Prevalence of Vibrio cholerae O1 serogroup in Assam, India: A hospital-based study

Cholera continues to be a major public health problem of growing concern in most developing countries1. It is endemic in India and whole of the Ganges basin from where it has spread in the form of epidemics and pandemics to several other parts of the world. Globally, cholera alone causes 120,000 deaths annually2. In Asia, Indian subcontinent contributes about 78 per cent of cholera cases3. The incidence of cholera is estimated to be 1.6 cases/1000 population per year or 40/1000 cases of acute diarrhoea in India45.

Depending on somatic antigens, Vibrio cholerae have more than 200 serogroups, of which O1 and O139 are outbreak/epidemic strains6. V. cholerae O1 has two well-established biotypes, namely, classical and El Tor, which are differentiated primarily based on phenotypic characters. Observation during the last few years shows that many of the phenotypic and genotypic tests have been proven to be inadequate for classifying V. cholerae O1 strains into their biotype7. Strains having conventional phenotypic properties of both classical and El Tor are designated as ‘hybrid biotype’ and another which is similar to the El Tor biotype by conventional phenotypic traits, but produces classical type CT are designated as ‘El Tor variant’7. This El Tor variant has caused several outbreaks/epidemics in Odisha, Chandigarh, Southern India, West Bengal, etc.8910111213. This study was undertaken to evaluate the changing trend of epidemiological profile, prevalent biotypes and serotypes of V. cholerae and their antimicrobial resistance pattern during 2003-2013 in Assam, India.

Material & Methods

This hospital-based study was carried out at Gauhati Medical College, Guwahati, Assam, India, for a period of 11 years from 2003-2013. The study protocol was approved by the Institutional Ethics Committee. All acute watery diarrhoea (AWD) cases referred to Gauhati Medical College Hospital and Fakhruddin Ali Ahmed Medical College Hospital from various districts of Assam were included in the study after applying the inclusion and exclusion criteria14. Written informed consent was obtained from patients or parents in case of children.

All AWD patients with or without vomiting and/or dehydration of any age group were included in the study. Those presenting with acute bloody diarrhoea (acute diarrhoea with visible blood in stool with or without fever and/or pain in abdomen), and patients with AWD who were on antibiotics were excluded from the study.

During the study period, 1779 stool samples/rectal swabs were collected. The stool samples were collected in sterile container and the rectal swabs were collected in Cary-Blair Transport Medium. Enrichment was done in alkaline peptone water (APW, pH 8.0) for 6-8 h. Before the samples were subcultured, hanging drop preparations were made to detect the typical darting motility of V. cholerae if present, both directly from the sample and after enrichment with APW. The subculture was done on MacConkey agar and thiosulphate-citrate-bile salt-sucrose agar medium. The isolates were identified as V. cholerae according to the standard laboratory methods using Gram stain, wet mount for motility, oxidase test, string test, triple sugar iron agar test and lysine iron agar test15. The culture media and reagents were obtained from HiMedia Laboratories, Mumbai.

For further confirmation, serotyping was done using V. cholerae O1 polyvalent antiserum followed by monovalent antiserum Ogawa and Inaba obtained from Denka Seiken Co., Ltd., Tokyo, Japan. Biotyping was done by standard conventional biotyping scheme using haemolysis of sheep blood, chick erythrocyte agglutination, Voges-Proskauer (VP) test and polymyxin B sensitivity with 50 μg disc15. Some of the representative strains were sent to National Institute of Cholera and Enteric Diseases (ICMR-NICED), Kolkata, for phage typing.

Antibiotic sensitivity was done as per the Clinical and Laboratory Standards Institute (CLSI) guidelines using antimicrobial agents (μg/disc) such as amoxicillin (10), co-trimoxazole (25), tetracycline (30), chloramphenicol (30), amikacin (30), ciprofloxacin (5), erythromycin (15) and furazolidone (50)16. Escherichia coli ATCC 25922 was used as quality control strain. Antibiotic sensitivity profile of each isolate was recorded as resistant, intermediate and sensitive based on the zone diameter according to CLSI guidelines. The antibiotic discs were obtained from HiMedia Laboratories Mumbai.

Data were collected and recorded using a predesigned questionnaire to study the trend of epidemiology of cholera during the study period. Year-wise proportions of cholera cases by age, gender, season and antibiotic resistance profile were analyzed to determine the trend. The descriptive statistical analysis was done manually and using statistical software Epi Info 7.1.2.0 (Centre for Disease Control and Prevention, USA).

Results

The AWD cases included in the study belonged to Dhubri, Barpeta, Nalbari, Kamrup (metro), Kamrup (rural), Nagaon, Dima Hasao, Cachar, Sonitpur, Karbi Anglong and Golaghat districts of Assam. The prevalence of cholera was 6.7 per cent (39 of 582 cases) in rural areas, 1.4 per cent (7 of 478 cases) in urban areas, 1.9 per cent (8 of 403 cases) in urban slums and 5.06 per cent (16 of 316 cases) in the tea gardens.

During 2003 to 2013, 1779 stool samples/rectal swabs were processed. V. cholerae was isolated from 70 (3.93%), Shigella flexneri from 30 (1.68%) and non-typhoidal Salmonella from seven (0.4%) cases. Other isolates included Aeromonas species in five (0.28%) and Staphylococcus aureus in pure culture in 10 (0.56%) cases. In 92.3 per cent of the AWD cases, normal enteric flora was found which predominantly included pure growth of E. coli in 1020 (57.3%) cases. However, serotyping of these E. coli isolates was not done to exclude enterotoxigenic and enteropathogenic E. coli. Vibrio cholerae isolation rate was highest in 2007 (8.7%) followed by 7.8 and 7.3 per cent in 2010 and 2011, respectively. In 2005 and 2006, no cholera case was found. S. flexneri was more prevalent in 2003 (10.6%) followed by 3.7 and 3.4 per cent in 2004 and 2011, respectively. Non-typhoidal Salmonella (3.9%) was isolated only in 2011. Table I depicts year-wise isolation of various enteropathogens.

Table I Year-wise isolation of enteropathogens in the study

Age-wise proportion of cholera cases is shown in Table II. The highest proportion of cholera was observed in the age group 0-10 yr (13.7%) followed by 11-20 yr (4.3%) and 21-30 yr (3.2%). In 2004, 2007-2009 and 2011 and 2012, 0-10 yr age group was the most commonly affected age group. Analysis of year-wise trend showed that in 2008, 2009, 2011 and 2012, 25 per cent (1 of 4), 80 per cent (4 of 5), 15.4 per cent (2 of 13) and 14.3 per cent (1 of 7) cholera cases were found, respectively, in the age group 0-5 yr from the cholera endemic region. However, no such discernible pattern was observed in other years of the study period. Of the 1779 cases, 993 (55.8%) were female and 786 (44.1%) were male. V. cholerae was isolated from 4.3 per cent (43 of 993) of the female cases and 3.4 per cent (27 of 786) of the male cases. The year-wise analysis revealed that proportion of cholera cases was more among the females in most of the years except in 2003 and 2007, where cholera was more prevalent among males than females (7.8 vs 3.7% and 10.6 vs 7.05%, respectively) (Figure).

Table II Year-wise proportion (%) of cholera cases in different age groups

Close

Figure

Year-wise and gender-wise distribution of acute watery diarrhoea (AWD) and cholera cases.

Export to PPT

Month-wise and year-wise AWD and cholera cases are presented in Table III. It was observed that during the period 2003-2009, most of the diarrhoea cases occurred in the monsoon season (July-October) which contributed 41.2 per cent (733 of 1779 cases) of the AWD cases as well as 71.4 per cent of the cholera cases (50 of 70 cases). The overall prevalence of cholera was highest 6.8 per cent (50 of 733 cases) during the monsoon. In 2013, only one cholera case was reported in monsoon season (Table III).

Table III Month- and year-wise trend of acute watery diarrhoea (AWD) and cholera

The predominant isolate in our study was V. cholerae O1 biotype El Tor serotype Ogawa (94.3%). In 2004, 2.9 per cent of the cases were found to be V. cholerae O1 biotype El Tor serotype Inaba. Two (2.9%) V. cholerae O1 isolates showed characteristics of El Tor biotype with some characteristics of classical variety (hybrid variety). These isolates were negative for VP reaction, sensitive to polymyxin B (50 μg), the traits typical for the classical biotype and were found to cause haemolysis of sheep erythrocytes, chick erythrocyte agglutination. These isolates were phage typed in NICED, Kolkata, and were reported to be phage type T-4 (Basu & Mukherjee) and T-27 (New scheme).

The antibiotic resistance profile was analyzed during the study period and found none of the isolates were consistently sensitive to any of the antibiotics. The most important finding was that 52.8 per cent of the isolates were resistant to tetracycline. However, a decline in the rate of tetracycline resistance was observed since 2010. The low-level resistance of 7.1 and 14.3 per cent was observed against amikacin and ciprofloxacin, respectively (Table IV).

Table IV Year-wise antibiotic resistance profile of the Vibrio cholerae isolates (resistance per cent)

Discussion

The present study showed high prevalence of cholera in rural areas (6.7%) followed by the tea gardens (5.06%) of Assam. This finding corroborated with that of the earlier studies carried out in Assam1718 which reported outbreaks of cholera in rural areas and tea gardens of Assam. Data from population-based diarrhoea surveillance in an endemic area of Kolkata revealed a cholera incidence of 2.2 cases per 1000 person-years19. During a 10-yr study period, the States having the highest number of reported cholera outbreaks were West Bengal, Odisha, Maharashtra and Kerala, which together accounted for 60 per cent of all reported outbreaks. Most of the cholera cases (91%) were identified in the States or union territories of Odisha, West Bengal, Andaman and Nicobar Islands, Assam and Chhattisgarh19. In India, there has been resurgence of cholera in many places such as Chandigarh and several north western States and outbreaks including major epidemics have occurred from time to time at various places891017202122. The attack and case fatality rate of cholera in Assam have been reported to be 11.6 and 0.8 per cent, respectively17.

Cholera contributed to 3.93 per cent of all AWD cases in the present study. Sur et al5 reported a similar prevalence rate (4%) of culture confirmed cholera cases among all the AWD cases in a community-based study in slums of Kolkata, India. Kaur and Lal23 also reported similar isolation rate of V. cholerae 2, 2.6, 6.7, 7.08, 0.9 and 2.6 per cent during 1992-1997, respectively, in a hospital-based study in north India. In 2005 and 2006, no cholera case was reported and during 2003-2013, most of the stool/rectal swab cultures revealed neither V. cholerae nor any other enteropathogen which might be due to the prior usage of antibiotics by the patients. However, some studies reported higher prevalence rates 14.03 per cent24, 20.8 per cent25 and 16.2 per cent26.

No discernible pattern was noticed in year-wise prevalence of cholera in different age groups. However, finding of 11.5 per cent cholera cases in the age group 0-5 yr from cholera endemic region was an important finding of the study. The prevalence rate of cholera in 0-5 yr age group found in this study was lower than that reported by others others (32.722 and 31.427 % of cholera cases). Cholera was seen more among females in most of the years in our study except the year 2003 and 2007, during which male predominated the female cases. Increased prevalence among females has been reported by Pal et al25, however, many studies have reported higher attack rate among males than females or no variation was reported in the attack rate between genders2627.

V. cholerae O1 biotype El Tor serotype Ogawa (94.3%) was the predominant isolate in the present study which was comparable to many studies done in India1718242527. Another important finding of the study was that 2.9 per cent of the V. cholerae isolates were hybrid biotype showing common characteristics of both classical and El Tor biotypes. However, VP reaction was variable and was not a recommended marker for the differentiation of classical and El Tor vibrios. The existing biotyping scheme has limitations and causes confusion as many of the hybrid biotype and El Tor variant strains have phenotypic and genetic changes.

The analysis of year-wise antibiotic resistance profile revealed that amoxicillin resistance decreased during 2003-2004, followed by an increase in resistance to 100 per cent during 2011-2013. A similar finding was also reported by Chander et al27. A declining trend of tetracycline resistance was observed during 2003-2007 ranging from 71.4 to 64.2 per cent followed by a high-level resistance (100%) in the year 2008, after which the resistance again declined to 28.5 per cent in 2012. Overall, 52.8 per cent V. cholerae isolates were resistant to tetracycline, which was an important finding. There are reports of tetracycline-resistant V. cholerae strains responsible for major epidemics of cholera in Latin America, Tanzania, Bangladesh and Zaire28. Hence, any level of resistance to tetracycline is significant and requires monitoring. In the study carried out in Odisha by Pal et al9, all clinical V. cholerae isolates were sensitive to gentamicin, azithromycin, chloramphenicol and tetracycline and resistant to ciprofloxacin, norfloxacin, ampicillin, streptomycin, neomycin, nalidixic acid, furazolidone and co-trimoxazole. A low-level resistance to tetracycline 4.34-15.38 per cent27 and one per cent5 has also been reported. Bhattacharya et al29 reported a sudden upsurge in tetracycline resistance among V. cholerae isolates, from one per cent in 2004 to 76 per cent in 2007 before decreasing to 50 per cent in 2009. Kar et al30 reported 100 per cent tetracycline-resistant V. cholerae O1 El Tor variant strains causing the cholera epidemic in Odisha.

In the present study, 34 per cent isolates were sensitive to co-trimoxazole and furazolidone in 2004; however, these were found to be the least effective antibiotics in vitro for the rest of the study period. Bhattacharya et al29 also reported an increase in resistance to furazolidone and trimethoprim/sulphamethoxazole during the study period. A declining trend of chloramphenicol resistance was also observed during the study period ranging from 100 per cent resistance in 2003 to 0 per cent in 2013. A low-level resistance to chloramphenicol has already been reported2122. Emergence of resistance to ciprofloxacin was observed in our study during 2008-2012, which ranged from 15.3 to 40 per cent. This correlated with the findings reported by earlier workers2730.

In conclusion, multiple drug-resistant V. cholerae O1 Ogawa strains were reported during 2003-2013 from Assam. Development of resistance to commonly used antibiotics indicates a serious public health concern because it complicates treatment by extending the duration of hospital stay for patients. Hence, a continuous surveillance for diarrhoeal disorders is necessary to control the future outbreaks of cholera in this region.