Oral non-typable Haemophilus influenzae enhances physiological mechanism of airways protection

Abstract

Oral immunotherapy with inactivated non-typeable Haemophilus influenzae (NTHi) prevents exacerbations of chronic obstructive pulmonary disease, but the mechanism is unclear. The aim of this study was to determine the mechanism of protection. This was a placebo versus active prospective study over 3 months in 64 smokers. The active treatment was three courses of oral NTHi given at monthly intervals, followed by measurement of bacteriological and immunological parameters. The results can be summarized: (i) NTHi-specific T cells increased in the placebo treatment group over time (P<0.05); (ii) the T cell response in the oral NTHi group started earlier than that in the placebo group (P<0.05); and (iii) serum NTHi-specific immunoglobulin (Ig)G had significantly greater variation in the placebo group (P<0.0001). The increase in antibody in placebos over time correlated with exposure to live H. influenzae (P<0.05) determined from culture of gargles; (iv) reduction in saliva lysozyme over time (P<0.05) was detected only in the oral NTHi treatment group. These data are consistent with T cell priming of gut lymphoid tissue by aspiration of bronchus content into the gut, with oral immunotherapy augmenting this process leading to enhanced bronchus protection. The evidence for protection was a stable IgG antibody level through the study in the oral NTHi treatment group, contrasting with an increase in antibody correlating with exposure of the airways to H. influenzae in the placebo group. Saliva lysozyme was a useful biomarker of mucosal inflammation, falling after oral NTHi consistent with a reduction in the level of intralumenal inflammation.

Fig. 1

Serum non-typeable Haemophilus influenzae (NTHi)-specific immunoglobulin (Ig)G. The mean serum NTHi-specific IgG for active (□) and placebo () treatments groups is shown. Error bars are standard error of the mean.

Fig. 2

Median change in serum immunoglobulin (Ig)G (EU/ml) for placebo subjects with Haemophilus influenzae detected in nil to one or two to six gargles. Relationship between H. influenzae carriage and serum non-typeable H. influenzae (NTHi)-specific IgG. The change (visit 7 minus visit 1) in median serum NTHi-specific IgG is plotted for active (□) or placebo () treatment groups for subjects with H. influenzae detected in nil or one gargle sample, or those with H. influenzae detected in two to six gargle samples. In the placebo group there were 23 subjects with nil to one and seven subjects with two to six positive gargles. In the HI-164-OV group there were 24 subjects with nil to one and six subjects with two to six positive gargles.

Fig. 3

Blood leucocyte stimulation in vitro with non-typeable Haemophilus influenzae (NTHi) antigen. The mean stimulation index for peripheral blood leucocytes in response to in vitro stimulation with 1 µg NTHi antigen/ml for placebo () and active (□) treatment groups is shown for each visit. The stimulation index (SI) for the active treatment group is significantly different (P < 0·05) at visits 4, 5 and 7 to that at baseline (visit 1). The SI for the placebo treatment group is significantly different (P < 0·05) at visits 4 and 5 to that at baseline (visit 1).

Fig. 4

Blood leucocyte stimulation in vitro with phytohaemagglutinin (PHA). The mean stimulation index for peripheral blood leucocytes in response to in vitro stimulation with PHA for placebo () and active (□) treatment groups is shown for each visit. The stimulation index (SI) for the active treatment group is significantly different (P < 0·05) at visits 3 and 4 to that at baseline (visit 1). The SI for the placebo treatment group is significantly different (P < 0·05) at visit 6 to that at baseline (visit 1).

Fig. 5

Saliva lysozyme. The saliva lysozyme concentrations for placebo () and active (□) treatment groups are shown for each visit.