This Month in Gastroenterology

Article Outline

• Mortality in Undiagnosed Celiac Disease
• Blood Urea Nitrogen in Acute Pancreatitis
• Multiple Inputs Modulate Antral G-Cell Gastrin Secretion
• Dysfunctional Autophagy After LAMP-2 Depletion in Alcohol Plus Toxemia-Induced Pancreatitis
• Copyright

Back to Article Outline

Mortality in Undiagnosed Celiac Disease 

Celiac disease is a well-recognized cause of malabsorptive diarrhea, which responds to a strict, gluten-free diet. The availability of noninvasive, serologic testing has revealed a higher prevalence of asymptomatic or minimally symptomatic celiac disease than previously recognized. Untreated symptomatic celiac disease is associated with substantial morbidity and increased mortality. It is unclear whether undiagnosed celiac disease is also a cause of increased mortality.

In this issue of Gastroenterology, Rubio-Tapia et al report on the results of a study aimed at determining the prevalence of celiac disease and its prognostic impact in the United States 50 years ago. The authors tested 3 cohorts of sera for tissue transglutaminase antibodies (tTGA) and endomysial antibodies (EMA). A diagnosis of celiac disease was made when both tTGA and EMA were positive; the result was considered equivocal when only tTGA was positive and negative with negative tTGA. A first cohort consisted of 9,133 young healthy men from whom serum was obtained at the Warren Air Force Base (WAFB) between 1948 and1954. A second cohort consisted of a complete sampling of the older (>50) population in Olmsted County during the years 1995–2003, of which 5558 men had a birth year comparable with the WAFB cohort. A third cohort consisted of a sampling of 7210 younger men in the general population (18–49 years old) from Olmsted County during the years 2006–2008.

The prevalence of undiagnosed CD was 4 times higher in the older present-day cohort (1/121) and 4.5 times higher in the younger present-day cohort (1/106) than in the WAFB cohort (1/652; P < .0001). Over 45 years of follow-up, the hazard ratio for mortality was nearly 4-fold higher for subjects with undiagnosed CD compared with seronegative persons (Figure 1). Mortality rates were similar among persons with equivocal serologic results and those with seronegative results.

• 
  • View Large Image
  • Download to PowerPoint

• Figure 1. 

  Survival over 45 years of follow-up in 14 subjects with undiagnosed celiac disease and 9,076 seronegative persons in the WAFB cohort.

This population-based study at different points in time confirms the increased mortality rate associated with undiagnosed celiac disease. Whether this warrants screening for celiac disease, and whether this increased mortality would diminish in response to a gluten-free diet, remains to be established. The study also shows a dramatic increase in prevalence of celiac disease over the last 50 years. The reasons for this increase are unclear, and both changes in food quantity or quality as well as increasing autoimmunity by changing patterns of childhood infection (the hygiene hypothesis) may contribute.

See page 88.

Back to Article Outline

Blood Urea Nitrogen in Acute Pancreatitis 

Acute pancreatitis is a self-limiting disease in the majority of case, but may progress to complicated disease with systemic inflammation and multiorgan dysfunction. Current guidelines recommend vigorous fluid resuscitation in the early treatment of acute pancreatitis, aiming at improvements in pancreatic microcirculation to reduce local tissue injury and systemic complications. From a clinical point of view, it is important to recognize these patients early in their disease process. A number of prognostic scores have been proposed, aiming at recognition of patients with an unfavorable outcome. These include parameters suggestive of hemoconcentration such as hematocrit and blood urea nitrogen (BUN).

In this issue of Gastroenterology, Wu et al report on an observational cohort study to compare the prognostic utility of serial measurement of BUN versus hemoglobin in the early prognostic assessment of acute pancreatitis. The authors extracted serial blood results from patients with acute pancreatitis from the Cardinal database between January 2003 and December 2006. They used statistical modeling to extract prognostic information from changes in BUN and hemoglobin in the first 48 hours of hospitalization.

The authors identified 5819 cases of acute pancreatitis with ≥3 laboratory result sets in the first 48 hours after admission, of which 95 cases (1.6%) were nonsurvivors. A significant difference in trends of BUN levels over the first 48 hours was found in survivors and nonsurvivors (P < .0001; Figure 2). No difference in changes in hemoglobin levels was found between both groups (Figure 2). Multivariate analysis showed that both admission BUN and change in BUN were independent predictors of mortality, even after controlling for demographic and hemoglobin variables (P < .0001). Irrespective of admission value, increases in BUN at 24 hours were associated with higher mortality, whereas decreases in BUN were associated with reduced mortality. The best predictive accuracy for BUN changes occurred between admission and the 24-hour time point.

• 
  • View Large Image
  • Download to PowerPoint

• Figure 2. 

  Repeated measure analysis with mixed effects model. (A) Mean BUN levels among survivors versus nonsurvivors. (B) Mean hemoglobin levels among survivors versus nonsurvivors. Mean levels along with 95% confidence limits are depicted.

This study demonstrates that changes in BUN are related to mortality in acute pancreatitis. BUN is a routine laboratory marker that is universally available; therefore, it is simple to incorporate into daily clinical management. The mechanisms that relate prognosis in acute pancreatitis to BUN remain to be established. These could involve both intravascular volume depletion with prerenal azotemia as progressive deterioration of renal function or a state of increased protein catabolism. In addition, It remains to be determined whether aggressive fluid resuscitation can improve outcome in patients with early elevations in BUN.

See page 129.

Back to Article Outline

Multiple Inputs Modulate Antral G-Cell Gastrin Secretion 

The hormone gastrin, made as a precursor then processed and secreted by stomach antral G-cells, stimulates gastric acid secretion and helps to regulate mucosal growth. Gastrin itself is autoregulated by gastric pH; low pH stimulates somatostatin release from antral D cells to reduce gastrin release. G-cells are the predominant neuroendocrine cell in the antrum, are located in the lower to middle third of antral glands next to capillaries, and have cytoplasmic extensions that allow the G-cell to access the lumen. It is not fully clear if other compounds or conditions, in addition to gastric acid and branched aromatic amino acids, can regulate gastrin secretion because no pure G-cell population has been studied systematically to explore this.

In the study by Kidd et al, gastric antral mucosal preparations from male Harlan Sprague Dawley rats were processed through Nycodenz density gradient centrifugation and acridine orange-loaded flow cytometry to obtain >97% pure G-cell fractions (as determined by gastrin transcript content levels and the near absence [<3%] of other neuroendocrine cells), that ranged in cell viability from 90% to 99%. Basal gastrin release (158.9 ± 70.6 ng/10⁶ cells) was stimulated 2- to 3-fold with the cyclic adenosine monophosphate (cAMP)-inducer forskolin, and was stimulated through G-protein–coupled receptors (as determined by IBMX, PMA, cholera toxin, and pertussis toxin treatment) and reduced up to 85% with inhibitors of the protein kinase A (PKA)/cAMP/mitogen-activated protein kinase (MAPK) and PI₃K/protein kinase C pathways. In addition to general activators of gastrin secretion (neural ligands, amino acids, and serotonin), motilin, the motilin agonist erythromycin, and bile salts stimulated gastrin release, as did the tastants sucralose, glucose, denatonium, and caffeine, and as did the villanoid receptor agonist capsaicin and luminal bacterial lipopolysaccharides (LPS), and all of which stimulated cAMP production (Figure 3). Furthermore, the G-cells express G-protein–coupled receptors such as α-gustducin by transcriptome and immunohistochemical analysis, and express Toll-like receptor 4. The somatostatin analog octreotide and histamine inhibited gastrin release. In addition to cAMP production, gastrin-releasing peptide, sucralose, denatomium, and S enteritidis LPS stimulated ERK phosphorylation, and only S enteritidis LPS stimulated nuclear factor-κB and c-Jun–N-terminal kinase pathways. Lowering the pH, as expected, significantly decreased gastrin release, but at both neutral (7.4) and acidic (5.5) pHs, mechanical force stimulated gastrin release (a finding that was more prominent in mucosal strips over cell cultures) that could be inhibited with a selective adenosine receptor (ADORA2B) antagonist.

• 
  • View Large Image
  • Download to PowerPoint

• Figure 3. 

  Compounds that cause gastrin release in isolated G-cells. (Left) Constituents of food including the “sweet” tasting amino acids (tyrosine), sweet and bitter tastants and pain/heat transducers (capsaicin) stimulate gastrin secretion. (Right) Bacteria also stimulate gastrin secretion through LPS.

The study indicates that in pure rat G-cells, which express G-coupled protein receptors to function as chemosensors, respond to multiple inputs including neural, endocrine, and luminal agents, and including tastants and bacterial LPS to secrete gastrin principally through PKA/cAMP/MAPK pathways. Additionally, mechanical forces stimulate gastrin release through adenosine receptors. It is likely that the G-cell with its sophisticated sensory capabilities plays a major role in function of the upper gastrointestinal tract.

See page 231.

Back to Article Outline

Dysfunctional Autophagy After LAMP-2 Depletion in Alcohol Plus Toxemia-Induced Pancreatitis 

Acute pancreatitis is triggered by pancreatic injury, such as repeated alcohol ingestion coupled with endotoxemia that is followed by an intense inflammatory response, but the pathogenesis is not completely understood. Pancreatic acinar cells undergo vacuolization in pancreatitis, the ultrastructure characteristics of which have not been fully described. Autophagy, the process of sequestering organelles and cytosolic components into vesicles (autophagosomes) that are then delivered to the lysosomes for degradation, has not been evaluated in pancreatitis. Autophagy is cytoprotective, and dysregulation of this process can be detrimental. Mice deficient in lysosomal-associated membrane protein 2 (Lamp-2), a protein that is required for proper fusion of lysosomes with autophagosomes, accumulate autophagic vacuoles in multiple cell types, including the pancreas. It is possible that the vacuolization in acinar cells in pancreatitis is caused by dysregulated autophagy.

In the study by Fortunato et al, pancreata from rats fed alcohol for 14 weeks followed by a single lipopolysaccharide (LPS) injection (simulating alcohol followed by endotoxemia) as well as from humans were studied for autophagy apoptosis, and necrosis. In rats, the combination of alcohol and LPS induced marked acinar vacuolization, as well as necrosis, edema, and inflammation as early as 3 hours after the LPS injection, which increased through 24 hours after LPS. Although alcohol-fed rats had an increase in apoptosis and necrosis, the alcohol–LPS combination increased necrosis a further 2.5-fold without a change in apoptosis, and was associated with reduced nuclear high mobility group box 1 (HMGB1) protein, typically released from necrotic nuclei. Half of the mitochondria were damaged, but this was not increased in alcohol–LPS rats compared with LPS-alone treated rats; however, alcohol–LPS rats increased the ratio of adenosine diphosphate/adenosine triphosphate (ATP) showing a reduced capacity to generate ATP in acinar cells, and likely predisposing them to necrosis. Caspase activation, a marker for apoptosis, was reduced 1.6-fold in alcohol–LPS rats, whereas cathepsin B and trypsin activity increased 1.65- and 1.7-fold, respectively, indicating a shift from apoptosis to necrosis with increasing destructive protein activities. Using electron microscopy, autophagosome volume increased 1.5-fold over controls in alcohol–LPS rats, but autolysosomes (formed by fusion of autophagosomes and lysosomes) was reduced 2-fold over controls (Figure 4). Lamp-2 expression as well as another lysosomal membrane protein Gramp-92, were depleted in alcohol–LPS mice (17-fold decrease over controls). Knockdown of Lamp-2 via siRNA techniques allowed the accumulation of autophagosomes in pancreatic acinar cells, and Lamp-2 expression was 30-fold less in human pancreatitis specimens over controls.

• 
  • View Large Image
  • Download to PowerPoint

• Figure 4. 

  Quantification of the volume of autolysosomes (expressed as percentage of the total acinar volume) from electron microscopy. Data plotted as mean values ± standard error of the mean of 4 individuals per group. *P < .0001 versus pair-fed 24 hours after LPS. **P < .0001 versus alcohol-fed rats; ***P < .0001 versus alcohol-fed rats 24 hours after LPS. ****P < .0044 versus alcohol-fed rats 24 hours after LPS.

This study indicates that in alcohol-LPS–induced pancreatitis in rats and in human pancreatitis specimens, lysosomal membrane protein depletion occurs that allows accumulation of autophagosomes and the depletion of autolysosomes (seen as vacuolization), the disruption of authophagy. This lysosomal protein depletion along with the shift from apoptosis to necrosis plays a key role in the onset of acute pancreatitis.

See page 350.