Measles continues to be a leading cause of childhood morbidity and mortality in developing countries and an outbreak threat in the majority of countries, despite the availability of an effective vaccine for over 40 years [1]. Globally, major reductions have been achieved in measles mortality, from an estimated 733,000 deaths in 2000 to 164,000 in 2008 [2].

Measles virus (MeV) is a negative-sense, single-stranded RNA virus in the genus Morbillivirus within the family Paramyxoviridae. MeV is serologically monotypic, but genetic variability exists among wild type strains [3]. The World Health Organization (WHO) currently recognizes 23 genotypes and one provisional genotype of MeV [4], [5] based on sequence analysis of the 450 nucleotides (nt) that code for the 3′ region of the nucleoprotein (N) gene (N450) and 1854 nt that code for the entire hemagglutin (H) gene [6], [7], [8], [9].

Molecular epidemiological studies of MeV have made significant contributions to measles control efforts, by providing a means to confirm the sources of virus or suggest a source for unknown-source cases as well as to establish links, trace the transmission pathway, allow discrimination of imported from indigenous measles cases, or classify suspected cases as vaccine reactions [10], [11], [12]. Molecular surveillance is most beneficial when there is continuous high quality surveillance and it is possible to observe the change in viral genotypes over time in a particular geographical region. Since this information can help to document the interruption of transmission of MeV, molecular surveillance can provide an important method to assess the effectiveness of vaccination programmes.

Measles is a vaccine-preventable disease, but is still a major killer of children worldwide. Measles caused many outbreaks in China before the 1965 introduction of liquid measles vaccine (MV) administered as 1 dose to infants aged >8 months. The National Expanded Program on Immunization (EPI) was established in 1978 to allow routine immunization with MV to cover the entire country [13], [14]. During 1986–2004, a 2-dose schedule using lyophilized MV at 8 months and at 7 years of age was introduced [6]. The targeted population of the routine immunization has been adjusted to 8 months and 24 months of 2 doses since 2005. When the Universal Childhood Immunization goals were attained, measles morbidity and mortality decreased markedly [13], [15].

The WHO Regional Committee of the Western Pacific Region (WPR) formally declared a measles elimination goal in 2005 and established a target date of 2012 for regional measles elimination [16]. In China, MeV surveillance has been established since 1993, and a serological surveillance laboratory network was established in 2001 to provide serological confirmation of measles infection by IgM testing and to support measles surveillance [14], [15]. To reach the goal of measles elimination by 2012 in China, the Ministry of Health (MOH) proposed a comprehensive vaccination and surveillance programme in 2006 (Measles Elimination Action Plan from 2006–2012 issued by MOH). The "Measles elimination action plan" defines maintenance of a high routine coverage of the first measles containing vaccine (MCV) dose at 8 months of age and the second MCV dose at 18–24 months combined with supplemental immunization activities among individuals aged 8 months to 14 years [14]. We have been studying the molecular epidemiology of MeV circulating in China since the 1990s [13], [15], [17], [18]. In this study, we reviewed the incidence of measles cases in China during 1991 to 2008 and analyzed the sequences of a collection of 1507 measles isolates collected between 1993 and 2008. Viruses of the single endemic genotype, H1, were found in different geographical areas of China over the period of at least 16 years and a major shift from H1b to H1a was documented. This study investigated the genotype and cluster distribution of MeV in China over a 16-year period to establish an important genetic baseline before MeV elimination in the Western Pacific Region (WPR).

This study summarizes the measles surveillance data and documents epidemic measles cycles occurring approximately every 3 to 5 years (1992, 1997, 2001, 2005, 2008). China experienced lower incidence for 11 years during 1994–2004 after two dose measles vaccine schedule were initiated in 1986. However, this was followed by measles resurgence during 2005–2008. Thus, the 3–5 years measles epidemic cycles were possibly associated with the fluctuation of immunity coming from the combination of natural infection and vaccine-induced immunity.

China has committed to eliminate endemic MeV transmission by 2012 [16]. It is critical to determine the genetic baseline of MeV in China before MeV elimination. This study investigated the genotype and cluster distribution of MeV in China over a 16-year period to establish the genetic baseline. Molecular epidemiology and phylogenetic analyses have become important tools in monitoring virus circulation and the progress of elimination efforts, and these baseline data will prove invaluable for classifying viruses found in China as either indigenous or imported. For example in 2009, based on the data presented here, genotype D9, D4 and d11 viruses were identified by molecular epidemiology to be non-indigenous strains and in combination with epidemiological investigations were shown to be imported from Thailand, France and Myanmar, respectively.[5], [23], [24].

Among the wild-type viruses of two genotypes detected in China between 1993 and 2008 (H2 and H1), genotypes H2 had been not found since 1993–1994. H2 genotype virus, Beijing.China94-1/H2 (later designated as the WHO reference strain) was isolated from an unvaccinated 22-year-old male from Beijing, but no epidemiologic data were available to indicate the source of infection. However, it was found to be closely genetically similar to contemporaneously circulating Vietnamese measles viruses [25], [26]. No H2 viruses were found in mainland China after 1994. The genotype A strain: MVi/Shandong,CHN/9353 should be a vaccine associated virus since no more wild type viruses were detected since recent decades. One virus, MVi/Shandong.PRC/14.07/1, was isolated from the throat swab specimen of a vaccinee, a 7 years old boy who received measles vaccine (S191) 7 days before rash onset. Although immunization histories were not available for the other three genotype A sporadic cases, the sequence information indicated that the three isolates likely came from the MeV used in the vaccination programme. Vaccine virus isolations can easily occur when outbreaks and vaccination campaigns occur simultaneously and 10% of S191 vaccinee have the side effect of fever and rash, therefore, more S191 vaccine viruses will be detected with the implementation of Supplementary Immunization Activities (SIAs) in China. Therefore, it is very important to differentiate the measles cases and vaccine associated cases rapidly to guide the control strategies.

The data presented here show that genotype H1 was the predominant circulating genotype throughout China continuously for at least 16 years. Genotype H1 has the quite great degree of intratypic variation among all genotypes, with sequences falling into two major clusters, previously designated as cluster 1 and cluster 2 [13], [15], [17] which corresponded to clusters H1a and H1b, respectively, in our study.

A 4.2% average nucleotide divergence was found between the two clusters, cluster H1a and H1b. These two clusters of viruses were co-circulating during 1993 to 2005, while cluster H1a viruses was continuously circulating after 2006. In order to rapidly increase the immunity level of population, block the spread of the virus and reduce the incidence, the province-specific SIAs had been conducted since 2004. The transmission of H1b viruses was possibly interrupted by vaccination in 2005 and that only the H1a viruses continued to circulate. With the implementation of province-specific SIAs in 2004–2009 and following synchronized nationwide SIAs throughout of China, the reported measles incidence decreased to a historically low level of 39.5–28.9 cases per million in 2009–2010. The imported D4, D9 and d11 genotype measles viruses were found, however H1 still is predominant genotype circulating in China mainland during 2009–2010(unpublished data). More and more imported non-H1 genotype measles viruses could be detected at the measles elimination stage in China in the future.

The molecular epidemiological pattern of measles viruses in China appears to be similar to mumps virus [27], human enterovirus 71 [28], and coxsackievirus A16 [29], where one predominant genotype continuously circulating in China since they were first detected in the 1990s. However, it is quite different from those of the rubella virus [30], where virological surveillance for rubella virus since 1979 showed that 5 different genotypes were cocirculating in China.

In this study, analysis of the predicted amino acid substitution in the 450 nt sequencing window of the MeV N gene showed no evidence of selective pressure in the past 16 years. Of course, the major target of neutralizing antibodies to measles virus in the H protein and it will be important to monitor genetic changes that may affect antigenic sites on this protein.

Although measles virological surveillance has been successfully implemented in China, and now included 30 of 31 provinces, systematic collection from each province and each prefecture should be encouraged to establish a baseline of complete chronological and geographical distribution of Chinese MeV for the entire country. Quickly identifying imported cases of measles will become more critical as measles elimination goals are achieved in China in the near future.