Mesalazine in the initial management of severely acutely malnourished children with environmental enteric dysfunction: a pilot randomized controlled trial

Environmental enteric dysfunction (EED, previously called environmental or tropical enteropathy) is an acquired syndrome of reduced intestinal barrier and absorptive function that is common among children living in settings where food security, safe water and facilities for hygienic waste disposal are absent or erratic [1]-[3]. It reflects an epidemiologic concept that lacks formal criteria for case definition but is considered to play a central role in the pathogenesis of stunting in early life. It may be a critical obstacle underlying the strikingly poor impact of even intensive nutritional and hygiene interventions on growth faltering [4]. Undernutrition (including acute malnutrition and stunting) remains the most important risk-factor for childhood mortality, responsible for an estimated 45% of all deaths among children under five years old worldwide, alongside huge burdens of short-term infectious morbidity and long-term loss of growth and developmental potential [5]-[10].

EED has features of an inflammatory enteropathy: histopathological changes of small intestinal villous atrophy and crypt hyperplasia are accompanied by lymphocytic infiltration of the lamina propria [11],[12]. Linear growth failure is associated with detectable products of Th1 signaling and neutrophil extravasation in stool [13],[14]. Systemic inflammatory activation results from breakdown in mucosal barrier function with consequent translocation of exogenous antigens alongside luminal bacteria and their products [5],[15]. It has been suggested that gastrointestinal luminal contents drive this inflammation via specific enteric pathogens, a global increase in pathogen burden (due to high feco-oral transmission) or altered microbiota [4],[16]-[18]. However, in all other inflammatory intestinal diseases of childhood, aberrant or inappropriate host response plays a critical role in the pathogenesis, even where an intraluminal trigger is evident [19]. In celiac disease, for example, although antigen-specific T cell activation is the hallmark, bystander induction of innate and other immune components contributes to tissue damage and sustained pathology [20].

A key unanswered question is whether intestinal inflammation in EED is an appropriate, adaptive response to excessive pathogen exposure that mitigates the impact of acute and chronic enteric infections or if it is maladaptive, with host-driven immunopathology contributing to mucosal barrier breakdown and growth failure. A maladaptive response would be a rational target for treatment with immunomodulatory or immunosuppressive agents such as the aminosalicylate mesalazine, which is commonly used to induce remission of mild to moderate pediatric ulcerative colitis and for maintenance of remission of ulcerative colitis and Crohn’s disease [21]-[23].

We conducted a pilot clinical trial with a primary objective of investigating the safety and acceptability of mesalazine as an adjunctive treatment for children with severe acute malnutrition (SAM). Acutely malnourished children are at very high risk of ongoing growth failure despite nutritional rehabilitation and their degree of intestinal barrier dysfunction has been correlated with outcome [24]. The trial was designed with intensive follow-up in order to detect any indication that the strategy was interfering with a beneficial, adaptive response. Secondary objectives were to collect hypothesis-generating data relating to inflammatory activation and growth.

This was a single-center double-blind randomized placebo-controlled trial on the use of mesalazine alongside standard medical and nutritional care in severely acutely malnourished children. Randomization was balanced 1:1.

From June to September 2013, 133 children with SAM and stunting were screened for eligibility. Forty-four children completed screening procedures and were eligible, all of whom were enrolled (Figure 1). The arms were well balanced in terms of clinical, anthropometric and demographic characteristics at baseline (Table 1).

This is the first controlled trial of a directly immunomodulatory agent in children with SAM. Despite being eligible for outpatient management with `uncomplicated’ SAM, intestinal and systemic inflammatory activation was very common. Baseline levels of fecal calprotectin were higher than has been reported from healthy control populations in high-income countries and sub-Saharan Africa, approaching those seen in active IBD [33],[34]. Treatment with mesalazine was safe. The frequency of acute illness episodes was similar in both arms, and there was no excess gastrointestinal morbidity to suggest a disadvantageous effect on gut mucosal homeostasis. The trial was not intended to be powered to detect rates of resolution of edema, recovery of acute malnutrition or linear growth. Individual growth trajectories revealed greater MUAC growth in the placebo arm to day 56 but because edema can alter MUAC and most of the children in this study were edematous at enrollment, the clinical significance is unclear.

In pediatric IBD, linear growth impairment correlates with cytokine-mediated depression of the growth hormone - IGF-1 axis, and a similar effect was evident among participants in this trial [35]-[37]. It is, therefore, encouraging that the mesalazine arm achieved modest reductions in multiple clinically-relevant and gut-specific biomarkers compared to placebo after 28 days treatment. If EED represented an essential level of inflammatory activation and immune-surveillance in the face of a high pathogen burden, we would have expected the opposite result: that mesalazine would have been associated with a loss of control at the gut mucosal surface, increased levels of bacterial translocation and systemic immune activation, alongside development or worsening of gastrointestinal symptoms. Instead, we have demonstrated that pharmacologically-mediated reduction in enteric inflammation is well tolerated, thereby providing the first clear evidence that EED is likely to be at least partly maladaptive and, consequently, a potential direct target for therapy. New therapies are urgently required because trials of nutritional support, hygiene-based interventions, probiotics and empiric treatment for small intestinal bacterial overgrowth or specific pathogens have failed to show clinically significant improvements in linear growth or EED activity in children [4],[38]-[40]. We suggest that failure to control host-driven inflammatory activation may have been a barrier to efficacy of these interventions.

An important limitation of our study is that enteric inflammation is inferred on the basis of indirect blood or fecal markers. A recent expert working group concluded on the basis of the myriad ethical and logistical challenges inherent in performing gastrointestinal endoscopy on young children, that `definition of environmental enteric dysfunction will need to rely on biomarkers alone, without biopsy data to connect functional and structural changes’ [1]. Importantly, the high levels of fecal calprotectin and systemic inflammatory activation demonstrated in this study are more consistent with inflammation in the colon than in the small intestine [41]-[43]. In view of the fact that sigmoidoscopy can be achieved far more easily than full colonoscopy and without the need for bowel preparation or deep sedation, if colonic inflammation were a major component of EED, it might be substantially more amenable to serial tissue-based monitoring in clinical trial settings. Biopsy studies of colonic architecture in children at risk of EED have not been reported to our knowledge.

The enrollment of children who were severely acutely malnourished as well as stunted was ethically appropriate because children with SAM are at the highest risk of continued growth disturbance, illness and death, meaning that they have the most to gain from any potential new intervention. However, even though these results suggest a maladaptive inflammatory enteropathy, it may be that ensuring optimal capacity to respond to a new pathogen challenge takes precedence over optimizing growth in the context of SAM, where vulnerability to major infection is intense. Targeting those with moderate acute malnutrition or non-acutely malnourished children would be likely to increase the chances of detecting any nascent IGF-1-mediated linear growth benefit and presents the most plausible scenario under which such interventions could be used in the field. In this regard, a limitation of the study is that the inflammatory enteropathy found in kwashiorkor (which was present in 73% of those enrolled) may be qualitatively different than that present in the context of other forms of acute malnutrition, and to the form of EED that appears to be prevalent with stunting. Results of the current trial should not be considered generalizable either to children with stunting but without acute malnutrition or to populations of moderately or severely malnourished children without high prevalence of kwashiorkor, and future studies in such groups will need to take a similarly cautious and thorough approach as we have here [12]. That said, such studies should consider the use of more intensive treatment schedules because although mesalazine is a good agent for maintenance of remission in IBD, it is less effective in induction. Difficulty in diagnosing tuberculosis in acutely malnourished children may preclude the use of systemic immunosuppressants, but longer courses or higher doses of mesalazine could reasonably be trialed. Furthermore, although EED occurs in all age groups, irreversible stunting mainly occurs in the first two years of life, and may even be apparent at birth [15],[44]. Treatment prior to the development of clear evidence of growth failure may be optimal, but this requires the development and refinement of point-of-care biomarkers for EED. Fecal calprotectin may be a suitable marker: point-of-care tests already exist, and evaluation against dual sugar absorption tests (currently the most well-established biomarkers of EED) should be a research priority. The possibility of targeting short-term medical treatments towards this key early window of vulnerability to linear growth failure, which is also the peak period for development of acute malnutrition and major infectious diseases, means that clinical benefit for individual children might be possible even if modification of environmental determinants was minimal. This reflects social and political realities in many of the settings where EED is endemic.

Our results have highlighted that elevation of liver enzymes is reasonably common in uncomplicated SAM (8% of those screened) and that current management strategies may be associated with a transient exacerbation. Fatty infiltration of the liver is a recognized feature of kwashiorkor and could be exacerbated by a sudden switch towards a plentiful and very high-fat diet [45]. That RUTF is safe and effective for use in the community is partly a factor of its high lipid content (which restricts bacterial growth) and it is likely that any risk associated with modest increase in liver enzymes is outweighed by the proven benefits of community-based care. However, there may be a subset of patients in whom this presents a particular problem, especially as RUTF is starting to be used more commonly in the more-unwell group of hospitalized severely acutely malnourished children.

The participants involved in this trial represented a uniquely vulnerable population. Detection of a severely acutely malnourished child during community screening was frequently an indicator of multiple social, nutritional and medical problems affecting the whole family. The provision of appropriate ancillary support was an essential component of our trial methodologies, made possible because of the comprehensive integration of medical and social services and long history of community engagement in Mathare by `German Doctors’, which made this setting a particularly appropriate one in which to conduct the trial. Where required, parents and siblings were provided with a daily food ration, free medical care and access to social support and advice. Study visits took place at participants’ home to minimize the opportunity cost of attending the clinic and community health workers performed regular home visits in order to provide support and assess for additional needs. The same benefits were afforded to all children screened for enrollment regardless of their eligibility.

In summary, we have reported the first controlled trial directly targeting the host inflammatory response in children with SAM. The trial is conceptually novel in considering EED as a maladaptive host response to environmental challenge and has provided pilot evidence of safety. Larger trials will be needed to assess efficacy but our finding of concordance of a range of inflammatory markers encourages the design of such studies.

KDJJ, BH-K, AMRL, CWM, MO, GF, SHM, PBS, JOW and JAB designed the study. KDJJ, BH-K, CWM, JM, HSN and MT conducted the clinical aspects of the trial. KDJJ, MMM and DO conducted the laboratory aspects of the trial. Data management was provided by MN and GF. KDJJ, MN, GF and JAB analyzed the data. KDJJ wrote the first draft of the manuscript. All authors contributed to subsequent drafts. All authors read and approved the final manuscript.