Poliovirus is transmitted via the oral-oral route or the fecal-oral route, through inhalation of droplets, or contamination of hands, food, utensils or water with nasopharyngeal secretions or fecal material from an infected individual [5]. Infection of naïve hosts with poliovirus results in viral replication and shedding in nasopharyngeal secretions for about 1–2 weeks and in stools for several weeks. Poliovirus replication stimulates the production of type-specific neutralizing antibodies in serum (Immunoglobulins M, G and A), and in nasopharyngeal and intestinal secretions (secretory immunoglobulin A), within 1–3 weeks [6].

Serum and mucosal antibody levels decline slowly with time and may become negative following initial infection, but are boosted upon new exposures or vaccination. The presence of antibodies in serum confers protection against paralysis even at low titers at the level of detection [7]. Even if antibodies decrease to undetectable levels, poliovirus-specific memory cells appear to provide life-long protection against paralysis. In contrast to humoral immunity, mucosal immunity is partial and wanes with time. Studies conducted in the United States in the 1950’s found that shedding of vaccine virus following administration of OPV occurred in a smaller proportion and for shorter duration among children and adults who were seropositive from natural infection before the challenge OPV dose than in seronegative individuals [8,9].

OPV contains live attenuated poliovirus strains (single or combinations of Sabin strains types 1, 2, and 3 depending on preparation) that induce humoral and mucosal immunity in a way similar to infection from wild poliovirus [6,10,11]. About 10–25% of fully OPV-vaccinated children, depending on time since last dose, will still shed virus following an OPV challenge, especially with a high dose [9,10,12–14]. OPV-vaccinated individuals also shed poliovirus for a shorter time and at lower viral concentration than naïve and IPV-vaccinated individuals [9–12].

Provision of several doses of OPV through routine immunization offers long-term protection against paralysis and reduces probability and duration of viral shedding upon re-infection with poliovirus, thus contributing to the interruption of poliovirus circulation [13]. Sabin strains, especially type 2, may also spread and indirectly vaccinate contacts [15], which enhances the field effectiveness of OPV. For example, after OPV introduction in Yaoundé, Cameroon in 1979, the incidence of paralytic poliomyelitis decreased by 85% although only 35% of children 12 to 13 months old received three doses of OPV [16]. The major drawbacks of OPV are that Sabin strains cause vaccine-associated paralytic polio (VAPP) in rare occasions (about 4.7 cases per million births) [17], and that prolonged transmission of Sabin viruses in areas with poor vaccine coverage facilitates reversion of attenuating mutations resulting in emergence of cVDPVs, which have re-acquired the neurovirulence and transmissibility of wild strains [18,19].

Trivalent OPV has been the vaccine of choice for global polio eradication since the initiative was launched in 1988 when the three serotypes were circulating and cases were occurring in more than 125 countries. Because Sabin type 2 is more immunogenic, susceptible individuals usually respond to type 2 after the first dose, and may need several doses to seroconvert (become seropositive) to types 1 and 3. Following 3 doses of trivalent OPV ≥ 95% infants were found to seroconvert and develop long lasting immunity to all three poliovirus serotypes in temperate countries [20]. However, OPV is less immunogenic in tropical countries, probably because of higher prevalence of factors that interfere with vaccine take such as malnutrition, diarrhea or infection by other enteroviruses [20–22]. Three doses of trivalent OPV were found to seroconvert a weighted average of 73% (range, 36%–99%) for type 1, 90% (71% to 100%) for type 2, and 70% (range, 40%–99%) for type 3 [21].

To overcome the inferior immunogenicity of trivalent OPV in tropical countries the GPEI used several strategies: a) reduce the relative content of type 2 with a formulation containing a 10:1:6 ratio for types 1, 2 and 3 [23]; b) deliver multiple doses of OPV through routine immunization (RI) and immunization campaigns conducted nationally and sub-nationally [24]; and c) in the mid-to late-2000s, re-introduce monovalent OPV types 1, 2, and 3, and bivalent (type 1 and 3) OPV. Clinical trials during 2000–2015 have confirmed the higher per-dose immunogenicity of monovalent OPVs compared to trivalent OPV in several tropical countries [25–29]. Seroconversion after 2 doses administered between birth and 3 months of age were 90% for monovalent OPV1, 90–97% for monovalent OPV2, and 74–84% for monovalent OPV3; after 3 doses of monovalent OPV1 seroconversion was greater than 95% [26]. Two or 3 doses of bivalent OPV elicited non-inferior responses to monovalent OPV1 or OPV3 [25–28].

Use of monovalent OPV1 in campaigns during 2005–2008 in poliovirus reservoirs in Northern India and Northern Nigeria significantly decreased transmission of type 1 wild poliovirus, although this strategy resulted in the resurgence of wild type 3 during 2009 [30]. The addition of bivalent OPV to campaigns since 2009 together with implementation of strategies to improve coverage, were crucial to achieve high levels of population immunity to both serotypes and eliminate the final chains of endemic wild poliovirus from both countries [31,32]. But replacement of trivalent with bivalent OPV in campaigns in countries with poor routine immunization, also contributed to the emergence of type 2 cVDPV before the global switch to bivalent OPV in 2016 [18]. The widespread transmission of type 2 cVDPV outbreaks observed during 2019–20 is a consequence of the reduction in mucosal immunity to type 2 caused by the global switch to bivalent OPV, and the inadequate quality of campaigns delivering type 2 OPV conducted before and after the switch. Lapses in poliovirus surveillance also allowed some cVDPV outbreaks to remain undetected at the time of the switch [19].

IPV is administered via the intramuscular, subcutaneous, or intradermal route, and contains chemically inactivated poliovirus types 1, 2 and 3. Wild poliovirus strains are included in Salk IPV and Sabin strains in Sabin IPV. IPV induces high levels of neutralizing antibodies in serum and elicits oropharyngeal mucosal immunity but has minimal effect on intestinal immunity in naïve individuals [10–12]. Vaccination schedules with IPV alone protect recipients from paralysis and reduce transmission of poliovirus through the oral-oral route (nasopharyngeal shedding), but are ineffective at stopping poliovirus transmission through the fecal-oral route, which plays a major role among young children and in settings with suboptimal sanitation [5].

IPV immunogenicity is affected by antigen content, presence of adjuvants, age at first dose, and interval between doses [20,33]. In clinical trials, a single dose of Salk IPV (containing 40:8:32 D Unit/per dose of types 1, 2 and 3) provided between 6 weeks and 4 months of age induced seroconversion in 19–57% infants for type 1, 26–80% for type 2, and 32–45% for type 3 [20,33]. Lower seroconversion rates were observed with IPV administration before 3 months of age because of interference of maternal antibodies and incomplete maturity of the immune system, although most of the seronegative infants were primed (i.e., exhibited an early immune response upon a challenge with a new vaccine dose) [34–36]. Two doses elicited seroconversion to all three serotypes in greater than 95% of infants and high antibody titers, when the first dose was given after 2 months of age and the second dose after a 4–6 month interval [36,37]. Primary vaccination with 3–4 IPV doses seroconverted 100% of children and elicited high titers at the end of the series [20,38,39].

Primary immunization with one or two doses of fractional IPV injected via intradermal route (0.1 mL or 1/5th dose) induced lower seroconversion rates and antibody titers than full-dose IPV, but differences were less significant with schedules that included three doses, or delay the first dose and include an interval between doses above 8 weeks [20,37,38]. Two fractional IPV doses administered at 3.5–4 months and at 8–9 months of age may induce greater than 95% seroconversion to the three serotypes [37]. The quality of the intradermal injection, which depends on the vaccinator technique and device used for injection, may also affect the immunogenicity of fractional IPV [40]. Loss of vaccine fluid on the skin and a small bleb size (below 5 mm) have been associated with inferior immune responses [40].

Use of both vaccines can offset the limitations in safety and efficacy of each one, although at a higher cost per child vaccinated. Primary immunization series with combined or sequential OPV-IPV schedules may achieve higher seroconversion rates and antibody titers for all serotypes than series with trivalent OPV alone, depending on the time of IPV administration, the type and total number of OPV doses received, and the efficacy of OPV in that setting. As shown on Table 1, clinical trials found that addition of 1 or 3 doses of IPV to trivalent OPV with the WHO schedule of 6, 10 and 14 weeks of age (with or without a birth dose) increased seroconversion rates and final titers for types 1 and 3 in most settings, but the effect was often not clinically or statistically significant for type 2 because tOPV also achieved high seroconversion rates and titers [39,41–45]. Provision of an IPV dose after several doses of bivalent OPV provided protection for type 2 poliovirus in 49–80% of naïve infants (Table 1), and boosted titers for serotypes 1 and 3 without significantly modifying the high seroconversion rates achieved with bivalent OPV alone [39,42,43].

A supplemental dose with IPV plus OPV may have more impact on humoral immunity than OPV alone in populations that are under-vaccinated because of suboptimal coverage or use of monovalent OPVs, and in populations with low OPV efficacy. As shown on Table 2, a supplemental dose of IPV was more effective than trivalent and monovalent OPV3 in closing remaining humoral immunity gaps to serotypes 1 and 3 after 2–5 doses of trivalent OPV in Cote d’Ivoire, Gambia, Oman and Cuba [46–49]. In areas with persistent wild poliovirus circulation in India and Pakistan, IPV closed immunity gaps to type 2 and 3 caused by intensive use of bivalent and monovalent OPV1 in campaigns [22,50]. In Pakistan, an IPV booster was more immunogenic than bivalent OPV among malnourished infants 9–12 months of age, who already had lower seroprevalence for the three serotypes at baseline compared with normally nourished infants [22]. Also in Pakistan, addition of IPV to trivalent OPV as a booster dose was more immunogenic than monovalent OPV2 among children seronegative after four doses of different polio vaccines (trivalent OPV, bivalent OPV and/or IPV) [51]. Seroconversion to type 2 poliovirus after the booster at 22 weeks was 29% (38/133) after trivalent OPV, 61% (31/51) after monovalent OPV2 and 100% (47/47) after trivalent OPV plus IPV. Compared with a full IPV dose, a supplemental dose with fractional IPV induces lower or similar changes in seroconversion and boosting of titers, depending on the device used, but lower final antibody titers (Table 2) [49,52].

A supplemental IPV dose may also boost intestinal local immunity better than an OPV dose among children with intestinal immunity waning after vaccination, who live in poliovirus transmission reservoirs with decreased OPV efficacy [13]. Administration of an IPV booster to 10-year-old children in northern India, who had received multiple OPV doses up to 5 years of age, reduced shedding of type 1 poliovirus one week after a bivalent OPV challenge by 82%, whereas a bivalent OPV booster reduced shedding by 51%, compared with children who had received no booster (poliovirus detected in 7%, 19%, and 39% of the IPV, OPV, and control groups, respectively). However, neither IPV nor OPV had effect on type 1 poliovirus shedding after the bivalent OPV challenge among infants 6–11 months of age who had recently received several doses of OPV (poliovirus detected in 6%, 12%, and 4% of IPV, OPV, and control groups respectively) [13]. Among Pakistani infants 9–12 months of age who were given a booster with bivalent OPV alone or bivalent OPV plus IPV after having received an average of 10 OPV doses, a very small proportion shed poliovirus 7 days after a challenge with bivalent OPV, without significant differences between study arms (type 1 poliovirus detected in 2–5% and type 3 poliovirus in 3–7% of infants) [22]. Among children aged 1–4 years from Vellore, India, who had received a median of 10 OPV doses, an IPV booster reduced slightly shedding for both type 1 and 3 serotypes after a bivalent OPV challenge compared with control children who received no booster. Serotype 1 poliovirus was detected in 12% of IPV-boosted children versus 19% of control children (P < 0.05); serotype 3 poliovirus was detected in 8% versus 26% (p < 0.0001) [14].

OPV has been the vaccine of choice for preventive and outbreak response campaigns. OPV is ideal for campaign delivery because it is inexpensive, easy to administer, and well accepted by children. National Immunization Days and “mop-up” campaigns can be performed through house-to-house visits in a short time by deploying large numbers of volunteers with minimum training [24]. House-to-house vaccination has been demonstrated to be crucial to achieve high coverage in countries of Africa and Asia [55]. The ability of OPV to spread to un-vaccinated individuals may increase campaign impact on population immunity [16], but is also responsible for new cVDPV emergences when RI and campaign coverage is very low [18]. The risk of re-seeding new type 2 VDPVs with OPV campaigns continues to increase since the withdrawal of trivalent OPV in 2016, as new birth cohorts without intestinal immunity to type 2 poliovirus and waning immunity among those born before 2016 increase the pool of individuals participating in transmission of OPV strains [19].

A review of data reported through the WHO’s Polio Information System (POLIS) as of March 2021, found that between December 2013 and October 2020, 78 campaigns delivered either full (n = 73) or fractional (n = 5) dose IPV (Table 3). IPV campaigns were implemented initially in response to some type 1 wild poliovirus circulation, and since 2017, in response to some cVPDV2 outbreaks as well. Because of limitations of the available data, some IPV campaigns may have been for catch-up vaccination in birth cohorts that missed IPV in RI, and not for outbreak response. Pakistan, Afghanistan and Nigeria conducted 76% of the IPV-containing rounds (Fig. 1); other countries with rounds included Angola (1), Cameroon (1), China (4), Ghana (1), India (1), Kenya (1), Malaysia (1), Somalia (1), and Syria (6) (Table 3). IPV was co-administered with OPV (bivalent, trivalent, or type 2 monovalent) in 31 campaigns (40%) and given alone in 47 campaigns (60%). The target age for IPV was limited to above 6 weeks of age in all campaigns, with 46 campaigns (59%) including children below 5 years, 26 campaigns (33%) including children below 2 or 3 years, and 6 (8%) covering other age groups.

In general, IPV campaigns were small (Table 3). The GPEI guidelines recommend that an outbreak response for a cVDPV2 outbreak includes a small rapid response campaign (round zero) targeting about 100,000 to 500,000 children; followed by two larger scale rounds targeting at least 1–4 million individuals and additional small mop-up rounds in areas with poor performance [54]. The median target population of IPV campaigns was below 300,000 children and covered <15 districts. Only 13 campaigns targeted ≥ 1 million children and four campaigns ≥ 5 million. Most campaigns (55, 72% of 76 with dates of implementation) were conducted within 7 days, and 17 (22%) within 8–17 days. Three campaigns covering 0.2–6.5 million children required 23 to 29 days; and the national campaign in Angola (1.2 million) was completed in three phases over 99 days (Table 3).

Use of an injectable vaccine for polio campaigns require administration by healthcare workers instead of volunteers and delivery through fixed-posts instead of house-to-house visits. Also, more resources and complex logistics are required to handle supervision and management of safe injection practices and adverse events, transport of vaccines in cold chain, and sharps disposal. Reports from campaigns conducted in Kenya (2013), India (2016), Pakistan (2015–19), and Syria (2017–18) provide operational lessons learned with delivery of IPV plus OPV in campaigns [56–61].

Reports from the first IPV campaign conducted in Kenya (2013) demonstrated that it is feasible to conduct a campaign delivering two different polio vaccines, and that the population accepts administration of both vaccines without major concern about potential side effects [61].

Scarcity of healthcare workers in many countries with polio outbreaks is the major factor limiting the size and duration of campaigns with injectable vaccines. Deployment of skilled vaccinators from non-affected areas to the outbreak response areas allowed quick implementation of IPV-OPV rounds in Kenya, India, and Syria [56,58,61]. Quick successful implementation was also attributed to prior campaign experience, either because of frequent planned campaigns (India, Pakistan) or because of recent outbreak rounds in the same area (i.e., Syria and Kenya used IPV after several OPV rounds)[56,58,59,61].

The cost of adding IPV to OPV for campaigns was estimated in two outbreak response rounds conducted in 2013 in Kenya [61]. The total cost per child vaccinated reached $3.27 for the December round delivering IPV plus trivalent OPV compared with $0.50 per child during the November round delivering trivalent OPV alone. The vaccine cost alone was $2.09 for an IPV dose and $0.14 for an OPV dose.

Although fractional IPV could reduce the cost per vaccine dose, stretch a limited supply, and reduce the risk of adverse effects of poor intramuscular injection technique, it has been used rarely in campaigns during 2016–2020. Reports of campaigns in India and Pakistan highlighted the importance of selecting experienced vaccinators and providing additional training in achieving intradermal injections of good quality [56,59]. Use of jet injectors and syringe adaptors that guide the needle to its appropriate position in the skin, were well accepted by vaccinators, and jet injectors increased vaccination acceptance among caregivers. However, ease of use, quality of intradermal injections observed, and (when studied), immunogenicity, varied with devices [52,57,60].

Information on the quality of IPV-containing campaigns reported on the POLIS database is limited. Only 27% of the 78 campaigns reported had administrative coverage data, and only 7 (9%) had information on missing children using lot quality assurance sampling (LQAS) surveys. Median administrative coverage was 99–100% in Pakistan and Afghanistan (range 65–111%), and only 68% (range 36–100%) in Nigeria (Table 3). Results from available LQAS surveys ranged from 0 to 100% of lots passing the 90% threshold for children with missed vaccination (results not presented), and, for campaigns that distributed IPV plus OPV, it was not possible to know which children received OPV only or OPV plus IPV.

Evaluations of IPV campaigns in Pakistan and Kenya described the challenges for vaccinators to reach target populations with fixed posts in rural areas. In Pakistan, a campaign delivering fractional IPV in four districts in Sindh had an estimated coverage of 90% (95% confidence interval = 88–92%), but social mobilization and coverage in outreach areas was likely lower than in urban areas [59]. In Kenya, a vaccine coverage survey found that IPV plus trivalent OPV reached more than 90% in refugee camps and surrounding communities, but vaccination sites had to become “temporary fixed posts” to reach vaccination targets, as caregivers requested vaccinators to bring vaccine closer to their homes. In a convenience sample of nomadic families settled near the villages and camps participating in the survey, <40% of children had received vaccine during the IPV round and during a previous round with trivalent OPV; the same children who had missed the trivalent OPV round were also missed by the IPV round [61].

IPV use in the United States in the 1950’s decreased the number of polio cases beyond expectations based upon the proportion of children vaccinated. The “herd effect” of IPV has been attributed to a direct effect on mucosal immunity in the nasopharynx and indirect effect in boosting intestinal immunity among individuals exposed to natural infection [62].

High coverage with IPV-only schedules has successfully prevented community transmission following importations for 20–50 years in countries with good sanitation with very few exceptions. Outbreaks in the Netherlands in 1978 and 1992 [63], and in Israel in 1988 and 2013 [64] caused between zero (Israel 2013) and 110 paralytic cases (Netherlands 1978). Most of the paralytic cases occurred among individuals who had not received IPV, and asymptomatic transmission among IPV-vaccinated individuals was documented. Administration of OPV, with or without IPV, was required to control the outbreaks [63,64]. Campaigns with IPV alone were conducted in Telangana, India in 2016, and in China in 2019 and 2020, following detection of VDPV2 in environmental samples (India and China in 2019) or in AFP case and contacts (in 2020 in another area in China). No new VDPV was detected after the campaigns [56,65].

Interpretation of the impact of adding IPV to OPV in campaigns on prevention of paralytic cases and slowing transmission often depends on assumptions used for statistical analysis or modeling, type of poliovirus, and period of time analyzed [2]. Shirref, Grassly, et al. looked at incidence of paralytic cases and prevalence of positive environmental samples 90 days before and after a campaign. For campaigns conducted during 2014–2016 in Nigeria, there was a slight reduction in type 2 cVDPV cases with IPV plus trivalent OPV whereas trivalent OPV alone had no effect. The incidence risk ratio (IRR) for cases detected 90 days before and after was 0.17 (95% CI 0.04–0.78, p = 0.02) for combined rounds and 0.59 (95% CI 0.18–1.97) for trivalent OPV rounds [2]. In Pakistan, campaigns that used IPV alongside any OPV during 2014–2017 had a significant impact on the detection of environmental samples of type 1 wild poliovirus (prevalence ratio 0.63, 90% bootstrap CI 0.47–0.81), but not on incidence of poliomyelitis cases (IRR 0.62, 90% bootstrap CI 0.23–1.14) [66]. Using a deterministic model of poliovirus transmission, Duintjer Tebbens et al. found that adding IPV to the first or second round in Nigeria would result in a small reduction in polio cases (4–6%) at a much higher cost, compared with mOPV2-only rounds. The incremental benefit of IPV remained limited after 2017 [67,68].