Zygomycosis, France | CDC EID

Volume 15, Number 9–September 2009
Research
Increasing Incidence of Zygomycosis (Mucormycosis), France, 1997–2006
Dounia Bitar, Dieter Van Cauteren, Fanny Lanternier, Eric Dannaoui, Didier Che, Francoise Dromer, Jean-Claude Desenclos, and Olivier Lortholary
Author affiliations: Institut de Veille Sanitaire, Saint Maurice, France (D. Bitar, D. Van Cauteren, D. Che, J.-C. Desenclos); Université Paris-Descartes, Paris, France (F. Lanternier, O. Lortholary); Institut Pasteur, Paris (E. Dannaoui, F. Dromer, O. Lortholary); and Hôpital Georges Pompidou, Paris (E. Dannaoui)

Suggested citation for this article

Abstract
We analyzed hospital records to provide a population-based estimate of zygomycosis incidence and trends over a 10-year period at a national level in France. Data showed an increasing incidence from 0.7/million in 1997 to 1.2/million in 2006 (p<0.001). We compared our data with those from the French Mycosis Study Group, a recently established voluntary network of French mycologists coordinated by the National Reference Center for Mycoses and Antifungals. We documented that incidence of zygomycosis increased, particularly in patients with hematologic malignancies or bone marrow transplants. The role of previous exposure to antifungal drugs lacking activity against zygomycetes could explain this increase but does not appear exclusive. Incidence also increased in the population of patients with diabetes mellitus. We conclude that observed trends reflect a genuine increase of zygomycosis cases in at-risk populations.

Zygomycoses are severe angioinvasive infections caused by common filamentous fungi, the zygomycetes. These ubiquitous opportunistic fungi can cause infections with high lethality in immunocompromised or diabetic patients. Whatever the route of infection (inhalation of airborne spores, ingestion, or direct skin inoculation), the hyphae invade blood vessels, causing tissue infarction and necrosis (1–4). In healthy persons, innate immunity is sufficient to prevent infection, except in cases of massive contamination after traumatic inoculation of contaminated soil (5). Patients with phagocytic dysfunctions caused by neutropenia or ketoacidosis, as well as patients with high iron serum concentrations, are at high risk of developing zygomycosis (4). These underlying conditions can influence clinical presentation and outcome (2,6,7). The rhinocerebral presentation is the most frequently reported localized symptom followed by pulmonary, cutaneous, cerebral, gastrointestinal, and disseminated infections (3,8). In the rhinocerebral or pulmonary forms, patient death rates are reported to be as high as 60% because of delayed diagnosis or delayed therapeutic management (7–9). Treatment strategies are based on high doses of any lipid formulation of amphotericin B, associated with large surgical resections when possible. Most triazole agents are not effective in vivo (10), except for posaconazole, which shows some efficacy both experimentally and in patients as second-line therapy (11–13). The clinical contribution of new iron chelating agents remains controversial (14,15).

Some reports have suggested an increasing incidence of zygomycosis on the basis of analysis of data from monocentric studies (7,10,14,16). Several explanations have been posited: longer survival of persons with severe hematologic malignancies or solid organ transplantations; increased knowledge of the infection and thus a better diagnosis yield; and finally, prolonged use of new antifungal drugs ineffective against zygomycetes as prophylactic or empiric treatment in bone marrow transplant recipients or other immunocompromised patients (10,17–20). However, the incidence of infection among the general population and infection trends over time has rarely been documented at a national level. To describe the demographic characteristics of patients and to estimate the incidence and case-fatality ratio (CFR) associated with the major underlying diseases, we analyzed the electronic hospitalization and death records of patients in France in whom zygomycosis had been diagnosed.

Material and Methods
Data Sources
To describe cases and to estimate incidence, we extracted all hospital records related to zygomycosis from 1997 through 2006 in metropolitan France, which includes mainland France and the island of Corsica. The French hospital information system, Programme de Médicalisation du Système d'Information (PMSI), is a managerial tool based on the systematic collection of standardized administrative and medical information for any new hospital admission. An estimated 95% of all public and private hospitals use this tool, including all third-level structures, i.e., university hospitals and other reference hospitals (21). An anonymous extraction of the dataset with limited sociodemographic information (age, sex, and residence area) can be made available for specific epidemiologic studies. Medical information includes the main cause of admission (principal diagnosis), the related medical conditions (with >20 entries for various associated diagnoses), duration of stay, modes of admission and discharge, and major medical activities performed during the stay. Diseases are coded according to the International Classification of Diseases, 10th revision (ICD-10). This coding system was stable during the study period. Conversely, the codes for medical or surgical procedures, determined at the national level, have evolved over time, with notable changes occurring in 2003.

To estimate the overall CFR, deaths that occurred during hospitalization, including those among readmitted patients, were identified through the PMSI. To take into account deaths that occurred after discharge, we also used data from death certificates available at the institution in charge of recording and analyzing death certificates, the CepiDc. The certificate is completed at the time of death by the attending physician (at hospital or home) who records the main cause of death (i.e., the direct, immediate cause), as well as underlying diseases that may have contributed to death. The certificate is then sent to the CepiDc, where it is coded according to ICD-10.

Case Identification
We used all PMSI records in which the ICD-10 codes B46 or B460 to B469 were identified; these corresponded to any clinical presentation of zygomycosis, or mucormycosis. To distinguish first admissions from rehospitalizations, we created a unique patient identifier by chaining the variables year of birth (derived from the patient's age and the date of admission), gender, and residence postal code. The same identifier was used to match PMSI and CepiDc data, to identify additional deaths and to check for duplicates. However, the CepiDc identifier was based on the exact year of birth, but in the PMSI we derived the year of birth from age at admission.

Case Definitions
We defined as unique cases those that had an identifier detected only 1 time in the PMSI during the 10-year period. Cases detected >1 time were classified as new cases at first occurrence in the database; subsequent ones were coded as readmissions. After checking for duplicates, we analyzed unique or new cases for incidence.

The zygomycosis cases were classified according to their clinical presentation, when available. When several body localizations were reported, we selected the most severe one, either in a single stay or in subsequent admissions. For instance, the association of unspecified localization (B46, B468, or B469) or gastrointestinal (B462) or cutaneous (B463) localizations, together with a pulmonary (B461), a rhinocerebral (B462), or a disseminated (B464) localization, was classified under 1 of the 3 latter categories.

We similarly analyzed the underlying diseases potentially associated with zygomycosis and focused on known risk factors. We combined the ICD-10 codes with medical procedures because some conditions, such as bone marrow transplantation, were coded as a procedure and not as a disease. To take into consideration patients with >1 underlying condition, we introduced a hierarchy. We first created the bone marrow transplantation (BMT) category, which included autologous and allogenic BMT, with or without graft versus host disease. To reduce underreporting bias, we looked for BMT occurrence in both incident and readmitted patients. Because of changes in the coding system regarding medical procedures, identification of autologous versus allogenic BMTs before 2003 was not possible. In a second category labeled hematologic malignancies, we included neutropenia or aplastic anemia, acute lymphoid or myeloid leukemia, lymphomas, and other disorders of the myeloid system. The third category, labeled nonhematologic immunodepression, included solid organ cancers, HIV infection, solid organ transplantations, and/or rejection of these transplantations. By convention, BMT patients (with or without graft versus host disease) who also had a neutropenia were classified under the BMT category. Finally, we analyzed diabetes types 1 or 2 patients under a separate category; for patients who had both an immunosuppressive condition and diabetes, the immunosuppressive condition was kept as the main risk factor.

Data Analysis
Taking into account population growth, we used the 1999 national population census to estimate annual incidence rates. To enable trends comparisons, we used the same approach for the specific incidence rates in patients with underlying factors (zygomycosis patients presenting these underlying factors in the numerator and total population in the denominator). Because of heterogeneous age distributions in the population according to geographic areas, standardization of incidence rates by gender and age groups was performed in each of the 96 districts (departments) of metropolitan France. Populations in the districts ranged from 73,500 to >2.5 million inhabitants.

For CFR, we considered a possible underreporting in the PMSI combined with inaccuracies in the registration of causes of deaths in the CepiDc. To estimate the number of deaths that could have been missed by both data sources, we performed a capture–recapture analysis following the method synthesized by Gallay et al. (22). In a 2 × 2 table, the numbers of deaths identified in one of the respective databases (n1,2 or n2,1, in which the subscript 1 indicates identified and 2 indicates not identified) or in both databases (n1,1) are entered in the respective 2 × 2 table's boxes. Under the hypothesis that both sources are independent, deaths that could have been missed by both sources (n2,2) can be estimated by the equation (n2,2 = n1,2 × n2,1 / n1,1). The total number of deaths is N = n1,1 + n1,2 + n2,1 + n2,2. The variance and 95% confidence intervals (CIs) are Var(N) = (n1,1 + n2,1)(n1,1 + n1,2)(n1,2)(n2,1)/(n1,1)3 and 95% CI = N ± 1.96Var(N), respectively.

The CFR estimates were impaired by the unavailability of accurate information about underlying medical conditions for deaths identified in the CepiDc. In addition, the approximation of the year of birth when using the PMSI dataset limited the CFR studies by age groups. We therefore restricted the analysis to the overall CFR by underlying diseases whenever possible and made no further assumptions on missing data.

Statistical Analysis
Statistical analysis was performed by using Excel (Microsoft Corporation, Redmond,, WA, USA) and STATA-9 (StataCorp LP, College Station, TX, USA) software. The Fisher exact or χ2 tests were used where appropriate to compare groups. Trends were assessed by using a Poisson regression. Where appropriate, we present results with 95% CIs and p values, considering p<0.05 as significant.

Suggested Citation for this Article

Bitar D, Van Cauteren D, Lanternier F, Dannaoui E, Che D, Dromer F, et al. Increasing incidence of zygomycosis (mucormycosis), France, 1997–2006. Emerg Infect Dis [serial on the Internet]. 2009 Sep [date cited]. Available from http://www.cdc.gov/EID/content/15/9/1395.htm

DOI: 10.3201/eid1509.090334

abrir aquí para acceder al documento CDC completo del cual se reproduce un 35%:
Zygomycosis, France | CDC EID