Enteral Nutrition in Crohn Disease: More Than Just Calories

It is now more than 30 years since a group of surgeons reported on the benefits of enteral nutrition (EN) in adults with Crohn disease (1). Patients awaiting surgery for treatment-resistant disease experienced disease remission while receiving an exclusive enteral amino acid formula that was being used to optimize their preoperative nutritional status. Ever since, researchers have been trying to tease out the mechanisms underlying this surprising observation. Although the mechanisms remain poorly understood today, clinicians have continued to report on the benefits of EN, particularly in children with Crohn disease. Despite large studies in adults favoring the use of steroids for the induction of remission in active Crohn disease (2), adults complying with a course of EN show it to be as effective as steroids (3). A meta-analysis of all children randomized to these two therapies also suggests at least equal efficacy in achieving a clinical remission (4). Although the longer-term effects of inducing a remission with EN have yet to be shown to be superior to a remission induced by corticosteroids, there remains no doubt about the short-term advantages of EN on growth, nutrition, and its excellent adverse effects profile. Neither therapy has ever been shown to be an effective maintenance treatment. Therefore, the heated debate that continues around which therapy should be used for induction of disease remission, particularly at diagnosis, seems rather trivial, given the importance of effective long-term disease control. However, with the currently available data, the choice of ideal therapy for a child with newly diagnosed Crohn disease appears remarkably straightforward. If both therapies have a similar efficacy, but one has beneficial effects on growth, leads to a rapid nutritional restitution, and has fewer adverse effects, most clinicians, and children, would have little difficulty in making that choice. Alongside the ongoing clinical debate, studies on the possible mechanisms of action have continued. While several different hypotheses have been invoked to explain the efficacy of EN, it is only more recently that certain aspects of these have become clearer. MECHANISMS OF ENTERAL NUTRITION Recent evidence is increasingly suggesting a central role for the gut flora in maintaining chronic intestinal inflammation in Crohn disease (5). Whether enteric flora is altered during and after therapy with exclusive EN has not been demonstrated in humans. However, it has been shown that feeding an exclusive amino-acid–based formula to germ-free T-cell receptor α-chain deficient mice (TCR α−/−) that spontaneously develop a Th-2 dependent colitis, prevents gut inflammation (6). Certain Bacteroides species are suppressed in the EN-fed mice, whereas most other bacterial species remain similar in both EN-fed and control-fed groups. Rather surprisingly there was no significant difference in the number of aberrantly produced CD4+ ββT cells in either the gut or the peripheral lymphoid system. Interestingly, EN-fed mice lost their spontaneous Th-2–type response, with no interleukin (IL)-4, IL-6, or IL-10 detected in cells isolated from the colonic lamina propria. The presence of a specific bacterium, B. vulgatus, was found to preferentially induce Th-2–type cytokine release by CD4+ ββT cells, which also had a more limited Vβ repertoire compared with those in normally fed mice. Most importantly from a mechanistic point of view, instillation of B. vulgatus into the rectum of EN-fed mice led to a Th-2–type cytokine profile identical to that in the control mice that experience spontaneous colitis. This study highlights the fact that even a single bacterium may induce gut inflammation in the genetically predisposed. Thus, it seems highly likely that specific components of the gut flora are responsible for the inflammatory response in predisposed individuals. Evidence seems to suggest that the addition of certain probiotic species has a protective effect in Crohn disease (7,8), implying that an imbalance in normal gut flora may be an important determinant of inflammation. Some progress has been made on the mechanisms by which enteral nutrition exerts its influence on the mucosal immune system. Several groups have been exploring the hypothesis that changes in luminal content, i.e., bacterial and/or dietary, can alter the expression of genes within the epithelium, which in turn then affect change within the mucosal immune system. Bacterial breakdown products such as butyrate may alter transcription of various genes by inhibition of the enzyme histone deacetylase (9). In the presence of other pro-inflammatory cytokines, butyrate also upregulates the important metalloprotease stromelysin-1, already directly implicated in the disruption of the extracellular matrix in active IBD (10,11). Therefore butyrate, a beneficial and usually harmless bacterial breakdown product and the preferred energy substrate of colonocytes, may in fact contribute to gut damage in conditions in which an imbalance in pro- and anti-inflammatory cytokines pre-exists. IL-8 and monocyte–chemoattractant–protein-1 are chemokines responsible for neutrophil attraction and are upregulated in active IBD (12,13). Both may be altered by dietary factors (9). Thus, there is evidence emerging that far from providing a simple protective barrier, the epithelium, in concert with certain cytokines, can itself mediate a chronic inflammatory state (14). The initial concept of “gut rest” was based on the notion that preventing any dietary antigens reaching the terminal ileum would lead to mucosal healing. The early success of elemental feeds appeared to support this theory; however, later studies have not substantiated it. Data from both adult and pediatric studies have shown no significant benefit to using an elemental formula over a whole protein/polymeric formula in achieving a clinical remission (2). This confirms that dietary antigen exclusion alone, so-called gut rest, is not the primary mechanism. In fact, the greater palatability and significantly lower cost of a whole protein formula make polymeric diets the optimal choice. The polymeric formula used in the study by Bannerjee et al. was the first reported to contain the anti-inflammatory cytokine TGFβ2 in greater concentration than other formulae (15). However, recently Fell et al. (16) documented the dramatic increase in ileal TGFβ1 mRNA after 8 weeks of therapy. Thus, the tempting hypothesis that the mucosal TGFβ is directly upregulated by a formula is less likely. Although TGFβ2 may still be having some direct anti-inflammatory effects on the gut mucosa, the increase in mucosal TGFβ1 may simply be an epi-phenomenon of a healed mucosa. Controlled studies would be necessary to specifically identify this. Direct evidence comparing the biochemical markers of inflammation with the physical indices of nutrition in children with active Crohn disease during their early response to EN is available in the study of Bannerjee et al. in this edition of the Journal. The authors are to be congratulated on this carefully conducted study, in which they use a clinical protocol to try and advance our understanding of the mechanisms of EN. Despite being an uncontrolled cohort study, it does allow important observations to be made on the earliest responses to an exclusive enteral diet. Previous data have shown a dramatic decrease in C-reactive protein by 14 days (16); however, this has never been accompanied by formal nutritional measurement. Assessment of the latter is clearly limited by the very nature of the anthropometric techniques. Although the best validated physical techniques are used, their sensitivity to nutritional change inevitably lags well behind changes at a biochemical and cellular level. Nonetheless, the close monitoring of the pro-inflammatory cytokine IL-6 and the erythrocyte sedimentation rate reveals the striking speed with which these return to normal values, intuitively suggesting factors other than simple caloric intake are likely to be at work. IL-6 is of particular importance in Crohn disease, where in addition to making T cells more resistant to apoptosis (17), it may be implicated in growth delay by inhibiting osteoblast differentiation (18) and in causing the osteopenia that results from unchecked disease activity (19). After promising results in animal models, trials of anti–IL-6 monoclonal antibodies are under way in adults with Crohn disease. IL-6 is also known to interfere directly with the insulin-growth-factor-1 (IGF-1) axis in juvenile rheumatoid arthritis (20). This is best demonstrated in the transgenic mouse model, where blockade of overexpressed IL-6, associated with low levels of IGF-1, leads to partial reversal of growth failure (21). IGF-1 levels are also known to be suppressed by active Crohn disease but increase during remission (22,23). Thomas et al. (22) confirmed this increase occurred after both an elemental feed and after corticosteroids; although, unlike corticosteroids, exclusive EN promoted short-term linear growth. Beattie et al. (23) documented significant increases in both IGF-1 and IGFBP-3 as long as 2 months after the end of the course of exclusive EN, and although some EN continues during food reintroduction, Bannerjee et al. confirm this effect with stable results 2 weeks after stopping exclusive EN. Without a “control” group of children receiving an ad libitum diet of similar caloric density, it is of course impossible to definitively rule out nutritional restitution per se as the primary benefit of EN. More than 20 years ago, Kirschner et al. (24) reported on seven children whose growth velocity improved significantly with only an additional 1,000 kcal/d, whereas some time later the same group showed that low levels of somatomedin-C (IGF-1) normalized with calorie supplements alone and even led to better growth (25). Growth velocity is a sensitive marker of long-term disease activity, with an effective clinical remission leading to optimal growth rates. Bannerjee et al. have provided hard evidence on something that was previously assumed but never documented. As such, they have significantly contributed to the available evidence on EN in Crohn disease. With the increase in IGF-1 occurring within 7 days of commencing exclusive EN, accompanied by the simultaneous reduction in IL-6, C-reactive protein, erythrocyte sedimentation rate, and Paediatric Crohn's Disease Activity Index, one needs little additional evidence to support this form of therapy as an ideal first-line treatment in children with active Crohn disease.