1 Department of Microbiology, Graduate School of Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan. anomoto@m.u-tokyo.ac.jp
2 Agency for Cooperation in International Health, 4-11-1 Higashi-machi, Kumamoto City, Kumamoto 862-0901, Japan.
One of the major goals of the WHO is elimination of polio, but the nature of rapidly evolving enteroviruses makes this task more complex than it initially appears.
Poliomyelitis is an acute paralytic disease caused by poliovirus, an enterovirus that belongs to the Picornaviridae (Fig. 1). Humans are the only natural hosts of poliovirus. After oral ingestion, the virus multiplies in the alimentary mucosa, and possibly in the tonsils and Peyer's patches. The virus then moves into the deep cervical and mesenteric lymph nodes and into the blood stream (viremia). The circulating virus invades the central nervous system (CNS) and replicates in neurons, particularly motor neurons. Paralytic poliomyelitis occurs as a result of neuronal destruction by lytic replication of poliovirus, although paralysis develops in less than 1% of those infected. Poliovirus transmission in communities occurs mainly via the fecal-oral route. Transmission is also thought to occur via droplets from the pharynx; this is believed to be the primary transmission route in industrialized countries. Neutralizing antibodies to blood-borne poliovirus prevent the development of poliomyelitis, which suggests that viremia is necessary for the spread of the virus to the CNS.
The genome of poliovirus consists of a single-strand, positive-sense RNA of 7500 nucleotides. VPg (open circle) is a small protein that is covalently attached to the 5' end of the genome; poly(A) is 3' terminal. The positions of initiation and termination of viral polyprotein synthesis are indicated by closed triangles. P1 denotes the viral capsid protein region; P2 and P3 denote the viral noncapsid regions. 2A and 3C are the virus-specific proteases; 3D is the virus-specific RNA polymerase.
Poliomyelitis vaccines Poliovirus is classified into three stable serotypes (types 1, 2 and 3), each of which can cause poliomyelitis. To control poliomyelitis, two strategies have been used to develop preventive vaccines for all three serotypes. One is a live vaccine: this strategy involves the isolation of attenuated viral strains, of which the Sabin strains are the safest and most effective for an oral poliovaccine (OPV)1,
2. This OPV is trivalent: it combines the type 1 Sabin strain (designated Sabin 1) with the Sabin 2 and 3 strains, and it is administered orally. The other vaccination strategy involves the preparation of inactivated wild-type polioviruses by formalin treatment to generate trivalent inactivated poliovaccine (IPV)3, which is administered by injection. Both OPV and IPV are excellent preventive vaccines and have been used effectively throughout the world.
Polio eradication initiative The only natural hosts of poliovirus are humans, and poliomyelitis is controlled by preventive vaccines. This situation is similar to that of smallpox, which was eradicated in 19804. In 1988, the World Health Assembly initiated a program for the global eradication of poliomyelitis by 20005. The major focus of this eradication effort has been to stop wild-type poliovirus transmission, and the OPV was chosen for this purpose.
The OPV, after oral administration, can replicate to a sufficiently high level in the alimentary tract to elicit neutralizing antibodies—secretory immunoglobulin A (IgA) and circulating antibodies—although the virus has little capacity to replicate in the CNS. Soon after OPV was licensed in 1962, it became the vaccine of choice for almost all countries. Compared to IPV, it is much easier to administer, cheaper to produce, provides substantial intestinal immunity (the secretory IgA produced by alimentary mucosa) against wild-type poliovirus infection and spreads to close contacts, thereby protecting a number of people who are not themselves vaccinated. Thus, OPV can prevent transmission of wild-type polioviruses in communities where infection occurs primarily via the fecal-oral route. However, OPV vaccination must result in stopping wild-type poliovirus transmission by any route.
Although the World Health Organization (WHO) missed the original target year of 2000, the polio eradication initiative has made remarkable progress. For example, from 1988 through 2000, the number of countries in which polio was endemic decreased from more than 125 to 20, and the estimated number of polio cases decreased from 350,000 to less than 3500 (ref. 6), which left many countries in the world free of wild-type poliovirus (Fig. 2).
Figure 2. Number of confirmed wild-type poliovirus cases in 2000.
Modified figure used with permission from the WHO newsletter Polio News.
However, in about one case per a few million doses of OPV distributed7, paralytic disease (vaccine-associated paralytic poliomyelitis, or VAPP) occurred among the vaccine recipients and, sometimes, their close contacts. Reversion from the attenuated to the neurovirulent phenotype possibly occurred upon passage through the OPV recipient. This problem is inherent to RNA viruses, whose mutation rates are fairly high (approximately 10-4). Because VAPP cases are extremely rare, however, it is believed that Sabin strains do not to easily revert to fully neurovirulent viruses. It is also thought that the occurrence of such cases could possibly be attributed to an especially high susceptibility of some patients to poliovirus that results from genetic polymorphisms.
Nonetheless, to avoid this small risk, many industrialized countries recently began to use IPV. IPV is relatively expensive to produce and must be given by injection instead of oral administration. Despite these drawbacks, it does not cause VAPP. In addition, it cannot spread to close contacts, as OPV does, because it is an inactivated product. IPV elicits circulating neutralizing antibodies, but not intestinal immunity. Thus, IPV appears to protect only those individuals that receive the vaccination and does not prevent poliovirus transmission via the fecal-oral route, even among the vacinees. Thus, vaccine coverage needs to be nearly 100% in order to control poliomyelitis with IPV. In addition, imported wild-type poliovirus, or even OPV, may circulate in countries with extensive IPV programs. Therefore, switching from an OPV to an IPV program before the global eradication of wild-type poliovirus could be dangerous. In industrialized countries, pharyngeal immunity conferred by IPV may prevent viral transmission because droplet infection from the pharynx is believed to be the major transmission route. It should be noted, however, that enterobacterial food poisoning continues to occur even in industrialized countries, which suggests the continued existence of a fecal-oral route of infection. In addition, IPV is less effective in tropical areas, although the reasons are unknown. In any event, IPV cannot be used in some parts of the world because of its higher costs, more difficult administration method and lower effectiveness compared to OPV.
Certified free of wild-type polioviruses For a region to be considered free of poliovirus, three years must have elapsed since the last case was reported. The concept of certification emerged during the last phase of smallpox eradication4. In that program, certification was given when only two years had passed without any smallpox cases being reported, despite continuing surveillance. The period of two years was based on the longest interval between a supposed last case and the actual true last case (about nine months); that period was then more than doubled for safety. For polio eradication, an additional year was added because of surveillance difficulties. It is fairly simple to carry out smallpox surveillance because all infected persons show characteristic symptoms. In the case of polio eradication, however, surveillance is much more complicated, due to the many asymptomatic polio cases (more than 99% of those infected) and the many acute flaccid paralysis (AFP) patients who are not infected with poliovirus8. To determine whether an AFP patient is infected, wild-type poliovirus must be isolated from a stool specimen taken from the individual. Thus, to achieve the last step of certification of global wild-type virus eradication will be much more difficult than it was for smallpox, and the additional one year required for the certification is not based on actual scientific data.
Certification of the absence of wild-type poliovirus is applied to individual WHO regions: these are the AMRO, American regional office; WPRO, Western Pacific RO; EURO, European RO; SEARO, South Eastern RO; EMRO, East Mediterranean RO; and AFRO, African RO. So far, only AMRO and WPRO have been certified (in 1994 and 2000, respectively). Some national programs may have difficulty declaring a last case because of the complicated strategy for the certification. Certification is especially difficult to accomplish in countries that contain strife-torn areas. Nevertheless, WHO began to plan for the post-eradication era because of the rapid progress in reducing the number of countries with endemic poliomyelitis. Originally, WHO policy had been that, when global eradication of wild-type poliovirus is certified, the worldwide OPV vaccination program will be stopped. This WHO policy is based on data from Hungary and Cuba, which suggest that, under conditions of mass OPV administration, the circulation of vaccine-derived polioviruses (VDPVs) is usually limited to a few months after OPV use is stopped9,
10. This policy, however, has been challenged by recent findings such as (i) some people excrete vaccine viruses for long periods of time11,
12, (ii) a recent poliomyelitis outbreak in Hispaniola (the island shared by the Dominican Republic and Haiti) was caused by VDPV13, (iii) difficulties in the containment of laboratory stocks and (iv) persons who supposedly received adequate OPV vaccination were infected with VDPV and showed paralysis15.
Long-term carriers So far nine people have been found to excrete vaccine viruses for periods of time ranging from a few months to more than ten years11. These people have certain hereditary immune-deficiency disorders. Some VDPVs excreted from these people had fully restored their neurovirulent phenotype12. Few studies have provided indications of the frequency of this phenomenon, and although the WHO believes these cases to be rare, it is probable that others have carried and excreted the viruses. Much more work needs to be done to determine the frequency of this phenomenon, these VDPVs may circulate and cause paralytic disease in communities after OPV vaccination has ceased.
Outbreaks due to VDPVs Analysis of a poliomyelitis outbreak (from July 2000 to July 2001) on Hispaniola shows that the virus responsible was a type 1 VDPV (Table 1)13. The genetic divergence in the viral capsid protein, VP1, coding sequence from the sequence in the original Sabin 1 strain indicates that it had been circulating for about 24 months before causing disease. Low vaccine coverage in this area may have allowed the virus to circulate for this prolonged period. This data is striking because it had been thought that poliomyelitis outbreaks occurred only through the introduction of wild-type poliovirus. Thus, high vaccine coverage and vigilant surveillance are absolutely required to prevent an outbreak caused by VDPVs.
The VDPV responsible for the outbreak in Hispaniola was Sabin 1 virus that had recombined with another unknown nonpolio enterovirus; the 5' half of the VDPV genome was derived from Sabin 1 and the 3' half from the nonpolio enterovirus. It is important that we determine whether the neurovirulence and transmissibility of the VDPV were due to mutations in the Sabin 1 nucleotide sequence of the genome or were a result of the recombination with nonpolio enteroviruses. Neurovirulence can be tested by giving back to the VDPV the corresponding 3' region of the original Sabin 1 virus, but estimating its transmissibility would be difficult.
There has been another case in Egypt, that was similar to that in Hispaniola (Table 1). A recent retrospective study of poliovirus isolates from Egypt has revealed that paralytic diseases from 1988 through 1993 were associated with type 2 VDPVs14. In this case, the VDPV had circulated for ten years (1982−1993). Low vaccination coverage appears to be the common risk factor that allowed VDPVs to circulate among susceptible populations and acquire wild-type characteristics. Considering that vaccine coverage is fairly low in many areas in the world, the number of VDPVs that were detected in the circulation was probably deceptively small. Thus, VDPV circulation should be subjected to more extensive surveillance. Alternatively, it is possible that recombination with particular nonpolio enteroviruses is exceedingly rare. If so, the enterovirus that recombines with the OPV to generate circulating VDPVs must be identified.
Containment of laboratory wild-type virus The WHO indicates that "containment of laboratory virus including destruction of unnecessary stocks of it" is a prerequisite for the elimination of the risk of a return of wild-type poliovirus. Containment is necessary to safeguard the inadvertent release of laboratory viruses. However, many laboratories in the world have fecal specimen stocks for diarrhea research that may contain wild-type poliovirus, which casts doubt as to the feasibility of such a stringent plan. If 100% containment or destruction of laboratory wild-type virus is necessary to stop vaccination, it is doubtful that vaccination programs will ever end: destruction of all laboratory samples seems unrealistic.
OPV might fail to prevent VDPV transmission In addition to the VDPV cases in Hispaniola and Egypt, a small outbreak occurred in the Philippines from March to July 2001 (Table 1)15. The VDPV responsible was type 1 VDPV, which had recombined with a nonpolio enterovirus, as happened in Hispaniola. In this case, however, the patients had presumably received adequate OPV vaccination. In addition, vaccination coverage in this area was not low (it was about 80%). Thus, the reasons for the outbreak are presently unknown. However, it is possible that the antigenicity (or immunogenicity) of this VDPV shifted during its circulation. Slight changes could have allowed the VDPV to be transmitted among people who had inadequate protective response to the VDPV, which resulted in cases—albeit very rare ones—of paralytic disease.
A possible polio-free world Let us suppose that we are able to overcome all difficulties and achieve a polio-free world that is also free of neutralizing poliovirus antibodies. In such a situation, we must take into consideration a return of wild-type virus through bioterrorism. It is not difficult to produce wild-type poliovirus artificially, because the 7500-nucleotide genome can be synthesized from the published sequences of wild-type poliovirus strains. It isn't obvious how we should handle such a situation, but it is worth noting the potential problem.
In addition, there will continue to be a niche for polio in the world of enteroviruses, which evolve rapidly. C-cluster enterovirus (C-CEVs)—which consists of C-cluster coxsackie A viruses (C-CAVs; CAV1, CAV11, CAV13, CAV15, CAV17, CAV18, CAV19, CAV20, CAV21, CAV22, CAV24 and CAV24v) and polioviruses (PV1, PV2 and PV3)—have been grouped together based on their genomic sequences16,
17. On the basis of disease syndromes caused in humans, however, C-CAVs and PVs differ greatly. The former cause respiratory disease like the major receptor group of rhinoviruses do, whereas PVs can cause paralysis. It is assumed that the difference in C-CEV pathogenesis is primarily determined by their cellular receptors. C-CAVs use intercellular adhesion molecule 1 (ICAM-1), as do the major receptor group rhinoviruses, whereas PVs uniquely use CD15516. Both ICAM-1 and CD155 are members of the immunoglobulin superfamily, and it is their immunoglobulin-like domains that are recognized by viruses. Remarkably, based on a phylogenetic analysis of nonstructural viral proteins, the coxsackie strains CAV11, CAV13, CAV17 and CAV18 are intermingled with, rather than separate from, the three poliovirus serotypes (Fig. 3). Type 1 poliovirus is actually more closely related to CAV18 than to type 2 poliovirus. This observation suggests that polioviruses may have emerged from a pool of C-CAVs by evolving toward a unique receptor specificity17. Thus, it is possible that in a world free of poliovirus and neutralizing poliovirus antibodies, C-CAVs would have a greater opportunity to switch their receptor specificity from ICAM-1 to CD155 and, thus, to evolve gradually into a new polio-like virus. It is possible to model such evolutionary processes, and the probability of the development of such a scenario should be tested, as we may have to develop novel vaccines against such new polio-like viruses.
Figure 3. Comparison of the nucleotide sequences obtained from viral RNA polymerase (3D)-coding region.
Nucleotide sequences (390 bases) were compared to each other. A−D indicate the individual genetic clusters. Reproduced with the permission of Elsevier Press from Pulli, T. et al. Virology212
, 30-38 (1995)16.
Summing up The WHO's polio eradication strategy follows that of smallpox eradication, which is the only global program to have successfully eradicated a vaccine-preventable disease. The common strategy used is elimination of the pathogen, certification of a geographic region and the halting of vaccination. With the benefit of hindsight, however, a difference between smallpox virus and poliovirus is now recognized. Although both viruses are strictly host-specific, infecting only humans, smallpox virus (a DNA virus) is genetically stable, whereas poliovirus (an RNA virus) is not, as shown by the recent episodes of the Hispaniola outbreak caused by VDPV. Thus, if polio eradication is to follow the smallpox strategy of halting vaccination after the global certification, it is of vital importance that we conduct worldwide collective studies to at least resolve the following issues. (i)What is the frequency of long-term carriers? (ii) Does IPV prevent poliovirus transmission in industrialized countries? (iii) How are circulating neurovirulent VDPVs born? (iv) Which nonpolio enterovirus is involved in the generation of neurovirulent VDPVs? (v) Are there any VDPVs whose immunogenicity has shifted? (vi) How can we contain wild-type laboratory virus? (vii) How long does protective immunity conferred by OPV or IPV last? (viii) Which genetic polymorphisms in humans confer high susceptibility to poliovirus? In addition to the WHO's proposed plan to implement vigorous OPV vaccination together with extensive surveillance to stop VDPV circulation, the results of these studies will provide useful information on how to safely end immunization programs worldwide.
However, there is an additional problem: the possible emergence of new polio-like viruses after the worldwide elimination of polioviruses. Thus, the reproduction study mentioned earlier should also be added to this list. The research we propose here will further strengthen global surveillance of emerging diseases. The polio dilemma provides an opening from which to investigate the prevention of newly emerging diseases. The participation of the world scientific community—as coordinated by the WHO, and including research groups from the vaccine industry—will be needed if the development of a new vaccine for polio-like diseases is to be contemplated.
It will take at least eight to ten years to come to a point at which we need to make a final decision on whether OPV immunization should be halted immediately after global eradication of wild-type poliovirus has been certified. In the meantime, we must study the nature of poliovirus. Making the decision too soon would be unwise. The cessation of OPV vaccination is a global experiment that we human beings have not yet experienced.
Fenner, F. et al. Smallpox and its Eradication (World Health Organization, Geneva, 1988).
World Health Assembly Global Eradication of Poliomyelitis by the Year 2000 (World Health Organization, Geneva, 1988).
Technical Consultative Group to the World Health Organization on the Global Eradication of Poliomyelitis. Clin. Inf. Dis.34, 72–77 (2002). | ISI |
Stutter, R. W. et al. in Vaccines (eds Plotkin, S. A. & Orenstein, W. A.) 364–408 (Philadelphia, W B Saunders, 1999).
World Health Organization Wkly Epidemiol. Rec.73, 113–117 (1998).
Domok, I. et al. Brit. Med. J.i, 743–746 (1962).
MasLago, P. et al. Int. J. Epidemiol.20, 1029–1034 (2001). | Article |
Wood, D. J. et al. Bull. World Health Org.78, 347–363 (2000). | PubMed | ISI | ChemPort |
Yoneyama, T. et al. in Progress in Polio Eradication: Vaccine strategies for the end game (ed. Brown, F.) Vol. 105, 93–98 (Basel, Karger, 2001). | ChemPort |
Centers for Disease Control and Prevention MMWR Morb. Mortal. Wkly Rep.50, 147–148 (2001).
Centers for Disease Control and Prevention MMWR Morb. Mortal. Wkly Rep.50, 41–42 (2001). | PubMed |
Centers for Disease Control and Prevention MMWR Morb. Mortal. Wkly Rep.50, 874–875 (2001). | PubMed |
Rieder, E. et al. in Progress in Polio Eradication: Vaccine strategies for the end game (ed. Brown, F.) Vol. 105, 111–122 (Basel, Karger, 2001). | ChemPort |
7500 nucleotides. VPg (open circle) is a small protein that is covalently attached to the 5' end of the genome; poly(A) is 3' terminal. The positions of initiation and termination of viral polyprotein synthesis are indicated by closed triangles. P1 denotes the viral capsid protein region; P2 and P3 denote the viral noncapsid regions. 2A and 3C are the virus-specific proteases; 3D is the virus-specific RNA polymerase.
