Incidence of Haemophilus influenzae type b meningitis during 18 years of vaccine use: observational study using routine hospital data

Despite the spectacular success of Haemophilus influenzae type b (Hib) conjugate vaccines in the developed world,¹ failures have become more common in the Netherlands and the United Kingdom, where the incidence of invasive Hib diseases at age 0-4 years per 100 000 increased from 0.66 in 1998 to 2.96 in 2001.² An immunological defect is not to blame.

This increase in failure coincided with the change from whole cell pertussis to combined three component acellular vaccine (DTaP-Hib). This combination lowers Hib antibodies.³ The clinical significance is unclear. Acellular vaccine cannot be blamed in the Netherlands, where whole cell vaccine was still used.⁴ Both countries have been giving Hib vaccination at age 2, 3, and 4 months; a booster at 11 months is given in the Netherlands. Another characteristic in the United Kingdom is concomitant meningococcal group C conjugate. Replacement of serotypes other than Hib in severe diseases has also raised concern.

Because H influenzae meningitis in the Greater Helsinki area has been recorded since 1939, we had a unique opportunity to investigate any impact of vaccination.

Methods and results

Three paediatric hospitals have been dealing with severe infections here. We studied the records of H influenzae meningitis in 1939-2003. Of 454 children aged 0-4 years in 1980-5, only 0.2% (two cases, one of type a and one of type g) were caused by types other than Hib.

Three Hib conjugates have been used in Finland—diphtheria toxoid conjugate (PRP-D) in 1986-9, mutant diphtheria toxoid conjugate (HbOC) in 1988-9, and from 1994 onwards tetanus, toxoid conjugate (PRP-T) in 1990-3. Only three doses of Hib—at ages 4, 6, and 14-18 months—have been given since 1988.

No cases were found in 1991-4 (figure). In the next nine years, three Hib meningitides were found: in 1995 (a 6 year old girl with three doses of PRP-D), in 1999 (a 4 year old boy with three doses of HbOC), and in 2000 (a 4 month old boy vaccinated once only with HbOC).

Figure 1.

Cases of Hib meningitis in the Greater Helsinki area (current population >1 million), 1939-2003. No reincrease in the incidence has been found since wide scale vaccination began in 1986. Hib=Haemophilus influenzae type b; Hif=H influenzae type f; Hi-NT=non-typable H influenzae. Three different Hib vaccines have been used over the years

There were, however, three more cases of H influenzae meningitis, two caused by Hif, one by a non-typable strain. No immunological defect was seen.

Comment

Despite fears that the effectiveness of Hib conjugate was declining, excellent protection from Hib meningitis has been maintained with only three doses across Scandinavia. In 2003, five vaccine failures were identified in Finland, but despite this, we think our approach—keeping at least two months between the first two doses and giving the third as a late booster—is a good compromise.

The acellular pertussis component has been blaimed for the Hib vaccine failures, but those have not become more common after implementation of pentavalent or hexavalent vaccines in Germany. The lower anti-polysaccharide concentration alone cannot be the only explanation. The Dutch schedule was changed from immunisation at 3, 4, and 5 months to 2, 3, and 4 months, and this small change in timing might have played a role, despite a late booster. Significance of the concomitant use of meningococcal group C conjugate in the United Kingdom is not yet settled.

What is already known on this topic

The official recommendation is four doses for the most widely used Haemophilus influenzae type b (Hib) conjugate vaccines; despite this, Finland has been using only three vaccine doses since 1988 with excellent clinical effectiveness

What this study adds

Two Hib vaccine doses in early infancy with a late booster is efficacious and may practically eliminate Hib meningitis

Attention should be paid to the clinical effectiveness of Hib conjugates rather than surrogate antibody measurements, which we often cannot interpret correctly

Epidemiological data are probably more relevant than antibody measurements, which are not always correctly interpreted. The surrogate antipolysaccharide antibody concentrations of 0.15 μg/ml for short term and of 1.0 μg/ml for long term likely protection (determined by our group)⁵—still taken by many as surrogate markers for clinical effectiveness—applied to Hib polysaccharide; for conjugate vaccines, we simply do not know what the “protective” antibody concentrations are. Instead, what we can quantify is a vaccine's clinical effectiveness. We think a later booster is important at the age when vaccine induced antibodies are declining.