This study was comprised of two groups of patients, one from the USA (State of Connecticut) and the other from Germany (State of Schleswig-Holstein).
The USA study population consisted of 30 hospitalized adult patients (16 M, 14 F) over 18 years of age (range: 28–87 years) with positive cultures for Streptococcus pneumoniae isolated from sterile body fluids. Subjects were identified consecutively in a prospective manner by the microbiology laboratories at the Hospital of Saint Raphael, Saint Francis Hospital, and Yale-New Haven Hospital between September 2004 and September 2005. Written informed consent was obtained from all subjects or their surrogates. The respective hospital Investigational Review Boards and the Yale Human Investigation Committee approved the study protocol. Classification of a case as meningitis among the USA cohort was based either on positive pneumococcal cultures grown from the cerebrospinal fluid (CSF) or from treating physician diagnosis of meningitis based on clinical symptoms and signs and lumbar puncture results plus positive blood cultures.
The German study population included 89 adult Caucasian patients (52 M, 37 F) over 18 years of age (range: 24–96 years) with positive blood, CSF, bronchoalveolar lavage, sputum/tracheal secretion, and/or pleural fluid cultures growing Streptococcus pneumoniae. Subjects were identified at three German hospitals in a consecutive and prospective manner between December 1998 and May 2004. The study was approved by the local ethics committee, and written informed consent was obtained from subjects or their surrogates. Cases of meningitis were defined by clinical presentation and a marked neutrophilic pleocytosis in the CSF and/or detection of pneumococci in CSF gram stain or culture.
Genomic DNA from patient serum was isolated per manufacturer protocol using DNAzol Reagent (Invitrogen) or from dried, whole blood-spotted filter paper using the QIAmp DNA Blood Mini kit (Qiagen). Genomic DNA was amplified by multiple displacement amplification (MDA) overnight according to previously described methods [10
]. The −794 CATT repeat and the −173 G/C SNP in the MIF
gene were analyzed at the Yale University Keck Affymetrix Facility using an ABI Prism 7900 machine, and the allelic data were analyzed with SDS v2.2 software. Levels of MIF in the plasma of subjects were measured by sandwich ELISA [11
The data were analyzed using GraphPad Prism software (GraphPad Software, La Jolla, CA), STATA 11 (StataCorp, College Station, TX), and Microsoft Excel (Microsoft, Redmond, WA). Differences between two means were calculated using independent two-tailed t-tests. Distributions were presumed to be normal unless noted in the text. The relationship between genotype and disease type was examined using the Fishers' exact test for contingency tables containing small numbers. Prevalence ratios are reported for the relationship between genotype and meningitis since this study was cross-sectional. A p-value of less than 0.05 was defined as statistically significant for all tests.