Secondary Logo

Journal Logo


A Twenty-One Year Surveillance of Adenoviral Conjunctivitis in Sapporo, Japan

Aoki, Koki M.D., Ph.D.; Tagawa, Yoshitugu M.D., Ph.D.

International Ophthalmology Clinics: January 2002 - Volume 42 - Issue 1 - p 49-54
  • Free

In Japan, the national epidemiologic surveillance of 21 infectious diseases started under the patronage of the Ministry of Health and Welfare of Japan in July 1981. 1 The program selected three epidemic eye diseases due to adenoviruses and enteroviruses—epidemic keratoconjunctivitis (EKC), pharyngoconjunctival fever (PCF), and acute hemorrhagic conjunctivitis (AHC)—to be reported to the surveillance committee. A number of eye clinics were appointed as index stations responsible for the population of certain areas. The number of stations was 297 in 1991. These stations annually have reported the number of patients and information on etiology and seasonal prevalence, which then is analyzed? The program offers feedback of information to the respective reporting areas. Adenoviral conjunctivitis occurs in 1 million patients per year throughout Japan, according to this nationwide surveillance system.

Clinicoetiological Epidemiology

Adenoviral antigen was studied using Adenoclone from 1992 and Adenocheck from 1996. 2 Isolation and a neutralization test were used when the screening test, such as a commercial kit, proved positive, to identify the adenovirus (Ad) serotype and genome type; polymerase chain reaction and restriction fragment–length polymorphism were performed when results from Adenockeck were negative. 3 The clinical diagnosis of adenoviral conjunctivitis is not difficult in severe cases of Ad8, Ad19, or Ad37 infection but, in mild cases caused by Ad3, Ad7, or Ad11, the aforementioned methods must be used for etiological detection. Ocular symptoms due to Ad4 shows moderate severity and a broad clinical spectrum from PCF to EKC. 4 The differential diagnosis of infection includes herpes simplex virus, enterovirus 70, coxsackie A24 variant virus, and Chlamydia trachomatis, especially the adult type. In these cases, the systemic symptoms also are important for differential diagnosis (Fig 1). 5

Figure 1.
Figure 1.:
Differential diagnosis of Viral conjunctivitis (n = 1,886). (Ad = adenovirus; EV = enterovirus; HSV = herpes simplex virus.)

Information on the monthly incidence of adenoviral conjunctivitis is useful to prevent further infection in hospitals and homes. The occurrence of infection is more frequent in the summer season for Ad3, Ad4, and Ad37. Chronological changes in Ad3, Ad8, and Ad37 showed peaks of incidence in 1980, 1983, 1986, 1991, and 1995. Conversely, Ad4 had a peak in 1985 and has been decreasing since that time. Ad19 increased in 1997 and 1998 (Fig 2).

Figure 2.
Figure 2.:
Adenovirus (Ad) isolates according to subgenus among 1,454 patients with conjunctivitis in Sapporo, Japan, 1978–1998.

Serological Epidemiology

The prevalence of antibodies against Ad3, Ad4, Ad7, Ad11, Ad19, and Ad37, which cause conjunctivitis, has been investigated. 6 A total of 250 serum specimens were collected from healthy volunteers during two periods, from 1982 to 1986 and from 1993 to 1997. The neutralizing antibody to each serotype was measured using Hep2 cells at 37°C, with a 90-minute reaction time and 3 days of observation. Serum for which the titers were more than four dilutions in pair serum between acute stage and convalescent stage were regarded as positive in this epidemiological survey. Ad3 in subgenus B and Ad1, Ad2, Ad5, and Ad6 in subgenus C composed more than 60% of the serotypes.

Antibody against Ad3 was positive in 68%, Ad11 in 37%, Ad4 in 29%, Ad8 in 19%, Ad19 in 16%, and Ad37 in 8% of healthy volunteers (Fig 3).

Figure 3.
Figure 3.:
Antibody frequency of adenovirus (Ad) neutralization test (n = 250).

A comparison of annual changes in neutralization test (NT) antibody in the two periods showed that the frequencies in Ad3 and Ad19 antibody in 1982 to 1986 were lower than those in 1993 to 1997. The frequencies of other serotypes—Ad4, Ad8, Ad11, and Ad37—were lower in the 1993 to 1997 period.

Reinfection of adenoviral conjunctivitis was observed in this survey. Among 1,104 patients in whom adenoviral conjunctivitis was etiologically confirmed, 6 patients showed reinfection. 7 The first case was a 6-year-old boy infected first with Ad8 and 9 years later with Ad4. In other cases, the serotypes (Ad3 and Ad8) repeated.

Molecular Epidemiology

The gene products carrying the antigenic determinant represent only a small part of the viral genome. This means that analysis of cross-reactions by serological tests gives information on only a few gene products. 8 Wadell and colleagues 8 studied the DNA restriction viral entities when serological reaction gave unclear results and thereby obtained information on the genetic variation within a serotype.

To characterize the long-term observations of human adenoviral strains associated with conjunctivitis in Sapporo, Japan, DNA reaction analysis was performed on Ad3, Ad4, Ad7, Ad8, Ad11, Ad19, and Ad37 isolates recovered from patients with epidemic conjunctivitis during a 10-year period. 9

This chapter reports our data on the serotypes more frequently related to epidemic keratoconjunctivitis. Ad4 has been classified as the only member of subgenus E, and shares several properties with subgenus B, but the fiber of Ad4 is related to subgenus C. Ad4 is more closely related to one of the archetypes of human adenoviruses than are other serotypes. Ad4 was reported to cause disease in only very few cases in countries from 1975 to 1994. Wadell and associates 8 reported genome types Ad4p and Ad4a in 1984. Since 1979, epidemics of PCF and AHC caused by Ad4 have occurred in Japan. We have analyzed the Ad4 strain since 1978 and concluded that the genome type of Ad4 isolates in Sapporo is Ad4a (Table). 10

Table 1
Table 1:
Chronological Changes of Adenovirus ( Ad ) Genome Type

We reported that 122 prevalent strains of Ad4a tended to change between 1985 to 1989 and were replaced by new subgenome types within 2 years. 11 Almost all the prevalent strains during the epidemic outbreak of Ad4a in 1985 belonged to a single subgenome type. However, in 1986, when an epidemic was not observed, as many as six subgenome types were found, and they subsequently were replaced with new subgenome types. This indicated that mutation of Ad4 DNA occurred frequently in these years.

We could determine the index cases and reservoirs of infection, which was not possible before enzymes recognizing six base pair sequences were used. In the study of 16 strains isolated from 1993 to 1995, our colleague, Kanai, 12 identified one subgenome type in 1993, and the same subgenome type was seen in 1995 when no epidemic was observed. From these findings, Ad4 DNA seems to have become better conserved and altered less frequently in recent years. According the Japanese ocular infection surveillance data, the incidence of Ad4 has been decreasing since 1993. 12

The 51 human adenovirus serotypes fall into six subgenera. Several adenoviruses of subgenus D exhibit tropism for the human conjunctiva, including Ad8, Ad19a, and Ad37 serotypes, which frequently are associated with epidemic infection.

Since the first isolation of Ad8 as an etiological agent of EKC in 1959 by Jawetz and Hamma, 13 many Ad8 genome types have been identified by application of restriction endonucleases and cleavage pattern analyses by international co-studies in Asian countries. Eight genomes have been discovered in Sapporo, Busan, South Korea, and Kaohsiung, Taiwan. Ad8a and Ad8b were first isolated in Sapporo; Ad8c, d, e, f, g, and h were detected in Kaohsiung from 1980 to 1994. Ad8c was most prevalent from 1980 to 1981, and Ad8e was most prevalent from 1983 to 1987. 14,15 Ad8h was most prevalent from 1990 to 1994 in Kaohsiung. We have analyzed 52 strains of Ad8 in this survey. In Japan, two genome types were reported. Ad8a was prevalent in the period from 1975 to 1986, and Ad8b was found after 1976. Our colleague K. Tanaka has studied the genome type of Ad8 in Sapporo from 1986 to 1996 (unpublished data). Isolates from 1991 and 1996 were similar to Ad8e. This genome type was isolated in Taiwan in a survey in 1981 and in Korea in 1983. We have reported a similar pattern of Ad8e with enzymes BamHI, Hind III and Sall; however, Sacl divided them into two subgroups.

Our study shows that the genome types Ad8a and Ad8b are endemic in Sapporo, Japan. Adequate After the first isolation, the genome type Ad8e, which was found in Taiwan and Korea, also was found to be endemic in Japan a decade after Taiwan (see the table) (unpublished data).

To date, few cases of the Ad19 genome type have been reported since the prototype strain was isolated in 1995 from a Saudi Arabian child with trachoma. Ad19a was first isolated in 1973 in Antwerp and was reported to be associated with EKC by Wadell and de Jong in 1980. The prototype strain was not associated with any outbreak, but in 1974 Ad19a caused severe keratoconjunctivitis in different parts of the world. 16 In Japan, Ad19a has been recovered since 1977. Surveillance of viral conjunctivitis over a 20-year period in Sapporo showed isolation of Ad19a every year from 1978 to 1998. Among serotypes Ad8, Ad19, and Ad37 belonging to subgenus D, the isolation rate for Ad19a used to be low; however, in the period of 1997 to 1998, Ad19a was isolated more frequently from patients with EKC (see the table).

The isolation of Ad8 and Ad37 suddenly increased in 1995 and 1996, while the frequency of Ad19a isolation was low. In the following years, the reverse was observed—an increase in Ad19a isolation while both Ad8 and Ad37 decreased. These isolates occurred in EKC cases, including many nosocomial infections, formerly caused primarily by Ad8. We analyzed isolates in the periods between 1992 and 1993 when Ad19a was low and between 1997 and 1998 when its isolation increased. We found a new genome type labeled Ad19a in cases of nosocomical infection in two hospitals located in Hokkaido and Tokyo. 17

In 1976, de Jong isolated a new antigenically related serotype, Ad37, from a patient with EKC. 14 Since the appearance of Ad37 in 1976, some ocular adenoviral strains exhibited difficulties experienced by some laboratories in distinguishing Ad19a by neutralization testing. In such cases, the use of DNA restriction analysis has proved helpful. In 1982, 46 strains isolated from 1977 to 1981 were shown by restriction analysis to be Ad37. 18 Now, Ad37 is one of the main causative agents of adenoviral conjunctivitis.

At present, Ad37p is the most common genome type in Japan but not in Taipei. New genome types have been reported. In Japan, we confirmed Ad37p in 1980 but, in 1991 and 1993, we detected in Sapporo a prototype—like strain and three new genome types, designated Ad37a, Ad37b, and Ad37c (see the table). 19


We have studied the clinicoetiological, serological, and molecular epidemiological features of adenoviral conjunctivitis under the auspices of the nationwide surveillance project in Sapporo, located in northern Japan. We were able to analyze the results of 1,454 cases of adenoviral, herpetic, and chlamydial conjunctivitis. We found that adenonovirus serotypes 8, 19, and 37 caused severe conjunctivitis without systemic symptoms, whereas serotypes Ad3, Ad7, and Ad11 caused mild conjunctivitis with systemic involvement. Ad4 showed a broad range of symptoms, from PCF to EKC. Adenoviral conjunctivitis had seasonal outbreaks in the summer in association with adequate temperature and humidity in Sapporo. Neutralization antibodies against Ad8, Ad19, and Ad37 were detected in fewer than 20% of cases, and so the incidence of epidemics with these serotypes may rise in the coming years. The main genome types of adenovirus in Sapporo were Ad4a, Ad8e, Ad19a, Ad19b, and Ad37p, a, and b.


1. Aoki K, Isobe K, Ohno S. Nationwide surveillance program of epidemic conjunctivitis in Japan. In: Bialasiewicz AA, Schaal KP, eds. Infectious disease of the eye. The Netherlands: AEolus Press, 1994: 309–316
2. Uchio E, Aoki K, Saitoh W, et al. Rapid diagnosis of adenovirus conjunctivitis on conjunctival swab by 10 minute immunochromatography. Ophthalmology 1977; 104: 1294–1299
3. Saitoh-Inagawa W, Oshima A, Aoki K. Rapid diagnosis of adenoviral conjunctivitis by PCR and restriction fragment length polymorphism analysis. J Clin Microbiol 1996; 34: 2113–2116
4. Saitoh-Inagawa W, Aoki K, Uchio E. Ten year's surveillance of viral conjunctivitis in Sapporo, Japan. Graefes Arch Clin Exp Ophthalmol 1999; 237: 35–38
5. Aoki K, Kato M, Ohtsuka H. Clinical and etiological study of adenoviral conjunctivitis with special reference to adenovirus 4 and 19 infection. Br J Ophthalmol 1982; 66: 776–780
6. Aoki K, Sawada H. Seroepidemiological study of conjunctivitis related adenovirus. Acta Soc Ophthalmol Jpn 1987; 91: 181–186
7. Aoki K, Sen R, Sawada H. Reinfection of viral conjunctivitis. Rinsho Ganka 1991; 45: 1133–1137
8. Wadell G, Hammarskjold M-L, Winberg G. Genetic variability of adenoviruses. Ann NY Acad Sci 1980; 354: 16–42
9. Guo D, Shinagawa M, Aoki K. Genomic typing adenovirus strains isolated from conjunctivitis in Japan, Australia and Philippines. Microbiol Immunol 1998; 32: 1107–1118
10. Liping T, Wang K, Aoki K, Isobe K. Genome analysis of Ad4 strains isolated from acute conjunctivitis in Japan. Jpn J Ophthalmol 1997; 41: 308–311
11. Itakura S, Aoki K, Sawada H, Shinagawa M. Changes in subgenome types of Ad4 isolated from patients with ocular disease between 1985 and 1989 in Sapporo. J Clin Microbiol 1991; 8: 1740–1743
12. Kanai H. Genome analysis with restriction endonucleases recognizing 4– or 5–base pair sequences of adenovirus type. Jpn J Ophthalmol 2000; 44: 463–466
13. Jawetz E, Hanna L, Thygeson P. A laboratory infection with Ad8. Am J Hyg 1959; 69: 13–20
14. Fujii S, Nakazono N, Ishii K. Molecular epidemiology of adenovirus 8 in Taiwan. Jpn J Med Soc Biol 1984; 37: 161–169
15. Ishii K, Nakazono N, Aoki K, et al. Comparative studies on etiology and epidemiology of viral conjunctivitis in three countries of East Asia–Japan, Taiwan and South Korea. Int J Epidemiol 1987; 16: 98–103
16. Wadell G, de Jong JC. Restriction endonucleases in identification of a genome type of adenovirus 19 associated with keratoconjunctivitis. Infect Immun 1980; 27: 292–296
17. Tanaka-Yokogi K, Itoh N, Usui N, et al. New genome type of adenovirus serotype 19 causing nosocomial infections of epidemic keratoconjunctivitis in Japan. J Med Virol 2001; 65: 530–533
18. de Jong JC, Wigand R, Wadell G. Adenovirus 37 identification and characterization of a medically important new adenovirus type of subgroup D. J Med Virol 1981; 7: 105–118
19. Chang C, Sheu M, Chern C, et al. Epidemic keratoconjunctivitis caused by a new genotype of adenovirus type 8 (Ad8)-chronological review of Ad8 in southern Taiwan. Jpn J Ophthalmol 2001; 45: 160–166
© 2002 Lippincott Williams & Wilkins, Inc.