SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Objective

To retrospectively study the incidence of chronic cutaneous lupus erythematosus (CCLE) in French Guiana (FG), South America, during the period 1995–1999.

Methods

Private and public physicians specializing in dermatology, rheumatology, and internal medicine were asked during the year 2000 about lupus cases. We reviewed hospitals' files in data-processing departments.

Results

Twenty new cases of CCLE, mostly discoid form, were identified during this 5-year period in this population of predominantly African descent. The average annual incidence of the disease was 2.59 per 100,000 inhabitants (95% confidence interval 1.5–4). However, our methodology could introduce underestimation of the incidence of the disease.

Conclusion

The average annual incidence of CCLE in FG appears to be low in this retrospective study, but is very similar to the only previously published data in the US.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Lupus erythematosus is a rare autoimmune disease. In the Guilliam classification, its specific dermatologic manifestations are classified into 3 different specific categories: acute cutaneous lupus erythematosus, subcutaneous lupus erythematosus (SCLE), and chronic cutaneous lupus erythematosus (CCLE). CCLE is frequent in systemic lupus erythematosus (SLE), but may also develop as an isolated disease. Discoid lupus erythematosus (DLE) itself is divided into classic, mucosal, hypertrophic/verrucous, and lichenoid DLE, and lupus tumidus, chilblain lupus, and lupus panniculitis are considered to be subtypes of CCLE (1). There are little epidemiologic data available about CCLE; only one population-based study is available, in Rochester, Minnesota, which found a mean annual incidence of 4.3 per 100,000 for all forms of nonsystemic lupus in the population of Olmsted County between 1965 and 2005 (2). French Guiana (FG) is an equatorial territory in South America with 157,000 inhabitants, where health care is free and easily accessible. We conducted an epidemiologic investigation to determine the incidence of CCLE in the population of FG, which is mostly of African descent.

Patients and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

FG is a French Overseas Department of 91,000 km2, part of the European Community, and located in the northern part of South America, in the Amazonian basin between the state of Amapa in Brazil and Suriname. The climate is tropical, hot and humid, with no wet season. The population of 157,000 inhabitants (49.6% women, 50.4% men) in 1999 is predominantly of African descent coming from the slave trade (Creole and Maroons are estimated to be 90% of the inhabitants of FG), with a minority of other ethnic groups: French whites, Native Amerindians, and Asians (3). This is a very young population (75% age <40 years) (4). The annual population growth rate was 3.6% between 1990 and 1999 mainly due to the positive imbalance between birth rates and death rates (75%), with an additional part from migration (25%) (5).

We used multiple sources to try to detect all cases of lupus: 1) all dermatologists, in both public and private practice (n = 7), 2) all rheumatology or internal medicine practitioners, in both public and private practice (n = 7), and 3) the 2 main hospitals' data-processing departments. Public practitioners could have 2 different types of files. The first type, shared by private physicians, concerns only outpatients. It could have different forms (electronic and paper, which concern all outpatients seen by physicians and could be specific prospective files for several diseases). These files were screened one by one if they were not electronic or prospective. The other type is electronic files shared by all inpatients, and these were screened by the 2 main hospitals' data-processing departments, which performed electronic research using International Classification of Diseases, Ninth Revision, Clinical Modification (codes 710.0 and 695.4), and International Classification of Diseases, Tenth Revision, Clinical Modification (codes L93, M32) classification. This codification is necessarily attached to all hospitalizations in the French system of health care. Patients were not contacted for the study.

We sought the following information in the medical files: type of lesion and localization (limited if involving the face and the neck, and disseminated if both above and below the neck); date of first lesion or, when unavailable, date of first consultation; results of any cutaneous biopsy (standard and immunofluorescence tests); and American College of Rheumatology (ACR) 1982 criteria for SLE (6). Only cases of CCLE beginning between 1995 and 1999 were included. Disease onset was recorded as the year of the first lesion or, when unavailable, of the first consultation. Diagnosis was considered definite if consistent skin biopsy was available and probable if there was no biopsy. Patients with SLE were defined as those who at diagnosis fulfilled 4 or more 1982 ACR criteria for SLE at the onset of their disease, and patients for whom skin biopsy was not consistent with the diagnosis were excluded: CCLE cases included were scored as “pure CCLE” and all other cases were recorded as SLE. The reference population was the 1999 Guyanese national census population (4). In the French constitution, gathering data on ethnicity and race is not allowed.

Incident cases between January 1, 1995 and December 31, 1999 were used to assess the crude average annual incidence (AAI), calculated per 100,000 inhabitants. We estimated 95% confidence intervals (95% CIs) with the normal approximation of the Poisson distribution. Age values are expressed as the mean ± SD. Statistical analysis was performed using Stata, version 8.0 (StataCorp, College Station, TX) for Macintosh. The Medical Committee of Cayenne Hospital approved the study. However, according to French law, retrospective studies do not require ethics committee acceptance.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

All of the physicians agreed to participate in the study. We identified 120 patients considered as having lupus; 69 of these were diagnosed as having SLE and 51 were diagnosed as having pure cutaneous lupus (Figure 1). Of the 51 patients identified as having pure cutaneous lupus, 2 could be initially considered as having SLE. There were 3 patients with SCLE, 1 with neonatal lupus erythematosus, and 2 with bullous lupus. One declared case was not lupus, and there was insufficient information concerning 3 cases to confirm diagnosis. Thirty-nine patients (32.5%) had initially pure CCLE. Four (10.25%) of these 39 patients initially included as pure CCLE subsequently developed SLE, according to the 1982 ACR SLE criteria. Of the 39 initially pure CCLE patients, disease onset was between 1995 and 1999 for only 20 patients (n = 18 women, n = 2 men). Their characteristics are described in Table 1. All of the included patients come from dermatology outpatient clinics of public practice (n = 13 exclusively from Cayenne Hospital, files screened one by one) and private practice (n = 7 files of 2 physicians also screened one by one; because the last dermatologist of private practice declared just 1 SLE patient and was certain to have seen no pure cutaneous lupus, we did not consult his files). No patient was discovered by more than one source. Fifteen were definite cases (n = 14 women, n = 1 man) and 5 were probable pure CCLE cases (12.5% and 4.1%, respectively, of the 120 patients declared as having lupus). Eighteen patients had DLE, 1 had lupus tumidus, and 1 had lupus profundus (90%, 5%, and 5%, respectively, of definite and probable cases combined). All of the cases of discoid lupus were the classic form; we found no hypertrophic/verrucous, lichenoid, or mucosal forms. The woman to man ratio was 9:1. The mean ± SD age at diagnosis was 32 ± 14.6 years (range 8–74 years). The AAI for CCLE was 2.59 per 100,000 inhabitants (95% CI 1.5–4) from January 1, 1995 to December 31, 1999, consisting of 1.95 per 100,000 inhabitants for definite cases (95% CI 1.1–3.2) and 0.64 per 100,000 inhabitants for probable cases (95% CI 0.2–1.5). For women, the AAI was 4.72 per 100,000 inhabitants (95% CI 2.8–7.5), including 3.4 per 100,000 inhabitants (95% CI 1.8–5.8) for definite cases and 1.3 per 100,000 inhabitants (95% CI 0.4–3.1) for probable cases. For men, the AAI for definite cases was 0.5 per 100,000 inhabitants (95% CI 0.06–1.9).

thumbnail image

Figure 1. Schematic representation of patient selection. SLE = systemic lupus erythematosus; SCLE = subcutaneous lupus erythematosus; NNLE = neonatal lupus erythematosus; CCLE = chronic cutaneous lupus erythematosus.

Download figure to PowerPoint

Table 1. Characteristics of the 20 French Guyanese patients with chronic cutaneous lupus erythematosus onset between 1995 and 1999*
PatientAge at diagnosis, yearsSexYear of onsetType of lesionLocalization of lesionsANAsSkin biopsyLupus band test1982 ACR SLE criteria (6) at diagnosisStatus of case
  • *

    ANAs = antinuclear antibodies; ACR = American College of Rheumatology; SLE = systemic lupus erythematosus; ND = not determined.

138F1995Discoid (classic)Disseminated++ND2, 11Definite
240M1995Discoid (classic)LocalizedNDNoND2Probable
317F1996Discoid (classic)LocalizedND+ND2, 3, 4Definite
423F1996PanniculitisDisseminated++ND5, 11Definite
539F1996Discoid (classic)Disseminated++2, 9Definite
620M1996Discoid (classic)Localized+ND2, 9Definite
733F1996Discoid (classic)LocalizedND+ND2Definite
843F1996Discoid (classic)LocalizedND+ND2Definite
98F1997Discoid (classic)Localized++ND2, 9, 11Definite
1022F1997TumidusDisseminated+NoND3, 9, 11Probable
1119F1997Discoid (classic)Localized++ND2, 9, 11Definite
1222F1997Discoid (classic)Localized+2Definite
1341F1997Discoid (classic)LocalizedND+ND2Definite
1474F1998Discoid (classic)Localized++ND2, 11Definite
1542F1999Discoid (classic)Localized+ND2Definite
1611F1999Discoid (classic)Localized+ND2Definite
1733F1999Discoid (classic)Localized+ND2Definite
1845F1999Discoid (classic)LocalizedNDNoND2Probable
1940F1999Discoid (classic)Localized+NoND2, 11Probable
2031F1999Discoid (classic)LocalizedNoND2, 9Probable

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Here we report epidemiologic data concerning CCLE in FG. Published series are mostly hospital based; furthermore, many studies predate the description of SCLE. Only one study has described the relative proportions of the different components of CCLE in the population (2). The widely accepted and classic Gilliam classification for lupus erythematosus has been used in our study, because the Dusseldorf classification was not validated in 2000; the latter excludes lupus erythematosus tumidus from the CCLE entity (7, 8). In our survey, classic DLE was the sole variant identified. Other forms of CCLE are rare in our African descent population: lupus profundus and lupus tumidus were much less frequent than DLE, each comprising only 5% of the probable and definite incident cases between 1995 and 1999. This is consistent with the scarcity of the literature concerning these two lupus lesions compared with that addressing DLE. We did not find any cases of chilblain lupus, probably because of the hot tropical climate. The mean age at onset of pure CCLE in our study was in accordance with the literature, which describes that lesions usually start between the ages of 20 and 40 years (1, 9, 10). The 9:1 sex ratio observed is intriguing: Durosaro et al did not find such differences in Rochester; they report data of approximately 3:2 and 3:1, which are close to previous studies (1, 2, 10), although Jacyk and Damisah found a woman to man ratio of 5:1 among Nigerian patients (11). However, considering all of the patients with pure CCLE in FG (definite, probable, and whatever the date; n = 39), the sex ratio was 3.5:1. Therefore, the extreme sex ratio among our patients seems to result from a short-term scarcity of incident male patients during the period 1995–1999. Five percent to 12% of cases described in published DLE cohorts subsequently became SLE, which is consistent with our observation in a population-based study (1, 2). Frequent loss to followup in our CCLE population makes it difficult to interpret this observation, despite the probable better detection of SLE according to systemic symptoms. The prevalence of antinuclear antibodies (ANAs) among tested cases was in accordance with the data reported by Callen and by Prystowsky and Gilliam (10, 12). However, the absence of standardized technical procedures used by laboratories for ANA testing and the limited number of patients weaken any conclusions.

With some reservations due to methodologic issues, our study shows that the AAI of CCLE seems low in FG; however, we can compare our study with only one epidemiologic study in the predominantly white population of Rochester, Minnesota (2). In this study, we have to compare with CCLE and lupus panniculitis, whose mean annual incidences between 1965 and 2005 were 3.56 (95% CI 2.94–4.18) and 0.07 (95% CI 0.00–0.16), respectively. Our results in FG seem slightly lower but not so different. This was a physician-based study; some cases could be excluded on the pathology results despite the clinical suspicion of the dermatologist, as in our study. Moreover, cases diagnosed as CCLE before 1979 could be SCLE. This could result in an overestimation of CCLE. However, our low rate was not a surprise because it has already been reported in African American populations compared with whites (1). However, Hochberg et al reported data suggesting a higher prevalence of discoid lesions in African American patients with SLE compared with whites (13). Moreover, Callen observed more African American DLE patients in his cohort despite an African American to white ratio as low as 1:6 (10). Currently, the literature does not contain sufficient or appropriate data to allow comparison of epidemiologic rates between races.

Our study has several limitations in the ascertainment of cases. The lupus band test was not used because this pathology procedure is difficult to obtain locally. Moreover, skin biopsy, particularly involving the face, is frequently refused by patients. The apparently benign character of most CCLE lesions, particularly for people living deep in the forest, probably contributes to an underestimation, especially for men. The retrospective methodology did not allow the recovery of all of the information on every patient or disease such as a detailed description of the clinical skin manifestations leading to diagnosis, the presence of ANAs, or precise histopathology results sometimes only described as compatible. In our study, this limited the possibility to ascertain the diagnosis by criteria, leading to suggest possible and proved categories, excluding patients with negative histopathology despite clinical suspicion.

There are also several limitations in the capture of all cases. The method of case capture could introduce some memory bias and then, an underestimation of the disease. We did not contact general practitioners; indeed, lupus, systemic as pure cutaneous form, is a rare disease exclusively diagnosed by specialists and rarely followed only by primary care physicians. Therefore, in local health organizations like in metropolitan France, it is unlikely that lupus could avoid at least one specialized assessment, whether in public or private practice; thus, we assumed that looking to general practitioner files was of very limited additional value. Additionally, the population is captive, with only 7 dermatologists available with 1) no road to travel to the neighboring countries (Amapa state, Brazil, and Suriname), 2) the barrier of Amazonian forest covering 90% of the area of French Guiana, and 3) the cost limitation of the travel to the French West Indies or metropolitan France. However, the national health insurance (Sécurité sociale) that centralizes all health care could support medical transfer if requested by the local reference specialist (Cayenne Hospital).

It is noteworthy that in the cooperation between physicians of different specialties, mixing private and public practitioners was an important factor contributing to complete the best retrieval of all cases of lupus as possible in the local conditions, including the more benign form usually underrepresented in the other studies dedicated to lupus. Finally, the small number of patients (n = 20) likely introduces another limitation factor in the clinical description and epidemiologic data.

In conclusion, this is the second study on CCLE epidemiology. The AAI seems to be low in our population of African descent, but our data seem concordant with those reported in the white population of Olmsted County, Minnesota.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

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 submitted for publication. Dr. Deligny 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. Deligny, Clyti, Sainte-Marie, Couppie, Arfi, Pradinaud.

Acquisition of data. Deligny, Clyti, Sainte-Marie, Couppie, Pradinaud.

Analysis and interpretation of data. Deligny, Clyti, Sainte-Marie, Couppie, Lê Thi Huong, Piette, Arfi, Pradinaud.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

The authors thank all of the physicians of FG for their help.

REFERENCES

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and Methods
  5. Results
  6. Discussion
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES
  • 1
    Sonntheimer RD, McCauliffe DP. Lupus-specific skin disease (Cutaneous lupus). In: WallaceDJ, HahnBH, editors. Dubois' lupus erythematosus. 7th ed. Philadelphia: Lippincott Williams & Wilkins; 2006. p. 576620.
  • 2
    Durosaro O, Davis MD, Reed KB, Rohlinger AL. Incidence of cutaneous lupus erythematosus, 1965–2005. Arch Dermatol 2009; 145: 24953.
  • 3
    Levinson D. The Americas: French dependencies. In: LevinsonD, editor. Ethnic groups worldwide: a ready reference group handbook. Phoenix (AZ): Oryx Press; 1998. p. 350.
  • 4
    Institut National de la Statistique et des Etudes Economiques (INSEE). Recensement de la population de mars 1999. URL: http://www.insee.fr.
  • 5
    Institut National de la Statistique et des Etudes Economiques (INSEE). Les projections de population aux Antilles-Guyane à l'horizon 2030. URL: http://www.insee.fr.
  • 6
    Tan EM, Cohen AS, Fries JF, Masi AT, McShane DJ, Rothfield NF, et al. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum 1982; 25: 12717.
  • 7
    Kuhn A, Richter-Hintz D, Oslislo C, Ruzicka T, Megahed M, Lehmann P. Lupus erythematosus tumidus: a neglected subset of cutaneous lupus erythematosus. Report of 40 cases. Arch Dermatol 2000; 136: 103341.
  • 8
    Kuhn A, Sonntag M, Ruzicka T, Lehmann P, Megahed M. Histopathologic findings in lupus erythematosus tumidus: review of 80 patients. J Am Acad Dermatol 2003; 48: 9018.
  • 9
    O'Loughlin S, Schroeter AL, Jordon RE. A study of lupus erythematosus with particular reference to generalized discoid lupus. Br J Dermatol 1978; 99: 111.
  • 10
    Callen JP. Chronic cutaneous lupus erythematosus: clinical, laboratory, therapeutic, and prognostic examination of 62 patients. Arch Dermatol 1982; 118: 4126.
  • 11
    Jacyk WK, Damisah M. Discoid lupus erythematosus in the Nigerians. Arch Dermatol 1979; 100: 1315.
  • 12
    Prystowsky SD, Gilliam JN. Discoid lupus erythematosus as part of a larger disease spectrum: correlation of clinical features with laboratory findings in lupus erythematosus. Arch Dermatol 1975; 111: 144852.
  • 13
    Hochberg MC, Boyd RE, Ahearn JM, Arnett FC, Bias WB, Provost TT, et al. Systemic lupus erythematosus: a review of clinico-laboratory features and immunogenetic markers in 150 patients with emphasis on demographic subsets. Medicine (Baltimore) 1985; 64: 28595.