Forty-seven children with clinically and immunologically documented and physician-verified birch pollen allergy were enrolled in the study. All had previously had symptoms of allergic rhinitis confined to the birch pollen season, and the specific test result to birch pollen had been positive, either with skin prick testing (positive result if the mean wheal diameter for birch pollen allergen was greater than 3 mm) or demonstration of specific IgE antibody in serum (according to the manufacturer’s reference values, the level for positivity was a specific IgE more than 0.35 kU/L). The symptoms of the patients during the previous pollen season included sneezing and runny or blocked nose, and the nasal symptoms were often associated with allergic eye symptoms, such as itching and redness of the conjunctiva. Patients with diagnosed asthma, habitual use of probiotics and/or prebiotics and recent use of antibiotics were excluded from the study.

The study was approved by the Ethics Committee of the Hospital District of South-Western Finland and informed, written consent was received from the participants and their parents. The study was registered at http://www.clinicaltrials.gov under the identifier NCT00746226.

The study material consisted of capsules containing 5 ×10⁹ CFU of a combination of 25% Lactobacillus acidophilus (L. acidophilus) NCFM^TM (ATCC 700396) and 75% Bifidobacterium lactis (B. lactis) Bl-04 (ATCC SD5219), Danisco Cultures, Madison, USA. The strains were selected based on their anti-inflammatory properties and Th1-type immune stimulating effects and their ratio was chosen to optimize these functions. The viability of the study material was determined at monthly intervals during the study by Danisco Cultures, Madison. This laboratory was not involved in the study, and no significant reduction in viability was observed (results not shown). The placebo consisted of identical capsules containing microcrystalline cellulose. The study material was randomized at the site of production by a person not involved in the study. The volunteers, researchers and biostatisticians remained blinded for the duration of the study and the analyses. The code was broken after completion of the statistical analyses.

Eligible subjects were assigned to receive one of the individually coded test products. The subjects or their parents were instructed to consume one capsule daily or to suspend its contents in a suitable liquid.

The patients attended the study unit 3 times during spring 2006 (Figure 1). The first visit was in March before the birch pollen season, and consumption of the study material started after this visit. The second visit was at the end of April-beginning of May during the birch pollen season and the third visit was in June at the end of the season, when consumption of the study product was finished.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Percentage of subjects with specific symptoms of birch pollen allergy among allergic children. The gray bar represents the alder pollen season and the black bar the birch pollen season in Turku (Finland) during the study period (> 100 pollen/m³).

Clinical examination of the patients was carried out at each study visit. At each visit, nasal smears were taken from both nostrils by gently rubbing the middle or inferior turbinate with a cotton covered stick, followed by fixing and staining with eosin and methylene blue. Eosinophil counts were determined semiquantitatively by microscope as follows: -, no eosinophils in any of the fields of view; +, a few scattered eosinophils at least in some fields of view; ++, eosinophils easily found in every field of view; and +++, every field of view abundant in eosinophils. Venous blood samples were drawn at each study visit. At the first and third visit, the blood eosinophil count was determined by automated blood cell counter (Advia 120, Bayer Health Care, Germany), and allergen-specific IgE antibodies in serum were measured by an immunofluorometric assay (ImmunoCAP^TM, Phadia AB, Uppsala, Sweden). All samples were initially screened using a mix of inhalation allergens (Phadiatop ImmunoCAP). The samples which were positive at screening were subsequently tested for birch pollen-specific IgE.

The concentrations of the cytokines interleukin (IL)-4, IL-5, IL-6, IL-10 and tumor necrosis factor (TNF)-α were measured by multiplex flow cytometric assay using a commercial Human Th1/Th2 Cytokine Kit (BD Immunocytometry Systems and BD Biosciences Pharmingen). The concentrations of transforming growth factor (TGF)-β2 were measured at the first and second visit and soluble CD-14 (sCD-14) at the first and third visit using commercial sandwich ELISA specific for these molecules (R&D Systems Europe Ltd., Abingdon, UK).

The presence of blocked or runny nose, respiratory difficulty, coughing, eye symptoms or eczema was recorded by the parents on the diary cards throughout the study. The number of days with each symptom during each of the study months was reported.

The major groups of fecal bacteria were analyzed using the FISH method as previously described[13]. In brief, fecal samples were suspended in PBS and homogenized. Bacteria were fixed with paraformaldehyde and hybridized with Cy3 indocarbocyanin-labeled oligonucleotide probe. Probes included Bac303 (5’-CCAATGTGGGGGACCTT-3’) for the Bacteroides-Prevotella group, Bif164 (5’-CATCCGGCATTACCACCC-3’) for bifidobacteria, His150 [5’-TTATGCGGTATTAATCT (C/T) CCTTT-3’] for clostridia of the C. histolyticum group and MUC-1437 (5’-CCTTGCGGTTGGCTTCAGAT-3’) for Akkermansia muciniphila-like bacteria[14], total cells were enumerated using an EUB338 (5’-GCTGCCTCCCGTAGGAGT-3’)-fluorescein (FITC)-labelled probe[1516].

Flow-cytometric analyses were performed using a BD^TM LSR II flow cytometer (Becton, Dickinson and Co., USA) equipped with 4 lasers (355, 405, 488 and 635 nm). We used the 488 nm laser at 15 mW. This standard instrument is equipped with 2 light scatter detectors which measure forward (FSC) and side scatter (SSC) and 4 fluorescence detectors detecting appropriately filtered light at green (FL1, 525 nm), yellow (FL2, 575 nm), orange (FL3, 620 nm), and red (FL4, 675 nm) wavelengths. To avoid cell coincidence, the flow rate was kept at the lowest setting (data rate 200-300 events per second). At least 30 000 events were recorded for each sample and all experiments were conducted in duplicate. Data were stored as list-mode files and analyzed off-line using the BD FACSDiva™ software version 4.1.1 (Becton, Dickinson and Co., USA).

Immediately prior to analysis, Flow-Count fluorospheres (Beckman Coulter, USA) were added to each sample. Absolute bacterial cell counts were determined following the manufacturer’s instructions, using the ratio of positive bacteria to fluorospheres counted using the following formula: cells/&mgr;L = [(cells counted)/(fluorospheres counted)] × fluorospheres/&mgr;L. To avoid loss of the signal intensity of hybridized cells, they were kept in the dark on ice at 4°C until the flow cytometry assay. Results were expressed as the numbers of cells hybridizing with the specific group-Cy3 probe and total bacteria EUB 338-FITC probe.

DNA extractions from pure cultures of the different microorganisms and fecal samples were extracted using the QIAamp DNA stool Mini kit (Qiagen, Hilden, Germany) following the manufacturer’s instructions. qPCRs were conducted as previously described[1011]. For characterization of the fecal microbiota PCR primers were designed targeting B. lactis according to Gueimonde and co-workers[17]. L. acidophilus NCFM^TM was quantified using primers as described by Ouwehand and co-workers[18]. The oligonucleotides were purchased from the Thermo Electron Corporation (Thermo Biosciences, Ulm, Germany). Briefly, PCR amplification and detection were performed with an ABI PRISM 7300-PCR sequence detection system (Applied Biosystems, UK). Each reaction mixture of 25 &mgr;L was composed of SYBR^® Green PCR Master Mix (Applied Biosystems, UK.), 1 &mgr;L of each of the specific primers at a concentration of 0.25 &mgr;mol/L, and 1 &mgr;L of template DNA. The fluorescent products were detected at the last step of each cycle. A melting curve analysis was made after amplification to distinguish the targeted PCR product from the non-targeted PCR product.

Microbial activity was determined by measurement of lactic acid, short chain fatty acids: acetic, butyric, lactic and propionic acids, and branched chain fatty acids: valeric, isobutyric, 2-methylbutyric and isovaleric acid. These were determined by gas chromatography as described previously by Holben et al[19]. The results were expressed in mmol/kg.

Changes in the immunological status of the intestine were monitored by measuring the concentrations of IgA and calprotectin from the soluble fraction of feces. For IgA measurements the frozen samples were thawed and extracted with BSA as described previously and stored at -20°C before analysis[20]. Concentrations of IgA were then determined with the ELISA according to the manufacturer’s instructions (E80-102, Bethyl Laboratories, Inc., Montgomery, TX, USA). The concentrations of calprotectin were determined with ELISA following the manufacturer’s instructions (Calpro AS, Oslo, Norway). The results were expressed as &mgr;g per g fresh weight.

Differences between the treatment groups were analysed using the Mann-Whitney U test. Differences within the treatment groups between the different time points were analyzed using the Wilcoxon signed rank test. Correlations between the level of fecal colonization of B. lactis or L. acidophilus NCFM^TM and immune or microbiological variables in the probiotic group during the second visit were determined with the Spearman rank correlation test. To determine differences between the groups and over time, the data was analyzed with a linear mixed effects model, having effects for time, group, and their joint effect. Individual baseline differences were taken into account by including the first time point as a covariate in the model. P-values below 0.05 were considered significant.

Baseline clinical characteristics of the 2 groups are presented in Table 1. Twenty patients in the probiotic and 21 in the placebo group completed the study. Three patients, 2 in the probiotic group and 1 in the placebo group, were excluded during the study because of newly developed asthma. Three patients, 2 in the probiotic group and 1 in the placebo group, decided to drop out of the study. The mean duration of breast-feeding was 7.5 mo (range 0-18 mo), there being 2 patients whom were not breast-fed.

The impact of the onset of the birch pollen season on allergy symptoms in the two study groups is depicted in Figure 1. The peak period of birch pollen exposure, April and May, resulted in specific symptoms such as nasal obstruction and runny nose. All the recorded symptoms increased during April and May (the height of the pollen season in Turku, Finland), and decreased again in June (P < 0.001).

At the start of the study, there was no difference in the observed symptoms between the 2 study groups. Runny nose tended (P = 0.078) to be reported in May by fewer subjects in the probiotic group (76.2%) than in the placebo group (95.2%). During June, fewer subjects reported nasal blocking in the probiotics group (11.1%) than in the placebo group (33.3%). This did not, however, reach statistical significance (P = 0.101). Concomitantly with the differences in nasal symptoms, the number of subjects with infiltration of eosinophils in the nasal mucosa increased in the placebo group during the April/May visit but an increase was not observed in the probiotic group (P = 0.013), where the fraction of subjects with nasal eosinophil infiltration remained unchanged (Table 2).

During May, the subjects in the probiotic group reported more days with eye symptoms, 12.5 d (interquartile range (IQR) 6-18 d), than subjects in the placebo group, 7.5 d (IQR 0-11.5 d), P = 0.066. In June, this difference was no longer detectable.

In addition to the increase in symptoms over the season, concentrations of birch pollen-specific IgE also increased in both groups from March to June, P = 0.0015 and P = 0.0002 for placebo and probiotic respectively (Table 3). Likewise, blood eosinophil numbers increased in both placebo (P = 0.03) and probiotic groups (P = 0.002), from March to June (Table 3). Concentrations of IL-6 decreased in both groups from March to April/May, P = 0.0004 and P = 0.0002 in placebo and probiotic groups, respectively (Table 3). Concentrations of IL-10 were reduced in the placebo group from March to April/May, P = 0.0495. For both groups, concentrations of TNF-α decreased from March to April/May, P = 0.0007 and P = 0.0005 in placebo and probiotic groups, respectively. Concentrations of TNF-α were higher in the placebo group at the start of the study, P = 0.0062, however, at the April/May visit no difference was observed, P = 0.17. Concentrations of interferon-γ and IL-2 were below the detection limit for both groups at all 3 time points tested. No differences in the concentrations of TGF-β2, IL-4, IL-5 and CD14 were detected (Table 3).

Concentrations of fecal IgA and calprotectin exhibited great individual variance (P < 0.0001 and P = 0.015, respectively). Because of this significant variation, the data were analyzed taking the baseline level into account. Neither the probiotic treatment, nor pollen season had any significant effects on the fecal calprotectin concentrations. However, fecal IgA concentrations were increased in the placebo group during April/May (P = 0.028), which was prevented by the probiotic intervention. In the latter group, the concentration remained stable during the entire study period, Table 4.

Administration of B. lactis Bl-04 and L. acidophilus NCFM^TM led to a significant increase in the fecal numbers of B. lactis (P = 0.0032) and L. acidophilus NCFM^TM (P = 0.0036), from March to April/May and numbers remained high until the end of the intervention in June, indicating good compliance by the subjects (Table 5). The differences were also significant compared to the placebo group during April/May; B. lactis (P < 0.0001) and L. acidophilus NCFM^TM (P = 0.0002), and June; B. lactis (P < 0.0001) and L. acidophilus NCFM^TM (P < 0.0001), Table 5.

A general decrease in numbers of the major bacterial groups Eubacterium, Bifidobacterium, Bacteroides and Clostridium accompanied the birch pollen season in both study groups (Table 5). This may not be explained by the increase in fecal numbers of B. lactis and L. acidophilus as the decrease in fecal numbers was manifested in both probiotic and placebo groups. In the placebo group, concentrations of 2-methylbutyric acid were reduced from April/May to June, P = 0.0501. However, no differences in concentrations of other measured fecal organic acids were observed between the 2 groups for any of the time points (Table 6).

Pollen allergies are increasing in affluent societies. This is thought to be related to a reduced microbial exposure. Probiotics could provide a safe form of microbial exposure possibly through modulation of the intestinal microbiota.

Studies have shown that selected probiotics may modulate the immune response and contribute to the relief or primary prevention of atopic dermatitis. Few studies have, however, investigated the potential of probiotics on pollen allergies. The present study indicates a potential for the specific combination of Lactobacillus acidophilus (L. acidophilus) NCFM and Bifidobacterium lactis (B. lactis) Bl-04 in the alleviation of birch pollen allergy in children.

The present randomized, placebo-controlled, double-blind study showed that the intestinal microbiota changes during the birch pollen season, and the combination of L. acidophilus NCFM and B. lactis Bl-04 was not able to counteract this change. Despite the prophylactic use of antihistamines, the probiotic combination did reduce an objective marker of pollen allergy (nasal eosinophilia) and indicated a trend for a reduction in subjective markers, nasal blocking and runny nose.

Probiotics may provide an alternative or complementary treatment for pollen allergies. A future study could investigate whether this would lead to a reduced use of antihistamines.

Probiotics are live microorganisms which, when administered in adequate amounts confer a health benefit on the host. Specific strains of probiotics have been observed to modulate the immune system and the composition and activity of the intestinal microbiota.

The study addresses an interesting and relevant issue on which data is scarce, and the findings may contribute to our understanding of the role of probiotics in allergic rhinitis.