Correspondence *Department of Internal Medicine, Erasmus University Medical Center, 's-Gravendijkwal 230, 3015 CE Rotterdam, The Netherlands

Email j.h.p.wilson@erasmusmc.nl

The case

A 22-year-old female presented to the emergency room with pitting edema of the legs, facial and abdominal swelling, diarrhea and a dry cough; these symptoms had developed 10 days earlier. When she was 2 years old the patient was found to have congenital hepatic fibrosis and portal hypertension, and aged 15 years she had been diagnosed with congenital disorder of glycosylation type Ib (CDG-Ib). This diagnosis was based on two key findings: a characteristic isoelectric focusing pattern of the patient's serum transferrin, and a very low level of phosphomannose isomerase activity in her fibroblasts (77 nmol/h per mg protein; normal range: 1250–2800 nmol/h per mg protein). There was no family history of congenital disorders of glycosylation (CDGs). The patient had been treated with oral mannose for some months but had stopped treatment because of associated diarrhea and abdominal pain. At the age of 20 years the patient experienced an episode of Salmonella sepsis with profuse diarrhea, which was followed by Clostridium difficile-induced diarrhea. She was found to have protein-losing enteropathy (PLE), which resolved after antibiotic treatment. In 2004, aged 21 years, the patient underwent endoscopic rubber band ligation of fundal and esophageal varices, following an episode of hematemesis.

At presentation the patient's abdomen was distended and shifting dullness was noted. Her bowel sounds were normal, and her liver and spleen were not palpable. Laboratory examination revealed hypoalbuminemia and pancytopenia (Table 1). Abdominal ultrasonography showed normal blood flow in both the portal vein and inferior vena cava. The patient was admitted to a hospital ward for further evaluation.

Table 1 Summary of the patient's relevant laboratory results, demonstrating the presence of hypoalbuminemia and pancytopenia.

Full tableFigures & Tables indexDownload Power Point slide (74K)

Over the following 3 months the patient experienced several inflammatory episodes with raised temperature, increasingly frequent diarrhea, abdominal pain and muscle pain. These episodes were associated with increased C-reactive protein levels and declining albumin levels. Repeated cultures of blood, urine and stools, however, failed to reveal any pathogenic organisms. Double-balloon enteroscopy of the jejunum and ileum showed some patchy edema, but on microscopic examination no lymphangiectasia or influx of inflammatory cells was noted. Ascites was also noted at this point. A fecal excretion test using ⁵¹Cr-labeled albumin confirmed the presence of PLE.

The ascites and edema responded only partially to spironolactone (Aldactone^®, Pfizer, New York, NY; 25 mg/day) and furosemide (40–80 mg/day). The patient was intermittently treated with albumin infusions, which temporarily increased her serum albumin level and reduced her symptoms, but no sustained improvement in clinical or laboratory parameters was noted. The patient was then given intravenous albumin alone (20 g/day) for 5 days, upon which her serum albumin level increased from 22 g/l to 30 g/l. It was at this point we decided to start a therapeutic trial of heparin therapy.

Upon cessation of albumin the patient was treated by infusion of unfractionated heparin (Leo Pharma, Ballerup, Denmark) at an initial dosage of 200 international units (IU)/h, which was gradually increased over 8 days to 400 IU/h. After 11 days of intravenous heparin therapy the patient was switched to subcutaneous injections of unfractionated heparin (5,000 IU twice a day for 7 days, followed by 7,500 IU twice a day). The dosage was monitored using heparin antifactor Xa levels, which remained well below the therapeutic range. A prolongation of the patient's activated partial thromboplastin time (APTT) and prothrombin time was noted before initiation of this therapy—most likely resulting from the decreased levels of coagulation factors (due to PLE) and reduced glycosylation of these factors.

The patient's APTT remained stable during both intravenous and subcutaneous heparin therapy, and the activity of several coagulation factors even increased (antithrombin from 18% to 49%; antiplasmin from 42% to 83%). The patient's serum albumin concentration increased to 35 g/l in response to heparin therapy, and her fecal -1 antitrypsin concentration decreased (Figure 1). Her ascites and edema began to resolve, the frequency of diarrhea decreased and she reported feeling better.

Figure 1 Graph showing the change in the patient's fecal -1 antitrypsin concentration in response to heparin therapy.

Abbreviations: H–iv, intravenous heparin; H–sc, subcutaneous heparin.

Full figure and legend (17K)Figures & Tables indexDownload Power Point slide (62K)

The patient was discharged after 7 weeks of heparin therapy but she continued to self-administer heparin subcutaneously at home, continuing the dosage of 7,500 IU twice a day. Due to a steady clinical improvement and stable concentrations of serum albumin, heparin therapy was finally terminated after a total of 13 weeks.

The patient remained well and serum albumin levels were stable for a further 16 weeks, at which point she reported tiredness and abdominal pain. Her serum albumin levels also started to decrease. In light of the adverse effects associated with chronic administration of unfractionated heparin (which include osteoporosis and adrenal insufficiency) and the fact that the drug has to be administered parenterally, the patient was started on a new treatment trial with oral mannose. The dose was slowly increased from 1.25 g four times a day to 2.5 g four times a day. Over time, however, the patient's diarrhea and malaise worsened and she became jaundiced. There was no change in the patient's transferrin isoelectric focusing pattern, and mannose therapy was, therefore, terminated after 5 months.

Subcutaneous heparin injections were reinstated (7,500 IU twice a day). Over the following 4 weeks there were no apparent benefits of this treatment and the patient increasingly complained of pain, hematomas and swelling around the sites of administration. It was suspected that heparin was building up subcutaneously rather than reaching the circulation, and a portacath was fitted to facilitate intravenous administration. Around 2 weeks later the patient's symptoms started to diminish, and her clinical status continued to improve over the following 6 weeks. The dose of heparin was gradually decreased to 5,000 IU twice a day, owing to the observation of a prolonged APTT. To date, the patient's status remains the same.

Discussion of diagnosis

CDGs are divided into two main groups: type I CDGs are caused by defects in the assembly of glycoproteins; whereas type II CDGs comprise defects in the trimming and processing of the glycoproteins.¹ The affected gene and enzyme and the clinical characteristics of the most common type I CDGs are presented in Table 2. PLE—the loss of plasma protein through the intestine—is a common feature of patients with CDG-Ib, as exemplified by the present case.^{2, 3}

Table 2 Summary of the genetic, enzymatic and clinical characteristics of the most common types of CDG.¹⁹

Full tableFigures & Tables indexDownload Power Point slide (92K)

The clinical manifestations of PLE have been recently described by Herfarth and colleagues in this journal.⁴ Lower extremity and facial edema, and fat, vitamin and carbohydrate malabsorption are all common. Laboratory abnormalities include hypoproteinemia, hypoalbuminemia and lymphocytopenia. The authors suggest fecal -1 antitrypsin measurement as a simple first-line diagnostic procedure for PLE. In the present case PLE was indeed confirmed by this method, and was quantified using a ⁵¹Cr-labeled albumin fecal excretion test.

Herfarth et al. describe mucosal injury and increased lymphatic pressure as the two mechanisms known to cause PLE.⁴ In a patient with CDG-Ib, portal hypertension leading to increased lymphatic pressure should be considered as a differential diagnosis for PLE. The case patient was known to have portal hypertension before presentation; however the abdominal ultrasound at presentation did not show portal vein thrombosis or an increase in collaterals, suggesting that a worsening of portal hypertension was highly unlikely to be the underlying cause of PLE. Although intestinal lymphangiectasia can occur in CDG-Ib patients, there was no evidence for lymphangiectasia upon enteroscopic examination in this case. We hypothesize, therefore, that the patient's PLE was caused by a loss of integrity of the intestinal wall due to a reduction in the number of glycoproteins on the surface of enterocytes. It has been shown that heparan sulfate proteoglycans (HSPGs), which are normally found on the basolateral surface of intestinal epithelial cells, are absent or mislocalized during episodes of PLE but reappear when the condition resolves, suggesting a functional link between HSPG loss and PLE.^{5, 6}

Treatment and management

Treatment of PLE should, in general, focus on the underlying disease. In patients with intestinal lymphangiectasia, fat intake should be reduced to diminish intestinal lymphatic flow, and octreotide, a somatostatin analogue, could also be administered.⁴ If the lymphangiectasia is localized, then a lymphovenous anastomosis to divert lymph to the venous system or resection could be considered. In patients with CDG-Ib, oral mannose therapy has been shown to reverse PLE.⁷ Mannose overcomes effects of the underlying enzyme defect, thereby normalizing protein glycosylation.

In the present case an alternative therapy was needed, as the patient had not tolerated mannose well in the past. The administration of heparin was largely experimental, although there are published case reports that document the use of heparin to treat PLE, mainly in children. For example, subcutaneous heparin injections have been reported to reverse PLE in some patients who have undergone the Fontan procedure.^{8, 9, 10} In another report, a patient with tricuspid hypoplasia who developed PLE failed to respond to medical or surgical treatment, but administration of heparin immediately decreased intestinal protein loss.¹¹ While the mechanisms of these improvements are unknown, the patients' favorable response is likely to be independent of heparin's anticoagulant activity.

Due to the absence of studies on heparin in patients with CDG, it is currently impossible to provide a dose recommendation. For the case patient, we decided to titrate heparin to an anti-Xa level below that required for full therapeutic heparin therapy in patients with thromboembolism, because of the decrease in clotting factors in patients with CDG. Moreover, the risk of bleeding in this patient was further increased because of liver cirrhosis and mild portal hypertension, which was associated with the presence of distended collateral veins, splenomegaly and thrombocytopenia.

Further guidance on the pathogenesis of PLE and the mechanisms underlying heparin therapy came from reports of other PLE patients with primary disorders such as CDG or systemic lupus erythematosus. Despite the diversity of the underlying conditions, the patients in these reports share several common features in that their PLE was episodic, and its onset often associated with viral infections^{5, 12} and inflammatory conditions.^{13, 14} These studies led us to try to identify key players in PLE pathogenesis and the mechanisms underlying heparin therapy.

Our studies in tissue culture and mouse models have shown that the loss of HSPG from the basolateral surface of intestinal epithelial cells has a central role in PLE pathogenesis by amplifying the effects of inflammatory cytokines and increased pressure.^{15, 16, 17, 18} These studies have also shown that soluble heparin reduces the leakage of proteins through confluent intestinal epithelial cell layers that have been stripped of heparan sulfate and exposed to inflammatory cytokines. The binding of cytokines to their receptors causes tight junctions between cells to loosen, leading to paracellular protein leakage. Cell-surface heparan sulfate and soluble heparin also bind inflammatory cytokines, however, and are thought to compete with the endogenous receptors, thus reducing protein leakage. These cytokine-quenching mechanisms could provide an explanation for the therapeutic effects of heparin in patients with PLE.

Conclusions

This Case Study describes a patient with CDG-Ib and recurrent episodes of PLE who was treated with unfractionated heparin. The patient showed considerable clinical improvement and stabilization of serum albumin levels, which persisted even after heparin treatment was stopped. This experience provides a basis for further therapeutic trials of unfractionated heparin in other patients with CDG who develop PLE and cannot be treated with mannose. Heparin therapy could also be effective in cases of PLE that are not caused by underlying CDG.

Acknowledgments

The authors are grateful to Dr Tom Kennedy of the University of Utah for his advice on heparin therapy. HHF and LB are supported by NIH Grants R01DK55615, R21HL078997 and The Children's Hearts Fund. Désirée Lie, University of California, Irvine, CA, is the author of and is solely responsible for the content of the learning objectives, questions and answers of the Medscape-accredited continuing medical education activity associated with this article.

References

1. Freeze HH and Aebi M (2005) Altered glycan structures: the molecular basis of congenital disorders of glycosylation. Curr Opin Struct Biol 15: 490–498 | Article | PubMed | ChemPort |
2. Pelletier VA et al. (1986) Secretory diarrhea with protein-losing enteropathy, enterocolitis cystica superficialis, intestinal lymphangiectasia, and congenital hepatic fibrosis: a new syndrome. J Pediatr 108: 61–65 | Article | PubMed | ChemPort |
3. Damen G et al. (2004) Gastrointestinal and other clinical manifestations in 17 children with congenital disorders of glycosylation type Ia, Ib, and Ic. J Pediatr Gastroenterol Nutr 38: 282–287 | Article | PubMed |
4. Herfarth H et al. (2007) A case of recurrent gastrointestinal bleeding and protein-losing gastroenteropathy. Nat Clin Pract Gastroenterol Hepatol 4: 288–293 | Article | PubMed |
5. Westphal V et al. (2000) Reduced heparan sulfate accumulation in enterocytes contributes to protein-losing enteropathy in a congenital disorder of glycosylation. Am J Pathol 157: 1917–1925 | PubMed | ChemPort |
6. Murch SH et al. (1996) Congenital enterocyte heparan sulphate deficiency with massive albumin loss, secretory diarrhoea, and malnutrition. Lancet 347: 1299–1301 | Article | PubMed | ChemPort |
7. Niehues R et al. (1998) Carbohydrate-deficient glycoprotein syndrome type Ib: phosphomannose isomerase deficiency and mannose therapy. J Clin Invest 101: 1414–1420 | Article | PubMed | ISI | ChemPort |
8. Kelly AM et al. (1998) Use of heparin in the treatment of protein-losing enteropathy after fontan operation for complex congenital heart disease. Mayo Clin Proc 73: 777–779 | PubMed | ChemPort |
9. Donnelly JP et al. (1997) Reversal of protein-losing enteropathy with heparin therapy in three patients with univentricular hearts and Fontan palliation. J Pediatr 130: 474–478 | Article | PubMed | ChemPort |
10. Rychik J (1998) Management of protein-losing enteropathy after the Fontan procedure. Semin Thorac Cardiovasc Surg Pediatr Card Surg Annu 1: 15–22 | PubMed |
11. Bendayan I et al. (2000) Heparin therapy and reversal of protein-losing enteropathy in a case with congenital heart disease. Pediatr Cardiol 21: 267–268 | Article | PubMed | ChemPort |
12. Lenz D et al. (2003) Protein-losing enteropathy in patients with Fontan circulation: is it triggered by infection? Crit Care 7: 185–190 | Article | PubMed |
13. Bruewer M et al. (2003) Proinflammatory cytokines disrupt epithelial barrier function by apoptosis-independent mechanisms. J Immunol 171: 6164–6172 | PubMed | ISI | ChemPort |
14. Chakravortty D and Kumar KS (1999) Modulation of barrier function of small intestinal epithelial cells by lamina propria fibroblasts in response to lipopolysaccharide: possible role in TNFalpha in inducing barrier dysfunction. Microbiol Immunol 43: 527–533 | PubMed | ChemPort |
15. Bode L and Freeze HH (2006) Applied glycoproteomics—approaches to study genetic-environmental collisions causing protein-losing enteropathy. Biochim Biophys Acta 1760: 547–559 | PubMed | ChemPort |
16. Bode L et al. (2006) Heparan sulfate plays a central role in a dynamic in vitro model of protein-losing enteropathy. J Biol Chem 281: 7809–7815 | Article | PubMed | ChemPort |
17. Bode L et al. (2005) Heparan sulfate depletion amplifies TNF-alpha-induced protein leakage in an in vitro model of protein-losing enteropathy. Am J Physiol Gastrointest Liver Physiol 288: G1015–G1023 | Article | PubMed | ISI | ChemPort |
18. Bode L et al. (2008) Heparan sulfate and syndecan-1 are essential in maintaining murine and human intestinal epithelial barrier function. J Clin Invest 118: 229–238 | Article | PubMed | ChemPort |
19. Online Mendelian Inheritance in Man [http://www.ncbi.nlm.nih.gov/sites/entrez?db=OMIM]

Contact the journal about this article

Subject areas under which this article appears: Small intestine | Therapy