Participants, Materials, and Methods

Study design Elderly people born in the birth cohort presumed to be at risk (1925–1945; hereafter, “the risk birth cohort”) and IPV recipients born in the adjacent birth cohort (1945–1950) were challenged with a standardized dose of monovalent OPV type 1 (MOPV-1) or type 3 (MOPV-3) and were followed for 8 weeks. An outline of the study design is given in figure 1. All inhabitants of Tiel from the risk birth cohort (n=6175) and a random sample of inhabitants from the adjacent birth cohort (n=850) were invited to participate in the screening. Tiel is a medium-sized town in the center of the country and has 40,000 inhabitants and average rates of vaccination coverage. The purpose of the screening was to assess levels of antibodies against poliovirus types 1–3. During screening, participants signed an informed-consent form, provided a serum sample (10 mL), and completed a questionnaire. Within 8 weeks, the participants received written information about their antibody levels. Those who were not selected for the second part of the study and had antibody levels that were inadequate to provide protection were advised to obtain a booster inoculation of diphtheria, tetanus, and inactivated poliovirus vaccine (DT-IPV)

For the second part of the study, the OPV challenge, participants were selected on the basis of their prechallenge serum neutralizing antibody titers, IPV vaccination histories, and exclusion criteria. Those with major medical problems (particularly immunodeficiency disorders) and with a history of OPV vaccination were excluded from the study. The selected participants were assigned to 1 of the 4 following groups: those seronegative for poliovirus type 1 with no history of vaccination (SN-1), those seronegative for poliovirus type 3 with no history of vaccination (SN-3), those seropositive for all 3 poliovirus types with no history of vaccination (i.e., those who were naturally immune [NI]), and those seropositive for all 3 poliovirus types with a documented history of a full series of IPV vaccinations and no history of OPV vaccination (IPVV). The NI and IPVV participants were randomly assigned to be challenged with either MOPV-1 (hereafter, “NI-1” and “IPVV-1”) or MOPV-3 (hereafter, “NI-3” and “IPVV-3”), so that 6 groups were used in the analyses. At the first appointment, prechallenge (day 0) serum samples were obtained, after which OPV was administered. Stool samples were obtained on days 3, 7, 14, 21, 28, 35, 42, 49, and 56 after challenge. Serum samples were obtained on days 7, 28, and 56 after challenge. On completion of their participation, the participants who belonged to the SN-1 and SN-3 groups were offered a full series of inoculations with DT-IPV. The Medical Ethics Committee of the Dutch Organization for Applied Scientific Research (Leiden, The Netherlands) approved the study proposal

Vaccines To keep the study feasibly small, only 2 vaccines, for poliovirus types 1 and 3, were used in the challenge experiments. Poliovirus types 1 and 3 are the most important types, having caused epidemics during recent decades. MOPV-1 and MOPV-3 (Oral-Virelon T1 Type 1 and Oral-Virelon T1 Type 3; Chiron Behring) contained 5.8×10⁵ TCID₅₀ of poliovirus type 1 and 1.0×10⁴ TCID₅₀ of poliovirus type 3, respectively, according to the manufacturer. Vaccine was received from the manufacturer 1 week before the start of the challenge study (April 1999), was stored at 4°C until use, was administered orally in 1-mL volumes, and was used within the shelf-life expiration date assigned by the manufacturer (October 1999)

Collection of samples At each of the 5 visits, venous blood (10 mL) was obtained for antibody determinations (neutralizing antibody titers and poliovirus type–specific IgA and IgM). Serum was separated and stored at −20°C until testing. Stool samples, collected to determine viral excretion, were collected at home and kept in the refrigerator until delivery or were sent immediately by regular mail to the National Institute of Public Health and the Environment. Virus isolation was performed by use of L20B cells, as described elsewhere [11]

Serological assays All serum samples were tested for poliovirus antibodies by use of an ELISA developed by Herremans et al. (PoBi assay; [12]). Titers obtained in the PoBi assay and the standard neutralizing antibody test (NT) are closely correlated [12]. Seronegativity for poliovirus types 1 and 3 was confirmed by a standard NT, as recommended by the World Health Organization (WHO) [11]. The Sabin strains of poliovirus types 1 and 3 were used for this purpose. Furthermore, a standard NT was used to confirm all PoBi assay results in the SN-3 group. In brief, serial 2-fold dilutions of serum samples and 100 50% cell-culture infective doses of virus were incubated in 96-microwell plates (Greiner) for 3 h at 37°C. After incubation, 1.75×10⁴ HEp-2C cells were added to each well. After 6 days of incubation at 37°C, the plates were read. The results are given as log₂ reciprocal titers, expressed as the reciprocal of the greatest dilution showing complete neutralization of the cytopathic effect of 100% cell-culture infectious doses. A titer of 1:8 (i.e., a log₂ titer of 3) or more is a generally accepted and frequently used indication of protective immunity. A rapid (i.e., memory) response was defined as a ⩾4-fold increase in neutralizing antibody titer occurring during the first 7 days after OPV challenge [13]

A total of 42 participants (10%) were excluded from the analysis because of misclassification. Misclassifications were caused by small differences between the antibody titers determined during screening and the prechallenge antibody titers. All serum samples were tested for poliovirus-specific IgA and for poliovirus-specific IgM by use of ELISAs, as described elsewhere [14, 15]

Results

Participation In total, 7025 people were invited to the screening, and 1847 people (26%) participated. The level of participation in the challenge study (for the entire period of 8 weeks) was much greater, with an average participation rate of 48%. Figure 1 shows the number of participants in each group and the corresponding participation rates

Serum neutralizing antibody response after OPV challenge The percentage of participants experiencing a ⩾4-fold increase in neutralizing antibody titers for poliovirus types 1–3 (hereafter, the “responders”) after challenge with either MOPV-1 or MOPV-3 are given in table 1 for SN-1, SN-3, NI-1, NI-3, IPVV-1, and IPVV-3 participants, stratified by prechallenge antibody titer. The postchallenge antibody responses were classified as rapid (⩾4-fold increase occurring within 7 days), slow (⩾4-fold increase occurring between days 8 and 56) and none (<4-fold increase during the study period)

Most of the seronegative participants challenged with OPV experienced a slow (i.e., primary) antibody response—60% of the SN-1 participants after MOPV-1 challenge and 68% of the SN-3 participants after MOPV-3 challenge. In contrast, 33% of the SN-1 participants and only 5% of the SN-3 participants experienced the expected rapid (i.e., secondary) antibody response after challenge. The remainder did not experience an antibody response to the challenge

Overall, 93% of the SN-1 participants experienced a ⩾4-fold increase in neutralizing antibody titers for poliovirus type 1 after challenge with MOPV-1, which was significantly more than the 33% of the IPVV-1 participants and the 46% of the NI-1 participants. The groups challenged with MOPV-3 showed a similar pattern but with a lower rate of ⩾4-fold increases in neutralizing antibody titer. In the 2 seronegative groups, lower prechallenge antibody titers were not associated with a higher rate of ⩾4-fold increases in neutralizing antibody titer

The serological data showed clear evidence of cross-reactivity. A substantial percentage of the participants challenged with MOPV-1 experienced a ⩾4-fold increase in neutralizing antibody titers for poliovirus types 2 and 3 as well. Similar results were found for the participants challenged with MOPV-3. The highest percentages of participants experiencing a cross-reactive ⩾4-fold increase in neutralizing antibody titer were observed for poliovirus type 2 in the 2 seronegative groups—56% of the SN-1 participants and 49% of the SN-3 participants. In addition, booster responses for the virus types other than the one used in the challenge were also found for the IPVV and NI participants (up to 26%)

IgM response Poliovirus-specific IgM was undetectable in all serum samples from the rapid responders and was almost absent in the serum samples from the slow responders experiencing a ⩾4-fold increase in neutralizing antibody titer on day 56 (table 2). Poliovirus-specific IgM was, however, detectable in serum samples from slow responders experiencing a ⩾4-fold increase in neutralizing antibody titer on day 28. The greatest prevalences were found in the 2 seronegative groups

IgA response As expected, poliovirus-specific IgA was absent in the majority of serum samples from seronegative participants before challenge and increased significantly after challenge, with a peak on day 28 (table 3). Of the IPVV and NI participants, the majority were already seropositive for IgA before challenge

Poliovirus excretion Quantitative data on virus excretion are shown in figures 2 and 3. Most seronegative participants challenged with OPV excreted virus for a considerable period of time (figures 2A and 3A ). In the SN-1 group, the highest percentage (81%) excreted poliovirus type 1 three days after challenge, whereas, in the SN-3 group, the highest percentage (75%) excreted poliovirus type 3 seven days after challenge. Although the maximum excretion rates for poliovirus type 3 were observed somewhat later, compared with those for poliovirus type 1, overall, the percentage of seronegative participants who excreted poliovirus type 3 was higher during the entire study period

Comparison of the excretion rates for poliovirus type 1 showed that rapid and slow responders excreted similar amounts of virus. For poliovirus type 3, the excretion rates of the rapid and slow responders showed a consistent pattern: the slower the increase in neutralizing antibody titer, the higher the percentage of individuals excreting poliovirus type 3. Furthermore, in both the SN-1 and SN-3 groups, there was no relationship between prechallenge titers and (1) duration of excreting or (2) the percentage of participants excreting virus (data not shown)

Figure 2B and 2C and figure 3B and 3C clearly show the effect of preexisting antibodies on virus excretion. As expected, only 17% of the IPVV-1 participants and 18% of the IPVV-3 participants excreted challenge virus, and the time of excretion was short (maximum, 28 days), compared with that in the 2 seronegative groups. Similar results were found for the NI participants. After challenge, only 22% of the NI-1 participants and 30% of the NI-3 participants excreted challenge virus, and the period of excretion was relatively short

Finally, the mean titers of virus isolated from stool were not significantly different between poliovirus type 1 excreters and poliovirus type 3 excreters when SN-1, SN-3, IPVV, or NI participants were compared. There was no relationship between the titers of virus shed in stool and day after challenge

Discussion

In the present study, we found clear evidence of memory immunity in 33% of the elderly population without detectable antibodies against poliovirus type 1 (i.e., the SN-1 group) and in 5% of the elderly population without detectable antibodies against poliovirus type 3 (i.e., the SN-3 group). The complete absence of IgM responses among rapid responders is supplementary evidence for the existence of memory immunity, because IgM antibodies are characteristic of primary antibody responses. In contrast to preexisting antibodies, memory immunity, as measured in persons seronegative for the type of virus they are challenged with, does not protect against virus excretion. These results have important implications for the poliomyelitis-eradication initiative, in particular for future immunization policies after eradication has been achieved

Our study was aimed at the elderly because lower seroprevalences were found in that age group [5]. To our knowledge, there are no published studies that have evaluated the antibody response among the elderly after artificial OPV challenge. All OPV challenge studies conducted to date have focused on children. However, it is important to mention that some of these studies have suggested that people with low or undetectable serum antibody levels are probably not in danger of developing clinical poliomyelitis, because they possess memory immunity [16, 17]. The concept of immunologic hyperreactivity, or immunologic memory, was invoked almost 50 years ago [18]. As this concept developed [19, 20], it was suggested that, when the level of neutralizing antibodies fell below detectability, immunologic memory would persist irreversibly, so that restimulation by vaccine or infection would result in a rapid and strong increase in antibody levels. This secondary response to infection was postulated to be rapid enough to protect against virus replication and paralytic disease. In other words, the individual who has immunologic memory, even if he or she does not have (detectable) circulating antibodies at the time of exposure, may respond sufficiently rapidly to block invasion of the central nervous system and prevent viral excretion in feces. In the present study, a rapid (i.e., memory) response was defined as a ⩾4-fold increase in neutralizing antibody titer occurring within 7 days after OPV challenge, which is the definition most commonly used [13]

Our study found few rapid responders—33% of the SN-1 participants and 5% of the SN-3 participants. However, it is likely that this is an underestimation of the true effect, because antibody responses considered to be indicative of a rapid response were evaluated on day 7 only. Our choice of day 7 was based on the definition of memory response and on the results of studies by Herremans et al. [21] and Rumke et al. [10], in which IPV was used as the challenge virus

Noteworthy is the small number of rapid responders observed after challenge with MOPV-3. Although one might argue that this finding was caused by the fact that the potency of the type 3 vaccine used in the present study was less than that recommended by the WHO for trivalent OPV, our results on virus-excretion patterns clearly show that this was not the case. After all, no differences in the virus-excretion patterns of the 2 seronegative groups were observed. Furthermore, the OPV type 3 in trivalent OPV is more highly dosed, compared with monovalent OPV type 3, to balance the dominance of OPV type 2 [22]. Finally, our results are consistent with those of an earlier study in children who received trivalent OPV; this study showed that rates of seroconversion are lowest for poliovirus type 3 [23]

The small number of antibody responses (⩾4-fold increase) among the IPVV and NI participants in the present study seems to indicate an excellent level of protection—there was no increase in neutralizing antibody titer because poliovirus replication was absent. In these participants, titers of antibody (both IgG and IgA) against poliovirus were already present before OPV challenge. As described elsewhere [24], secretory IgA provides a local barrier to poliovirus infection, although the level of antibody that provides protection is not known. The presence of IgA in the serum of the IPVV participants can be explained by the fact that poliovirus was still endemic during the initial period of their lives, and so infection with poliovirus might have occurred. Because IPV vaccination alone is not able to induce IgA, previous mucosal priming with live virus is needed. Subsequently, IPV is able to boost IgA responses, which explains the higher prevalence of IgA found in the IPVV participants, compared with that in the NI participants [21]

The prevalence of antibody responses was somewhat greater for the NI participants, which can be explained by the lower prechallenge antibody levels. Previous studies [8, 9] have suggested that higher prechallenge antibody levels prevent enteric reinfection by a vaccine virus after a challenge dose is administered, whereas lower prechallenge antibody levels permit boosting infection to take place

In most previous challenge studies that used monovalent vaccine, serological responses were described only for the type of virus used for the challenge. In the present study, the existence of cross-reactivity has been clearly shown. In the seronegative groups, as many as 66% of the participants were found to experience a primary response for 1 of the 2 poliovirus types other than the challenge type. Although most participants experienced a primary response, we even found a noticeable number of participants who experienced a rapid response for a poliovirus type other than the challenge type. The high level of cross-reactivity observed in the present study is not a PoBi assay artifact, because all results in the SN-3 group and a sample of the results in the SN-1 group (25%) were confirmed by a standard NT. Moreover, this surprising level of cross-reactivity has not previously been observed with the PoBi technique [12]. A potential explanation is that elderly people have previously come into contact with the other types, given that poliovirus was endemic during their lives. These cross-reactive responses we measured could partly be secondary responses

A unique feature of the present study was the challenge with OPV instead of the more commonly used IPV; challenge with OPV simulates a more natural route of infection and provided us with the opportunity to monitor virus excretion in feces as a parameter of protection. Failure to detect viral multiplication in the intestine (by excretion of virus in stool) after challenge with poliovirus indirectly demonstrates the presence of intestinal immunity. Conversely, fecal shedding of virus comparable with that seen in a nonimmune individual suggests absence of intestinal immunity and neutralization of the virus. One of the most striking results of the present study is that similar fecal virus-excretion patterns were found for the rapid and slow responders in the 2 seronegative groups. Furthermore, ∼80% of the seronegative participants excreted virus after challenge with OPV, and, for 85%–100% of those who showed a memory response, virus excretion continued for up to 8 weeks after challenge. The effect that preexisting antibodies has on virus excretion was clearly shown: only ∼20% of the IPVV participants and ∼25% of the NI participants excreted challenge virus, and the period of excretion was relatively short

Extrapolation of our findings to the general population results in estimates that at least 6% (for poliovirus type 1) and 15% (for poliovirus type 3) of the elderly population of The Netherlands is at risk for infection [5]. In addition to the individual risk of infection, the risk of the circulation of poliovirus among the population should be emphasized, because the vast majority of those protected by memory immunity will excrete virus for a substantial period of time after exposure. Moreover, virus excretion will be considerably higher after exposure to wt poliovirus, compared with that after exposure to OPV

It can be debated whether protection against virus excretion (as measured in the present study) is correlated with protection against disease. Although the implications of our results will manifest themselves only in the event of the reintroduction of the virus—and the risk of reintroduction is small—in the event of an epidemic, we would consider vaccination for those born between 1925 and 1945 who live in the Bible Belt of The Netherlands

In summary, on the basis of the rapidity of the antibody response and the absence of IgM, we conclude that there is clear evidence for the existence of memory immunity in 33% of the elderly population without detectable antibodies against poliovirus type 1 and in 5% of the elderly population without detectable antibodies against poliovirus type 3. However, the majority of seronegative individuals, whether showing evidence of memory immunity or not, are a threat to the poliomyelitis-eradication initiative, because they are not protected against virus replication after challenge with OPV