Exploring the link between gastrointestinal complications and over-the-counter analgesics: current issues and considerations.

Nonsteroidal anti-inflammatory drugs (NSAIDs) are one of the most widely used classes of drug worldwide, with as many as 8% of the global adult population taking prescription NSAIDs at any given time. It is well documented that NSAID-related gastrointestinal (GI) effects are a significant cause of morbidity and mortality. Data from Australia alone suggest that each year there may be as many as 2300 serious GI complications and 300 deaths. Risk factors for NSAID-related GI complications have been established and can be broadly divided into two groups: uncontrollable factors, which include elements such as age, gender, comorbidity, and a history of GI conditions, and controllable factors, such as the dose, type, and duration of NSAID treatment and cotherapy and possibly Helicobacter pylori infection. The increasing trend toward self-medication raises questions about the potential for GI complications with OTC doses of NSAIDs. Although there is some evidence that the frequency of GI complications with OTC doses may be less than that seen with prescription doses, it still exists. At present, there is insufficient evidence to determine the population attributable risk associated with their widespread use. Emerging data and factors that confound their interpretation are discussed. Given that the data are limited, a clear picture of the true extent of GI complications with OTC NSAIDs is not yet available. In light of our current understanding and because paracetamol continues to demonstrate a favorable side-effect profile, it remains a first-line analgesic for everyday pain.

Effect of venesection on bone mineral density in an eugonadal woman with haemochromatosis.

BACKGROUND: A 41-year-old premenopausal woman with newly diagnosed haemochromatosis was found to have osteopenia on screening bone mineral densitometry. METHODS AND RESULTS: Liver biopsy showed grade 3 haemochromatosis with an hepatic iron index of 4. Investigation for secondary factors for osteopenia revealed no cause. The patient was clinically and biochemically eugonadal. Following venesection of 8 L blood (4 g iron) over 17 months and calcium supplementation, her bone density rose significantly. Neck of femur bone density increased by 6.0% over 13 months and lumbar vertebral bone density increased by 7.2%. There are no previous reports of response of bone density to venesection in eugonadal patients or in women with haemochromatosis.

Cyclic AMP-dependent anion secretion in human small and large intestine.

Cyclic AMP-dependent Cl- secretion is the major secretion pathway in human intestine. The aim of the present study was to examine mechanisms involved in cAMP-dependent anion secretion in human small and large intestine. Surgical resection specimens from both jejunum and distal colon were studied under short circuited conditions. Addition of the phosphodiesterase inhibitor IBMX induced an increase in the short-circuit current (Isc) equivalent to the net increase in Cl- secretion. The Isc was inhibited by diphenylamine decarboxylate (DPC; Cl- channel blocker), bumetanide (basolateral Na+/K+/2Cl- cotransporter), BaCl2 (basolateral K+ channel) and Cl- free buffer in both segments and indomethacin (cyclo-oxygenase inhibitor) in colon alone. Diphenylamine decarboxylate appears to directly inhibit secretion in jejunum, although its inhibitory effect is possibly mediated by inhibition of cyclo-oxygenase in the colon. A small component of IBMX-stimulated Isc was inhibited by acetazolamide. Cyclic AMP-dependent secretion is largely apical Cl- secretion, although a small component appears to be HCO3. Secretion is dependent on basolateral K+ channels and Na+/K+/2Cl- cotransporters and, in the colon, is inhibited by indomethacin, implying a role for cyclo-oxygenase metabolites. The chloride channel blocker DPC inhibits secretion in both areas. This class of compounds may have potential for treatment of secretory diarrhoea.

Modulation of host response to Escherichia coli o157:H7 infection by anti-CD18 antibody in rabbits.

BACKGROUND/AIMS: Escherichia coli O157:H7 infection induces diarrhea, severe colitis, and colonic electrolyte transport abnormalities characterized by decreased Na absorption and Cl secretion. The aim of this study was to examine the role of the host inflammatory response in inducing distal colonic transport changes during infection with E. coli O157:H7. METHODS: New Zealand white rabbits aged 10 days were infected with E. coli O157:H7 strain EDL933 (plasmid+, verotoxin 1+, verotoxin 2+). Studies were performed daily from day 1 to day 5 postinfection and compared with uninfected controls (10 days old). Distal colonic ion transport was studied in vitro under short-circuited conditions in Ussing chambers, and tissue inflammation was assessed by mucosal myeloperoxidase activities and mucosal neutrophil (polymorphonuclear neutrophil [PMN]) counts. In a second study, PMN infiltration was inhibited by an anti-CD18 (leukocyte adhesion molecule) monoclonal antibody, IB4, and histology and transport were studied on day 5 postinfection. RESULTS: Infection with O157:H7 induced diarrhea and inhibition of Na absorption by day 3. CI secretion occurred on day 5, coincident with tissue infiltration with PMN. Pretreatment with IB4 prevented histological damage and tissue infiltration with PMN, and it inhibited the transport abnormalities induced by infection alone. CONCLUSIONS: Infection with O157:H7 reduces Na absorption and stimulates Cl secretion in the distal colon. Disruption of the epithelium and changes in colonic electrolyte transport during enterohemorrhagic E. coli are mediated by the host inflammatory response.

Decreased uptake of L-cystine by duodenal brush border membrane vesicles from patients with cystinuria.

BACKGROUND: The pathophysiology of cystinuria remains unclear. Decreased absorption of L-cystine across brush border membranes of small intestinal and renal proximal tubular epithelial cells is likely but has not been directly demonstrated. AIMS: To compare the rates of L-cystine transport by isolated duodenal brush border membranes of normal individuals and patients with cystinuria. METHODS: Distal duodenal biopsies were taken from normal individuals and patients with cystinuria. Brush border membrane vesicles (BBMV) were prepared using magnesium aggregation and differential centrifugation and the rates of L-cystine transport into the vesicles measured using a rapid filtration technique. RESULTS: Rates of L-cystine transport by BBMV from patients with cystinuria were reduced at 5 minute (p = 0.003) and 30 minute (p = 0.053). Time points, indicating that L-cystine absorption across brush border membranes is abnormal in cystinuria.

The effect of enterohemorrhagic Escherichia coli O157:H7 on intestinal structure and solute transport in rabbits.

BACKGROUND: The effect of enterohemorrhagic Escherichia coli O157:H7 infection on intestinal morphology and solute transport was examined. METHODS: New Zealand white rabbits, aged 10 days, were infected with E. coli strain EDL933 (O157:H7 containing the 60-megadalton plasmid-encoding adhesion factors VT1 and VT2) and compared with controls. Small and large intestinal histology and solute transport were studied 5 days after inoculation. Ion transport in the distal colon was also examined in animals infected with different strains encoding a combination of pathogenic factors. RESULTS: Infection with EDL933 induced diarrhea and mucosal disease in the colon, inhibited colonic Na+ absorption, and stimulated of Cl- secretion, but had no impact on the small intestine. Infection with strains A7785-C3A (O157:H7, plasmid-, VT1+, VT2+) and 85-170 (O157:H7, plasmid+, VT-) induced similar transport changes to EDL933. C600/1 (E. coli K-12, plasmid+, VT1+) decreased Na+ and Cl- absorption only. CONCLUSIONS: Abnormalities of colonic structure and ion transport could account for diarrhea production, but pathogenic factors other than the 60-megadalton plasmid-encoding adhesion factor and verotoxins appear to be involved in enterohemorrhagic E. coli infection.

Does iron affect osteoblast function? Studies in vitro and in patients with chronic liver disease.

In order to study the role of trace elements as potential osteoblastic toxins, we measured bone aluminum, copper, and iron in 106 ambulant patients with histologically proven liver disease. We used analytical and histochemical methods and we correlated our results with serum biochemistry, forearm and spinal bone density, and dynamic bone histomorphometry. Patients with chronic liver disease had higher iron-stained perimeters than control subjects (P less than 0.001). However, the mean iron-stained perimeter was no greater than 5% of the total mineralized bone perimeter and did not correlate significantly with either the osteoblast perimeters or bone formation rates. The mean concentration of bone iron were 2.5 times (P less than 0.01) greater in the patients than in the controls although 80% of the patients fell within the normal range. There was a weak negative correlation between bone iron and the osteoblast perimeters (R = 0.18, P = ns) and between bone iron and bone formation (R = -0.30, P less than 0.05). There were 57 patients (56% of the total) with diminished bone formation, but only 16 had elevated bone iron concentrations. In a regression analysis, age, hypogonadism, and serum albumin concentrations were the most important predictors of osteoblast perimeters and bone formation rates. In vitro experiments using rat osteoblast-like osteosarcoma cells showed that an iron concentration of 400 mumol/liter was required to diminish cellular proliferation and function. Iron concentrations are elevated in the bones of patients with chronic liver disease. However, there is at present insufficient evidence that this metal is responsible for the osteoblast dysfunction seen in these patients.(ABSTRACT TRUNCATED AT 250 WORDS)

Abnormal epithelial transport in cystic fibrosis jejunum.

Abnormal epithelial electrolyte transport has been identified in a range of cystic fibrosis (CF) organs and appears to account for the various clinical manifestations of the disease. The aim of this study was to further define the Cl- secretion defect in CF jejunum. Excised jejunum was obtained from 11 CF patients and 12 controls. Transport studies were performed on stripped epithelium in vitro under short-circuited conditions in Ussing Chambers. 3-Isobutyl-1-methylxanthine (IBMX) (300 microM) significantly increased Cl- secretion in control (-2.3 +/- 0.6 to -3.3 +/- 0.7 mueq.cm-2.h-1; P less than 0.01, paired t test; n = 5 subjects) but not in CF jejunum (-0.5 +/- 0.3 to -0.1 +/- 0.4; n = 4). However in contrast to control jejunum, net Na+ absorption in CF jejunum was higher in the IBMX (1.3 +/- 0.5 mueq.cm-2.h-1) compared with basal periods (0.6 +/- 0.3; P less than 0.05, paired t test). IBMX stimulation of tissue adenosine 3',5'-cyclic monophosphate (cAMP) was similar in both control and CF jejunum. A range of secretagogues known to induce secretion in mammalian intestine, including dibutyryl cAMP (DBcAMP), DBcGMP, Ca2+ ionophore A23187, and the protein kinase C activator 4 beta-phorbol 12,13-dibutyrate, failed to induce secretion in CF jejunum. In conclusion, CF jejunum failed to exhibit Cl- secretion and also demonstrated abnormalities of Na+ absorption. These results support the view that the defect lies at a site distal to the intracellular messengers. Moreover, these abnormalities of intestinal electrolyte transport may account for some of the gastrointestinal manifestations of the disease such as meconium ileus and distal intestinal obstruction syndrome.

Effects of testosterone and venesection on spinal and peripheral bone mineral in six hypogonadal men with hemochromatosis.

To measure the effect of testosterone replacement and venesection on spinal and peripheral bone mineral we prospectively studied six hypogonadal men and six eugonadal men with idiopathic hemochromatosis for 24 months. Venesections were performed every week on all patients, and intramuscular injections of testosterone were administered every 3 weeks to the hypogonadal men only. Bone mineral was measured by quantitative computed tomography in the spine and by single-photon absorptiometry in the forearm. During the 24 month period of observation serum testosterone concentrations and serum ferritin levels became normal. In the hypogonadal men mean lumbar spine bone mineral increased by 13.1 +/- 4.9% (95% CI, 0.5-25.6) and mean forearm bone mineral increased by 4.7 +/- 3.8% (95% CI, -5.1 to 14.6). In contrast in the eugonadal men treated over the same period, mean lumbar spine bone mineral decreased by 3.5 +/- 2.8% (95% CI, -10.9 to 3.8, P less than 0.01) and mean forearm bone mineral remained virtually unchanged (0.07 +/- 0.9%; 95% CI, -1.7 to 3.1, P less than 0.05). These data suggest that bone mineral increases in the lumbar spine and in the forearm in hypogonadal men with hemochromatosis treated by testosterone replacement and venesection.

Phenylthiocarbamide taste sensitivity in chronic peptic ulcer.

Ability to taste phenylthiocarbamide is genetically determined and has been investigated as a possible genetic marker for disease. This study examined phenylthiocarbamide taste sensitivity in gastric and duodenal ulcer disease. The study sample included 164 patients with gastric ulcer, 134 with duodenal ulcer, and 299 community controls. Eight concentrations of phenylthiocarbamide in distilled water were obtained by binary serial dilution. The lowest concentration distinguished by taste from distilled water defined taste threshold. Bimodality of threshold distributions distinguished nontasters from tasters. Comparisons of patients with controls gave odds ratios of nontaste in gastric ulcer and duodenal ulcer of, respectively, 0.7 (P greater than 0.1) and 1.3 (P greater than 0.03). The power of detecting at least a twofold difference between patients and controls in the odds of nontaste was 80%. Nontaste was more common in duodenal than in gastric ulcer patients (odds ratio = 2.0, P = 0.02). Taste sensitivity was unassociated with other genetic factors related to ulcer--ABO blood group, secretor status, and serum pepsinogen 1 level. The difference between gastric and duodenal ulcer patients in the ability to taste phenylthiocarbamide may be genetic; however, this study's inability, despite substantial power, to detect at least a twofold difference between patients and controls suggests that if phenylthiocarbamide taste sensitivity is a genetic factor in peptic ulcer, the relationship is weak.

Osteoporosis and skeletal fractures in chronic liver disease.

In order to determine the prevalence and severity of hepatic osteodystrophy by non-invasive means we compared 115 consecutive ambulant patients with histologically proven chronic liver disease to 113 age and sex matched control subjects. Methods used included the assessment of fracture prevalence rates, spinal radiography, and measurements of bone mineral density in the spine and the forearm. Spinal and peripheral fractures were more prevalent in the patients than in the control subjects (p less than 0.03 and p less than 0.01 respectively). The type of the underlying liver disease did not significantly affect the fracture prevalence rates, but alcoholic patients sustained more peripheral fractures than patients with other hepatic disorders (p less than 0.05). The bone mineral densities of the spines and the forearms were significantly reduced in male patients of all age groups and in female patients aged 60 years or more (p less than 0.001 for men and p less than 0.01 for women for both measurements). The prevalence rates of spinal and forearm osteoporosis were twice as high among patients with liver disease than in control subjects regardless of the definitions used. The presence of cirrhosis and hypogonadism were major risk factors for development of both spinal (Beta coef = 0.190 and 0.176; SE = 0.079 and 0.086 respectively) and forearm osteoporosis (Beta coef = 0.20 and 0.29; SE = 0.073 and 0.80 respectively). Spinal bone density was the predominant determinant of spinal fractures (Beta coef = -0.007; SE = 0.001), while hypogonadism (Beta coef = 0.363; SE = 0.075) and cirrhosis (Beta coef = 0.185; SE = 0.068) were the major predictors of peripheral fractures. The concentrations of serum calcium and serum vitamin D metabolites and the use of corticosteroids were apparently without effect on the prevalence of skeletal fractures or bone density.

Serum vitamin D metabolites are not responsible for low turnover osteoporosis in chronic liver disease.

We measured the concentrations of vitamin D-binding protein (DBP), total 25-hydroxyvitamin D, total 1,25-dihydroxyvitamin D [1,25-(OH)2D], and free 1,25-(OH)2D in sera of 107 patients with histologically proven chronic liver disease. Bone density measurements and dynamic skeletal histomorphometry were also performed. Osteoporosis, as defined by arbitrary criteria, was found in 42 patients (39%), while no patient had osteomalacia. Serum concentrations of vitamin D-binding protein, 25-hydroxyvitamin D, total 1,25-(OH)2D, and free 1,25-(OH)2D were reduced in patients with cirrhosis, but not in the noncirrhotic patients. Bone formation rates, which were low in 55 patients (51%), were correlated with liver functions, but not with the concentrations of either vitamin D metabolite. A subgroup of 44 patients with low serum 1,25-(OH)2D concentrations and low bone formation rates failed to show an appropriate increase in serum bone Gla protein after 1,25-(OH)2D3 administration even though serum concentrations of 1,25-(OH)2D rose normally. These data suggest that the bone disease in patients with hepatic disorders is not related to the serum concentrations of vitamin D metabolites or the effect of these metabolites on osteoblast function.

Osteoporosis in hemochromatosis: iron excess, gonadal deficiency, or other factors?

STUDY OBJECTIVE: To define the prevalence, severity, type and pathogenesis of osteopenia in idiopathic hemochromatosis. DESIGN: Prospective study conducted over 18 months. SETTING: Tertiary care center. SUBJECTS: Twenty-two men with idiopathic hemochromatosis and 20 age-matched controls. There were 5 hypogonadal patients, 9 eugonadal nonvenesected patients, and 8 eugonadal venesected patients. MEASUREMENTS AND MAIN RESULTS: All patients and controls were evaluated by spinal radiography, spinal and forearm bone mineral density estimations, dynamic skeletal histomorphometry, and serum biochemistry. Ten patients (45%; 95% CI, 24% to 68%) had osteoporosis as defined by spinal and forearm bone density measurements. Trabecular bone volumes were significantly reduced in the patients (the difference in means between patients and age-matched controls was 3.9%; CI, 1.3% to 6.7%). No patient had osteomalacia. Hypogonadal men had lower bone mass measurements than eugonadal men (radial bone density: beta coefficient = -20.5; CI, -29.2 to -11.8; trabecular bone volume: beta coefficient = -7.1; CI, -10.8 to -3.3). Osteoid and osteoblastic surfaces and bone formation rates were significantly greater in the eugonadal venesected compared with the eugonadal nonvenesected persons (P less than 0.05 for all measurements). CONCLUSIONS: A significant decrease in bone density is seen in idiopathic hemochromatosis, particularly when hypogonadism is present. Low serum free-testosterone concentrations rather than the calciotrophic hormones determine bone mass in this condition.

Ethanol reduces bone formation and may cause osteoporosis.

INTRODUCTION: The etiology of ethanol-associated osteopenia is not fully understood. In order to define the role of ethanol in the pathogenesis of hepatic osteodystrophy, we compared two groups of alcoholic patients with histologically established alcoholic liver disease. PATIENTS AND METHODS: Twenty-eight patients currently drinking ethanol ("drinkers") and 12 claiming not to have consumed any ethanol for at least six months ("abstainers") were enrolled in the study. In addition, 35 non-alcoholic control subjects without clinical or biochemical evidence of liver disease were also studied. Bone mineral density and various biochemical and hormonal values were measured in each subject; iliac crest biopsies were taken under local anesthesia in the patients and under general anesthesia in the control subjects. RESULTS: Forearm bone mineral densities, spinal bone mineral densities, and iliac crest cancellous bone areas were significantly lower in the alcoholic patients compared with control subjects (p less than 0.01 for all measurements), but these values did not differ between the drinkers and the abstainers. The drinkers, however, had significantly less osteoblastic activity than the abstainers, as assessed by dynamic bone histomorphometry (p less than 0.001). Serum bone Gla-protein concentrations were higher in the abstainers than in the drinkers (p less than 0.001). No differences were seen relating to histologic parameters of bone resorption, although the alcoholic patients who had lower serum free testosterone concentrations than the control subjects also had higher urinary hydroxyproline excretion rates. CONCLUSION: These data suggest that ethanol may be responsible for osteoblastic dysfunction resulting in diminished bone formation and reduced bone mineralization.

Hepatic osteodystrophy. Static and dynamic bone histomorphometry and serum bone Gla-protein in 80 patients with chronic liver disease.

To study the pathogenesis of osteoporosis in patients with chronic liver disease, we performed dynamic bone histomorphometry and measured serum bone Gla-protein in 80 patients with various types of chronic liver disease. These results were compared with results obtained in 40 healthy controls. Mean trabecular bone volume and mean trabecular thickness were significantly reduced in both men and women with chronic liver disease (p less than 0.001 for both measurements in men and p less than 0.01 for both measurements in women). Osteoporosis as defined by histologic parameters was present in 17 (21%) patients with no significant differences in prevalence rates among the various hepatic disorders. No patient had histologic evidence of osteomalacia, although mineralization lag times were prolonged (p less than 0.01 for men and women). Bone formation rates were significantly reduced in 46 (57%) patients, and unlike the static measurements, were related to the type and severity of the underlying liver disease. Patients with alcoholic liver disease, hemochromatosis, and cholestatic liver disease had lower bone turnover rates and osteoblastic surfaces (p less than 0.001 and p less than 0.05, respectively) than patients with chronic active hepatitis. Furthermore, the presence of hepatic cirrhosis was associated with diminished bone formation and lower osteoblast surfaces. Serum bone Gla-protein levels were significantly correlated with bone formation rates and osteoblast surfaces (r = 0.585 and r = 0.434, respectively). A reduction in osteoblast surfaces has not previously been demonstrated in liver disease. This reduction and the associated impairment of osteoblastic activity may contribute to the pathogenesis of osteoporosis and can be assessed by the measurement of serum bone Gla-protein.

Fibrolamellar carcinoma as a cause of bile duct obstruction.

Obstructive jaundice due to growth within bile ducts of hepatocellular carcinoma is uncommon and usually a manifestation of advanced, lethal tumour. We report a case of fibrolamellar carcinoma of the liver presenting with obstructive jaundice, caused by tumorous permeation of the left hepatic duct with migration of tumour fragments into the common bile duct. Immunocytochemical and ultrastructural features are described. Two and a half years after complete surgical resection the patient is free of tumour. The importance of accurate diagnosis of such tumours is emphasized.