Michael D. Levitt

Association between fecal hydrogen sulfide production and pouchitis

PURPOSEThe beneficial effect of antibiotics in pouchitis suggests that an unidentified fecal bacterial product causes this condition. A candidate compound is hydrogen sulfide, a highly toxic gas produced by certain fecal bacteria, which causes tissue injury in experimental models. We investigated hydrogen sulfide release and sulfate-reducing bac-terial counts in pouch contents to determine whether hy-drogen sulfide production correlates with pouchitis.METHODSDuring incubation at 37°C, the production of hydrogen sulfide, methylmercaptan, carbon dioxide, and hydrogen were studied using fresh fecal specimens obtained from 50 patients with ileoanal pouches constructed after total proctocolectomy for ulcerative colitis (n = 45) or for familial adenomatous polyposis (n = 5). Patients with ulcerative colitis were divided into five groups: a) no history of pouchitis (pouch for at least 2 years; n = 8); b) past episode(s) of pouchitis but no active disease for the previous year (n = 9); c) pouchitis in the past year but presently inactive (n = 9); d) ongoing antibiotic treatment (metronidazole or ciprofloxacin) for pouchitis (n = 11); e) currently suffering from pouchitis (n = 8).RESULTSRelease of hydrogen sulfide when pouchitis was active (6.06 ± 1.03 μmol g−1 4 h−1) or had occurred in the past year (4.71 ± 0.41 μmol g−1 4 h−1) was significantly higher (P < 0.05) than when pouchitis had never occurred (1.71 ± 0.43 μmol g−1 4 h−1) or had been inactive in the past year (2.62 ± 0.49 μmol g−1 4 h−1). Antibiotic therapy was associated with very low hydrogen sulfide release (0.68 ± 0.29 μmol g−1 4 h−1). Pouch contents from familial adenomatous polyposis patients produced significantly less hydrogen sulfide (0.75 ± 0.09 μmol g−1 4 h−1) than did any group of nonantibiotic-treated ulcerative colitis patients. Sulfate-reducing bacterial counts in active pouchitis (9.5 ± 0.5 log10/g) were significantly higher than in those who never experienced pouchitis (7.38 ± 0.32 log10/g), and these counts fell dramatically with antibiotic treatment. No statistically significant differences in carbon dioxide and hydrogen were observed among the groups not receiving antibiotics.CONCLUSIONSPouch contents of patients with ongoing pouchitis or an episode within the previous year released significantly more hydrogen sulfide than did the contents of patients who never had an attack of pouchitis and those with longstanding inactive disease. The response to therapy with metronidazole or ciprofloxacin was associated with marked reductions in hydrogen sulfide release and sulfate-reducing bacteria. These results provide a rationale for additional studies to determine whether the high sulfide production is a cause or effect of pouchitis. The lower hydrogen sulfide production by pouch contents of familial adenomatous polyposis vs. patients with ulcerative colitis suggests a fundamental difference in gut sulfide metabolism that could have implications for the etiology of ulcerative colitis as well as the pouchitis of patients with ulcerative colitis.

Comparison of serum amylase pancreatic isoamylase and lipase in patients with hyperamylasemia

We compared results of measurements of total serum amylase, pancreatic isoamylase, and lipase measurements in patients with hyperamylasemia. Serial measurements of these three enzyme levels in patients recovering from acute pancreatitis indicated that pancreatic isoamylase and lipase were elevated above normal to a greater extent and remained elevated much longer than did the total amylase. This finding indicates an appreciable sensitivity advantage of the pancreatic isoamylase and lipase over total amylase measurement during the recovery phase of pancreatitis. Comparison of pancreatic isoamylase and lipase levels in selected sera indicated a good correlation (r=0.84) between these two measurements in patients who did not have macroamylasemia. Lipase was normal in sera with amylase elevations due solely to salivary isoamylase. Thus, in nonmacroamylsemic sera, pancreatic isoamylase and lipase appear to be roughly interchangeable markers of the level of pancreatic enzymes in the blood. An advantage of the lipase assay is that this enzyme is normal in hyperamylasemia caused by macroamylasemia, whereas the inhibitor assay indicates that the pancreatic isoamylase is elevated. Development of automated assays for either pancreatic isoamylase or lipase should lead to the routine use of one of these assays in place of the present reliance on total amylase measurements in the diagnosis of pancreatitis.

Follow-up of a flatulent patient.

SummaryThis paper describes a low-flatulence diet developed by an extremely flatulent patient. Based on meticulous recording of each passage of flatus, the patient employed an elimination diet to determine what foods were responsible for his gas production. The diet reduced the frequency of his gas passage from 34±7 to 17±5 times per day (normal: 14±6) and similar reductions were observed by two other flatulent patients during adherence to the diet. The rectal gas of each of these subjects largely consisted of two gases (CO2 and H2) which result from bacterial fermentation of carbohydrates. The diet, which is low in lactose and wheat products, presumably minimizes the quantity of carbohydrates delivered to the colonic bacteria.

Pneumatosis cystoides intestinalis and high breath H2 excretion: Insights into the role of H2 in this condition

Patients with pneumatosis cystoides intestinalis have been reported to excrete excessive H2 because of a lack of H2-consuming intestinal bacteria. This study describes a patient with bacterial overgrowth and pneumatosis of the small intestine whose colonic flora avidly consumed H2 but whose small bowel flora produced but did not consume H2. There is no commonly accepted mechanism whereby excessive luminal H2 causes intramural gas. An explanation is proposed in which an initial, transitory source of intramural gas is distinguished from the mechanism that results in the persistence of the gas. Independent of the initial source of gas, rapid diffusion of H2 from the lumen into an intramural gas bubble would cause N2, O2, and CO2 to diffuse from the blood into the bubble. As a result, the bubble would expand and then persist indefinitely as long as H2 continued to diffuse from the lumen to the intramural gas collection.

Prevalence and nature of hyperamylasemia in acute alcoholism

Determination of serum amylase activity in 100 consecutive patients admitted to an alcohol detoxification unit revealed hyperamylasemia in 39 cases. Further clinical evaluation of 15 of the 39 alcoholic patients with hyperamylasemia was unremarkable except for bilateral enlargement of the parotid glands in two cases. Nine of the 15 patients demonstrated markedly low amylase to creatinine clearance ratio; however, macroamylase complexes were not detected in the sera of any patients. Serum isoamylase separation revealed that the mean salivary isoamylase for the 15 alcoholic patients was significantly (P<0.05) elevated as compared to the control values. Individually, the salivary-type isoamylase was clearly elevated in ten patients while pancreatic type isoamylase was elevated in four. These data indicate that elevated serum amylase activity occurs frequently in alcoholic patients. Hyperamylasemia in a large number of alcoholic patients is nonpancreatic in origin and may be related to the injurious effect of ethanol on salivary glands and other tissues.

Evaluation of an inhibitor assay to determine serum isoamylase distribution

We tested a technique to distinguish salivary from pancreatic isoamylase using a wheat protein which inhibits salivary isoamylase. The inhibitor technique accurately reflected the preponderance of pancreatic or salivary isoamylase in sera which had been “spiked” with human pancreatic or salivary isoamylase. Comparison of the results of either cellulose acetate electrophoresis or isoelectric focusing showed an excellent correlation (r=0.99) in 43 hyperamylasemic sera which did not contain macroamylase. Normal values (1sd) of total serum amylase activity and the upper limit of normal for serum pancreatic isoamylase was 166 IU/liter. The inhibitor assay provides a simple and accurate means of differentiating salivary from pancreatic hyperamylasemia.

Shaking of the intact rat and intestinal angulation diminish the jejunal unstirred layer

A sizeable pre-epithelial diffusion barrier (unstirred layer) is present during perfusion of the rat jejunum. In the present study, three rapidly transported compounds, CO, [14C]warfarin, and glucose (5.5 mmol/L), were used as probes to assess the ability of manipulations to reduce the unstirred layer. This layer was 700-800 microns thick in a 30-cm jejunal segment perfused in conventional fashion on the abdominal wall. Placement of four sharp angulations in the segment or replacement in the abdominal cavity reduced the maximal unstirred layer to 200-400 microns. Increasingly rapid shaking of the anesthetized, intact rat on a platform shaker produced progressively thinner unstirred layers. At 250 revolutions per minute, the maximal layer ranged from 32 to 68 microns for the three probes and may have been appreciably less if the epithelium offered appreciable resistance. Shaking yields a > 15-fold reduction in unstirred layer resistance and provides a means for measuring this resistance and for obtaining more accurate assessment of the true in vivo transport Michaelis constant (Km) of any compound.

Storage of breath samples for H2 analyses

Breath hydrogen tests are limited by leak of gases from breath samples from standard glass or plastic syringes. These losses may be substantial when these samples are stored for long periods of time. We describe a simple manipulation involving the use of a mineral oil or water seal at the end of the syringe barrel that markedly reduces the leakage of H2 and other gases to negligible levels.

Gas chromatographic method for the determination of the lower volatile alcohols in rat blood and in human stool specimens on a fused silica capillary column

A method is described for the simultaneous quantitation of the lower volatile alcohols in stool specimens and rat blood. The addition of potassium carbonate to the assay mixture markedly increased the sensitivity in the detection of these compounds. The method is shown to be simple and reproducible and is suitable for following the metabolism of ethanol in human stool specimens.