Complications and outcome of percutaneous endoscopic gastrostomy in different patient groups

A retrospective study of percutaneous endoscopic gastrostomy (PEG) was performed to evaluate the complications of PEG and determine the role of prophylactic antibiotics in preventing early wound infection and to evaluate the outcome of patients in different groups (neurological disease, head injury, AIDS). Percutaneous endoscopic gastrostomy was performed on 50 patients between March 1991 and November 1993 and the survey was completed in July 1994. The average time to PEG placement post-cerebrovascular accident or head injury was 5 weeks. No deaths were attributable to the procedure. Four of 50 (8%) patients died in the first 30 days (early mortality). Wound infection (early eight, late five) was the most common complication: two of 13 patients receiving prophylactic antibiotics and six of 37 not receiving antibiotics had early wound infection. At completion of follow-up 19 of 26 patients with neurological disease had died (median survival = 370 days), three resumed oral feeding and four continued PEG feeding. All head injury patients (n = 6) were alive and had resumed oral feeding; five returned home. All AIDS patients (n = 7) died within 12 months (median survival = 138 days), although improved nutritional state was found after gastrostomy feeding. In conclusion, PEG placement is a useful procedure to assist feeding in multiple patient groups. Wound infection is a common but not life threatening complication. The need for prophylactic antibiotics to prevent early wound infection has not been proven. The 5 week delay in PEG insertion may contribute to lower early mortality. The 100% survival rate in head injury patients may reflect their young age and absence of underlying medical illness. In AIDS patients, improved nutritional state is not known to translate into better quality of life or prolonged survival.

A simple filtration method to remove plankton-associated Vibrio cholerae in raw water supplies in developing countries

Plankton to which cells of Vibrio cholerae O1 and/or O139 were attached was introduced into 0.5% Instant Ocean microcosms maintained at 25 degrees C. The bulk of the plankton and associated particulates was removed with a filter constructed from either nylon net and one of several different types of sari material, the latter being very inexpensive and readily available in villages in Bangladesh, where V. cholerae is endemic. V. cholerae was enumerated before and after filtration to evaluate the efficiency of the filtration procedure. The results obtained indicate that 99% of V. cholerae, i.e., those cells attached to plankton, were removed from the water samples. Epidemic strains of V. cholerae O1 and O139 from various geographical sources, including Bangladesh, Brazil, India, and Mexico, were included in the experiments. Removal of vibrios from water by this simple filtration method was found to yield consistent results with all strains examined in this study. Thus, it is concluded that a simple filtration procedure involving the use of domestic sari material can reduce the number of cholera vibrios attached to plankton in raw water from ponds and rivers commonly used for drinking. Since untreated water from such sources serves as drinking water for millions of people living in developing countries (e.g., Bangladesh), filtration should prove effective at reducing the incidence and severity of outbreaks, especially in places that lack fuel wood for boiling water and/or municipal water treatment plants. The results of this study provide the basis for determining such reductions, which are to be carried out in the near future.

Deaths in visceral leishmaniasis (Kala-azar) during treatment

Twenty-seven out of five hundred and fifty three patients hospitalized for visceral leishmaniasis (Kala-azar) died during treatment with sodium antimony gluconate. Data from these patients were evaluated to find out the cause of death. Eight patients had associated diseases such as pulmonary tuberculosis (3), severe malnutrition (1), acute gastroenteritis (1), spleenic infarction (1), acute renal failure (1) and atrial septal defect (1) which could be attributed to death. Twelve patients developed spontaneous haemorrhages from nose, gums and gastrointestinal tract and died, despite of adequate supportive measures. Seven other patients who were improving slowly with antimony therapy died unexpectedly. Though, cause of death could be explained in some patients with associated disease conditions, it could not be explained in others as significant clinical manifestations, haematological, biochemical and electrocardiographic alterations were not evident prior to death. Our impression is that mortality in Kala-azar patients during standard antimonial therapy is more related to the drug rather than the disease process.

Serogroup conversion of Vibrio cholerae non-O1 to Vibrio cholerae O1: effect of growth state of cells, temperature, and salinity

Recently, we reported the occurrence of seroconversion from Vibrio cholerae non-O1 to V. cholerae O1, but little is known about the environmental and physiological factors influencing seroconversion. We investigated effects of temperature (4, 25, and 35 degrees C) and salinity ( < 0.05 and 10%0.), as well as the stage of growth of cells, on serogroup conversion. Seroconversion of V. cholerae occurred under various environmental conditions. However, the rate of seroconversion in natural water ( < 0.5% salinity) and synthetic seawater microcosms (10%0 salinity), employing cells harvested from stationary phase culture, was approximately 2 logs higher than cells harvested from cultures in the logarithmic phase (i.e., 10(5) versus 10(3) per 10(10) cells. Thus, the physiological state of the cells, and to a lesser degree, temperature and salinity, is an important factor in the conversion of V. cholerae from non-O1 to O1 serogroup.

Potential virulence of viable but nonculturable Shigella dysenteriae type 1

We examined a virulent strain of Shigella dysenteriae type 1 after induction into the viable but nonculturable (VBNC) state for its ability to (i) maintain the Shiga toxin (stx) gene; (ii) maintain biologically active Shiga toxin (ShT); and (iii) adhere to intestinal epithelial cells (Henle 407 cell line). PCR was used to amplify the stx gene from VBNC cells of S. dysenteriae type 1, thereby establishing its presence even when cells are in the VBNC state. VBNC S. dysenteriae type 1 ShT was monitored by the enzyme-linked immunosorbent assay with mouse monoclonal antibodies against the B subunit of ShT and affinity-purified rabbit polyclonal antibodies against ShT. We used the Henle 407 cell line to study the adhesive property of VBNC S. dysenteriae type 1 cells in a series of tissue culture experiments. Results showed that VBNC S. dysenteriae type 1 not only maintained the stx gene and biologically active ShT but also remained capable of adhering to Henle 407 cells. However, S. dysenteriae type 1 cells lost the ability to invade Henle 407 cells after entering the VBNC state. From results of the study, we conclude that VBNC cells of S. dysenteriae type 1 retain several virulence factors and remain potentially virulent, posing a public health problem.

Serogroup conversion of Vibrio cholerae

Vibrio cholerae serogroup O1 can be detected in the environment in a viable but nonculturable form, whereas V. cholerae non-O1 cells can be readily cultured during interepidemic periods in geographical regions where cholera is endemic. In the present study, pure cultures of V. cholerae non-O1 cells contained O1 cells when examined by immune-fluorescence microscopy. Laboratory microcosms were used to examine the outgrowth of the O1 cells in cultures of non-O1 V. cholerae. One O1 cell per 10(6) non-O1 cells could be detected by direct fluorescent-monoclonal antibody staining but only after incubation of the non-O1 culture for 48 h. Individual O1 cells were not detected in cultures incubated less than 48 h. Hybridization study, using a polymerase chain reaction (PCR) amplified fragment of the O-antigen of V. cholerae O1 as a probe, revealed the existence of a homologous gene in a microcosm sample of V. cholerae non-O1 containing serogroup-converted cells. The mechanism by which O1 cells can occur in cultures of non-O1 V. cholerae most likely resulted from spontaneous mutation of gene(s) encoding the O-somatic properties and (or) chemical, physical, or biological changes in the environment inducing expression or repression of the controlling gene(s). These findings have important implications for the epidemiology of cholera and the environmental source(s) of toxin producing V. cholerae O1.

Enterobacterial repetitive intergenic consensus sequences and the PCR to generate fingerprints of genomic DNAs from Vibrio cholerae O1, O139, and non-O1 strains

Enterobacterial repetitive intergenic consensus (ERIC) sequence polymorphism was studied in Vibrio Cholerae strains isolated before and after the cholera epidemic in Brazil (in 1991), along with epidemic strains from Peru, Mexico, and India, by PCR. A total of 17 fingerprint patterns (FPs) were detected in the V. cholerae strains examined; 96.7% of the toxigenic V. cholerae O1 strains and 100% of the O139 serogroup strains were found to belong to the same FP group comprising four fragments (FP1). The nontoxigenic V. cholerae O1 also yielded four fragments but constituted a different FP group (FP2). A total of 15 different patterns were observed among the V. cholerae non-O1 strains. Two patterns were observed most frequently for V. cholerae non-01 strains, 25% of which have FP3, with five fragments, and 16.7% of which have FP4, with two fragments. Three fragments, 1.75, 0.79, and 0.5 kb, were found to be common to both toxigenic and nontoxigenic V. cholerae O1 strains as well as to group FP3, containing V. cholerae non-O1 strains. Two fragments of group FP3, 1.3 and 1.0 kb, were present in FP1 and FP2 respectively. The 0.5-kb fragment was common to all strains and serogroups of V. cholerae analyzed. It is concluded from the results of this study, based on DNA FPs of environmental isolates, that it is possible to detect an emerging virulent strain in a cholera-endemic region. ERIC-PCR constitutes a powerful tool for determination of the virulence potential of V. cholerae O1 strains isolated in surveillance programs and for molecular epidemiological investigations.

A gene for the enterotoxin zonula occludens toxin is present in Vibrio mimicus and Vibrio cholerae O139

The presence of the zonula occludens toxin (ZOT) gene, which encodes an enterotoxin produced by serotype O1 strains of the pathogenic bacterium, Vibrio cholerae, in addition to cholera toxin, was investigated in selected strains of V. mimicus and the new pandemic V. cholerae non-O1 serotype O139. The zot gene was detected by polymerase chain reaction (PCR) amplification, using sets of primers based on the sequence of the V. cholerae O1 zot sequence. PCR amplification of genomic DNAs of both cholera toxin gene (ctx) positive and ctx- strains of V. mimicus detected the presence of zot gene. An AccI-EcoRV V. cholerae zot gene fragment designed to overlap PCR products was used as a probe. Southern hybridization studies confirmed that the PCR fragments from V. mimicus and V. cholerae O139 were strongly homologous to the V. cholerae O1 zot gene. The zot gene was found with 3 of 5 strains of V. mimicus of which only one strain harbored the ctx gene. The presence of a zot gene in ctx- toxigenic V. mimicus indicates a possible role of ZOT in the toxigenicity of this species. We conclude that, in addition to ctx, V. mimicus and V. cholerae O139 have the potential to produce ZOT.

Transferable drug resistance (R-factor) among the enterobacteriaceae in urinary tract infections: a study at an urban hospital in Bangladesh

Prevalence and patterns of drug resistance were studied among Enterobacteriaceae, isolated from the cases of urinary tract infection (UTI) attending at an outpatient department of an urban hospital in Dhaka. Out of 90 Enterobacteriaceae isolated, 95.5% were resistant to the different antimicrobials tested. The most common resistance pattern observed was against ampicillin (A), trimethoprim (Tm), sulphamethoxazole (S), tetracycline (T) and chloramphenicol (C) in all four genera of Enterobacteriaceae. Transferable drug resistance (R(+)-factor) was detected in 68.5% Escherichia coli (E. coli), 60% Klebsiella, 66.6% Proteus and 50% Citrobacter strains. By using a resistance transfer factor (RTF) mobilizing strain, resistance factors were transferred from 3 (11.3%) out of 26 non-autotransferable strains. The study revealed that transferable drug resistance is common in organisms isolated from UTI in Bangladesh.

Copper sulphate poisoning

A number of cases of copper sulphate poisoning from the southern region of Bangladesh are presented. There was a high mortality (24.9%) in comparison to other forms of poisoning. Hepatotoxicity, acute renal failure and gastrointestinal bleeding were the main complications. This degree of copper sulphate poisoning in one particular area is cause for alarm.

Studies on hemolytic action of a hemolysin produced by Vibrio mimicus

Some properties and mechanism of action of a hemolysin (VMH) produced by an enteropathogenic Vibrio mimicus strain was examined. VMH was heat-labile and inhibited by addition of divalent cations, including Ca2+, Mg2+ and Mn2+. The hemolysis by VMH was inhibited by incubating with gangliosides, suggesting that the ganglioside was the binding site on the erythrocyte membrane for VMH. Existence of a galactose moiety on reducing end of the ganglioside molecule and a sialic acid on the galactose moiety was suggested to be important for the binding of VMH molecule. Colloid osmotic manner of the hemolysis by VMH was demonstrated.

Dissecting microscopic (DM) and light microscopic (LM) study of the intestinal mucosa of Bangladeshi population

A total of 67 specimens of proximal intestinal mucosa, one from each subject, were examined under dissecting microscopy (DM) and light microscopy (LM). Majority of the subjects were adults and males. The overall per-oral biopsy failure was 13. DM was found to be useful to demonstrate gross villous atrophy. It also helped to orientate the specimens properly before fixation for embedding and sectioning. DM and LM findings did not concord in 17% of the specimens. In all of these specimens the mucosal surface which appeared normal under DM revealed grade-2 change under LM. The present study showed that dissecting microscopy (DM) of the intestinal mucosa is not an alternative for histologic examination under LM. However, scanner of light microscope (SLM) can be used as an alternative for DM.