Immunogenicity of sequential inactivated and oral poliovirus vaccines (OPV) versus inactivated poliovirus vaccine (IPV) alone in healthy infants: A systematic review and meta-analysis

ABSTRACT

Background: The emergence of vaccine-associated paralytic poliomyelitis has become an ongoing burden of poliomyelitis. During this special period from OPV to IPV-only immunization schedule, we did a meta-analysis to compare the immunogenicity of sequential IPV and OPV versus IPV alone in healthy infants.

Methods: This systematic review and meta-analysis was registered at international prospective register of systematic reviews (PROSPERO), and the number was CRD42017054889. We performed it as described.

Results: Finally, 6 articles were qualified for our review. The results showed that seroconversion rates against all 3 serotype polioviruses were non-inferior and Geometric mean antibody titers (GMTs) were superior in sequential schedules compared with IPV-only schedule. Thus, the sequential vaccination schedules could induce a stronger immunogenicity.

Conclusions: To decrease vaccine-associated and vaccine-derived poliomyelitis, it is a reasonable option to select sequential schedules during this special transition from OPV to IPV-only immunization schedule, which coincides with the current WHO recommendations.

KEYWORDS:

• Inactivated poliovirus vaccines
• Oral poliovirus vaccines
• Immunogenicity
• sequential schedule
• vaccination

Introduction

Poliomyelitis, caused by wild poliovirus (type 1, 2 and 3), is an infectious disease threatening human health. Vaccination is the only effective way to eradicate poliomyelitis. With the introduction and widely use of live Oral Polio Vaccine (OPV),¹ which contains a live-attenuated vaccine-virus (Sabin virus), the global epidemic cases of poliomyelitis have led to a 99.99% sharply decrease (from about 350,000 cases in 1988 to 22 reported cases in 2017)² and the serotypes 2 and 3 of wild poliovirus have experienced an eradication.³ Successful elimination is closer than ever before, and wild type 1 poliovirus remains the endemic strain which circulates only in Afghanistan, Pakistan and Nigeria.⁴ Despite the achievements of OPV in reducing the spread of paralytic poliomyelitis, the emergence of vaccine-associated paralytic poliomyelitis (VAPP)⁵ and vaccine-derived poliovirus (VDPV)¹ has become an ongoing burden caused by the use of live-attenuated vaccine strains.⁶ VAPP is a rare but serious adverse event due to genetic mutations and recombination during the replication of attenuated OPV viruses in the gut. The vaccine viruses rapidly accumulate such mutations and genetic rearrangements, reverting to neurovirulence and transmissibility,⁷ especially keen to infect patients with immunodeficiency.⁸ In addition, VDPVs are genetic variants (OPV-derived polioviruses) with the potential to cause paralytic poliomyelitis and can transmit in regions with low population immunity.^9,10 The nucleotide sequence of VDPV varies from the original OPV with the divergence >1%. However, in order to improve sensitivity for early prevention from its outbreaks, this threshold has been adjusted to 0.6% for type 2 VDPVs (VDPV2).¹

Trivalent OPV (tOPV) contains a mixture of live-attenuated polioviruses of all three serotypes, which has been the predominant vaccine against poliovirus used for routine immunization before April 2016. However, the type 2 component of tOPV has caused more than 90% of all cVDPV cases and up to 38% of all VAPP cases in recent years.¹¹ In addition, the burden related to poliomyelitis is mainly resulted from vaccine rather than wild poliovirus recently.⁵ Therefore, the Global Polio Eradication Initiative (GPEI) implemented a switch from tOPV to bivalent OPV (bOPV) containing serotypes 1 and 3 poliovirus, and the switch means that cVDPV and VAPP related to type 2 vaccine components no long occur. To achieve the goal of eliminating VAPP and VDPVs completely requires introduction an inevitable but wise choice¹²: inactivated poliovirus vaccines (IPV). IPV is an inactivated but not a ‘live’ attenuated vaccine, which consists of all three wild-type poliovirus strains, so it carries no possibility to appear VAPP or VDPVs. IPV can induce human bodies to produce protective antibodies against all three types of poliovirus. In case of infection, these antibodies can prevent the virus spreading to central nervous system and protect against poliomyelitis. After OPV cessation, IPV will be the only vaccine available in routine vaccination schedule to eliminate the occurrence of VAPP and cVDVPs. GPEI is planning to quit OPV as soon as possible after Wild poliovirus (WPV) has been certified as eradicated.¹¹ The American Academy of Pediatrics (AAP) also recommended the all-IPV immunization schedule in early 2000.¹³ The polio vaccination policy was changed from OPV to exclusive IPV successfully in the high standards of hygiene and sanitation areas such as United States,¹⁴ Sweden, Iceland, Finland and the Netherlands¹⁵ and subsequently eliminated polio in these industrialized countries. An ever-increasing amount of industrialized, polio-free countries choose IPV as the vaccine. However, IPV-only schedules would not be sufficient to halt poliovirus transmission in regions of poor sanitary condition because of its absence of strong mucosal immunity¹⁵ and close contacts immunization through a secondary spread.¹⁶ Additionally, IPV is too expensive (over five times more expensive than OPV) for global use,¹⁷ especially in developing countries, because administering the IPV requires sterile injection equipment and procedures, as well as trained health workers.^16,18 However, it is inspiring that Gavi supports the Polio Endgame through the introduction of IPV into routine immunization programmes,¹⁹ and there were more than 40 million children immunized with the help of Gavi (containing the poorest countries around the world) by the end of 2016.

Considering those disadvantages for IPV-only schedule, it is difficult to abolish OPV completely and feasible for phased removal of OPV. The sequential IPV-OPV / OPV-IPV vaccination schedules may play a significant role during the transition, although the WHO strategy is to substitute OPV with exclusive IPV for protection against polio ultimately.²⁰ It is consistent with the recommendations of WHO position paper,^21,22 which suggests staged withdrawal of OPV beginning with type 2 component and global introduction at least one dose of IPV into routine vaccination schedules. Under these circumstances, policy makers need to know if such sequential schedules are competent to prevent poliovirus transmission. There was only one study²³ directly assessed the immunogenicity of sequential schedules compared with IPV-only schedule, but median titres comparison for each group was not be implemented. Therefore, it was necessary for further study on this topic. Our meta-analysis is an integrated review, which compared the humoral immunity responds by not only seroconversion rate but also GMTs against poliovirus of those two schedules. In this review we aimed to determine whether sequential vaccination programs could be with a high seroconversion and successfully protect against polio infection in this special period. Thus, it could accelerate an evidence-based decision-making process.

Results

Our search strategy yielded 1800 potentially relevant publications, and there were 1315 records reserved after duplicates removed. Afterwards, we eliminated 1302 articles by screening titles and abstracts for not matching inclusion criteria. Ultimately, 13 papers underwent a full-length review, and of those, 6 studies^23–28 were qualified for our meta-analysis. The ineligible records contained 2 researches^29,30 which included duplication of participants with their own other papers; 3 publications^31–33 in which the vaccination schedule was simultaneous IPV and OPV instead of the sequential scheme. In addition, there was an article³⁴ without a clear amount of samples involved in analysis (its withdrawal rate was not shown). And the results of another one³⁵ were not comparable, because its sequential vaccination groups were inoculated more than the IPV-only group (Figure 1).

Figure 1. Flow chart of study selection.

Methodological quality of qualified studies

Quality assessment of two included randomized controlled trials (RCTs)^23,25 is shown in Table 1. They were both rated as “high” methodological quality. However, the four included case series were all rated as “poor”, because the follow-up rates of three studies^24,27,28 were not up to 90% and the other were never reported or explained.²⁶

Table 1. Methodological quality of included RCTs.

references	Blinding	Random sequence generation	Allocation sequence concealment	Selective reporting	Incomplete outcome
Asturias, E. J. et al.	high	low	low	low	high
O'Ryan, M. et al.	high	low	low	low	low

¹ all analyses were performed in a masked manner except the study participant families and the physicians, because there were obvious variations between OPV and IPV vaccination.

² some incomplete outcome data, intent-to-treat analysis is an inappropriate approach.

Study characteristics

As displayed in Table 2, 4 of 6 qualified studies were published in recent 5 years and the remaining 2 studies^24,25 were published earlier. A half studies were performed in China and the others came from different countries. There were only 2 RCTs^23,25 reported their study period. Moreover, participants were infants aged below 13 weeks in 5 studies. Although the vaccination time varied from study to study, all participants had completed their vaccination program within 24 weeks. The blood collection time detailed was one month after every participant fulfilled the vaccination schedule. Furthermore, sequential IPV and OPV schedules in our selected studies included one or two doses of IPV followed by two or one dose of OPV.

Table 2. Study characteristics.

First author	Year	Country	Study period	Design type	Sample size	Participant Age	Vaccination time	Blood collection time	sequential Vaccination schedule
O'Ryan, M	2015	Chilean	April 25 to August 1, 2013	RCT	528	8 weeks (± 7 days)	8, 16, 24 weeks	28 weeks		I-bO-bO and I-I-bO
Lu, L.	2015	China	NR	case series	339	60-89 days	2, 3, 4 months	5 months		I-O-O and I-I-O
Asturias, E. J.	2007	Guatemalan	April to November 2004	RCT	292	6–11 weeks	2, 4, 6 months	7months		I-I-O
Liu, H. B.	2013	China	NR	case series	81	NP	2, 3, 4 months	5 months		I-O-O and I-I-O
Zhang, L. W.	2014	China	NR	case series	184	60-89 days	2, 3, 4 months	5 months		I-O-O and I-I-O
Faden, H.	1990	U.S.	NR	case series	86	6-10 weeks	2, 4, 12 months	13 months		I-O-O and I-I-O

NP = not reported; I = IPV; O = OPV; bO = bOPV.

Neutralizing antibody seroconversion

As described in selected studies, it was considered that seroconversion presented a positive and protective neutralizing antibody level if titers ≧1:8. Both sequential IPV/OPV and IPV-only groups developed very high seroconversion rates with almost 100% against serotypes 1 and 3 one month after the full vaccination schedules and without significant heterogeneity between groups (Table 3). Besides, seroconversion rates against serotype 2 were both 100%, but it should be noted that a study of O'Ryan et al.²³ was not included, because they used bOPV without type 2 poliovirus component. In summary, both sequential IPV and OPV schedules and IPV alone could induce strong serum immunogenicity in healthy infants.

Table 3. Meta-analysis for seroconversion rates between groups.

Groups		Unweighted rate n/N (percentage)	Odds radio (95% CI, p value)	I^2	Cochrane Q
Serotype 1	Sequential group	932/934 (99.79%)	1.12 (0.13, 9.73), p = 0.92	0%	0.92
	IPV-only group	574/576 (99.65%)
Serotype 2	Sequential group	583/583 (100%)	Not estimable	Not applicable	Not applicable
	IPV-only group	398/398 (100%)
Serotype 3	Sequential group	923/933 (98.93%)	0.61 (0.18, 2.08), p = 0.43	0%	0.86
	IPV-only group	572/574 (99.65%)

GMTs

Afterwards, we performed meta-analysis to compare the immunogenicity between such two groups through the quantitative index GMTs when data was available (two authors^23,24 were contacted, but didn't provided useful information.). Ultimately, there were three eligible studies^26–28 performed in China included. As well known, antibody titer is exponentially increased, so it is unsuitable for direct meta-analysis. It was the value that the logarithm of the raw data to base 2 being used in our further analysis. As shown in Figure 2, the sequential vaccination schedules (IPV followed by OPV) could induce much stronger serotype 1 antibody titers than IPV-only scheme. After exponential conversion, the mean difference (95%CI) was 671.76 (639.40, 704.12). And there was no significant heterogeneity existed between the two groups with Cochran Q test P = 0.85; I² = 0%. Similarly, the meta-analysis forest plots of GMTs in serotype 2 (Figure 3) and serotype 3 (Figure 4) were also finished, which produced the same results as type 1: the weighted GMTs were higher in the sequential group compared with IPV-only group, and the mean differences were 731.50 (604.48, 858.52) and 378.92 (293.73, 464.10) respectively. And there was no great heterogeneity in above two analyses across groups (Cochran Q test P = 0.95; I² = 0% as well as Cochran Q test P = 0.89; I² = 0%).

Figure 2. Meta-analysis plot of polio GMT against serotype 1.

Figure 3. Meta-analysis plot of polio GMT against serotype 2.

Figure 4. Meta-analysis plot of polio GMT against serotype 3.

Subgroup analysis

As mentioned, both sequential IPV/OPV and IPV-only groups induced very strong immune response and without great heterogeneity against all 3 serotypes after completing vaccination programs. Therefore, it was no need and significance for the further subgroup analysis by seroconversion rates. However, considering only three articles could be used for GMTs meta-analysis, those articles were all performed in China and the vaccination procedures were all IPV followed by OPV at the same age, there was no further subgroup analysis performed in this review. We summarized the GMTs of diverse sequential vaccination series (IPV-OPV-OPV vs. IPV-IPV-OPV) in these researches. We found that GMTs were very high after above both vaccination regimens, whereas for the former program, they were slightly higher against serotype 1, and slightly lower against serotype 2 and 3.

Adverse event

As showed in our eligible studies, both sequential IPV/OPV and IPV-only vaccination strategies were well tolerated, with no fatality. Although several infants experienced adverse events and considered to be related to studying vaccines, none aggressive VAPP or VDPV was reported. The relevant adverse events included irritability, fever, local reaction (pain and induration at the vaccination site) and so on.

Discussion

To our knowledge, there was no meta-analysis previously to compare the immunogenicity of sequential IPV and OPV schedules with IPV-only schedules in healthy infants. Compared with IPV-only schedules, we found that seroconversion rates against all 3 serotype polioviruses were not inferior in sequential schedules, and the proportions after three doses were close or up to 100% with no statistical difference between groups. However, the GMTs of seroconversion reached higher levels in sequential schedules than in IPV-only schedules. Thus, sequential schedules could induce a stronger immunogenicity.

With the success in global eradication of wild poliovirus, evidence-based strategies for polio endgame should be agreed on and implemented. Currently, the burden related to poliomyelitis is mainly caused by vaccine rather than wild poliovirus.³⁶ The Polio Eradication and Endgame Strategic Plan 2013–2018 set by GPEI recommended that it was necessary to introduce at least one dose of IPV worldwide in their routine immunization programs and eventually withdrew all OPV by 2019 in order to eliminate all poliomyelitis, including VAPP, and consistently keep eradication of all polioviruses.³⁷ Sequential IPV and OPV immunization schedules seem to optimal option during this transition period. Our meta-analysis also provides an important clinical evidence for the stronger immunogenicity of sequential schedules than IPV-only schedules. To speed up the universal IPV introduction, it is necessary to develop more-affordable IPV options.³⁸ Possible approaches include developing microneedle patch or intradermal device, using fractional IPV dose, reducing the quantity of IPV doses in routine immunization, using adjuvanted vaccines to reduce doses of IPV and optimizing the production process (using of non-infectious seed strain materials). Of course, this is also a GPEI's multi-pronged research agenda to meet this objective. Resik et al.^39,40 demonstrated that fractional dose of IPV (fIPV) delivery using needle-free injector devices is feasible and highly accepted. And subsequent study has⁴¹ revealed that a single fractional dose (1/5 of the full dose) of IPV provides lower seroconversion rates compared with a single full dose, but after 2 fractional doses the rates are non-inferior to those after 2 full doses. In addition, 2 fractional doses of IPV, given intradermally at 6 and 14 weeks, provide higher seroconversion rates than a single full dose given at 14 weeks. Besides, Rivera and his colleagues⁴² found that three aluminium hydroxide adjuvanted vaccines with different doses (1/3, 1/5 and 1/10) of IPV were non-inferior to IPV in percentage of seroconversion. In combination with the results of our meta-analysis, future RCTs can consider exploring humoral and intestinal immunity of those more-affordable IPV options in sequential schedules. Meanwhile, their safety also needs to be closely concentrated on.

There was no comparison between such two schedules about their intestinal immunogenicity, which is reflected by the amount of virus shed and duration of shedding after challenged with corresponding serotype OPV. However, several earlier studies demonstrated that IPV used alone can only provide weak intestinal immunity,^43,44 but can achieve a substantial boost in intestinal immunity in the children under 5 years old who received mixed IPV and OPV vaccination program.^45,46 This finding indicated that sequential schedules were more likely to play an important role in accelerating elimination and preventing outbreaks of poliomyelitis. However, further high-quality studies are needed to support this.

There was no case of poliomyelitis induced by naturally circulating wild-type 2 poliovirus having been reported for more than 15 years. But with the withdrawal of type 2 components in tOPV, IPV should be added to bOPV for a sequential vaccination schedule in order to protect new birth cohorts from outbreaks of poliomyelitis. This measure has already been implemented in some regions, containing Chilean,²³ Latin American,⁴⁶ India⁴⁷ and Pakistan,⁴⁸ and it has been testified inducing excellent intestinal and humoral immune response against type 2 poliovirus. Furthermore, Sáez-Llorens et al.²⁰ found that there was an alternative option to use a novel monovalent high-dose inactivated poliovirus type 2 vaccine (mIPV2HD) for primary protection and outbreak response during the poliomyelitis eradication phase in area at risk of type 2 poliovirus. Because the seroconversion rates and median antibody titers against serotype 2 were both significantly higher in mIPV2HD group than in standard IPV group. To make policy decisions more flexible in choosing a vaccination schedule to eliminate type 2 related VAPP and VDPVs, more investigations are needed for further confirmation.

This meta-analysis had some limitations. There were several variations from the previous plan listed below: 1) it is no longer an ineligibility criterion that the participants administered with other vaccines in the progress of research, because it was unreasonable to quit other routine vaccine regimens during this study. 2) Considering slightly loosening selecting criterion did not affect the quality of review,⁴⁹ case series were also selected in this review rather than just RCTs on account of lacking sufficient eligible RCTs, which suggests that more high-quality studies are required to further confirmation. 3) As mentioned earlier, in-depth subgroup analyzes were not conducted due to the limited literature data selected in our review; so more high-quality researches are indispensable for obtaining more accurate conclusions.

In conclusion, seroconversion rates against all 3 serotype polioviruses were non-inferior and GMTs were superior in sequential IPV and OPV schedules compared with IPV-only schedules. In order to decrease vaccine-associated and vaccine-derived poliomyelitis and take into account the challenges faced with complete withdrawal OPV, it is a reasonable option to select sequential IPV and OPV schedules during this special transition.

Methods

Search methods

This systematic review and meta-analysis was registered at the international prospective register of systematic reviews (PROSPERO), and the number was CRD42017054889. A comprehensive literature retrieval of the Pubmed, Embase (only articles in this database) and Cochrane Central Register of Controlled Trials (CENTRAL) was conducted on November 14th, 2016. The search strategy was without any methods filters and not limited by language and date. We used the search strategy for pubmed below, and adapted it for other databases. Search terms included medical subject headings (MeSH) and free words found in title or abstract. “Poliovirus Vaccine, Inactivated” was our first MeSH, studies identified were combined using the set operator “or” with its free words (Inactivated Poliovirus Vaccine /Vaccine, Inactivated Poliovirus /Salk vaccine /Vaccine, Salk). The second MeSH: “Poliovirus Vaccine, Oral”, which were combined with its free words (Oral Poliovirus Vaccine /Vaccine, Oral Poliovirus / Sabin Vaccine / Vaccine, Sabin) in the same way. Afterwards, we combined the two groups of terms with the operational character “and” to complete our literature search.

Selection criteria

Eligibility criteria were established in advance by the authors and were shown as follows: 1) Studies in which the populations were healthy infants who were not in morbid states and of which the outcomes contained proportion of seroconversion to the three poliovirus serotypes were included. 2) We selected the publications with both sequential IPV-OPV or OPV-IPV schedules and IPV-only vaccination schedule. 3) Only studies with each group no less than 10 participants were selected to minimize bias inevitably related to individual small case series. 4) Only English or Chinese articles were not being removed as well. 5) In addition, RCTs and case series but only completed publications were included in this review.

Study selection and Data collection

Two independent reviewers (Wen Yin and Guihua Tang) screened the title, abstract or full text and determined their eligibility of retrieved literatures using predefined inclusion criteria. Discrepancies were resolved through consultation and discussion, asking for guidance, if necessary.

Data were extracted using a pretested and standardized data extraction form by the above two reviewers respectively. We have tried our best to contact the corresponding authors for the unpublished GMTs of qualified studies. If there were any differences, a supplementary discussion to consensus was carried out. Information collected included demographic information; vaccination schedules; blood sampling date; seroconversion rates and antibody titers; and adverse events related to vaccination.

Methodological quality assessment

Methodological qualities of Randomized controlled trials were assessed using the Cochrane risk of bias tool.^50,51 The items were evaluated including blinding, sequence generation, allocation sequence concealment, selective outcome reporting, dealing with incomplete outcome data and other potential threats to validity. They were rated as high, low or unclear risks respectively. Whereas case series were appraised using the 8 guidance criterions from the Centre for Reviews and Dissemination checklist⁵² (Table 4). Researches were identified as “good” quality if matched all listed criteria, they were considered to be “satisfactory” when the answer is “yes” to checklists 2 and 4–7, and rated as “poor” if the answer is “no” to any criteria listed for “satisfactory”. Results of methodological quality appraisal did not affect the eligibility of the studies.

Table 4. Quality assessment items for case series.

Quality criterions from the Centre for Reviews and Dissemination checklist
1. Were selection/eligibility criteria adequately reported?
2. Was the selected population representative of that seen in normal practice?
3. Did the study report relevant prognostic factors?
4. Were patients recruited consecutively?
5. Were patients recruited prospectively?
6. Were at least 90% of those included at baseline followed up?
7. Was loss to follow-up reported or explained?
8. Was an appropriate measure of variability reported?

Statistical methods

The immunogenicity of diverse polio vaccination program was calculated and judged by the rates of seroconversion and GMTs. We analyzed the data with Review Manager (version 5.3) and used the Mantel-Haenszel method to calculate odd ratios with 95% confidence intervals (CI) to summarize the seroconversion rates and GMTs of the three serotype poliovirus respectively as well as predefined subgroups across trials. Considering the most conservative (a more correct) estimate will be given, we performed this meta-analysis based on the random-effects model. Heterogeneity among studies or subgroups was evaluated using the Cochran Q test and the I² statistic, while Cochrane Q < 0.10 or I² > 50% was considered to significant heterogeneity.

Conflicts of interest

None.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.