The role of hepatitis B and hepatitis C viral infections in the incidence of hepatocellular carcinoma in Sudan

TRANSACTIONS OF THE ROYAL SOCIETY OF TROPICAL MEDICINE AND HYGIENE (2001)95,487-491

The role of hepatitis B and hepatitis hepatocellular carcinoma in Sudan

C viral infections

in the incidence

of

R. E. Omer’,*, P. Van? Vee3, A. M. Y. Kadaru3, E. Kampman’, I. M. El Khidir4, S. S. Fedail’ and ‘Department of Crop Protection, Faculty of Agriculture, University of Khartoum, Sudan; ‘Division of Human F. J. Kok* Nutrition and Epidemiology, Wageningen University, The Netherlands; ‘Department of Medicine, Faculty of Medicine, I? 0. Box 102, University of Khartoum, Sudan; 4Khartoum National Health Laboratory, Virology Department, I? 0. Box 287, Khartoum, Sudan; jNationa1 Centre for Gastrointestinal and Liver Disease, l? 0. Box 311, Khartoum, Sudan Abstract In Sudan, the incidence ofhepatocellular carcinoma (HCC) is high and increasing. Hepatitis B virus (HBV) and hepatitis C virus (HCV) are important risk factors of HCC. This study aims to assess the role of HBV and HCV infections in the incidence of HCC in 2 regions of Sudan. A case-control study was conducted in 1996-98 among 150 HCC patients and 205 controls from 2 regions in Sudan. Their demographic characteristics as well as food habits and chronic conditions have been investigated. In this study, 115 cases and 199 controls were tested for hepatitis B surface antigen (HBsAg) and for HCV antibodies. Strong positive associations were found between HBV or HCV, and HCC with odds ratios (ORs) 9.8 (95% CI 5.118.9) and 8.3 (95% CI 2.3-29.9), respectively. After adjustment for age, by logistic regression, the ORs for HBVandHCVwere 16.1 (95% CI 7.4-34.9) and4.5 (95% CI l.l-18.6), respectively. Further adjustment for region, education level and job type did not appreciably affect the results. Given a prevalence of HBV and HCV of 7.0% and 1.5% among controls, about 57% of all HCC cases can be attributed to these viral infections. Hepatitis infections seem to be important risk factors for HCC in Sudan. Keywords:

hepatitis B virus, hepatitis C virus, prevalence, hepatocellular carcinoma, risk factors, Sudan

Introduction Hepatocellular carcinoma (HCC) is the sixth most frequent tumour in the world. Although it is relatively rare in western Europe, North America and Australia, it is considered to be one of the most common malignancies in sub-Saharan Africa, South-East Asia and China (MUIR, 1992). In spite of the fact that the tumour was underdiagnosed because of lack of autopsies and biopsies, HCC has been reported to show an increasing incidence in the past 2 decades in The Sudan (ZAKI, 1991). DAOUD et al. (1969) reported 35 cases of HCC (2.7% of all malignancies) seen in Khartoum hospitals in the year 1968; more recently, medical workers at Khartoum hospitals have reported much higher numbers, especially from western Sudan (MOHAMADIEN, 1993). It is well-established worldwide that HCC is strongly associated with chronic infection with hepatitis B virus (HBV; IARC. 19941. HBV infections have also been ieportkd in Sidanese patients with HCC (ITOSHIMA et al., 1989). In some Asian countries, such as Taiwan, Hong Kong, mainland China, and countries within subSaharan Africa, HBV prevalence is high and considered to be a major cause of HCC, with HBV acting either on its own or in combination with aflatoxins (JENG & TSAI, 1991; COURSAGET et al., 1992; LEUNG et al., 1992; WILD et al., 1992; BIJKH et al., 1993). Several case-control studies suggest that hepatitis C virus (HCV) may have a role in the aetiology of HCC; in most of these a significant association between antiHCV-positive cases and HCC was found (KEW et al., 1990; DI BISCEGLIE et al., 1991; KAKLAMANI et al., 1991; TANAKA et al., 1991; YU et al., 1991; CHUANG et al., 1992; ZAVITANOS et al., 1992; Yu et al., 1997). The importance of HCV as an aetiological factor of chronic liver diseases including HCC was also observed in Egypt (WAKED et al., 1995). In contrast, in Senegal, where HBV remains the main viral cause of HCC, the epidemiological association between HCV and HCC was not proven (IQ et al., 1996). In order to assess the potential contribution of HBV and HCV infections to the incidence of HCC in Sudan,

Address for correspondence: Dr P. Van? Veer, Division of Human Nutrition and Epidemiology, Wageningen University, John Snow Building, Dreijenlaan 1,6703 HA Wageningen, The Netherlands; phone f31 317 485105, fax f31 317 482782, e-mail [email protected]

we compared the hepatitis B surface antigen (HBsAg) and anti-HCV status for HCC cases and controls collected from 2 areas in Sudan. Methods A case-control study was conducted, during a period of 2 years (September 1996-September 1998), in West and Central Sudan representing regions of higher and lower incidence of liver cancer, respectively. West Sudan (North Kordofan State) is an area of 230 000 km’, and has a population size of 1 500 000. Central Sudan (Gezera State) is an area of 152 000 km2, and has a population size of 2 500 000. All cases and controls were recruited from these regions; subjects’ consent was obtained after they had been informed. Case recruitment A total of 150 HCC cases between 2 l-70 years were recruited from the 2 study regions, in a period of 2 successive years; from 115 of these, blood samples were obtained and tested for HBsAg and for HCV antibodies, missing blood samples were attributable to study logistics only. Cases were recruited from 5 out of 6 Khartoum (capital) hospitals (Tropical Medicine Hospital, Oumdurman Hospital, Khartoum Hospital, Soba Hospital and Ibn Sena Hospital). In these hospitals liver cancers are diagnosed, following referral from the regional hospitals. The admission lists from the departments of Internal Medicine in the above-mentioned hospitals were screened weekly to identify new admitted cases and the responsible doctor and patients were contacted. For all 115 cases, the diagnosis of HCC was verified clinically, by liver function tests and by histopathological examination of a liver biopsy. Control-group recruitment As hepatitis infection is one of the present study questions, and since subjects with hepatitis and other diseases may be overrepresented in hospitals, the control group for this study was enrolled from the general population. The controls were community-based subjects, recruited from the same catchment areas as the cases; the number of controls was chosen proportional to the population size of the 2 areas. Therefore, 4 localities were chosen from West Sudan and 6 localities in Central Sudan. From each locality one sugar shop was chosen randomly; these sugar shops hold a complete registry of all inhabitants in the locality and registration by the shops

R. E. OMER ETAL.

488

conjugated to horseradish peroxidase was added to the micro-well and incubated for 30 min at 37°C. This was followed by subsequent washing for 5 times to remove the unbound conjugate. The specimen was then incubated for 15 min after the addition of an enzyme detection system composed of o-phenylene-diamine (OPD) and hydrogen peroxide: 50 ~JLof 4N sulphuric acid was then added to stop the reaction. The colour intensity was measured with a micro-well reader (nhotometer) using a 492-nm filter. Positive samples were then confirmedby CHIRON RIBA HCV 3.0 SIA.

is required for obtaining subsidized sugars, a product that is highly appreciated in regions of Sudan. Twenty households were taken randomly from the shop list. Since liver cancer is more prevalent among men, the selection from the households was made in a ratio 1:3 women to men. A total of 199 controls were enrolled (80 from West and 119 from Centre), aged 25-70 years. All chosen subjects agreed to cooperate in the study. Blood samples Blood samples from cases were centrifuged immediatelv at about 1200 e for 10 min. and the sera were then kept at -20°C for HEV and HCV testing. Blood samples from controls were kept immediately after venepuncture in a foam fridge containing cooling materials at 4”C, to be transported from the locality to the regional hospital laboratory where they were separated using the same technique as mentioned above. The separated blood sera were then kept at -20°C until they were transferred in a foam fridge at
Background data The orally administered questionnaire was designed to identify the potential confounders and other risk factors of HCC. The data for cases and controls were collected by the same investigator (R.E.O.) and entered in the database using EpiInfo 5. Data analysis Prevalence of hepatitis B and C were compared between cases and controls, followed by calculation of crude odds ratios (ORs). 1’ testina and determination of confidence intervals. Subsequently, analyses were stratified by the potential confounders age, region, gender, education and job type. Finally, logistic regression analysis was used (SAS) to establish the association between hepatitis B or C and HCC accounting for potential confounders such as age, region, education and job type. Variables were considered as potential confounders when their inclusion in the model altered the OR for HBV or HCV by 2 10%. Population attributable risk was calculated using the formula for the population excess risk relative to the rate in the population, adapted to case-control studies (DOS SANTOS SILVA, 1999).

Hepatitis analysis The samples were analysed at the National Health Laboratory, Virology Department. Test kits for hepatitis B and C and their confirmation tests were obtained from Organon Teknika, Boxtel, The Netherlands. HBsAg was determined by Hepanostika HBsAg test. In short, 100 pL of (undiluted) sample and control were pipetted into assigned wells. Each well was then soaked and washed 4 times with phosphate buffer after it had been incubated at 37 f 2°C for 1 h. TMB substrate (100 pL) was next pipetted into each well. The strips were then incubated at 18-25°C for 30 min, after which 100 p.L of sulphuric acid (1 mol/L) was added to each well to stop the reaction. The absorbency of the solution in each well was read at 450 f 5 nm. Of the positive samples, 25% were randomly selected and checked by the Hepanostika HBsAg Uni-form micro-ELISA system. All checked samples were confirmed as positive. Detection of hepatitis C virus (HCV) was done by ORTHO HCV 3.0 ELISA Test System with Enhanced SAVe. For this assay a diluted test specimen was incubated in a test well at 37°C for 30 min. Then it was washed 5 times to remove unbound serum proteins. Two hundred microlitres of murine monoclonal antibody Table 1. Main controls

characteristics

.

Results Table 1 shows the main characteristics of HCC cases and controls from the 2 study regions West and Central Sudan. The prevalences of hepatitis B and C are higher among cases-than controls; the prevalence of either hepatitis B or C is also hiaher (cases 53%. controls 8%). Cases were older than c&trojs, more often lived in West rather than Central Sudan, were less educated and more frequently farmers and labourers. No difference was

of hepatocellular

Hepatitis virus infections Hepatitis B Hepatitis C Hepatitis B or C Design-related Age (years) <40 40-59 60+ Region West Gender Male

carcinoma

cases and

Cases ;;;I . 0

Controls b=& @‘f’ . 0

49 (43) 13 (11) 61 (53)

2 I;; 16 (8)

‘;; Potential risk factor

I<,”

variables 11 (10)

38 (33) 66 (57)

z: g;

74 (64)

80 (40)

23 (12)

88 (77)

150 (75)

Potential confounders Education Illiteracy

80 (70)

68 (34)

‘“bF=er + labourer Alcohol history

66 (57) 45 (39)

70 (35) 61 (31)

489

HEPATITISB,CANDLIVERCANCERINSUDAN observed between cases and controls for a history of alcohol consumption. To identify potential confounders, Table 2 shows the association between viral hepatitis and age, region, gender, education and job type. HBV prevalence is higher among people aged < 40 years, and, among the Western population, less educated people, and farmers and labourers. Alcohol history was not related to HBV. The prevalence of HCV among controls was much lower than the prevalence of HBV. Although the few data do not permit conclusions there is no reason to suspect that the prevalence of HCV over strata of potential confounders is different than the distribution of HBV. Strong associations of HBV and HCV with the incidence of HCC are observed (Table 3). Age appears a clear negative confounder for HBV, hence the OR rises from 9.8 (crude) to 16.1 (adjusted); however, age acts as a positive confounder for HCV and HCC. Region, education and job type are no clear confounders for HBV; region and education appeared to be negative confounders for HCV, although confidence intervals are very wide. After adjustment for age, the variables region, education and job type were not substantial confounders anymore. Thus, the age-adjusted ORs of 16.1 for HBV and 4.5 for HCV provide reasonable estimates of the association between hepatitis infections and HCC in this population. Taking the prevalence of HBsAg and anti-HCV in the controls (7% and 15%, respectively) and assuming the adjusted ORs (16.1 and 4.5, respectively) are equivalent to the relative risk, the estimated fraction of the liver cancers in the population that is attributable to HBV and HCV is 52% and 5%, respectively. Discussion In this study we found that HBV and HCV infections are stronalv associated with HCC in Sudan. The data suggested that HBV infection has a stronger association with HCC than HCV infection. The prevalence of HBV was also found to be higher (7.0%) than the prevalence of HCV (1.5%) among this study population. Regarding selection of cases in our study, the use of registration lists of 5 hospitals in Khartoum for case recruitment enabled all cases from the 2 regions to be included in the study, except for those who might have died before they could be included in the study; furtherY,

Table 2. Potential risk factors for liver cancer and HBV or HCV infection (Sudan, 1996-98)

Potential risk factor Overall Age (years) <40 40-59 60+ Region West Central Gender Male Female Education Illiteracy Literacy Job type Farmer + labourer Others Alcohol history Yes No

HBV (n = 199) Positive (%) I4 (7)

HCV (n = 199) Positive (%) 3 (1.5)

7 (9) 7 (7) 0 (0)

0 (0)

7 (9) 7 (6)

2 (2)

10 (7) 4 (8)

2 (1) 1 (2)

1 (1)

2 (2) 1 (1)

8 (12) 6 (5)

2 (3) 1 (1)

9 (13) 5 (4)’

2 (3) i (ij

5 (8) 9 (7)

0 (0) 3 (100)

more few were missed during the 2 periods of control recruitment accounting for a total of 4 weeks during the 2-year study period. Moreover, 1 of the 5 hospitals is a National Centre for Gastrointestinal and Liver Diseases, so that HCC patients from all over the country are sent to that hospital for further diagnoses and better treatment. From 5 of 6 hosuitals in Khartoum 150 cases were obtained during the 2 study years. This number exceeds the number of 42 liver and gallbladder cancer cases reported from Khartoum hospitals by DAOUD et al. (1969) during the period 1957-65. Selection of controls was done at a random basis, with numbers of controls from West and Central Sudan proportional to the population size; these regions are the catchment areas that provided cases in the study. From those regions several sugar shops (see Methods) were chosen randomly. As social, cultural and demographical characteristics are homogeneous within the 2 regions, the randomly chosen sugar shops are thought to yield a representative sample of the subjects living in the region at large. Age, place of origin, education level and job type were evaluated as possible confounders of the association between HBV or HCV and HCC. The potential bias was excluded by Mantel-Haenszel analysis, followed by logistic regression that accounted for several confounders simultaneously. In fact, age appeared to be the only relevant confounder; alcohol drinking and gender were not considered as confounders since they were not related to HCC in the data. Regarding hepatitis virus infections, a study conducted in the year 1989 in Sudan showed that HBsAg and HBcAb (200X) were detected more often in HCC patients than in controls (ITOSHIMA et al., 1989). The summary ORs for HBsAg positivity and anti-HCViHCV RNA positivity from 32 published case-control studies included in a meta-analysis by DONATO et al. (1998) were found to be 13.7 and 11.5, respectively. In areas where HBV infection was low to intermediate endemic and HCV infection is predominant among HCC cases, the summary ORs for HBsAg and anti-HCV/HCV RNA positivity were 8.5 and 16.8, respectively. In the same study, in a total of 14 studies from sub-Saharan and southern African countries where HBV is endemic, summarv ORs of 16.7 for HBsAe and 6.2 for anti-HCViHCV RNA were observed. Our Results for HBsAg and antiHCV positivity coincide with the findings of DONATO et al. (1998) for sub-Saharan countries with high HBV prevalence. Despite the ethnic and cultural heterogeneity of the different regions of Sudan, this suggests that our results may also apply to a large geographical area including most other regions of Sudan. Given the observed prevalence of hepatitis and the ORs in this study, HBV mieht account for 52%. and HCV would account for 5% o?HCC in Sudan. ’ In addition, the study by DONATO et al. (1998) suggested synergism between the 2 viruses, i.e., the increase in HCC risk when both HBV and HCV infections were present is higher as compared with each infection alone. However, since in our study HBV and HCV infections coincided in onlv 1 case and 1 control, we were not able to study this synergism especially because of the verv low nrevalence of HCV (1.5%). The average age ofHCC cases in areas where HBV is endemic is lower than among cases from areas where HCC is predominantly due to HCV infection (DONATO et al., 1998). Similarly in our study, the mean age of HCC patients who were positive for HBsAg was 55 i 10 years while it was 65 f 5 years for cases positive for anti-HCV. This different age distribution, along with the lower OR for HCV. miaht be related to different ages at henatitis virus infect& or differences in pathtgenesis. .Thus studies are needed to determine the age of infection and mode of transmission of hepatitis viruses in the Sudanese population.

490

R. E. OMER ETAL.

Table 3. Crude and adjusted odds ratios (ORs) for hepatitis virus infection and liver cancer (Sudan, 1996-98)

Crude OR Adjusted for: Age Region Education

Job type Adjusted for: Age + region Age + education Age + job type

B or hepatitis

HBV

HCV

98 (5.1-18.9)

8.3 (2.3-29.9)

16.1 9.0 8.7 8.5

(7.4-34.9) (4*6- 17.6) (4.4-17.4) (4.4- 16.7)

4.5 10.1 13.1 7.6

15.0 (6.9-32.7) 14.5 (6.7-31.6) 14.8 (6.8-32.3)

C

(1.1-18.6) (2.7-37.4) (3.4-50.2) (2.1-27.8)

5.5 (1.3-23.9) 6.0 (1.4-24.8) 4.5 (1.0-19.3)

Values are odds ratios (ORs) with 95% confidence intervals in parentheses.

Apart from the role of HBV and HCV in the aetiology of HCC, clues have also been obtained from studies in other African and Asian countries with similar demographical, social and cultural characteristics. In most of these countries, HBV is believed to act in combination with other environmental and dietary exposures such as aflatoxins. In these areas, where food and feed stuffs are highly contaminated with allatoxins, the interaction between aflatoxin exposure and chronic HBV infection has been suspected to play an important role in the incidence of HCC. For instance, it is thought that HBV increases HCC risk by either integrating in the DNA or causing proliferation of cells with a mutated ~53 gene, which is seen in areas where there is excess exposure to aflatoxin in foods. Therefore, the strong association we observed with HBV may partially result from synergism with aflatoxin exposure, which is common in some regions in Sudan, including west Sudan (OMER et al., 1998). In conclusion, hepatitis virus infections, especially by HBV, are important risk factors for HCC in Sudan. Either alone, or in combination with exposure to environmental carcinogens such as aflatoxins, they may account for about 57% of HCC in Sudan. Acknowledgements The authors thank the staff of hospitals and outpatient clinics in Khartoum, North Kordofan State and Gazira State. We also thank the staff of the virology department at Khartoum Nutritional Health Laboratory for co-operation in the conduct of this study. Sudanese Standards and Meteorology Organization ‘SSMO’ gave financial support for the conduct of this study; scientific support and exchange of personnel were financed by Wageningen University. References Bukh, J., Miller, R. H., Kew, M. C. St Purcell, R. H. (1993). Hepatitis C virus RNA in southern Africa blacks with hepatocellular carcinoma. Proceedings of the NationalAcademy of Sciences of the USA, 90, 184% 185 1.

Chuang, W. L., Chang, W. Y., Lu S. N., Su, W. P., Chen, S. C., Hsieh, M. Y.,Wang, L. Y., You, S. L. & Chen, C. J. (1992). The role of hepatitis B and C viruses in hepatocellular carcinoma in a hepatitis B endemic area. Cancer, 69, 2052-2054.

Coursaget, I’., Leboulleux, D., Le Cann, I’., Bao, 0. & CollSeek, A. M. (1992). Hepatitis C virus infection in cirrhosis and primary hepatocellular carcinoma in Senegal. Transactions of the Royal Society of Tropical Medicine and Hygiene, 86, 552-553.

Daoud, E. H., El-Hassan, A. M., Zakova, N. & Zak, A. (1969). Aspects of malignant disease in the Sudan. In: Cancer in Africa, Cliford, I’., Linsell, C. A. & Timms, G. L. (editors). Nairobi: East Africa Publishing House, pp. 43-47. Di Bisceglie, A. M., Order, S. E., Klein, J. L., Waggoner, J. G., Sjogren, M. H., Kuo, G., Houghton, M., Choo, Q. L. & Hoofnagle, J. H. (199 1). The role of chronic viral hepatitis in hepatocellular carcinoma in the United States. American Journal of Gastroenterology, 86, 335-338.

Donato, F., Boffetta, P. & Puoti, M. (1998). A meta-analysis of epidemiological studies on the combined effect of hepatitis B and C virus infections in causing hepatocellular carcinoma. International

70urnal of Cancer. 75. 347-354.

DOSSantos S&a. I. 11999). Ca&erEdidemiologv:

Principles and Methods. Lyon: IARC, pp. 381:38j.~ --IARC (1994). Hepatitis viruses. Lyon: IARC. LARC Mono-

graphs on the Evaluation of Carcinogenic Risks of Humans, no. 59. Itoshima, T., Fedail, S. S., Souliman, I., Ali, A. K., Higashi, T. & Tsjit, T. (1989). Hepatitis B virus markers in patients with schistosomiasis, liver cirrhosis and hepatocellular carcinoma in Khartoum, Sudan. Acta Medical Okayama, 43,241-244. Jeng, J. E. & Tsai, J. F. (1991). Hepatitis C virus antibody in hepatocellular carcinoma in Taiwan. Medical Virology, 34, 74-77.

Ka, M. M., Herve, P., Leguenno, B., Ndiaye, M. F., Diop, T. M., Diop, B., Dangou, J. M., Bao, 0. & Brechot, C. (1996). Low prevalence of anti-hepatitis C virus antibodies in hepatocellular carcinoma in Senegal. Annals of Gastroenterology and Hepatology, 31,329-332.

Kaklamani, E., Trichopoulos, D., Tzonou, A., Zavitsanos, X., Koumantaki, Y., Hatzakis, A., Hsieh, C. C. & Hatziyannis, S. (199 1). Hepatitis B and C viruses and their interaction in the origin of hepatocellular carcinoma. Journal of the American Medical Association, 265, 1974- 1976.

Kew, M. C., Houghton, M., Choo, Q. L. & Kuo, G. (1990). Hepatitis C virus antibodies in southern Africa blacks with hepatocellular carcinoma. Lancet, 335, 873-874. Leung, N. W., Tam, J. S., Lai, J. Y., Leung, T. W., Lau, W. Y., Shiu, W. & Li, A. K. (1992). Does hepatitis C virus infection contribute to hepatocellular carcinoma in Hong Kong? Cancer, 70, 40-44.

Mohamadien, A. ( 1993). Biochemical tumor markers in HCC and other liver disease. MD thesis, University of Khartoum, Sudan. Muir, C. S. (1992). Cancer incidence in five continents. Lyon: IARC. IARC Science Publications, no. 120, pp. 25-30. Omer, R. E., Bakker, M. I., Van? Veer, P., Hoogenboom, R. L., Polman, T. H., Alink, G. M., Idris, M. O., Kadaru, A. M. & Kok, F. J. (1998). Aflatoxin and liver cancer in Sudan. Nutrition and Cancer, 32, 174-180. Tanaka, K., Hirohata, T., Koga, S., Sugimachi, K., Kanematsu, T., Ohryohji, F., Nawata, H., Ishibashi, H., Maeda, Y., Kiyokawa, H., Tokunaga, K., Irita, Y., Takeshita, S., Arase, Y. & Nishino, N. (1991). Hepatitis C and hepatitis B in the etiology of hepatocellular carcinoma in the Japanese population. Cancer Research, 5, 2842-2847.

Waked, I. A., Saleh, S. M., Moustafa, M. S., Raouf, A. A., Thomas, D. L. & Strickland, G. T. (1995). High prevalence of hepatitis C in Egyptian patients with chronic liver disease. Gut, 31, 5-7. Wild, C. P., Shrestha, S. M., Anwar, W. A. & Montesano, R. (1992). Field studies of aflatoxin exposure, metabolism and induction of genetic alteration of HBV infection and hepatocellular carcinoma in Gambia and Thailand. Toxicology Letters, 64165, 455-461. Yu, M. W., You, S. L., Shang, A. S., Lu, S. N., Liaw, Y. F. & Chen, C. J. (1991). Association between hepatitis C virus antibodies and hepatocellular carcinoma in Taiwan. Cancer Research, 51, 5621-5625. Yu, M. C., Yuan, J. M., Ross, R. K. & Govindarajan, S. (1997). Presence of antibodies to the hepatitis B surface antigen is associated with an excess risk for hepatocellular carcinoma

HEPATITIS

B, C AND LIVER

CANCER

491

IN SUDAN

among non-Asians in Los Angeles country, California. Hepawlogy, 25,226228. Zaki, M. (199 1). Aetiological factors in HCC. MD thesis, University of Khartoum, Sudan. Zavitanos, X., Hatzakis, A., Kaklamani, E., Tzonou, A., Toupadaki, N., Broeksma, C., Chrispeels, J., Troonen, H., Hadziyannis, S., Hsieh, H. A. & Trichopoulos, D. (1992).

TRANSACTIONS

OF THE ROYAL

1 Short Report

SOCIETY

OF TROPICAL

MEDICINE

(

Distribution of rotavirus G and P types in North and South Indian children with acute diarrhoea in 1998-99 G. Kang I*, T. Raman’, J. Green’, C. I. Gallimore’ and D. W. G. Brown* ‘Department of Gastrointestinal Sciences, Christian Medical College and Hospital, Vellore 632004, India; ‘Enter& Neurologikal and Respiratory Virus Laboratoy, Central Public Health Laboratory, 61 Colindale Avenue, London NW9 5HT, UK Abstract VP7 and VP4 genotypes of 82 rotavirus strains from children with acute diarrhoea during November 1998-January 1999, in 4 Indian towns, were determined-by reversetranscriution PCR. Overall. 68182 (83%) could be VP7- and 52182 (63%) VP4-typed. Geographical differences in rotavirus circulation have implications for future vaccination strategies. Keywords: rotavirus, genotypes, I’CR, diarrhoea, children, geographical distribution, India Group A rotaviruses cause acute gastroenteritis in children worldwide, but are especially significant causes ofmorbidity and mortality in developing countries. They are classified into serogroups (A-G), based on the antigenic specificity of the inner capsid protein VP6 and mto further types based on the 2 outer capsid nroteins. VP7 and VP4. In recent vears. reverse transcription uolvmerase chain reaction (RT-PCR) based on genbtyping of VP7 (G-typing) and VP4 (P-typing) has replaced older serological techniques in molecular epidemiological studies, G-genotypes and serotypes are concordant but P types are not and have a double nomenclature because serotype designations are not available for all genotypes. The commonest G types in human infections worldwide areGl-G4 andthecommonest Ptypes, the PlA[B] and PlBl41. but unusual isolates have been found in different parts of the world (KOSHIMURA et al., 2000) from different patient groups, with symptomatic and asymptomatic infections. A previous study from India studied the geographical variability of rotaviral strains, using a single-round PCR, and identified G9 strains in North India (RAMACHANDRAN et al., 1996). We now report the use of a nested RT-PCR and the identification of G9 rotaviruses in South India, with distinct variations in rotaviral circulation in 4 Indian towns during a single season. L

_,

*Author for correspondence; phone +91 416 222102 x2052, fax +9 1 416 232035, e-mail [email protected]

Association between hepatitis C virus and hepatocellular carcinoma using assaysbased on structural and non structural hepatitis C virus peptides. CancerResearch,52, 536445367. Received 18 September 2000; revised 3 April 2001; accepted for publication 5 April 2001

AND

HYGIENE

(2001) 95,491-492

This study was carried out during the winter (November 1998-January 1999) in 4 Indian towns (Vellore and Mysore in South India, Jalandhar and Yamunagar in North India). One primary care physician in each town was requested to collect samples from children aged <5 years, presenting with acute diarrhoea during the study period. A total of 365 faecal samples were analysed for identification of enteric pathogens and 82 were positive for rotaviruses by electron microscopy, latex agglutination (MERITEC, Meridian Diagnostics, Cincinnati, Ohio) and ELISA (Dako-EIA, Dakopatts, Copenhagen). The age distribution of the children ranged from 4 months to 5 years (mean 24 months) and was similar in all 4 towns. Genotyping of rotaviruses was performed using random priming as previously described (ITURRUAGOMARA et al., 1999). Sequencing of representative G types and all unusual isolates using an automated sequencer and sequence comparison was carried out as et al., 2000). described earlier (ITURRIZA-GOMARA Electropherotyping of viral RNA (BROWN et al., 1988) was done for representatives of each G and P type, and for all isolates that could not be typed by PCR. Rotaviruses were identified in 82 (22.5%) of 365 faecal samples from children during the winter of 1998-99, with other bacterial and parasitic enteric pathogens in 7 1 (19.5%) samples, mainly Campylobacter spp. (n = 14, 4%), Shigella spp. (n = 14, 4%) and Cyptosporidium (n = 31, 8.5%). The G typing PCR typed 68 strains (83%), and 52 (63%) were P-typed. Ineachtown, 275% strains typed by the G and P PCRs, with the exception of Mysore where 7 strains (32%) had a first-round product in the P typing PCR that did not amplify in the second round. The distribution of G and P types is shown in the Table. The commonest G: P combination seen was Gl : PlA[B] (n = 14, 60% of all typed Gl strains) and this was followed by the unusual combination Gl : PlB[4] (n = 9, 40% of all typed Gl strains) which was found in Vellore and Jalandhar. P2A[6] strains were found in Mysore and Jalandhar, in combination with G type 9 (n = 1) and unknown (n = 1). A total of 8 strains, 7 (32%) from Mysore and 1 from Vellore (6%) (4 G9, 2 Gl, 1 mixed, 1 G undetermined) gave a product in the first round of the nested PCR, but did not amplify in the second round, despite attempts at specific priming. Four of 7 G9 strains were sequenced. On comparison of the entire 1062 bp of the VP7 gene, there was 99100% sequence homology betweenthe strains, and these were 96-99% similar to strains from the UK and the USA as has been previously reported, but were more distantly related to 116E, the Indian asymptomatic et al., 2000). Electropherisolate (ITURRIZA-GOMARA otvnine of 6 G9s found ‘short’ uattems in all isolates. unhke The ‘long’ pattern reported*for 116E. Other G and P untypable strains from North India had both ‘short’ and ‘long’ patterns. In this study, Gl rotaviruses were the most frequently identified strains in 3 of 4 towns. In Vellore, the proportions of Gl and G4 were almost equal, similar to a previous hospital-based study from Vellore in the early 1980s (BROWN et al., 1988). Previous reuorts from North India have shown the ‘presence of G-types l-4 and 9 (HUSAIN et al., 1996; RAMACHANDRAN et al.,