Vaccine Characteristics

Oral poliovirus vaccine has been used in the United States since 1963. The vaccine contains serotype 1, serotype 2, and serotype 3 Sabin strains of poliovirus. OPV is easy to use and convenient. It induces both a humoral and a mucosal antibody response. Contact with children excreting vaccine viruses allows for transmission to and immunization of unimmunized individuals. OPV can be made in large volumes relatively inexpensively.

Research on poliovirus vaccines continues, although it is unlikely that there will be major changes in oral polio vaccines because poliovirus elimination is within sight and substantial changes in vaccine formulation would require extensive and expensive evaluation to demonstrate improved safety and at least equivalent efficacy. Minor changes are being made, however. For example, a new stabilizer may be used to provide greater thermostability, which is important in tropical countries where it is difficult to maintain vaccines at a constant temperature.

Current Use in the United States

The 1994 Report of the Committee on Infectious Diseases of the American Academy of Pediatrics states a preference for OPV because it induces intestinal immunity, is simple to administer, is well accepted by patients, results in some immunization of unvaccinated contacts, and has eliminated disease caused by wild-type poliovirus in the United States. In the past, the CDC's ACIP has also endorsed the use of OPV for routine vaccination.

In the United States, OPV is given at ages 2 and 4 months with DTP and Hib vaccines, at ages 6 to 18 months with hepatitis B vaccine, and at 4 to 6 years with either DTP or diphtheria and tetanus toxoids and acellular pertussis vaccine (DTaP). Adults who are incompletely immunized against polio can receive either OPV or IPV if they are at risk of exposure to the disease, e.g.: travelers to areas where endemic poliomyelitis persists; members of communities or special population groups experiencing disease caused by wild-type poliovirus; laboratory workers handling specimens potentially contaminated with or containing poliovirus; health care workers in contact with patients who may be excreting poliovirus; and child-care workers in contact with vaccinated infants and young children. Parents who are incompletely immunized should receive a supplemental dose of vaccine at the time of or prior to their infant's immunization.

Those in whom OPV is contraindicated include individuals with immunodeficiency disorders (including HIV infection, combined immunodeficiency, abnormalities of immunoglobulin synthesis, leukemia, and lymphoma or other generalized malignancy), patients receiving immunosuppressive therapy, and those with household contacts receiving such therapy. In addition, subsequent siblings of immunodeficient children should not receive OPV, and previously unvaccinated adults generally should not receive OPV. Such contraindications, if observed, should prevent some recipient and contact cases of polio.

Immunity

Data about the mucosal immunity induced by OPV are available. A U.S. study by Onorato and colleagues (1991) examined children who had been immunized with either IPV or OPV and who were then subsequently challenged with a monovalent type 1 virus used in OPV, to simulate exposure to wild-type poliovirus. After the challenge, virus shedding was higher from IPV-immunized children than from OPV-immunized children. Sutter and Patriarca (1993) looked at similar data. Virus shedding from naturally immune, OPV-vaccinated, IPV-vaccinated, and susceptible control or naive children was assessed. OPV-vaccinated and naturally immune children had similar levels of virus shedding. Virus shedding from children who received IPV was higher but was not nearly as high as shedding from susceptible hosts. The systemic response to the first dose of OPV is primarily to type 2 virus, and most children develop an antibody response to all three types after two doses.

Safety

The major disadvantage of OPV is the occurrence of VAPP. Over the last 12 years in the United States, the incidence of VAPP has been about 8 to 10 cases per year. Some of those individuals may develop post-polio syndrome (a recurrence or worsening of symptoms, possibly including paralysis) in subsequent years, just as have many of those who contracted polio from the wild-type poliovirus. In this regard IPV has a major advantage: no cases of VAPP occur following its use.

Reversion of some bases of the virus to a more neurovirulent strain occurs after OPV vaccination with all three serotypes, but particularly serotype 3. These “revertant” viruses are not wild-type, and the virus recovered from patients with vaccine-associated cases of polio is not always a revertant. Other genetic changes in the virus are sometimes seen as well.

The risk of VAPP in immunodeficient children is 3,000 times that in normal children and occurs primarily in children with disorders of B-cell function, agammaglobulinemia, and Bruton's X-linked agammaglobulinemia. Vaccine-associated polio appears to be less frequent in developing countries, where OPV is given at an earlier age than in the United States. It is possible that maternal antibody may be protective against vaccine-associated polio. In addition, record-keeping in developing countries is often deficient, and there may be no means for distinguishing VAPP from wild-type polio disease.

An IOM study committee on adverse events associated with vaccines found that evidence was inadequate to accept or reject the claim that transverse myelitis was associated with OPV (IOM, 1994). The committee found that there was evidence favoring acceptance of a causal relationship with Guillain-Barré syndrome. However, this association has been brought into question with subsequent publication of data on the episode that the committee cited. The committee also found that the evidence established a causal relation between OPV and paralytic and nonparalytic polio.

With the exception of earlier findings of simian virus 40 (SV40) contamination in OPV and IPV vaccine prepared from monkey kidneys, there has been no conclusive evidence of other adventitious agents in OPV. Adverse effects of that contamination have not been documented, but it is possible that follow-up might not have been sufficiently long-term or comprehensive to detect them. Other research describing the extraction from human tissue of genetic material resembling SV40 (Carbone et al., 1994) or with significant homology to an African green monkey cytomegalovirus isolate (Martin et al., 1995) has led the authors to speculate that the source of this material might be viral vaccines. Since the time of the SV40 episode, vaccine has been prepared in the kidneys of monkeys born in colonies where adventitious agents are screened out rather than from monkeys caught in the wild. The screening for adventitious agents, including retroviruses, by both manufacturers and FDA, is rigorous. Safety testing is not only for adventitious agents but also for neurovirulent poliovirus.

Vaccine Characteristics

Two IPV products are currently licensed in the United States. Both are known as enhanced-potency IPVs because they contain higher amounts of antigen than the IPV originally licensed in the 1950s. The manufacturing methods and sources of these two vaccines differ, with one produced on Vero cells manufactured in France and the other grown on MRC-5 cells (a human diploid cell line) manufactured in Canada. The vaccine is obtained from supernatants of cell cultures that are filtrated and concentrated. Impurities are removed and the virus is inactivated with formalin. After 6 days of inactivation, at which time the virus is no longer viable, viral clumps are removed by a second filtration process, and inactivation is continued for 6 more days.

Future plans include combinations of diphtheria and tetanus toxoids and acellular pertussis vaccine (DTaP), Haemophilus influenza type b vaccine, hepatitis B virus vaccine, and IPV.

Current Use in the United States

IPV use has been limited in the United States, with IPV indicated for use in immunocompromised individuals and their household contacts, unimmunized adults at risk of exposure to poliovirus, and individuals who choose not to be immunized with OPV. Approximately 300,000 doses of IPV are distributed annually in the United States.

Immunity

Both IPV and OPV result in good humoral immunity. Better mucosal immunity results with the use of OPV. Use of the inactivated vaccine is not associated with VAPP, but neither does it provide the potential benefit of indirect immunization of contacts. IPV provides some mucosal immunity and decreases poliovirus excretion. Upon challenge with OPV virus, excretion of pharyngeal OPV virus is reduced from 75 percent of individuals lacking antibody to 25 percent of those with high levels of antibody. Excretion in stools is reduced less. Recent outbreaks of poliomyelitis caused by wild-type virus have occurred equally in countries with IPV- and OPV-vaccinated populations. They occur mainly in unvaccinated groups, notably religious groups who decline vaccination. Both vaccines protect the immunized populations living near these groups.

In a study by McBean and colleagues (1988), children received one of the two types of IPV vaccines, or OPV vaccine, at 2, 4, and 18 months of age. Antibody responses were measured after the second dose and before and after the third dose. Most children (92 to 100 percent) developed neutralizing antibodies after two doses. The geometric mean titers were similar (or higher for OPV, for type 2) after two doses, but after three doses GMTs were higher in IPV recipients than in OPV recipients.

Blatter and colleagues (unpublished data) administered either a combination IPV-DTP vaccine, or IPV plus DTP administered separately at 2 and 4 months of age. Children received IPV boosters at 18 months of age. More than 95 percent of the infants developed neutralizing antibodies to poliovirus after two doses; 100 percent of those boosted with IPV had neutralizing antibody after the third dose. In another study, all children received a combination DTP-IPV vaccine at 2 and 4 months. At 6 months, either DTP-IPV, DTP plus OPV, or DTP-IPV plus OPV was administered. Neutralizing antibody developed in 99 to 100 percent of children following the third dose. The highest geometric mean titers were seen in the group receiving both IPV and OPV at 6 months of age.

In efficacy trials in Senegal and India (unpublished data), efficacy rates of 90 percent were observed with less than three doses of IPV, and Canadian authorities estimate 96 percent efficacy after three doses.

In a discussion about the possible differences between the effect of the currently used enhanced-potency IPV and the older IPV on gastrointestinal immunity, in light of higher serum antibody concentrations, workshop participants noted that immunoglobulin G antibody decreases pharyngeal excretion. The higher the titers achieved with the enhanced product, the greater the effect is likely to be, at least on the pharynx.

Safety

In 1955, the IPV manufactured by Cutter Laboratories caused 260 cases of paralytic poliomyelitis because there was residual active virus in the vaccine. This was rectified by the introduction of filtration and by prolonging the inactivation with formaldehyde. Ninety million doses of IPV were sold by Pasteur Mérieux from 1981 to 1994. In 1994 alone, 12.7 million doses of IPV were sold by all producers.

In France from 1992 to 1994, 5 million doses of tetanus toxoid/IPV, 12 million doses of DT-IPV, and 8 million doses of pediatric DTP-IPV were distributed. Passive reporting of adverse reactions to the manufacturer, the French government, and the French Pharmaco-vigilance System showed 5 or 10 cases of neurologic disease, mainly the Guillain-Barré syndrome. The rates were lower than the expected background incidence. Anaphylaxis was not reported. A similar experience was reported in Canada, where 700 million doses of monovalent IPV alone were distributed.

It was noted that the birth cohort size in some countries that have used an IPV-only strategy might be too small to detect rare adverse events. For example, at 800,000 to 900,000 per year, the French birth cohort is about one-fourth the size of that in the United States.

IPV Experience in Western Europe

Finland has used IPV since 1960. In 1984, the introduction of wild-type virus resulted in nine cases of polio, with subclinical poliovirus infection detected in other individuals. The epidemic occurred because an ineffective IPV induced a conversion rate of only 70 percent against the responsible strain (type 3). The outbreak was terminated by mass immunization with OPV. Enhanced inactivated polio vaccines were developed to prevent recurrence of the problem. In 1986, Finland reinstated immunization against polio with the enhanced IPV.

Sweden has used IPV successfully since 1957. Polio outbreaks have not occurred there, despite the introduction of wild-type virus in a religious group that does not accept vaccination and despite the Finnish outbreak in 1984. Iceland has used IPV since 1956, and no cases have been reported since 1960. Norway used IPV from 1956 to 1966, used OPV from 1966 to 1979, and has used IPV since 1979. Six cases of vaccine-associated paralytic poliomyelitis and six cases of wild-type poliomyelitis occurred during the period when OPV was used. Seven cases were imported from 1979 to 1994, without spread.

The Netherlands also has used IPV since 1957. That country experienced two introductions of wild-type virus in a religious sect that prohibits immunization. Both times the infection spread to members of the same religious sect in Canada. However, the virus did not spread beyond the religious sect in the Netherlands.

In France, IPV was used from 1956 to 1965, OPV was used from 1965 to 1983, and IPV has been used since 1983. The last vaccine-associated case of poliomyelitis occurred in 1983, and the last case of polio caused by a wild-type strain occurred in 1989. It appears that wild-type polio was eradicated in France by using an IPV schedule.

Denmark has used a combined IPV-OPV schedule (3 doses of IPV followed by 3 doses of OPV) since 1968. One case of VAPP was recorded in 1969, and one case of wild-type polio was reported in 1976.

General Comments

IPV can be combined with other parenterally administered vaccines, it is stable in adverse temperature conditions, and it has an excellent safety record. A disadvantage of IPV is that contacts of vaccinees are not immunized and intestinal immunity is less than with OPV, although herd immunity (protection of a population from a disease even though some individuals have not been immunized) is present when coverage is adequate. The persistence of immunity is probably similar for IPV and OPV.

Pasteur Mérieux is currently the only manufacturer of IPV and would monopolize the market until IPVs in development by other manufacturers are licensed. The amount of vaccine immediately available could not meet the demand in the United States if the schedule were immediately changed from OPV to IPV. The price per dose of IPV is greater than that for OPV; however, some savings are achieved because wastage per vial may be reduced. The price of IPV will be reduced as more manufacturers enter the market, and as the volume of sales increases, the price could become comparable to that of OPV.

Public Health Experience

When OPV was introduced in the United States in 1961–1964, most children who received the vaccine had already received some doses of IPV. Cases of VAPP have been reported rarely in OPV recipients who previously received IPV.

In Canada between 1965 and 1988, the various provinces used IPV, OPV, or a sequential series of IPV and OPV. Surveillance for vaccine-associated disease during this period of time revealed four vaccine-associated cases in recipients. All four cases occurred when the provinces were using exclusively OPV.

The province of Ontario had used only IPV except for a special mass OPV campaign in schools in 1962. A serologic survey was done in the school 6 to 7 years later, and it showed that children who had received supplemental doses of OPV had much higher seroprevalence levels than those who had received IPV alone.

Denmark has had a continuous program of sequential IPV-OPV immunization since 1968. IPV is given at 2, 6, and 15 months of age, and OPV is then given at 2, 3, and 4 years of age. Since 1968, one case of vaccine-associated paralytic poliomyelitis occurred in a child who had received only one previous dose of IPV, and one case of disease caused by wild-type virus occurred in a child who had received three doses of IPV but no OPV.

Effect of a Sequential Series on the Occurrence of Vaccine-Associated Paralytic Poliomyelitis

Prior vaccination with IPV reduces but does not prevent shedding of OPV virus when OPV is subsequently administered. A study by Henry showed that 90 percent of nonimmune infants given a single dose of OPV excreted vaccine virus in their stools on day 5 postvaccination and 80 percent were still excreting the virus on day 20. In children who previously receive three or four doses of IPV, the rate of poliovirus excretion following a dose of OPV is relatively high at first, but the rate of excretion falls off fairly rapidly, along with the risk of spread of the virus that could cause contact cases of VAPP.

Delaying OPV until the second year of life, as could happen in a sequential schedule, carries established benefits and potential risks. Such a delay could prevent some cases of VAPP in children with a primary immunodeficiency disorder, because the immune abnormality might be diagnosed before OPV is scheduled to be given. It cannot be assumed that all of the risk would be eliminated in an IPV-OPV regimen, however. The average age at diagnosis of Bruton's X-linked agammaglobulinemia, for example, is 2.9 years. A risk of delaying OPV administration could be that fecally transmitted pathogens are more likely to be spread to others from a child still in diapers and able to walk (in a day care setting, for example) than from a child who is not ambulatory. In addition, there is some concern that children in undervaccinated areas are less likely to complete their immunization schedules when vaccinations are extended into the second year of life. For example, a substantially larger proportion of children receive their third dose of DTP (given during the first year of life) compared with the proportion of children who receive their third dose of OPV (currently given during the second year).

Denmark, which uses an IPV-OPV schedule, has a VAPP rate of 1.3 per 100 million person-years, substantially lower than the rate of 3.2 per 100 million person-years in the United States. Hinman and colleagues (1988) estimated that, in a birth cohort the size of that in the United States, 2.6 cases of VAPP would occur with an IPV-OPV schedule, compared with 7.4 cases with an OPV schedule, a reduction of about 65 percent.

Antibody responses to IPV suggest that the response after two doses may be sufficient to result in a diminution in the risk of vaccine-associated poliomyelitis. For example, a study during an outbreak of type 1 wild polio in Senegal in the mid-1980s showed that prior vaccination with the enhanced-potency IPV provided protection to 36 percent of individuals receiving one dose and to 89 percent of individuals receiving two doses. Although the statistical confidence intervals around these percentages were large, the data suggest that two doses would probably be effective in preventing disease.

Because the systemic response to the first dose of OPV is primarily to type 2 poliovirus, to ensure mucosal immunity to all three types in a sequential schedule, at least two doses are required.

Immunogenicity and Shedding

A study in Buffalo compared three doses of OPV, three doses of IPV, one dose of IPV followed by two doses of OPV, and two doses of IPV followed by one dose of OPV. The vaccine was administered at 2, 4 and 12 months of age, and serum was drawn at 2, 4, 5, 12, and 13 months of age. No difference in immunogenicity was found between the groups. Seroconversion ranged from 96 to 100 percent in all four groups after the second dose and 100 percent after the third dose.

A study in Baltimore compared immunogenicity and fecal shedding of vaccine virus in five groups. One group received three doses of IPV, another group received three doses of OPV, and three groups received an IPV-OPV schedule that consisted of two doses of IPV followed by one, two or three doses of OPV. All of the groups had excellent seroconversion rates and good geometric mean titers after the third dose. Three months after completing the primary series, all subjects received a single dose of OPV. Fecal shedding of vaccine virus was then determined. In patients who had previously received three doses of IPV, shedding peaked early and then declined. While one dose of OPV prevented some shedding, two or three doses of OPV were better. There was no significant difference in fecal shedding of virus in individuals who had previously received two or three doses of OPV.

Data from one author suggesting a slight increase in the reversion of OPV to the wild-type genotype (at some bases of the virus) in children who receive OPV after previously receiving IPV are disputed by others as artifactual due to inappropriate comparison populations. It was suggested that reversion after an OPV only dose or following IPV-OPV should be the comparison point. Because there is a decrease in virus shedding in patients who previously received a dose of IPV, the overall net effect is uncertain.

Summary

All combinations of IPV and OPV are sufficiently immunogenic. A sequential IPV-OPV immunization schedule would probably prevent four to six of the cases of VAPP that occur in the United States annually. Two doses of IPV are required to prevent VAPP, and two doses of OPV are required for optimal mucosal immunity. The cost of a sequential IPV/OPV schedule would be higher than OPV only given current IPV pricing. A CDC-sponsored study showed that as demand for IPV increases and cost per dose of IPV decreases a break even point would be reached.

One schedule would be to give IPV at 2 and 4 months of age followed by two doses of OPV during the second year of life. By delaying OPV until the second year of life, it might be possible to prevent a few cases of VAPP from occurring in immunodeficient individuals. However, it could also delay the induction of adequate intestinal or mucosal immunity until the second year of life. This would be a problem only if wild-type poliovirus were introduced into the community. The first dose of OPV could be given at 6 months of age, with the second dose given at 15 months of age, but immunodeficient children could inadvertently be vaccinated. A three-dose IPV schedule during the first year of life followed by one or two doses of OPV in the second year of life could be used when combination vaccines become available. However, this would be slightly more expensive.

Vaccine-Associated Paralytic Poliomyelitis

The assumption was made that under an OPV strategy, the number of cases of VAPP would continue at about the same rate as that reported by CDC for the period 1980 to 1992. A strategy of all IPV or no polio vaccine would result in no cases of VAPP, under the assumption that IPV is always completely inactivated. With the parental choice approach, the number of cases of VAPP would range somewhere between no cases and that resulting from the all-OPV vaccine strategy, depending on what proportion of parents chose which vaccine. No data on this possibility are available for the United States. For an IPV-OPV strategy, the assumptions are more complex. Cases among immunodeficient individuals would be estimated under the assumption that IPV has no effect on the probability of VAPP for those with immunodeficiencies and that such a deficiency would not be detected at the time of OPV receipt. For immunocompetent recipients, it is necessary to know how IPV can protect against VAPP following subsequent OPV doses. The model uses serum antibody titers. Implicit in the assumption is that the reason most cases of VAPP occur after the first dose of OPV is that those who are immunized with at least one dose of OPV (and do not get VAPP) are protected for the succeeding doses of OPV. Factors in predicting the number of contact cases of VAPP include the percentage of recipients shedding virus, the duration of excretion, the amount of virus excreted, and the characteristics of the virus excreted.

Susceptible Population

The key parameters in estimating the number of susceptible individuals are vaccine coverage and rate of exposure to vaccine virus. The model treats immunization as an event occurring during the first year of life in a portion of the population. Some people are not protected by the vaccine because of vaccine failures or host characteristics. Those who are not so immunized are susceptible to polio. In succeeding years, susceptible individuals may come into contact with vaccine virus, resulting in either immunization, VAPP, or continued susceptibility. The model currently takes each yearly cohort to age 60.

The number of susceptible individuals for the OPV and IPV-OPV options can be calculated directly from the model. For IPV, the number of susceptible individuals is derived by adding the number not vaccinated to the number vaccinated but not immunized; this is dependent on the coverage level. For the no polio vaccine option, the number of susceptible individuals is considered the total population of all unvaccinated cohorts.

Remaining Questions

Major questions with respect to estimating the risk of VAPP through existing modeling techniques remain unanswered:

• Has there been a recent decline in immunocompetent individuals among cases of VAPP?
• What protection against VAPP is given by circulating antibodies?
  • - What antibody level is required?
  • - Is this host-related?
• What are the characteristics of shed virus?
  • - Are there product differences?
  • - Do studies on revertants conflict?

Questions with respect to predicting the number of susceptible individuals also remain:

• How would immunization coverage rates change with a more complicated schedule?
• What would be the coverage rates for OPV under a strategy with greater use of IPV?
• What is the “true” exposure rate to the vaccine strain of virus (including both vaccinees and contacts)?

Discussion

Several participants suggested that this modeling approach could be promising but that before publication or wide-spread dissemination, the approach should be circulated to a variety of experts for further input, particularly regarding the assumptions that have been made.