Diarrhea during early infancy (children less than 2 years of age) is relatively common, generally mild, and often self-limited. Most frequently, this form of diarrhea in infancy is of allergic or infectious origin, and usually not associated with any significant long-term sequelae. In contrast, anatomical disorders, such as gastroschisis, necrotizing enterocolitis (NEC), or acute volvulus can cause short-bowel syndrome (SBS) and long-lasting diarrhea after surgical resection.1 Another class of persistent and severe diarrhea presenting in the first weeks of life results from monogenic disorders and is termed congenital diarrhea and enteropathies (CODEs). CODEs are typically associated with feeding intolerance and malabsorption. Both CODE and SBS require significant dietary and therapeutic interventions, including specialized formulas or parenteral nutrition (PN) to sustain appropriate growth, electrolyte, and nutrient balance.
The availability of genetic diagnostic testing before 2008 was limited, and infants with many of the non-surgical severe diarrheal disorders often carried the catch-all diagnosis of “chronic diarrhea of unknown etiology.” These infants followed a diagnostic odyssey and clinical course associated with high levels of morbidity and prolonged expensive hospitalizations. In recent years, our understanding of the underlying genetic basis of these disorders, as with other rare Mendelian diseases, has been revolutionized by the availability of next-generation sequencing. These technologies have enabled the elucidation of the genetic basis of an increasing number of monogenic disorders causing CODEs, and most so far involve intestinal epithelial and/or immune function. However, these disorders are rare and clear diagnostic and therapeutic pathways for the care of these patients have not been established.
Here, we outline a prioritized approach for the diagnosis and evaluation of diarrhea in infants. The approach recognizes recent advances in next-generation sequencing, stem cell biology, gene editing, and other aspects of the rapidly evolving field of precision medicine.
Previous classifications have often divided diarrhea into osmotic and secretory, but these terms can be misleading. Therefore, some new terms are proposed.9
The term osmotic diarrhea has been used traditionally to refer to diarrhea resulting from unabsorbed solutes or nutrients; however, all diarrhea involves osmotic forces. Therefore, we prefer to use the more precise term, diet-induced diarrhea. Diet-induced diarrhea is characterized by an elevated stool osmotic gap (>100 mOsm). Examples include glucose or disaccharide malabsorption.
The term secretory diarrhea is also imprecise. The term describes the underlying pathophysiology of the diarrheas caused by active ion secretion into the intestine, but it does not describe the watery high-salt diarrheas caused by defects in intestinal sodium absorption (e.g., as seen in the congenital sodium diarrheas and in some viral infections). Neither can the term secretory be used to describe all the diarrheas with a low stool osmotic gap (<50 mOsm) (Table 1), because a low stool osmotic gap typically results from a combination of enhanced anion-driven fluid secretion and loss of Na+-driven fluid absorption. We prefer to use the term electrolyte-transport-related diarrhea. Examples include congenital chloride or sodium diarrheas.
Lastly, diarrhea that is obviously neither secretory nor osmotic, or has an intermediate stool osmotic gap (50–100 mOsm), has been referred to as “mixed.” Intermediate values for the stool osmotic gap occur frequently and are generally caused by a combination of diet-induced diarrhea and electrolyte transport-related diarrhea resulting from different dietary intakes at the time of testing.
Diarrhea presenting early or immediately postnatally in term or premature infants should prompt evaluation for congenital enteropathies, NEC, or anatomical abnormalities (Figure 1). For less-severe diarrhea presenting later in infancy, the initial workup should focus on investigating common acquired etiologies, such as infections, cow’s milk protein allergy, or food protein-induced enterocolitis syndrome (Figure 1).
In both high-income and middle-/low-income countries, viral pathogens are the most likely agents responsible for infectious diarrhea in infants. Rotavirus, cytomegalovirus, adenovirus, and norovirus should all be considered. Less common in high-income countries but a significant factor globally are enteric bacterial pathogens, such as Salmonella enterica, Shigella spp, Campylobacter jejuni, and pathogenic Escherichia coli.2,10 Persistent viral diarrheas, or diarrhea after administration of the rotavirus vaccine, may be the first sign of primary immunodeficiency or related to autoimmune enteropathies, such as immunodysregulation polyendocrinopathy, enteropathy X-linked.11 Bloody diarrhea with no evidence of infection should be followed by either a change to extensively hydrolyzed formula and/or a maternal dairy exclusion diet for a potential diagnosis of cow’s milk protein colitis.12 Food protein-induced enterocolitis syndrome, a primarily T-cell-mediated inflammatory disorder triggered by a variety of food antigens, can present in infancy with acute-onset diarrhea, usually after dietary introduction of solid foods.13,14
Evaluation of severe early-onset diarrhea by radiographic small-bowel follow-through studies should be used to identify intestinal malrotation with intermittent volvulus, or congenital short gut. Hirschsprung’s disease can rarely present with diarrhea and should also be considered. Obstructive symptoms without an identifiable lead point are typically associated with chronic intestinal pseudo-obstruction. NEC can present with watery or bloody diarrhea, usually in association with prematurity, abdominal distension, feeding intolerance, and temperature instability. Initial evaluation and management involves serial abdominal films to investigate the presence of bowel wall gas, cessation of enteral feeding, and antibiotic administration.6,15
Increased suspicion for CODE can be gleaned from the history and clinical course. Important information to gather includes prenatal history and testing; age at onset of symptoms; nature of symptoms; extraintestinal manifestations; nutrition and diet history; and a full family history, including any evidence of consanguinity and ethnicity (Figure 1). Specific populations have relatively high incidence rates of congenital enteropathies, such as the Finnish, Ashkenazi Jews, Navajo Native American, and those originating from the Arab Gulf regions.16–19
In patients undergoing evaluation after 1 month of age, the age of onset of the diarrhea is a key historical detail, with very-early-onset increasing the chance of CODE. Infectious and immune-related conditions generally have a symptom-free postnatal period of at least a few weeks before clinical symptoms are apparent. It should be noted that, in some conditions, the volume of diarrhea may be so severe that less-experienced parents and health care providers may confuse diarrhea for urine, delaying recognition and appropriate consultation.
Initial testing should include a complete blood count, serum electrolytes, inflammatory markers, liver function tests, immunoglobulin levels, a lipid panel (triglycerides, cholesterol), fat-soluble vitamins, a coagulation profile, and zinc level. Additional immunologic investigation should be considered if there is suspicion of immune dysfunction, such as specific T-and B-cell subset analysis.
Accurate stool testing is a key component of the diagnostic evaluation for congenital diarrheas. Collection of stool samples can be challenging due to the presence of mixed urine and stool in the diapers of neonates, and by rapid loss of stool water content due to absorption into the diaper material. Use of a urine catheter for a few days may allow for accurate stool sampling and volume measurement. Ideally, fresh stool samples obtained immediately after excretion should be tested (Table 1). A key component in the evaluation of stool testing is an exact quantification of dietary intake at the time of testing.
The workup for CODE is initiated after the exclusion of acquired diarrhea. As an initial diagnostic evaluation, it is helpful to grossly characterize the stool appearance in suspected CODE (Figures 1 and Figure 2). Although sometimes difficult, most stools can be broadly differentiated into 3 main categories: watery, fatty, and bloody stools, and this allows for clear prioritization of initial testing. Watery stool is characterized by a high liquid content—often with very little form—that can be mistaken for urine. Fatty stools are usually foul-smelling, can have a bulky or “fluffy” appearance, are pale in color, and/or are spot fecal fat-positive. Bloody stools contain gross blood mixed in with stools. The presence of large volumes of bright red blood or melena should trigger evaluation of vascular or anatomic gastrointestinal bleeding as well as infection.
It should be noted that genetic testing (see section Genomic Testing) can occur in parallel or early in the diagnostic algorithm, especially if there are clear factors to suspect a monogenic diarrheal disease, such as significant consanguinity; a family history of gastrointestinal disease in infancy; and clinical indicators, such as diarrhea severity and neonatal onset.
After documentation of diarrheal output with normal feeding, evaluation should include a no-feeding (nil per os) trial of at least 24 hours, with assessment of stool output and electrolytes (Figure 2, top). If diarrheal volume is unchanged or is minimally changed after fasting, this points to electrolyte-transport-related diarrhea. Significant improvement in diarrheal output after termination of enteral feeds points to diet-induced diarrhea. Subsequent evaluation should aim to elucidate whether a specific nutrient is malabsorbed or if the patient has a generalized malabsorptive diarrhea. A feeding trial with carbohydrate-free (Ross CarbohydrateFree; Abbott, Macquarie Park, NSW, Australia) or fructose-based formula (Galactomin-19; Nutricia, Wiltshire, UK) leading to significant improvement, along with reduced stool pH and elevated stool-reducing substances with carbohydrate-containing formula, supports the diagnosis of carbohydrate malabsorption. Elucidating the specific abnormality of carbohydrate assimilation can be performed with specific monosaccharide (glucose and fructose), disaccharide (sucrose, lactose, and maltose) dietary challenges and assessed by changes in stool volume and/or breath hydrogen testing.20 Disaccharidase activity assays for lactase, sucrase, maltase, and palatinase performed on proximal small bowel biopsies may be helpful to diagnose disaccharidase defi-ciency.21 However, these enzymatic assays are often unreliable due to poor sampling or in the setting of inflammation or villus atrophy due to secondary disaccharidase deficiency.
If there is no clear evidence of a selective carbohydrate malabsorption, diagnostic evaluation should begin with esophagoduodenoscopy and flexible sigmoidoscopy with biopsies for histologic analysis; if not contraindicated by the clinical status of the infant.22 Biopsies should include samples for both routine histology and electron microscopy, as well as for measuring mucosal disaccharidase activity.
Early endoscopy and biopsy differentiating normal from an abnormal villus to crypt ratio,23 and/or an inflammatory predominance allows for significant streamlining and prioritization of the evaluation, planning for dietary interventions, and initiation of genetic testing.
Parallel assessment for a possible protein-losing enteropathy is important, as it may be indicative of a compromised epithelial barrier, suggesting autoimmune enteropathy, or newly described monogenic disorders, such as DGAT1 and CD55 deficiency. The presence of elevated levels of α1-antitrypsin in the stool and low serum albumin, IgG, and lymphopenia are consistent with protein-losing enteropathy.24,25
Evaluation of fatty diarrhea is by spot fecal fat testing, including neutral and split fat and, if available, quantitatively by 72-hour stool fat collection (Figure 2). Stool elas-tase is a useful initial test, as it can help distinguish between conditions resulting from pancreatic insufficiency and those caused by intestinal fat malabsorption, although fecal elas-tase can often be falsely low (false positive) with highvolume diarrhea. The presence of fat-laden enterocytes in histologic sections along with serum lipid abnormalities can point to disorders of fat transport and metabolism, such as chylomicron retention disease and abetalipoproteinemia. Pancreatic insufficiency is confirmed by the responsiveness of diarrheal symptoms to enzyme replacement therapy.26
The presence of gross blood implies significant colitis and further evaluation should include stool inflammatory markers and endoscopy (Figure 2). The presence of inflammatory changes in histologic sections should precipitate further investigation of infantile very-early-onset inflammatory bowel disease, autoimmune enteropathy, or primary immunodeficiency. It should be noted that not all forms of immune-dysregulation-related enteropathies are associated with bloody stools.27
Initial evaluation of H&E-stained sections should focus on overall intestinal epithelial architecture, namely villus to crypt ratio, abundance of the epithelial cell type and structure, and the immune cell composition in the lamina propria and intraepithelial compartments.28
Immunohistochemical staining of differentiated cell populations may be helpful to confirm H&E findings and/or a specific diagnosis (Figure 3, Supplementary Table 1). Staining may be limited by the small amount of available tissue acquired during endoscopy. The assessment includes staining for enteroendocrine (chromogranin/synaptophy-sin), Paneth (lysozyme), and goblet cells (periodic acid-Schiff). Identification of proliferating cells can be done using Ki67, proliferating cell nuclear antigen, or pHisH3. Immune cell types can be initially assessed using specific surface markers for B cells (CD20), T cells (CD3/CD4/CD8), macrophages (PU.1/CD68/F4/80), and plasma cells (CD138); however, identification of specific immune cell subsets requires more extensive and specialized staining.
If histologic assessment suggests abnormal epithelial architecture, initial immunohistochemical staining in all cases should include CD10/villin (microvillus inclusion disease [MVID]), periodic acid-Schiff (DGAT1), MOC31 (CTE), and frozen-section staining with Oil Red O if lipid trafficking disorders are under consideration.29,30
Further evaluation includes electron microscopy to assess the presence and relative size and location of the microvilli and to identify intracellular microvillus inclusions or abnormal vesicular structures suggestive of disorders of intracellular trafficking.
Immunolocalization of specific transporter, structural, or intracellular proteins, such as (DGAT1, PCSK1, and others) may provide increased diagnostic validation. Figure 3 depicts the typical histology of MVID, CTE (epithelial cell adhesion molecule [EPCAM]), abetalipoproteinemia, and autoimmune enteropathy (immune-dysregulation-related enteropathies). All studies, however, should be performed in concert with genetic analyses.
Advances in next-generation sequencing technologies promise to shorten the diagnostic odyssey for many CODE patients. Although the clinical diagnostic algorithm provides a framework for evaluation and prioritized testing, in many cases where the diagnosis of CODE is highly suspected but the specific etiology is not identified or requires confirmation, either targeted genetic testing (Sanger sequencing) or whole-exome sequencing can identify the genetic cause and allow for appropriate early treatment.
In selective populations with a high prevalence of known specific genomic variants, or when the diagnostic evaluation is strongly suggestive of a specific disorder, such as the characteristic epithelial tufts seen on biopsy in EPCAM mutations, Sanger sequencing should be considered for rapid diagnosis and treatment. For example, there are a number of relatively common founder CODE gene mutations, including Mexican/Arab: EPCAM (c.491+1G>A and c.498insC),18,31 Ashkenazi Jews: DGAT1 (IVS8+2 T>C);32 Navajo: MYO5B (p.Pro660Leu),33 and Finns: SLC26A3 (p.Val317del).34
In cases of a suspected CODE, where the diagnosis based on clinical evaluation is unclear, it is now standard of care to perform whole-exome sequencing to identify a possible causative genetic mutation.35–38 It should always be considered, however, that whole-exome sequencing may not detect genetic defects in genes with poor coverage, large insertions and deletions, and mutations in regulatory and splice or intronic regions. Therefore, in certain cases with a high likelihood of a monogenic disorder, whole-genome sequencing, or RNA sequencing should also be considered. Microarray comparative genomic hybridization is a rapid and frequently used method to assess significant copy number variation and changes in whole chromosomes, and discovery of large deletions (>200 kb) and duplications.39
SBS occurs more frequently in premature neonates (25/ 100,000 live births) and is associated with either anatomical intestinal defects or NEC. Many of these neonates will have malabsorption early on and be dependent on PN to sustain normal growth and development.1,40
Disorders such as gastroschisis41 may result in diarrhea by various mechanisms, including impaired motility,42 surgical formation of ostomies that shorten bowel length, or bacterial overgrowth. Patients with gastroschisis can also acquire SBS as a result of volvulus that can occur before or after birth.
Intestinal atresia results from defects in intestinal development early in gestation due to vascular anomalies or inherited defects in luminal development.43 Atresia can be single, with reasonable complement of distal bowel, or there can be multiple atresias, with very limited bowel length. Next-generation sequencing has resulted in the identification of several severe inherited forms of intestinal atresia, including TTC7A and RFX6 (discussed Disorders of Epithelial Trafficking and Polarity).44,45
NEC is predominantly an illness that affects preterm infants who are initiating enteral feeding. The pathogenesis is not clearly defined, but includes vascular abnormalities, bacterial dysbiosis, and abnormal immune responses, with the final common pathway being intestinal ischemia.7 This can vary in severity, but uncommonly can include gangrene of nearly the entire small and large intestines. Diagnosis currently relies on careful clinical observation and plain films. Early biomarkers have been investigated but are not yet used in routine clinical practice.46 As with the aforementioned conditions, diarrhea and nonselective malabsorption result from SBS.
Monogenic diarrheal disorders can be broadly classified into 5 major categories (Figure 4, Table 2) reflective of a common pathophysiology, although there remains overlap among a number of these categories.47 Epithelial cell defects are the hallmark of the first 4 CODE categories and range from defects in epithelial transporters, enzymes, and metabolism to defects in epithelial trafficking and polarity and enteroendocrine cell dysfunction. The clinical presentation is almost always within the first several months of life and is associated with high-volume watery diarrhea.
The fifth category encompasses monogenic disorders that cause dysfunction of the immune system, which result in a wide spectrum of both intestinal and extraintestinal manifestations. This broad array of monogenic entities includes genetic defects also classified as infantile-onset inflammatory bowel disease (children less than 2 years of age), autoimmune enteropathy, or primary immunodeficiency.48 For a general description for the more common specific disorders, please see the CODE disorders page.
Alterations in epithelial transport proteins represent some of the most prevalent and well-known of the congenital diarrheas (Figure 4A). These encompass variants of pure electrolyte transporters, such as the Cl/HCO3–exchanger DRA (SLC26A3), which result in congenital chloride diarrhea34; the Na+/H+exchanger NHE3 (SLC9A3), which results in congenital sodium diarrhea; and electrolyte-nutrient co-transporters, such as SGLT1 (SLC5A1),49 which result in loss of both sodium and glucose absorption. Alterations in regulatory proteins, such as the guanylin receptor GC-C (GUCY2C)50 or other transporters, such as those mediating intestinal sodium-coupled bile salt re-uptake (SLC10A2),51 or primary excessive bile acid production, can induce secondary electrolyte transport defects in epithelial cells and excessive fluid loss.52 These disorders generally exhibit a structurally intact epithelium and brush border with a normal villus to crypt ratio. Excessive stool chloride levels are found in congenital chloride diarrhea (SLC26A3)53 whereas high stool sodium levels are found in congenital sodium diarrhea (SLC9A3, GUCY2C, SPINT2). Both of these disorders usually present immediately at birth with polyhydramnios often present in utero. In contrast, glucose-galactose malabsorption exhibits a diet-induced dehydrating diarrhea initially evident after initiation of feeding that ceases after institution of a glucose-galactose free diet.
Alterations in a number of important enzymes involved in both nutrient absorption as well as epithelial cell metabolism result in severe diarrhea (Figure 4B). Defects in brush-border enzymes involved in carbohydrate digestion, such as lactase and sucrose-isomaltase, result in a diet-induced diarrhea with onset after intake of carbohydrate-containing formula or food. These include the relatively common lactose intolerance caused by reduced function of LPH,54,55 which also may be acquired after gastroenteritis or due to prematurity, vs the very rare cases of total loss-of-function mutations in LPH56 or in the gene for sucrase-isomaltase (SI).57 Bi-allelic mutations of SI result in a loss of sucrose or isomaltase, or both enzyme activities, and will result in diarrhea on a diet containing sucrose and/or starch, isomaltose, and maltose. Disorders of these brush-border enzymes exhibit grossly normal intestinal histology after biopsy.
A more recently described CODE characterized by an electrolyte transport-related diarrhea, emesis, protein-losing enteropathy, and growth failure induced by enteral intake of lipids was found to be due to a loss-of-function mutation in DGAT1, which is involved in cellular triglyceride formation.32 Initial studies have indicated some loss in brush-border microvillus structure in patient biopsies, although it is unclear whether this persists in the absence of enteral lipids.
Other disorders of fat transport or metabolism result from mutations in proteins involved in fat absorption across the epithelium, such as microsomal triglyceride transfer protein resulting in abetalipoproteinmeia, apolipoprotein B resulting in hypobetalipoproteinemia, or chylomicron retention disease (SAR1B). These disorders classically show lipid-laden vacuoles within enterocytes (Figure 3D).
A number of disorders of epithelial trafficking and polarity lead to early-onset diarrhea, usually appearing in the first months of life. All disorders carry an autosomal recessive inheritance and have been described in various ethnic groups. The diagnosis is based on typical pathologic findings and/or extraintestinal manifestations, followed by confirmation of the diagnosis via genetic testing. The two most well-described disorders are MVID and CTE.
MVID results from mutations in the cytoskeletal motor protein Myosin 5b (MYO5B), which results in defective apical membrane recycling in intestinal epithelial cells.33,58 This results in the pathognomic structural abnormalities of the epithelial apical membrane, including loss of micro-villi59 leading to abnormal CD10 and villin staining of intestinal biopsies and intracellular microvillus inclusions seen in electron micrographs (Figure 3E-H).59 Patients present with profuse dehydrating diarrhea in the absence of enteral intake but worsened with feeding. A similar but milder phenotype to MYO5B mutations can be also found after loss of the trafficking protein syntaxin 3 (STX3).61
CTE results from loss of function in the epithelial signaling and adhesion protein EPCAM and causes a severe sodium-losing diarrhea that usually presents from birth to 3 months of life.31,62 The diarrhea in CTE does occur in the absence of feeding, however, it is generally much worse with enteral intake. The classical findings on biopsy include the presence of pathognomic surface epitheial “tufts” seen on H&E stains of biopsies, as well as the lack of EPCAM (MOC31) immunostaining (Figure 3I–L). “t
Other disorders involving epithelial structural defects include recently described mutations in the gene TTC7A,44 which leads to loss of the apicobasal polarity of the enter-ocyte, crypt-base apoptosis, crypt and villus atrophy and chronic inflammation, and TTC37 mutations,63 which result in mild to severe villus atrophy and variable inflammatory infiltrate on biopsy.
Disorders classified as enteric endocrinopathies result from either a loss of proper enteroendocrine cell (EEC) fate or generalized defects in processing of gut hormones. Collectively, these disorders result in a generalized mal-absorptive diarrhea that requires PN for the first several years of life, although diarrheal symptoms persist perhaps indefinitely. Each disorder is also associated with a unique set of systemic endocrinopathies that allow for anticipatory guidance of physicians and families alike.
The first disorder described involving enteroendocrine dysfunction resulted from bi-allelic loss-of-function mutations of Neurogenin3, a basic helix-loop-helix transcription factor, required for enteroendocrine and β-cell develop-ment.64,65 Patients present with a primarily diet-induced diarrhea that is not specific to any single nutrient and their intestinal biopsies reveal a normal crypt to villus ratio, with selective loss of all types of EECs.
Another EEC cell disorder was found to result from loss-of-function mutations of PCSK1, coding for prohormone convertase (PC1/3), a protease that is required for the biosynthetic processing of hormone precursors into their fully functional forms.66 Infants present with diarrhea that is phenotypically similar to NEUROG3 mutations but associated with a wider range of systemic endocrinopathies, including adrenal insufficiency, hypothyroidism, and diabetes insipidus-like picture.
Other EEC-related disorders include mutations in RFX6,67 a transcription factor that functions both up-and downstream of Neurogenin3 and mutations in ARX, a ho-meobox transcription factor that results in selective reduction of GLP-1 and cholecystokinin-expressing EECs.68,69
Recent genetics studies have rapidly increased the number of monogenic disorders, including FOXP3,70,71ICOS,72 IL10R,73 TRIM22,74 and ARPC1B,38 which cause dysregulation of the immune system and subsequently inflammation and enteropathy in the intestine. Immune-mediated intestinal disorders present with a wide variety of manifestations, but all have bloody or watery diarrhea, and are frequently associated with systemic disease and multi-organ involvement. Many of these disorders have been classified as infantile-onset inflammatory bowel disease, and have features characteristic of later-onset inflammatory bowel disease, such as bloody diarrhea, colitis, perianal disease, and ulceration, for example, IL10R.73,75Others have predominantly villus atrophy without ulceration and present with profuse watery diarrhea and malnutrition, for example, ICOS and FOXP3, and have also been termed autoimmune enteropathies. Most of these diseases are curable with bone marrow transplantation.
One of the disorders that presents with primarily watery diarrhea without bleeding or perianal disease is immuno-dysregulation polyendocrinopathy, enteropathy X-linked syndrome, which results from mutations in the FOXP3 gene.70,71 The intestinal pathologic features vary from complete villus atrophy with apoptosis in a graft-vs-host appearance, to loss of goblet and Paneth cells, with mild inflammation. Similarly, mutations in the gene encoding ICOS,72 a T-cell co-stimulatory protein, presents with watery predominantly diet-induced diarrhea starting after a few months of life with biopsy findings of villus atrophy, crypt apoptosis, and an inflammatory infiltrate (Figure 3A-C).
Whole-exome/genome sequencing has led to a plethora of new or previously unconfirmed gene mutations associated with CODE. After confirmation of a variant in a novel gene, the next critical step is to confirm a plausible biological mechanism through functional assessment of gene function. This requires careful phenotyping of patient’s cells, physiological assays of cellular function in both model systems and patient-derived tissue and cells. Hints at gene function may first be discerned from the clinical and morphologic data acquired during the initial stages of diagnosis and treatment, and can direct initial tissue and cellular phenotyping.
Initial phenotyping can be carried out through characterization of key epithelial structural, transport, and cellular trafficking proteins in patient-derived tissue sections. However, complete analysis of immunostaining is often limited by the small amount of paraffin-embedded biopsy tissue available. This characterization has been facilitated by the advent of multiplex immunofluorescence imaging plat-forms,76where a single section can be used to stain with up to 50 antibodies through iterative staining and imaging procedures (Supplementary Figure 1A). These methods have the ability to provide a detailed phenotyping of structural defects and barrier function; epithelial cell polarity; patterns of epithelial, immune, and neural cell differentiation; and identification of immune cell subsets.
Recent advancements in deriving intestinal organoids from human induced pluripotent stem cells and isolating multipotent intestinal stem cells from human crypts has provided unprecedented opportunities to model CODE disorders in a dish.77–80 Patient or CRISPR/Cas9-generated intestinal stem cells or induced pluripotent stem cells also provides opportunities to perform high-throughput assays with intestinal epithelium. Specifically, functional analysis of barrier function using live intestinal enteroids plated on filter supports in 2-dimensional formats can be achieved using electrical and biochemical measures of passive intercellular and transcellular transport of small solutes (ions and membrane impermeant small solutes). Both 2-and 3-dimensional enteroid cultures can be used for assessment of active (energy-dependent) trans-epithelial ion and solute transport using fluorescence ratio-imaging, enteroid swelling, and in some cases electrical approaches (Supplementary Figure 1B).
Many monogenic causes of the CODE primarily affect epithelial structure and the organization of plasma and intracellular membrane compartments. These genes affect membrane trafficking, which can be effectively and comprehensively assessed in enteroids and cells using assays for IgG intracellular transport by the Fc-receptor, FcRn, and transferrin transport by the rapidly recycling transferrin recep-tor.81 Both probes assess function for all endosomal compartments adapted to and serving the polarized epithelial cell phenotype (Supplementary Figure 1C).
Infants with CODE pose a clinical challenge that requires a structured diagnostic approach to allow early and correct diagnosis. The relatively uniform presentation of early severe diarrhea with food intolerance and the large number of potential etiologies often makes it difficult to prioritize investigations. The updated algorithm in this article harnesses the recent progress in the understanding of the pathogenesis of these disorders, along with advances in genetic testing and immunohistochemistry techniques. The algorithm suggests an initial exclusion of common causes for diarrhea in infancy, followed by early endoscopy for tissue diagnosis, focusing on the crypt to villus ratio as a simple first-line diagnostic tool. Specific immunohistochemistry staining in addition to the traditional H&E analysis further enhances the diagnostic process. The new algorithm also employs early genetic testing, highlighting the fact that the vast majority of the CODEs with sustained symptoms are monogenic disorders. Faster and accurate diagnosis of neonatal and infantile diarrheas should improve patient care, shorten length of stay, provide better prognostication to children and their families, and will provide clinical and pathological information for improved genotype and phenotype characterization and association.
Many of the epithelial-specific CODE disorders still require either life-long PN or allogenic intestinal transplantation.82 Transplanted patients require life-long high-dose immunosuppression that is associated with significant risk for acute and chronic opportunistic infections, allograft rejection/loss, and subsequent malignancy. Cell-based therapies of CRISPR/Cas9-corrected intestinal stem cells and autologous intestinal epithelial transplantation may be a viable option in the coming years, but will require numerous advancements in several areas, including developing ablation, implantation, and engraftment protocols.83
Further research via multicenter collaborations is essential for both understanding the pathogenesis of CODE disorders and for developing new therapies targeted at these rare but severe diseases. Investigation of CODEs will likely enhance our understanding of the basic biology of the intestine greatly, and may provide insights into the pathogenesis and treatment of more common gastrointestinal diseases.