Portal de Periódicos da CAPES

Sepsis Associated with Dermatologic Hospitalization Is not the Cause of High Mortality of Bullous Pemphigoid in Europe

2005; Elsevier BV; Volume: 124; Issue: 3 Linguagem: Inglês

10.1111/j.0022-202x.2005.23628.x

1523-1747

I. García‐Doval, Alberto Conde‐Taboada, Manuel J. Cruces Prado,

Platelet Disorders and Treatments

confidence interval standard deviation To the Editor: Colbert et al., 2004Colbert R.L. Allen D.M. Eastwood D. Fairley J.A. Mortality rate of bullous pemphigoid in a US Medical Center.J Invest Dermatol. 2004; 122: 1091-1095Abstract Full Text Full Text PDF PubMed Scopus (86) Google Scholar have hypothesized that sepsis induced by hospitalization for bullous pemphigoid can be the cause of the difference in mortality rates between US and Europe. After their report, we performed a study to describe the prognosis of our patients and to check whether the time course of events supports their hypothesis. We collected all bullous pemphigoid diagnoses from the dermatopathology files of our secondary care department (1998–2003, n=31). Inclusion criteria were: newly diagnosed patients, coming from the area of 300,000 inhabitants assigned to our hospital, with histopathology findings in agreement with the diagnosis of pemphigoid, and meeting the French Bullous Study Group criteria for diagnosis of bullous pemphigoid (Vaillant et al., 1998Vaillant L. Bernard P. Joly P. et al.Evaluation of clinical criteria for diagnosis of bullous pemphigoid. French Bullous Study Group.Arch Dermatol. 1998; 134: 1075-1080Crossref PubMed Scopus (154) Google Scholar; Joly et al., 2004Joly P. Courville P. Lok C. et al.Clinical criteria for diagnosis of bullous pemphigoid: A reevaluation according to immunoblot analysis of patient sera.Dermatology. 2004; 208: 16-20Crossref PubMed Scopus (63) Google Scholar) that we incorporated in clinical practice in 1998. We performed direct immunofluorescence in 10 patients with positive results in all of them. Three patients who did not fulfill diagnostic criteria, and two who were not new diagnoses, were excluded from the study. We collected data on all hospital admissions for bullous pemphigoid (all to the dermatology department) and patient evolution. Admissions because of other disorders were not considered to be relevant to the tested hypothesis and were not registered. All patients included in the study (n=26) or their families were contacted either in the clinical setting or by telephone call. Statistical analyses were performed using SPSS software (version 10). To calculate standardized mortality ratio (SMR), we used population data obtained from “Instituto Nacional de Estadística” (Galicia, 1998–1999). The mean age of patients at diagnosis was 77 y (range: 57–96 y). They were treated with oral corticosteroids (n=14), or topical corticosteroids (n=12). For patients treated with oral prednisone, the mean daily dose was 34 mg at start of therapy (standard deviation (SD): 9.8) (range: 20–50). In those alive at the end of therapy, the mean length of therapy was 20 mo (SD: 12). All patients in the topical corticosteroid group started with a dose of 30 g per d 0.05% clobetasol propionate (Clovate, Celltech Pharma Spain) in a single application. Most of the 26 patients suffered from other disorders: cerebrovascular disease (9), dementia (8), diabetes (5), neoplasia (4), cardiac disorders (4), decubitus ulcers (4), and limb ischemia (2). Only six patients had no other disorder but pemphigoid. At the time of data collection 15 patients were alive and 11 were dead. The mean and median follow-up were 22 and 14 mo. Death rates were 16% at 3 mo (95% confidence interval (CI), 7%–39%), 21% at 6 mo (95% CI, 10%–44%), and 40% at 1 y (95% CI, 23%–63%). Age- and gender-adjusted mortality rate at 1 y was two times higher in pemphigoid patients than in the general population of our area (SMR: 2.15; 95% CI, 1.07–3.84; p=0.016). Hazard rate plotting showed that the risk of death was higher in the first year after diagnosis. Almost all patients were hospitalized shortly after the diagnosis of bullous pemphigoid (only one patient was admitted 3 mo later). The time course of events is plotted in Figure 1. The mean time from patient discharge from the dermatology department to death was 13 mo (SD: 11) (range: 0.5–36). We found no differences in the number of days of hospitalizations between patients dying or surviving (t test, 95% CI: -10 to 5 d; p=0.46). Two patients died 2 wk after discharge from hospital: one had metastatic adenocarcinoma and the other died of pneumonia. The rest of the patients died at least 6 wk after discharge. The causes of death were neoplasia (2), sepsis (2), pneumonia (2), cerebrovascular accident (1), myocardial infarction (1), and unknown (1). The prognosis of bullous pemphigoid is an area of debate (Colbert et al., 2004Colbert R.L. Allen D.M. Eastwood D. Fairley J.A. Mortality rate of bullous pemphigoid in a US Medical Center.J Invest Dermatol. 2004; 122: 1091-1095Abstract Full Text Full Text PDF PubMed Scopus (86) Google Scholar; Swerlick and Korman, 2004Swerlick R.A. Korman N.J. Bullous pemphigoid: What is the prognosis?.J Invest Dermatol. 2004; 122: XVII-XVIIIAbstract Full Text Full Text PDF PubMed Scopus (14) Google Scholar). Our findings support, in a different setting, previous descriptions of pemphigoid as a severe disease, with around 40% of patients dying within 12 mo. Other relevant findings were that hospital admission to dermatology is routine in our setting after a diagnosis of pemphigoid, readmissions are exceptional, and this high mortality does not take place within periods of dermatological hospitalization or close to them. Our data fail to support the hypothesis of hospitalization as a cause of increased incidence of sepsis and mortality, as the long interval between hospitalization and death cannot be properly explained. We do not think that selection bias is the cause of the high mortality observed in our series. To avoid this, we recruited every patient with a biopsy-proved diagnosis, not just patients with generalized disease. We excluded the rare patients not coming from the area assigned to our hospital (Dermatological National Health Service care is divided in areas in Spain). An indicator of the absence of a marked selection bias was that co-morbidities of our patients were similar to those in other series. Moreover, we have probably received the majority of patients from our area. If we accept an incidence of seven cases per million per year (Wojnarowska et al., 2002Wojnarowska F. Kirtschig G. Highet A.S. Venning V.A. Khumalo N.P. Guidelines for the management of bullous pemphigoid.Br J Dermatol. 2002; 147: 214-221Crossref PubMed Scopus (127) Google Scholar), approximately 13 cases would be expected during the study period. The fact that many of our patients were treated with topical corticosteroids could also be a cause of bias. This therapy, however, has been shown to actually improve prognosis (Joly et al., 2002Joly P. Roujeau J.C. Benichou J. et al.A comparison of oral and topical corticosteroids in patients with bullous pemphigoid.N Engl J Med. 2002; 346: 321-327Crossref PubMed Scopus (475) Google Scholar); therefore, it should not account for the high rate of mortality observed. The main limitation of our study is the lack of immunofluorescence confirmation of the diagnosis in many patients. But we have used biopsy information, and clinical criteria that have been well validated. Our data confirm the grim prognosis of bullous pemphigoid patients in European series, but refute Colbert's et al hypothesis of this mortality being caused by sepsis associated with the different hospitalization pattern. We thank Juan Méndez Zunzunegui for his constructive comments and review of the manuscript.