Pediatric Protein-Losing Enteropathy

Updated: Jul 25, 2017
  • Author: Simon S Rabinowitz, MD, PhD, FAAP; Chief Editor: Carmen Cuffari, MD  more...
  • Print

Practice Essentials

Protein-losing enteropathy (PLE) is not a single disease but a symptom. Although it occurs in multiple conditions through various pathophysiologic processes, the end result is the loss of serum proteins into the GI tract. Protein losses from other regions of the GI tract are also considered PLE.

Hypoalbuminemia can result from protein loses through the skin, the kidneys, or the respiratory tract as well as decreased synthesis in the face of normal turnover. 

PLE is generally secondary to 1 of 3 mechanisms: 

  • Lymphatic obstruction
  • Inflamed mucosa
  • Molecular changes in the epithelial barrier in the absence of other signs of pathology 




Although PLE implies an intestinal disease associated with the small bowel, the term "protein-losing enteropathy" is commonly used to also include loss of protein from the colon, stomach, and, rarely, the esophagus. Some authors have used the term protein-losing gastroenteropathy. While there is nothing unique about PLE in children, the relative prevalence of various etiologies is different in children from that described in adults.

PLE can be either a primary manifestation or a subclinical component of various diseases. Historically, patients with hypoalbuminemia of unknown cause were referred to as having idiopathic hypoproteinemia, edema disease, or nephrosis. These patients had neither a decrease in the production of albumin (ie, no signs of malnutrition or hepatic disease) nor an increase in albumin losses from the respiratory tract, kidneys, or skin.

In 1949, Albright et al demonstrated an increase in protein turnover in patients with PLE. In 1958, Citrin et al were the first to use radiolabeled tracers to demonstrate the actual loss of a protein-containing fluid into the GI tract. However, a major advance was made when Crossley and Elliot demonstrated that measurement of alpha1-antitrypsin (A1-AT) levels in the stool was a practical, reproducible, and inexpensive investigation to diagnose PLE. This approach has identified various conditions that have subclinical PLE as a component of the disease process. With the application of newer genetic screening techniques, the number of distinct entities that can lead to PLE continues to grow.



No single mechanism can account for the loss of protein into the GI tract seen in a wide range of underlying clinical conditions. Several molecular changes in epithelial cells have been shown to yield PLE by increasing the permeability to serum proteins. [1, 2, 3, 4]  Modification of the epithelial matrix component, by congenital molecular abnormalities, by dysfunctional lymphatic drainage or by inflammation, offers an intriguing and unifying hypothesis for the many causes of PLE that merits further investigation.

In vitro analyses have demonstrated that loss of these proteoglycans not only directly causes PLE but also potentiates the effects of other reputed factors such as inflammatory cytokines and increased lymphatic pressure. [5]  In addition, infants and children with various forms of congenital glycosylation defects, another potential mechanism for loss of heparin sulfate proteoglycans, also have PLE secondary to increased intestinal permeability. [6]

PLE must always be distinguished from the loss of protein through other organs, most commonly the kidney and skin. One potential clue is that renal losses of protein are usually limited to smaller proteins such as albumin, whereas the GI tract and skin losses are less discriminating. In addition, hypoalbuminemia may be secondary to synthetic dysfunction or excessive catabolism rather than the result of increased losses. Synthetic dysfunction itself can be seen in liver disease or as a result of inadequate precursors (ie, malnutrition or malabsorption).

For practical purposes, the disease processes that cause PLE can be grouped into the following 3 major categories: (1) lymphatic obstruction or defects in structural integrity; (2) mucosal erosion or ulceration; and (3) epithelial cell dysfunction in the absence of macroscopic compromise.

Obstruction of lymphatics from any cause can produce increased pressure throughout the lymphatic system of the GI tract. This results in the stasis of lymph and, if the pressure is high enough, the loss of lymphatic fluid rich in albumin and other proteins from the lacteals in intestinal microvilli into the lumen of the GI tract. Alternatively, compromise of the lymphatic channels themselves can also result in leakage. If the loss of albumin exceeds the rate of synthesis, hypoalbuminemia and, eventually, edema develop. In addition to the loss of albumin, other important components of lymph are also lost into the bowel, including lymphocytes, immunoglobulins, and hydrophobic molecules such as cholesterol, lipids, and fat-soluble vitamins that yield other complications.

Lymphopenia is a common finding associated with PLE due to primary intestinal lymphangiectasia, Whipple disease, or constrictive pericarditis. In cases of PLE associated with lymphatic obstruction, alleviating the obstruction corrects the lymphopenia. A decrease in the circulating levels of immunoglobulins is also a feature of lymphatic obstruction, but because the synthetic machinery remains intact, response to antigenic challenge is usually adequate. In patients with lymphatic obstruction, fat malabsorption may develop secondary to losses from the lymphatics. In these patients, failure to thrive, poor weight gain, and deficiencies in the fat-soluble vitamins (ie, A, D, E, K) can also occur.

A wide variety of infectious diseases and noninfectious diseases can produce inflammation and ulceration of the GI mucosa resulting in PLE. Each of these processes has a unique pathophysiology. Similar to lymphatic obstruction, these inflammatory pathologies may also be associated with hypogammaglobulinemia. [1]

The Fontan procedure is a palliative surgical procedure performed in patients with a functional or anatomic single ventricle. It creates a venous pathway that directs the inferior vena caval flow into the pulmonary arteries, resulting in the entire systemic venous flow returning passively into the pulmonary arteries. [7] This creates a system where a single ventricle pumps blood into separate, in-series systemic and pulmonary circulations, thereby relieving cyanosis. PLE has long been recognized as a complication after cardiac surgery, especially in patients who have had the Fontan procedure. This complication is known to carry a high mortality rate. Research in this area remains active; however, the exact pathophysiology of the protein loss in this setting has still not been elucidated. [8]  Numerous publications have provided data to support various hypotheses including early elevations of postoperative central venous pressure, [9]  low pulmonary vascular compliance, [10]  and elevated serum hepatocyte growth factor. [11]  Retrospective series have defined the additional risks seen in patients with PLE after Fontan procedure (see Prognosis section below).

A case-control study compared 8 patients who underwent the Fontan procedure with PLE to 8 patients who underwent the Fontan procedure without PLE. The patients with PLE had immunologic abnormalities similar to patients with PLE and intestinal lymphangiectasia who had not had a Fontan procedure (eg, severely low CD4 counts with mildly decreased CD8 counts, hypogammaglobulinemia, depressed cell mediated immunity). These authors postulated that dysfunctional lymphatic drainage was a key contributor to Fontan-related PLE. [2]

A survey-based study also identified clinical features that suggest inadequate lymphatic drainage may play a role in post-Fontan PLE. Several patient specific factors associated with the diagnosis were noted, including the following:

  • Abdominal swelling may be the best symptom to use in screening for early signs of PLE 
  • Patients with PLE were more likely to have had prolonged chylous chest tube drainage.

This study also found that most patients with PLE had been treated with only one specific therapy. Multiple therapies are often needed for treatment. These authors recommended that “best practice” guidelines should be developed to assure successful management. [3]

One case report surprisingly suggests that post-Fontan PLE may be a consequence of lactase deficiency. In this report, a 12-year old girl with PLE shortly after undergoing the Fontan procedure was found to have an intestinal lactase deficiency based on her history. Although the patient failed a therapeutic trial of cortisone and heparin as well as other cardiac interventions, her symptoms improved after beginning a lactose-free diet. Importantly, her serum protein values and electrolytes normalized. The authors suggest that dietary treatment of lactose intolerance, a common condition whose incidence increases with age, may improve the outcomes in selected patients. Not mentioned by the authors is the fact that Lactaid milk often has lower fat than regular milk, which may have played a role in their outcome.

Another unique barrier dysfunction also is reported to yield PLE. An infant with PLE was found to have a mutation in PLVAP, which is a cationic, integral membrane protein expressed in endothelial cells. The same phenotype had been previously described in a PLVAP knock-out mice. The child presented with intractable secretory diarrhea that was felt to be secondary to the absence of PLAVP, which is responsible for organizing the diaphragms in endothelial cells. [4]

In addition to dysfunctional glycosylation of membrane proteins, this becomes the second ultrastructural defect that can account for PLE. Whether these are examples of a more generalized phenomenon that could account for an appreciable proportion of the PLE which is noted in the absence of both mucosal erosions and lymphatic leakage, remains to be determined. [4]




United States

No published data have reported an accurate incidence or prevalence of PLE in any parts of the United States.


No published have reported an accurate incidence or prevalence. The incidence is highest in areas with significant infectious enterocolitis. A recent multicenter European review of over 3000 patients with Fontan procedure describes a prevalence of 3.9%. [12] These authors used stringent criteria; three fourths of the patients had effusions and edema. Other studies have reported an even higher prevalence after this surgery.


Morbidity and mortality is dependent on the diseases that are the cause of the PLE and the availability of prompt recognition and treatment. In the European cohort of Fontan patients described above, medical treatment was ineffective in 75%, with a mortality of 46%, and surgical treatment was ineffective in 81%, with a mortality of 62%.


No racial predilection is noted. For some entities, especially infections of the GI tract, the prevalence is higher in the developing world and in those races more commonly found there.



Most cases of PLE secondary to mucosal inflammation are able to be resolved with effective therapy for the underlying enteropathy. Conversely, most of the inheritable forms of PLE are part of a larger constellation of deficits related to a genetic mutation. Therefore, the most common setting for PLE investigations is among patients who have had the Fontan procedure with several centers publishing retrospective series. A 20-year study from Germany followed 434 patients who had total cavopulmonary connection between May 1994 and March 2015. Although these patients had less problems than patients who underwent the Fontan procedure (eg, tachyarrhythmia, need for revisions, thromboembolism), PLE, liver dysfunction, and exercise limitations remained problematic. [13] ​ A prospective series from Children's Hospital of Philadelphia examined 33 patients undergoing the Fontan procedure, preoperatively, early postoperatively, and intermediately postoperatively (3-9 mo). [14] Although none of these patients showed consistent PLE in the time frames reported, 6 episodes of elevated stool A1-AT were identified, and 5 of those 6 episodes were associated with significant hemodynamic disturbances that required intervention.

A study from Michigan examining immune abnormalities prospectively in 16 patients after the Fontan procedure compared 8 patients with PLE to 8 without PLE. [15] Patients who underwent the Fontan procedure who had PLE had extensive quantitate immune abnormalities, including CD4 deficiency. These were similar to patients who did not undergo the Fontan procedure who had PLE secondary to lymphatic abnormalities. Of note, most of the 8 children who had PLE after Fontan procedure had negative titers for measles, mumps, and rubella vaccinations.

A retrospective study compared 96 patients with PLE waiting for a heart transplant who underwent the Fontan procedure and to 260 patients with PLE without the Fontan procedure also waiting for a heart transplant. [16] In this large multicenter cohort, the diagnosis of PLE was not associated with increasing waiting list mortality or posttransplant morbidity or mortality.

Another series consisting of 42 patients with protein-losing enteropathy followed at the Mayo Clinic noted decreased survival among patients with Fontan pressure, decreased ventricular function, higher pulmonary vascular resistance, lower cardiac index, and lower mixed venous saturation compared with survivors. [17] However, the authors concluded that although protein-losing enteropathy remains difficult to effectively treat in this population, survival has improved with advances in treatment.