE patients, a total of 431 CLE patients with a confirmed diagnosis of CLE were included in the final analysis (Figure 2).

Figure 2.

Study population. Photoprovocation confirmed the diagnosis in 64 (32.2%) of 199 patients with suspected cutaneous lupus erythematosus (CLE). PLE = polymorphic light eruption.

The mean ± SD MED, MTD, and IPD of the 431 CLE patients (98.9 ± 33.7 mJ/cm2, 50.5 ± 23.9 J/cm2, and 40.32 ± 23.3 J/cm2, respectively) were not significantly different from the mean ± SD MED, MTD, and IPD of the PLE patients (99.4 ± 33.5 mJ/cm2, 46.1 ± 23.0 J/cm2, and 33.5 ± 19.5 J/cm2, respectively). Moreover, no significant differences in the MED, MTD, and IPD were observed between CLE patients with a positive or negative photoprovocation. The cumulative doses of UVA or UVB in CLE patients with a positive photoprovocation compared with patients with a negative photoprovocation were not significantly different (mean ± SD UVA: 282.59 ± 55.56 J/cm2 and 286.17 ± 55.71 J/cm2, respectively, and mean ± SD UVB: 344.54 ± 136.62 mJ/cm2 and 375.59 ± 146.03 mJ/cm2, respectively). No significant differences in the MED, MTD, IPD, and the mean cumulative irradiation doses (UVA and/or UVB) in standardized photoprovocation were observed across different CLE subtypes (data not shown).

Photoprovocation results.

A positive photoprovocation using a standardized protocol was seen in 266 (61.7%) of 431 CLE patients (Figure 3A). No significant differences between men and women (Figure 3B) or patients between ages <40 years and ≥40 years were observed (Figure 3C). In the control group, 187 (59.4%) of 315 PLE patients demonstrated a positive photoprovocation, demonstrating no significant difference in the number of positive photoprovocation results in CLE and PLE patients.

Figure 3.

A, Comparison of photoprovocation results in cutaneous lupus erythematosus (CLE) and polymorphic light eruption (PLE) patients. B, Photoprovocation results for male and female CLE patients. C, Photoprovocation results in CLE patients ages <40 and ≥40 years. D, Photoprovocation results in different CLE subtypes. * = P < 0.05; ** = P < 0.001; ACLE = acute CLE; SLE = systemic LE; SCLE = subacute CLE; CCLE = chronic CLE; ICLE = intermittent CLE; CLE undiff. = undifferentiated CLE.

The subtype analysis revealed that ICLE patients showed the highest frequency of positive photoprovocations; 92 (74.8%) of 123 patients exhibited characteristic UV-induced skin lesions (Figure 3D). Furthermore, characteristic CLE skin lesions were induced in 52 SCLE patients (67.5%), 65 patients with undifferentiated CLE (62.5%), 12 ACLE patients (48%), and 45 CCLE patients (44.1%). ICLE patients tested positive significantly more often than ACLE (48%) and CCLE (44.1%) patients (P = 0.008 and P < 0.001, respectively). Moreover, SCLE patients demonstrated a positive photoprovocation significantly more often than CCLE patients (P = 0.002). Patients with undifferentiated CLE tested positive significantly more often than CCLE patients (P = 0.008).

Mean ± SD time for the development of a positive photoprovocation.

CLE patients exhibited a mean ± SD time between UV exposure and a positive photoprovocation of 6.82 ± 5.77 days, which was significantly longer than the mean ± SD time for PLE patients (2.48 ± 2.24 days; P < 0.001) (Figures 4A and C). The earliest photoprovocation result occurred in SCLE patients (mean ± SD 6.27 ± 4.36 days) compared with patients with other CLE subtypes (Figure 4B). However, the differences between the subtypes were not statistically significant. The latency for the development of skin lesions in all CLE subtypes was significantly longer compared with that of the PLE patients (ACLE versus PLE, P = 0.021; SCLE versus PLE, P < 0.001; CCLE versus PLE, P < 0.001; ICLE versus PLE, P < 0.001; and undifferentiated CLE versus PLE, P < 0.001).

Figure 4.

Time of onset of photoprovocation in cutaneous lupus erythematosus (CLE) and polymorphic light eruption (PLE) patients (A) and time of onset of photoprovocation in different CLE subtypes and PLE (B). White lines inside the boxes represent the median. Circles indicate outliers between 1.5 and 3 times the interquartile range. Triangles indicate outliers >3 times the interquartile range. C, Detailed representation of the time of onset of a photoprovocation reaction in CLE and PLE. Arrows indicate the mean and SD for each disease. ** = P < 0.001; * = P < 0.05; ACLE = acute CLE; SCLE = subacute CLE; CCLE = chronic CLE; ICLE = intermittent CLE; CLE undiff = undifferentiated CLE.

Reproducibility of photoprovocation in CLE patients.

In 35 CLE patients, >1 photoprovocations were performed during the course of the disease analyzed retrospectively in the study. The mean ± SD age of these patients was 44.1 ± 11.6 years at the first photoprovocation and 49.9 ± 11.6 years at the second photoprovocation. The mean ± SD time between the first and second photoprovocation was 5.8 ± 4.3 years; the median time was 5 years (range 1–17 years). The percentages of positive reactions of the first and second photoprovocation were similar (80.0% and 71.4%, respectively) (Figure 5A). Patients with a positive result in the first photoprovocation exhibited a positive result in the second test significantly more often (P = 0.013). However, 7 (25%) of the 28 patients who were positive in the first photoprovocation did not exhibit a positive reaction in the second photoprovocation. Four (57.1%) of the 7 CLE patients who were negative in the first photoprovocation exhibited a positive reaction in the second photoprovocation.

Figure 5.

A, Reproducibility of photoprovocation in the 35 cutaneous lupus erythematosus (CLE) patients with 2 distinct applications during the course of the disease. B, Photosensitivity by patient history versus photoprovocation results in CLE. C, Photoprovocation results of CLE patients in different irradiated areas. ** = P < 0.001; * = P < 0.05.

Four (11.4%) of the 35 patients who underwent 2 photoprovocations during the time of the analysis received a third photoprovocation. Two of these patients exhibited a positive reaction in all 3 tests, and 1 patient was positive in the first 2 photoprovocations but negative in the third. The fourth patient was negative in the first but positive in the second and third photoprovocations.

History of photosensitivity compared with photoprovocation results.

In the present study, 116 (56.9%) of 204 CLE patients were aware of an adverse effect of sunlight on their disease, but 88 patients (43.1%) denied any effect of sun exposure on their disease. Data on the history of photosensitivity were not available for the remaining 227 patients (52.7%). Patient history of photosensitivity did not vary greatly between the various CLE subtypes; ICLE patients reported reactions to sunlight most often (41 [60.3%] of 68), followed by ACLE (6 [54.5%] of 11), SCLE (20 [57.1%] of 35), CCLE (21 [52.5%] of 40), and undifferentiated CLE (28 [56.0%] of 50) patients. Eighty (69.0%) of the 116 CLE patients with a positive history of photosensitivity exhibited a positive photoprovocation; however, 44 (50.0%) of the 88 patients with a negative history of photosensitivity also demonstrated a positive photoprovocation (Figure 5B).

Areas of irradiation.

The area of the photoprovocation was specified in the charts of 168 CLE patients. The upper back was the most commonly used area for photoprovocation (56.5%), followed by the shoulders (16.1%), the extensor aspects of the arms (13.1%), and the buttocks (9.5%). Other areas, such as the chest and thighs, were used in 8 patients (4.8%). A positive photoprovocation on the upper back was observed in 79 (83.2%) of 95 CLE patients (Figure 5C), which was not significantly higher compared with positive results on the shoulders (20 [74.1%] of 27) and arms (15 [68.2%] of 22). However, photoprovocation on the upper back and shoulders was positive significantly more often than photoprovocation on the buttocks (5 [31.3%] of 16 CLE patients; P < 0.001 and P = 0.010, respectively).

Serologic features.

The results of the ANA assays were available for 120 CLE patients (27.8%), and 68 patients (56.7%) were positive (titers ≥1:160); 38 (55.9%) of these 68 patients exhibited a positive photoprovocation. Thirty-two (61.5%) of the 52 ANA-negative patients also exhibited a positive reaction. No significant differences in ANA, anti-Ro/SSA, anti-La/SSB, or anti-RNP antibodies were observed in CLE patients with positive or negative photoprovocation results in any of the different CLE subtypes (data not shown). Anti–double-stranded DNA (anti-dsDNA) antibodies were observed in 68 of the 120 ANA-tested CLE patients, and 10 (15.9%) of these 63 patients showed a positive photoprovocation. Patients with anti-dsDNA antibodies exhibited a negative photoprovocation result (70.0%) significantly more often than patients without anti-dsDNA antibodies (34.0%; P = 0.042).

DISCUSSION

Standardized photoprovocation in CLE has become an established and safe diagnostic tool for clinical and investigational studies in different European centers (21). To our knowledge, the present retrospective photoprovocation study includes the largest number of CLE patients to date. Moreover, an analysis of patients undergoing multiple photoprovocations for the evaluation of the reproducibility of results during the course of the disease has not been reported previously. We demonstrate that CLE patients with an initial positive photoprovocation were significantly more likely to exhibit a positive response in a subsequent application. However, 25% of patients who exhibited a positive first photoprovocation had a negative result in a second procedure, and more than half of the patients with a negative photoprovocation were subsequently positive during the course of the disease. No variables were found that could have influenced the difference in the distinct photoprovocations, such as medication or different areas of photoprovocation. In the present analysis, only 14 of 431 CLE patients were noted as receiving medication considered to be a factor affecting photosensitivity, such as antimalarials and/or steroids, likely due to the fact that photoprovocation is primarily performed in patients without systemic medications to prevent an impact on the outcome of the results (see Supplementary Table 1, available in the online version of this article at http://onlinelibrary.wiley.com/doi/10.1002/acr.21867/abstract). However, the conclusions from these data must be drawn carefully due to the low number of cases. Nevertheless, the results suggest that a single photoprovocation does not exclude photosensitivity throughout the entire course of the disease in a CLE patient and provide indirect evidence for a varying photosensitivity in some patients (4). Therefore, the limitations to sun exposure and the recommendations on sun protection should not be less stringent in CLE patients with a negative photoprovocation, since an exacerbation of this heterogeneous disease can neither be predicted nor ruled out entirely during the course of the disease. The explanation of changes in the response to standardized photoprovocation over time remains unknown, but disease activity in CLE at the time of photoprovocation may play a role.

The present study also demonstrates that photoprovocation using a standardized protocol can induce skin lesions in over 60% of CLE patients, confirming our previously published data on photoprovocation in various disease subtypes (19). Moreover, it has been suggested that patient history of photosensitivity does not necessarily correlate with photoprovocation results. The data on patient history of photosensitivity vary greatly in studies published in the literature (4, 25, 26), but a recent well-designed study on the prevalence of self-reported photosensitivity in a US population observed that 68% of the CLE patients reported prior or ongoing photosensitivity (27). Our present analysis revealed that 56.9% of CLE patients had a positive history of unusual reactions to sunlight. The varying populations of patients between the studies might explain the small difference (i.e., 28% of the CLE patients in the study from the US were African American). In addition, CLE patients may not be aware of their photosensitivity because lesion formation appears with a latency of up to 3 weeks after exposure to UV irradiation (4, 19). Moreover, photosensitivity may be overreported by patient history due to the inability of patients to distinguish CLE from other photodermatoses, such as PLE, which appears more frequently in CLE patients than in the normal population (28, 29). Therefore, the patients' variable knowledge of the disease characteristics may explain some of the reported differences in the frequency of photosensitivity. In addition, the predominant skin type and geographic region where the study was performed might contribute to the reported differences; however, these reasons remain speculative.

The present study confirms that a patient's history of photosensitivity is not predictive of a positive photoprovocation (19, 28). In total, 116 (26.9%) of 431 CLE patients fulfilled the ACR photosensitivity criterion prior to photoprovocation based on a positive history of photosensitivity. After photoprovocation, 302 (70%) of 431 CLE patients were photosensitive due to either a positive photoprovocation or a positive patient history. Interestingly, 36 patients with a positive history of photosensitivity exhibited a negative photoprovocation using a standardized protocol. Therefore, the development of an objective evaluation of photosensitivity in CLE to investigate photosensitivity as an important disease feature was required due to the apparent inaccuracy of photosensitivity using patient history. However, it cannot be excluded that these discrepancies may also be due to false-negative results of photoprovocation, which might occur due to the limited areas of irradiation in CLE patients. In our opinion, photoprovocation using the standardized protocol is very likely the most sensitive tool to diagnose photosensitive CLE and has been proven to be a safe and reproducible method showing comparable results across multiple centers (21).

In past decades, photoprovocation protocols have been optimized and standardized after the initial photoprovocation by Epstein et al in 1965 (18, 19, 30, 31). The previous results of photoprovocation studies in CLE patients were variable and ranged from 24–83% positivity (17, 31, 32). These discrepancies may largely be explained by the variability of photoprovocation protocols and the differences in the types of CLE patients. For example, Sanders and coworkers (33) irradiated primarily CLE patients with suspected photosensitivity, and 93% of the patients exhibited a positive reaction to UV light. This percentage is higher than the 61.7% of patients with a positive photoprovocation in the present study. Therefore, a standardized photoprovocation protocol has been used in a multicenter trial to evaluate and compare the results of photosensitivity in different European centers (21). No significant differences in photoprovocation between study sites were observed, and these comparable results support the safety and reproducibility of this method in diagnostic testing and clinical trials.

The latency of the onset of skin lesions after UV irradiation may be a helpful indicator to distinguish between CLE and PLE using standardized photoprovocation. The onset of CLE skin lesions is characterized by a latency period of up to 3 weeks, but PLE skin lesions usually appear within 4 days of UV irradiation (19). The present study confirms a significant difference in the mean latency of the reaction onset. However, some CLE patients developed specific lesions within the first few days after irradiation, and some PLE patients did not exhibit skin lesions until 1 week after the last day of photoprovocation. Therefore, it is highly important to follow the guidelines for the evaluation of photoprovocation, e.g., to confirm the clinical picture using histologic findings.

A correlation of the photoprovocation result with specific autoantibodies, particularly anti-Ro/SSA and anti-La/SSB antibodies, has been reported previously (34). A previous study by our group also demonstrated that patients with a positive photoprovocation revealed anti-Ro/SSA and anti-La/SSB antibodies more often than patients with a negative result, but this difference was not statistically significant (19). Tang and coworkers demonstrated a significant correlation between photosensitivity and the levels of anti-RNP antibodies in SLE patients (35). Also, a correlation between positive photoprovocation results and positive antibody titers in CLE patients has been suggested (19, 33); however, no statistically significant correlations between seropositivity for autoantibodies and photoprovocation results were observed in the present study. These results are consistent with a recent multicenter evaluation of a standardized photoprovocation protocol in 47 CLE patients (21). Surprisingly, an inverse correlation between the presence of anti-dsDNA antibodies and photoprovocation results was observed in these patients with different CLE subtypes. The data are likely attributed to the small sample size of the subanalysis because only 10 patients exhibited positive anti-dsDNA antibodies.

Photoprotective measures, such as high-potency sunscreens, are a primary strategy for the prevention of skin manifestations in CLE. The protective effect of high-potency sunscreens has been evaluated in clinical studies using a photoprovocation regimen to achieve standardized results (22, 36–38). Photoprovocation was utilized in a double-blind, intraindividual comparative trial to evaluate 3 different sunscreens in 11 LE patients (37, 38), and these sunscreens exhibited different efficacies. These results were confirmed in a retrospective analysis in which one of the sunscreens was applied to 47 CLE patients (36). A recent randomized, double-blind, vehicle-controlled, intraindividual, comparative study demonstrated that the application of a broad-spectrum sunscreen with high protection factors prevented the appearance of lesions in photosensitive CLE patients (22). Therefore, photoprovocation using the standardized protocol is an optimal model for the evaluation of preventive measures in CLE patients.

In summary, the present analysis of a large study population supports the importance of standardized photoprovocation in CLE patients. The main limitation of this study is its monocentric and retrospective design, which may be subject to a patient selection bias and a possible information bias when obtaining subjective data from patients, such as the patients' history of photosensitivity. Furthermore, the PLE patients included as controls had a high percentage of women and were younger compared to the CLE patients. Also, patients with ICLE seem to have been overrepresented in this study; however, this was due to the fact that the diagnosis of patients with ICLE is often unclear, and photoprovocation as a diagnostic tool is carried out more often compared to other subtypes. Nevertheless, the results of this study suggest that photosensitivity should never be excluded in CLE patients with a negative history of photosensitivity because the reaction to UV light may change during the course of this heterogeneous disease, and that history should always be evaluated in combination with phototesting. Therefore, we recommend that a standardized photoprovocation should be performed to evaluate or confirm photosensitivity in CLE patients and to increase the awareness of the deleterious effect of sunlight on the disease.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Kuhn had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Ruland, Haust, Ruzicka, Kuhn.

Acquisition of data. Haust, Stilling, Metze, Kuhn.

Analysis and interpretation of data. Ruland, Stilling, Amler, Kuhn.

Acknowledgements

We thank Dr. N. Neumann, Department of Dermatology, University of Duesseldorf, Duesseldorf, Germany, for providing access to patient material and Dr. Percy Lehmann, Department of Dermatology, Allergology and Environmental Medicine, HELIOS Klinikum Wuppertal, Germany, for his contribution to photoprovocation.

Ancillary