**5. Affecting factors coexisting in various studies and why inconsistent results in some studies?**

Most of the studies have been conducted in small numbers of patients and results have varied considerably, even with the same strain of probiotics. When compared to the hallmark Finnish Kalliomaki study, there were also a number of other key differences between these studies that could contribute to the disparity in clinical findings. **Firstly (**biological difference of probiotic strains**)**, different probiotic species were used in various studies. Lctbs rhamnosus GG is the strain that has been most studied. Probiotic doses have also varied considerably between the studies. Although there are noted biological differences between strains, various strains have been observed to have different immunologic effects both in vivo and in vitro [35]. Thus, the effects of preparations differ markedly, and the concept itself might be misleading in these studies and could be changed: continuous change in the supplemented strain could lead to continued immunologic stimulus and sustained and stronger effects. As a result, on the basis of these studies, immunologic effects expressed as chronic low-grade inflammation were more pronounced with LGG alone rather than with the mixture of 4 strains used in some prevention studies [57]. This might explain the more sustained preventive effect observed earlier. The results of various studies also demonstrate that findings from any probiotic bacteria cannot be extrapolated to other probiotic bacteria [20]. **Secondly** (use of probiotics in prenatal/intrauterine or postnatal or weaning period), Finnish mothers commenced supplementation during pregnancy, whereas some supplementation began in the first days of life. Prenatal/intrauterine use of probiotics would imply a direct immune effect in utero rather than any effect on postnatal colonization. However, it seems unlikely that supplementation for only a few weeks in the antenatal period alone would account for such significant differences in study outcomes, although this remains a possibility. **Thirdly (**intervention methods**)**, in most studies all babies received the supplement directly, regardless of feeding method, whereas in the Finnish and other studies, the mother took the probiotics if babies were breastfed. Therefore, the Finnish probiotic group and others included breast-fed infants who did not receive probiotics directly in addition to the bottle-fed infants who received probiotics for 6 months. **Fourthly** (postnatal assessment time), some researchers assessed clinical outcomes at 12 months of age, whereas the effects on AD in the Finnish or other studies were reported at 2, 4 and subsequently at 7 years of age. AD typically begins in the first year of life, and it is possible that more children could become affected in their second year of life. And in young infants, the immune system is still developing. There is still a possibility to direct it toward tolerance. In older children, the allergic phenotype is already established, and here one may only be able to relieve the symptoms. **Fifthly** (atopy risk and host factors of targeted group), some studies were performed in high-risk population all had maternal allergic disease confirmed by SPT, whereas the Finnish and other population included children with maternal, paternal, or sibling allergy. This may lead to some population being of slightly higher risk, compared with the Finnish and other population at the same age. Explanations for varied study results include host factors such as genetic susceptibility, environmental factors such as geographic region and diet [98]. Genotyping of study patients in relation to different genes predisposing to allergic diseases may help to find patients that might especially benefit from probiotic intervention. For example, 2 independent mutations in the gene encoding the epidermal protein filaggrin have been shown to be strong predisposing factors for childhood eczema [99]. Of note, these same mutations have recently been demonstrated to be associated not only with eczema-associated asthma susceptibility but also with asthma severity independent of eczema status. More generally, any means to better stratify or select defined subpopulations of subjects (e.g. patients with food allergy as a separate group) would help in clarifying the potency and limits of probiotic interventions against atopic diseases. **Sixthly (**other methodological set-up**),** there may be still due to differences in the clinical and methodological set-up (additional treatments such as topical treatment or feeding hydrolyzed infant formulae and importantly, different probiotic preparations or formulations or combinations). Other differences include age (6–18 months) of the studied children and an intervention period (16 weeks) of time. **Seventhly (**nomenclature of the disease**),** to date, randomized clinical trials of probiotics in allergic diseases have mostly focused on children with eczema and atopic eczema. The definitions of the disease have recently been revised by an international expert group [17-19]. In many of the studies published before the revision of the nomenclature, different definitions have been used, making direct comparisons between the studies difficult. The severity of AD at the start of an intervention may influence the outcome as well, as you imagine.

#### **6. Safety**

374 Atopic Dermatitis – Disease Etiology and Clinical Management

efficacy. For this meta-analysis of randomized controlled trials describing the efficacy of probiotics in AD, a comprehensive search was performed of databases through January 2008. Eleven studies were identified, and data from 10 studies (n: ≥678) were available to analyze. There was an overall statistically significant difference favoring probiotics compared with placebo in reducing the SCORAD index. Children with moderately severe disease were more likely to benefit. Duration of probiotic administration, age, and type of probiotic used did not affect outcome. Data from this meta-analysis suggest a modest role for probiotics in pediatric AD and the effect is seen in moderately severe rather than mild disease [94]. Lee et al meta-analyzed 10 double-blind randomized controlled clinical trials. And they found out that current evidence was more convincing for probiotics' efficacy in

However, in a Cochrane Review by Boyle et al concluded that the evidence suggests that probiotics are not an effective treatment for eczema, and probiotic treatment carries a small risk of adverse events [96]. Another review of 13 studies of probiotics for treating established eczema by Williams at al did not show convincing evidence of a clinically

Taken together, most of these metaanalytic studies show a mild-moderate benefit over placebo for the treatment and/or preventionof AD. However, several of the studies still show no benefit. Some probiotics appeared to be more effective than other probiotic preparations and in children with more severe disease. It seems that duration of probiotic administration, age, and type of probiotic used did not affect outcome. Although there was a reduction in clinical eczema score in infants, this effect was not consistent between studies and caution is advised in view of methodological concerns regarding included studies.

**5. Affecting factors coexisting in various studies and why inconsistent** 

Most of the studies have been conducted in small numbers of patients and results have varied considerably, even with the same strain of probiotics. When compared to the hallmark Finnish Kalliomaki study, there were also a number of other key differences between these studies that could contribute to the disparity in clinical findings. **Firstly (**biological difference of probiotic strains**)**, different probiotic species were used in various studies. Lctbs rhamnosus GG is the strain that has been most studied. Probiotic doses have also varied considerably between the studies. Although there are noted biological differences between strains, various strains have been observed to have different immunologic effects both in vivo and in vitro [35]. Thus, the effects of preparations differ markedly, and the concept itself might be misleading in these studies and could be changed: continuous change in the supplemented strain could lead to continued immunologic stimulus and sustained and stronger effects. As a result, on the basis of these studies, immunologic effects expressed as chronic low-grade inflammation were more pronounced with LGG alone rather than with the mixture of 4 strains used in some prevention studies [57]. This might explain the more sustained preventive effect observed earlier. The results of various studies also demonstrate that findings from any probiotic bacteria cannot be extrapolated to other probiotic bacteria [20]. **Secondly** (use of probiotics in prenatal/intrauterine or postnatal or weaning period), Finnish mothers commenced supplementation during pregnancy, whereas some supplementation began in the first days of life. Prenatal/intrauterine use of probiotics would imply a direct immune effect in utero rather than any effect on postnatal colonization. However, it seems unlikely that supplementation for

prevention than treatment of pediatric AD [95].

worthwhile benefit [97].

**results in some studies?** 

Probiotics available as food ingredients or dietary supplements that contain microorganisms have been used extensively in food processing for years, with a long history of safety and no adverse effects on metabolism. However, when considering the safety of probiotics, potential adverse effects include systemic infections, altered metabolism, and gene transfer [100]. A recent report has identified Lctbs septicemia in 2 children with short bowel syndrome who were receiving LGG supplementation for control of bacterial overgrowth

The Role of Probiotics in Atopic Dermatitis Prevention and Therapy 377

immune system by probiotics in hosts with immune dysfunctions, such as individuals genetically predisposed to autoimmunity, has raised some concerns too. With respect to the association between bacterial antigens and autoimmune responses and the adjuvant activity of lactic acid bacteria strains, the involvement of lactic acid bacteria in the pathogenesis of some models of autoimmunity in experimental animals and possibly in humans has been suggested [109]. Thus, from a safety point of view, the potential of probiotic bacteria (especially the immunostimulatory strains), to induce destructive inflammation or autoimmunity needs to be investigated. For instance, it has been experimentally demonstrated that Lctbs casei cell wall components (given intraperitoneally) are able to

As mentioned above, there is a large amount of conflicting data on the preventive/therapeutic effects of probiotics in AD. Results from metaanalyses and systematic reviews that combine results of studies from different types of probiotics to examine the effects in any disease should be interpreted with caution. One may quickly recognize the degree of heterogeneity among the different probiotic studies. As mentioned, very few studies were similar in design. Several different probiotic strains with different dosing regimens were used, and many studies showed similarity in efficacy to placebo shortly after probiotic therapy was discontinued. Some probiotic studies suggest short-term statistically significant improvement in SCORAD scores and no sustained benefit from continued ingestion. Therefore, subgroup analysis became critical in understanding the outcomes of the studies. Not all children receiving the probiotic agent benefited, but subsets of these patients, mainly those with moderate disease activity and IgE-associated disease (atopic eczema), seemed to have benefited the most. There are also difficulties of recognizing etiology and pathogenesis of AD in which have many mechanisms involved. Similarly, with various strains, especially e.g. in Kopp and Taylor et al's study, development and/or stimulation of Th2-mediated immune responses have been described [87,88]. Additionally, if probiotics are used in patients with ADs for any reason –therapy or prevention- cautionary approach ought to be taken. Thus, probiotics cannot be recommended generally for primary prevention of ADs. Any probiotics should not be used especially in immune-compromised children; even they have at risk for ADs. Finally, there is insufficient but fairly promising evidence to recommend the addition

Involvement of commensal enteric microflora and its components with strong immunoactivating properties in etiopathogenetic mechanism of multifactorial diseases, including atopic diseases has been recently suggested. Regulation of intestinal microflora composition (e.g. by probiotics) offers the possibility to influence the development of mucosal and systemic immunity as well as it can play a role also in prevention and treatment of AD. Progress has been made by the identification of receptors and pathways through which gut microbes influence development of the immune system. Such mechanistic data have moved a field that was once regarded as being on the scientific fringe to the mainstream, and support increased funding to advance this promising area of research in the hope that it might deliver

induce cardioangitis (an autoimmunity-associated heart disease) in mice [110].

of probiotics to foods for prevention and treatment of AD [111].

**8. Five-year view and future expectations** 

the long awaited answer of how to safely prevent AD.

**7. Conclusion** 

[101]. Land et al recently reported LGG probiotic sepsis occurring in immunocompromised infants and children [102]. A medically fragile infant 6 weeks of age became septic with a strain of LGG that was being provided as a supplement. Molecular DNA fingerprinting confirmed that the LGG probiotic supplement was the bacterial isolate from the infant. Neonatal sepsis and meningitis that were apparently associated with the administration of a probiotic supplement were also reported [103,104]. Children with abnormal immune function, premature infants, and those with indwelling central lines should use these products with caution, because many species such as Lactobacilli, streptococci, and enterococci are potential opportunistic pathogens. Owing to the theoretical risk of immunomodulation, especially in immunocompromised hosts or those with autoimmune disorders, few reports of probiotic-related disease have been reported [105].

Bifidobacteria have also been consumed in infant formulas for ≥15 years worldwide and have not been associated with any pathologic or adverse event. In particular, studies have documented safety and adequate growth with Bfdbm lactis in infants from birth and in vulnerable populations, including preterm infants, malnourished infants, and infants born to mothers with HIV disease. From the safety point of view, according to current available information, Bifidobacteria, particularly Bfdbm lactis, has a uniquely strong safety profile, making it a good probiotic candidate for newborns and young infants [106]. Lactobacilli, particularly Lctbs rhamnosus (LGG), also seem generally safe and may be a probiotic appropriate for older infants and children [107]. Until adequate data are available for each specific probiotic bacterium, use of probiotics in general cannot be recommended in immunocompromised populations. However, as safety is better documented for specific bacteria, we may be able to use them in certain populations (such as premature infants) that may stand to benefit the most from probiotic use.

Another consideration is that cow's milk protein allergy is one of the common food allergies in infants. Culture conditions used in growing several probiotic products may contain cow's milk protein. There have been reports of severe adverse reactions when pediatric patients with cow's milk protein allergy ingested probiotics. Therefore, caution should be used in prescribing such probiotic products in sensitized children to avoid significant reactions. There is also a study worth mentioning by Taylor et al using Lctbs acidophilus daily for the first 6 months of life in newborns of women with allergy. The presence of Lctbs in the body at 6 months of age was associated with increased risk of subsequent cow's milk sensitization as well [87]. Nevertheless other studies have examined the effect of probiotic consumption on sensitization to several allergens (e.g. peanut, hen's egg, soy, wheat, milk, cat, dog), as determined by specific IgE production or skin prick test. The authors could not find a difference before and after the treatment. Interesting another trial by Kopp et al demonstrated that supplementation with LGG during pregnancy and early infancy in affected children it might be associated with an increased rate of recurrent episodes of wheezing bronchitis [88]. However, a recent study was done by Kukkonen et al evaluating airway inflammation in probiotic-treated children at 5 years in 1018 children. Early intervention with probiotics and prebiotics did not affect airway inflammation later in childhood [108].

Furthermore, similarly certain probiotics are known to stimulate Th1 immunity, which has been suggested as one of the mechanisms by which they can suppress Th2-mediated allergic diseases. However, this presumed excessive immunostimulation might aggravate or induce Th1-mediated immune responses and diseases such as type 1 diabetes, multiple sclerosis; and it might cause an additional safety issue [105]. Consequence of over-activation of the immune system by probiotics in hosts with immune dysfunctions, such as individuals genetically predisposed to autoimmunity, has raised some concerns too. With respect to the association between bacterial antigens and autoimmune responses and the adjuvant activity of lactic acid bacteria strains, the involvement of lactic acid bacteria in the pathogenesis of some models of autoimmunity in experimental animals and possibly in humans has been suggested [109]. Thus, from a safety point of view, the potential of probiotic bacteria (especially the immunostimulatory strains), to induce destructive inflammation or autoimmunity needs to be investigated. For instance, it has been experimentally demonstrated that Lctbs casei cell wall components (given intraperitoneally) are able to induce cardioangitis (an autoimmunity-associated heart disease) in mice [110].
