Serotypes, biotypes and antimicrobial susceptibilities of Haemophilus influenzae isolated from invasive disease in children in Casablanca

Haemophilus influenzae can cause a variety of invasive diseases in children. These infections are responsible for sequelae[1] and important mortality[2–4]. Resistance of H. influenzae to a variety of antimicrobial agents has been reported with increasing frequency in recent years[5–6], although in countries where vaccination against H. influenzae serotype b has been introduced, an important decline in the incidence of invasive disease has been noticed[7–8].

Recommendations for vaccination policies and antibiotic treatment must be guided by local epidemiology. Recent data on invasive infections of H. influenzae in Africa are very rare. The aim of this study was to determinate the serotype, biotype and the levels of antibiotic resistance of H. influenzae responsible for invasive disease at the Casablanca IbnRochd University hospital over a period of 2.5 years.

From June 1994 to December 1996, 117 strains of H. influenzae were isolated from children aged 0–15 years with invasive disease, who were admitted to the Ibn Rochd University Hospital of Casablanca. Among the 117 strains, 109 were serotyped, 112 were biotyped, 115 were tested for beta-lactamase production and 85 were tested for their susceptibilities to cefotaxime, chloramphenicol, erythromycin and rifampin. Serotyping was carried out by testing the supernatant of a broth culture in tryptic soy broth supplemented with Fields Enrichment (BBL) by countercurrent-immuno-electrophoresis with polyvalent and monovalent antisera (Difco Laboratories, Detroit, MI, USA). Biotype was determined by the study of ornithine decarboxylase and indole production and urea utilization with the API 10 s system (BioMerieux, Narcy L'etoile, France) or Rosco tablets and urea indole medium (Sanofi-Pasteur, Marnes La Coquette, France). Beta-lactamase production was detected with a nitrocefin-impregnated disc (Carr Scarborough Microbiologicals, Decatur, Georgia, USA). Quality control for betalactamase production was carried out with Staphylococcus aureus ATCC 29213. The minimum inhibitory concentrations were determined on Mueller–Hinton chocolate agar by the agar dilution method with an inoculum of 10⁴ CFU per spot. The breakpoints used were those recommended by the National Committee for Clinical Laboratory Standards[9]. H. influenzae ATCC 49247 was used as quality control. Data entry and analysis were carried out using Epi Info 5 (WHO).

From the 117 isolates of H. influenzae responsible for invasive disease, 74.4% were recovered from cerebrospinal fluid. The other sites were blood (16.2%), articular fluid (6.8%) and respiratory tract specimens (2.6%). These infections were caused mainly by serotype b (99.1%). The nontypable isolates were rare in our series; only one case was isolated from articular fluid. The age of the patients was available for 113 children and 89.3% occurred in children that were less than 2 years of age ( Fig. 1). In our series, all biotypes could be responsible for invasive diseases in children, but at a different magnitudes: biotype I (57.1%), II (13.4%), VII (9.8%), VI (5.3%), VIII (5.4%), IV (3.6%), III (4.5%) and V (0.9%). Three isolates (2.6%) produced a beta-lactamase. These strains belonged to serotype b, and were responsible for meningitis. One of these strains was also resistant to chloramphenicol, although chloramphenicol was still highly active (97.8%). Intermediate resistance (MIC 0.5 4 mg/L) or resistance to erythromycin (MIC > 4 mg/L) was found in 85.9% of the isolates. No isolate was resistant to parenteral third-generation cephalosporins. Intermediate resistance to rifampin was found at a low rate (3.5%).

Figure 1. Percentage of cases according to age (n = 113)

The exact incidence of invasive H. influenzae disease in Casablanca, cannot be determined by this study, because other children could have been hospitalized in other hospitals. One important result was that most of our cases occurred in children under the age of 2 years. This feature is similar to the results reported in developing countries where H. influenzae infections occur earlier than in industrialized nations[1–3,7]. The predominance of serotype b in children has already been reported in other countries[1–8,10] and had major implication for the vaccine formulation. The predominance of biotype I is in accordance with the findings of other authors[]. The very low percentage of betalactamase producers reported here is in contrast with other reports[4–6]. In Morocco, a ‘conference de consensus’, organized in 1996[11], was based in part on these results. Its recommendations were that amoxicillin should be the first choice for children suffering from meningitis (neonates excluded). The wide variations of the percentages of betalactamase producers depending on geographical location[4–6,12] stress the need for local data to provide recommendations for antibiotic treatment. The incidence of multiresistant strains (mainly to ampicillin and chloramphenicol) is usually low[], but it is important to monitor.

Although these results do not represent all the invasive cases of H. influenzae in Casablanca during the period of the study, they are a baseline of surveillance for monitoring the impact of vaccination. The number of cases and the age of occurrence of the disease and the well-documented efficacy of the vaccine allow us to recommend the introduction of vaccination into the routine immunization schedule of children in our country.