We have initiated studies throughout Kenya. Each location represents a population with unique characteristics and exposure to bacterial pathogens that have yet to be fully described.  For example, Machakos  is a prosperous highland crop growing area; Entasopia , a southern center for rural Masai pastoralists; and Mathare , a vast Nairobi shantytown.  Study locations are rotated throughout Kenya in an effort to obtain information from a wide variety of climates and communities. These and plans to establish a surveillance network are collaborations with government clinics and NGOs, such as the African Medical and Research Foundation (AMREF), who are often able to obtain specimens from remote locations that do not have access to advanced laboratory based technology. 

Objectives

1. Determine the causes of acute diarrhea in selected, high incidence sentinel populations in Kenya.
2. Establish a surveillance network to determine antibiotic resistance   patterns and to detect emerging resistance. and 
3. Conduct studies to elucidate the mechanisms of resistance and develop tools to detect toxin genes.
4. Describe the pathogenic features of bacterial isolates other than toxins.

Recent Accomplishments

• First isolation of an invasive, toxigenic E. coil from this region: an E. coli O6:HNM positive by PCR for both shigatoxin-1 and VirG from Machakos was cytotoxic in Vero cell assay and invasive in the Sereny test.  
• First isolation of Shigella dysenteriae type 12 from Kenya made during an outbreak of dysentery at Loyangalani.
• Isolation of E. coli O157:H7 Masai in Southern Kenya, who have limited contact with visitors from outside Kenya.  This pathogen has been documented only once before, in the coastal town of Malindi.

Recent Studies

Machakos

Machakos District Hospital lies about 50KM east of Nairobi, on a high , semi-arid plateau devoted crops such as vegetables and coffee and animal husbandry. During march- September 1999, specimens were collected from 264 patients presenting with diarrhea. Case definition was the passage of three or more unformed stools in a 24-hour or any one stool containing blood and / or mucous. Among the 264 patients meeting these criteria, 40 presented with bloody diarrhea.

After serotyping , E.coli isolates were examined  by PCR for the presence of shigatoxin 1 and shigatoxin 2 genes;Shigelle dysenterae type 1,3 of which were shigatoxin (stx) gene positive. All 3 stx positive isolates produced this toxin in vero cells. Other serotypes of Shigella recovered during the study period included : S.boydii 3.8%, S.flexneri 5.7% and S.sonnei  6.4%. Pathogenic E.coli was isolated from 74(28%) of the patients. Among the E.coli isolates were 2 E.coli 0157:H- and 1E.coli 06:HNM.These were shigatoxin1(stx) gene positive for VirG  and were cytotoxic to Vero cells. 

This table summarizes our findings on antibiotic susceptibility:

Percent of each isolates resistant to given antibiotic. TE=Tetracycline, AM=Ampicillin, CH= Chloramphenicol, GE=Gentamycin, ER=Erythromycin, NA= Nalidixic Acid, SXT= Trimethprim-Su;famethoxazole

During this study , children 0-5 years old had the highest incidence of diarrhea caused by shigatoxin producing E.coli (STEC).Considering STEC alone , these isolates were responsible for 5.0% of bloody diarrhea cases and 2.7% of all E.coli isolated fom the Machakos study area. All STEC isolates carried the stx1 gene however , no isolate was positive  for stx2. The strong association of stx2 and the development hemolyticurmic syndrome (HUS) and the low incidence of this toxin during our study may be one explanation for why this syndrome is seldom reported from East Africa. The most interesting finding in this study, was the recovery of E.coli 06:HNM that carried both, stx1 and VirG. We believe this is the first finding of a pathogenic E.coli capable of both toxin production and invasion.

Due to the low recovery of STEC during this study it is important to consider other possibilities for why HUS and this group of organisms is seldom reported. For instance, based on the low recovery of STEC, it is possible that STEC do not represent significant threat in this region. This study does not indicate further investigation of this organism 's distribution  is warranted in Kenya because it could be a more significant threat to populations with more risk factors for acquiring STEC infection.

Entasopia

Entasopia is a village about 100 km south of Nairobi, on the floor of RiftValley. It is a gathering point for semi-nomadic Masai cattle herders , whom we anticipated would have higher than normal exposure to ruminants and lower than normal antibiotic use. The study ran from March to November 1999, and 165 patients were seen at the rural health center where the study took place.

Diarrhea was defined as the passage of three or more unformed stools in a 24-hour period or any one stool containing blood and/or mucous. Specimens were placed in Cary-Blair transport media and transported to our laboratory in Nairobi. Once in Nairobi, each specimen was inoculated onto selective media and incubated for 24hours at 37C. Identification of isolates was by standard biochemical reactions. Isolates identified as Shigella sp. Salmonella sp. or E.coli were serotyped using commercial anti-sera. Each specimen was also examined on site for parasites.

E.coli isolates were examined by multiplex PCR for the presence of shigatoxin1(stx1) and shigatoxin 2 (stx2),intimin(eae) invasiveness(VirG), and aggregative (Eagg) genes. These isolates were then examined for the ability to produce cytotoxic effects on Vero cells and for adherence to HEp-2 cells.

Shigella dysenteriae isolates were also examined y PCR for the shigatoxin gene and positive specimens were tested under in vitro conditions using Vero cells. All bacterial isolates were tested by disc diffusion for antibiotic susceptibility.

The most common cause of acute bacterial diarrhea among the masai is due to E.coli  and STEC (defined as the presence of either stx1 or stx2) represented over 22% of the E.coli isolates. The close associations of the masai with ruminants along with certain cultural practices are possible explanations for this finding. E.coli 0157 was detected  only twice during the study period indicating that non-0157STEC may play more significant role than E.coli 0157 in infection in this setting. The virulence properties of our isolates (Table1) show that the majority of isolates were eae negative, carry stx2 and are capable of adherence to HEp-2 cells in vitro. Two isolates were eae negative and did not exhibit vitro adherence and are believed to be non-pathogenic as these patients had a concurrent infection with G. lamblia. The remaining eae negative isolates all showed in vitro adherence and carried the stx2 gene. Vero cells assays showed that three isolates were capable of producing toxin. Based on this information, we  believe these were the only three isolates capable of causing human disease. The remaining five STEC isolates are most likely animal pathogens that were acquired through close contact of the patients with animals and their waste products.

Even though E.coli were the most significant cause of bacterial diarrhea during this study , parasitic causes of diarrhea were more common than any bacterial cause. Giardia lamblia was the most common parasitic cause of diarrhea, accounting for over 35% of all diarrhea cases. Other parasites observed during this period are E.hystollica (15.2%), S.mansonii(10.9), Hookworm (5.5%) and T. trichuria(4.2%).

The antibiotic data (Table2) indicates all isolates were resistant to three or more antibiotics, including at least one first line drug for Kenya. However no isolate showed resistance to nalidixic acid. This is in contrast to other developing nations, where nalidixic acid resistance is present. The resistance data indicates that overall, antibiotic resistance was lower among this population than others previously studied. Due to the economic conditions in Kenya, it is imperative that surveillance efforts on drug sensitivity continue , nalidixic acid is the last affordable drug to treat enteric infections in Kenya.

Table1.

 Table2.

Antibiotics resistance profiles of isolates. All shigella sp. and E. coli were resistant to 3 or more antibiotics including at least one of recommended first line antibiotics in Kenya.

Mathare 4B

Mathare 4B is a peri-urban area of Nairobi, with a population of approximately 200,000. The area is characterized with poor sanitary conditions and limited access to healthcare. Mathare 4B was selected following many cases of undiagnosed bloody diarrhea at a clinic. It is anticipated that predominant bacterial causes of diarrhea will be different from those observed in a rural setting and may show a different pattern of antimicrobial susceptibility. The main interest is epidemiology of campylobacter sp. and non-0157STEC in population.