For personal use only. Not to be reproduced without permission of The Lancet. ARTICLES THE LANCET • Vol 356 • October 28, 2000 1461 Summary Background A nucleotide change from U to C at position 472 in the 5' non-coding region of the type 3 poliovirus is associated with increased neurovirulence. Moreover, the proportion of type 3 polioviruses containing this mutation (472- C revertants) correlates with the neurovirulence of a particular sample. We used mutant analysis by PCR and restriction- enzyme cleavage (MAPREC) to estimate the neurovirulence of environmental samples obtained from Toyama prefecture, Japan. Methods Sewage and river water were collected between October, 1993, and September, 1995, and concentrated samples were inoculated into three different cell types. Isolated type 3 viruses were analysed to determine whether they were derived from the live oral poliovirus vaccine strain; they were then tested for neurovirulence by MAPREC. Results 29 type 3 strains were isolated—all of which were vaccine-derived. 16 (55%) comprised between 2% and 91% 472-C revertants by MAPREC and were expected to have high neurovirulence. The remaining strains included less than 0·25% revertants, and were regarded as attenuated viruses. Both types were isolated about 3 months after routine oral poliovirus vaccine administrations in May and October. Three strains isolated from river water were of the virulent type. Interpretation Our results emphasise that there is an environmental risk of vaccine-associated paralytic poliomyelitis as long as live oral poliovirus vaccine is not replaced by inactivated polio vaccine. Lancet 2000: 356: 1461–63 Department of Virology ll, National Institute of Infectious Diseases, Gakuen 4-7-1, Musashimurayama, Tokyo, 208-0011, Japan (H Yoshida MMedSci, T Miyamura MD); Japan Poliomyelitis Research Institute (H Horie PhD); and Department of Virology, Toyama Institute of Health (K Matsuura BS) Correspondence to: Dr Hiromu Yoshida (e-mail: hyoshida@nih.go.jp) Introduction WHO adopted a target of the year 2000 for eradication of poliomyelitis in the western pacific region. No wild strain has been isolated from paralytic patients in the western pacific region since March, 1997, and an imported wild strain was the cause of the 1999 case in China. 1–3 However, to eradicate poliomyelitis, we must address the fact that as long as immunisation with live oral poliomyelitis vaccine continues, the risk of vaccine-associated paralytic poliomyelitis will still exist. 4,5 In 2000, the USA started to replace oral poliomyelitis vaccine by the inactivated polio vaccine to reduce this risk. 6 Some have reported that when position 472 in the 5' non-coding region of type 3 poliovirus mutates from U to C, the neurovirulence increases. 7 Sequencing analysis at position 472 of the genome of excreted viruses showed that the substitution occurred gradually while the virus replicated in the human gut after administration of oral poliomyelitis vaccine. 8–14 Moreover, the study (which involved mutant analysis by PCR and restriction-enzyme cleavage [MAPREC]—a technique developed by Chumakov and colleagues as a quality-control test for oral poliomyelitis vaccine) showed that the proportion of type 3 viruses containing C at position 472 (472-C revertants) correlated with neurovirulence in monkeys. 15–20 MAPREC is useful not only as a quality-control test for oral poliomyelitis vaccine but also for laboratory surveillance based on virological diagnosis. The study and characterisation of vaccine strains excreted from the human gut and in the environment is important for the eradication of poliomyelitis and vaccine- associated paralytic poliomyelitis. 21–23 Oral poliomyelitis vaccine is given twice annually (in May and October) in Toyama prefecture, Japan. Environmental surveillance in sewage and river water was carried out from October, 1993, to September, 1995. As a result, 16 type 1, 31 type 2, and 31 type 3 polioviruses were isolated. 24 In the present study, we analysed the isolated type 3 viruses by MAPREC to estimate their neurovirulence. Methods Sewage and river water (1 L and 700–800 mL, respectively) were collected twice monthly at the sewage plant and the river-monitoring points in Toyama, and were concentrated by the filter adsorption/elution method. 24 The concentrated specimens were inoculated into rhabdomyosarcoma, primary monkey kidney, or Vero cells, and incubated at 36ºC. Type 3 isolates identified by polio pooled serum were used for further study. The proportion of 472-C revertants was examined by the method described by Chumakov and colleagues. 15 Amplified cDNA containing a virulent base C was identified by restriction-fragment-length polymorphism (RFLP) analysis with MboI. The F313 type 3 strain was used as a reference strain for the attenuate phenotype, 25 and the Leon strain was used as the reference virulent- Characterisation of vaccine-derived polioviruses isolated from sewage and river water in Japan Hiromu Yoshida, Hitoshi Horie, Kumiko Matsuura, Tatsuo Miyamura Articles