In 1988, the World Health Assembly launched the Global Polio Eradication Initiative. Live attenuated oral poliovirus vaccine (OPV) was selected by WHO for routine use and national immunization days. OPV might require several doses to induce immunity, but then it provides long-term protection against paralytic disease through durable humoral and mucosal immunity [1]. The use of OPV, however, is associated with some rare adverse events, including the occurrence of vaccine-associated paralytic poliomyelitis (VAPP) [2] and the emergence of vaccine-derived polioviruses (VDPVs) [3]. VAPP occurs at very low rates (∼1 case per 600,000 first-dose recipients) [4], and it is estimated that approximately 498 (range 255–1018) VAPP cases occur worldwide each year [2]. VAPP risk is 3000–7000 fold higher in persons with primary immunodeficiencies, notably agamma- and hypogammaglobulinemia, exposed to OPV [5]. In a small proportion of immunodeficient patients, Sabin strains can replicate in the intestine for several months or years [6–8], in contrast to the short (typically 3–4 weeks) period of replication in immunocompetent persons following their first OPV dose [9].

Because the poliovirus genome typically evolves at an average rate of 1 % per year, prolonged replication or circulation of viruses derived from OPV is recognized by the accumulation of genomic nucleotide substitutions [6]. VDPVs are operationally defined as vaccine-related isolates having >1 % nucleotide divergence from the corresponding type 3 or type 1 Sabin strain in the genome coding for the major capsid surface protein, VP1 (or >0.6 % in case of type 2 polioviruses). By contrast, OPV-like isolates have limited divergence from their parental OPV strains and are ubiquitous wherever OPV is used. The occurrence of immunodeficiency-related-VDPV (iVDPV) is typically consistent with at least 1 year of poliovirus replication since the administration of the first OPV dose. Nevertheless, evolution rates as fast as 2 %/year have been observed in the early phases of patients with iVDPV [10]. Since the introduction of OPV in 1961, more than 70 subjects with primary immunodeficiencies have been found excreting iVDPVs worldwide; the majority of these immunodeficiencies were detected after the onset of VAPP [1]. Type 2 iVDPVs are the most prevalent (64 %), followed by type 1 (21 %) and type 3 (15 %) [1]. By contrast, VAPP in immunocompetent OPV recipients and household contacts is most frequently associated with type 3 (71 %) followed by type 2 (26 %) poliovirus [5, 6].

Albania is considered polio free since 1997. The last polio outbreak of 1996 was caused by an imported wild type 1 poliovirus. Vaccination with OPV was maintained until March 2014, when a sequential schedule using inactivated polio vaccine (IPV) followed by OPV was adopted. The Acute Flaccid Paralysis (AFP) surveillance, established in 1997, has always met the WHO standard performance: two VAPP cases were detected until now, both affecting immunocompetent children after the first OPV dose [4].

Italy is considered polio-free since 1982, when the last indigenous wild polio cases occurred. Three imported wild polio cases from Iran, India and Libya (the last in 1988) were reported afterwards [11]. Since then, only rare VAPP cases (two of which were household contacts) have occurred in Italy [12, 13] before the adoption of Salk vaccine (IPV) in 2002 [14, 15]. The last VAPP case in Italy occurred in the year 2000 in a one-year old immunodeficient child who received three doses of OPV [16] and was detected through the AFP surveillance in force in Italy since 1997 according to the WHO guidelines [17].

We report the case of an Albanian infant affected by congenital agammaglobulinemia, who developed VAPP at the age of 5 months after receiving two doses of OPV in his country, and was then referred to Italy for further treatment. We isolated from sequential stool samples Sabin-like type 3 polioviruses, which accumulated 1 % of mutations in the VP1 viral genome in 7 months. We herein describe the genetic evolution of the poliovirus during the entire period of excretion, and discuss the implications of chronic virus excretion for the global polio eradication strategy.

This is the first case of VAPP in immunodeficient child detected in Albania through the AFP Surveillance system and the first iVDPV case investigated in Italy since the occurrence of the indigenous iVDPV case reported by Buttinelli and colleagues in 2003 [16].

Sabin OPV vaccine has been adopted by the WHO for the Global Polio Eradication Initiative (GPEI) and decades of experience have shown that it is safe and effective in preventing poliomyelitis. However, it is known that attenuated viruses may revert to neurovirulent phenotypes upon replication in the human gut, rarely causing VAPP in immunocompetent individuals [2]. Immunodeficient persons vaccinated with OPV have a 3000–7000 time higher risk of developing VAPP, especially when affected by agammaglobulinemia or hypogammaglobulinemia [32]. Moreover, in immunodeficient subjects attenuated OPV viruses replicate in the gut and can persist for long periods [15, 33–35]. Excreted viruses rapidly accumulate mutations and genetic rearrangements, increasing the chance of reversion to neurovirulence and transmissibility. For these reasons, long-term excretors of poliovirus and immunodeficient patients are an additional concern for the WHO, as they could constitute a polio reservoir in the post-eradication era, and infect both unvaccinated and immunocompromised persons [8, 36, 37].

In this work, we report a detailed molecular analysis of six sequential type 3 polioviruses isolated from an Albanian child affected by congenital agammaglobulinemia who developed flaccid paralysis 1 month after receiving the second dose of OPV in his country. The ITD of the strains confirmed the Sabin-like characteristic of all PV3 isolates.

In middle- and high-income countries, VAPP occurs more frequently after receiving the first dose of OPV and can affect unvaccinated contacts. Poliovirus type 2 (Sabin or iVDPV) is the prevalent isolated serotype in immunodeficient individuals, while type 3 only counts for 15–32 % of isolations [1, 2, 38]. Our patient developed VAPP after receiving the second dose of OPV, at the age of 5 months. Interestingly, he had been breastfed during the first 4.5 months of life. Enteroviruses are cleared from the host mainly by antibody-mediated mechanisms. Secretory IgA can neutralize viruses by recognizing specific capsid antigens, reducing infectivity and promoting viral clearance [39, 40]. The patient’s mother had been vaccinated with OPV, which stimulates durable mucosal immunity through IgA secretion. We speculate that passive transfer of maternal IgA via breast-feeding could have played a role in limiting virus proliferation in the gut of our patient after the first dose of OPV. It is also known that the presence of anti-polio antibodies in the blood prevents the central nervous system (CNS) invasion [41]. In our case, we can assume that the physiological postnatal decreasing of maternal antibodies, associated with breast-feeding interruption in the absence of self secretory-IgA, led to a rapid viral replication in the gut, and subsequently invasion of the CNS, causing VAPP at the age of 5 months.

The lack of isolation of poliovirus type 1 or 2 Sabin-like strains might be related to reduced replication of these serotypes in the human gut. We can nonetheless hypothesize a greater passive protection against these two serotypes conferred by maternal antibodies.

Sequencing of the 5’NCR and VP3 regions evidenced a U > C mutation at nt 472 in the domain V of the IRES and at the amino acid residue 91 (Phe > Ser): both correlate in PV3 with reversion of the attenuated phenotype, confirming their importance for neurovirulence and possibly for adaptation and replication in the gut [24, 25, 40]. The additional mutation in 5’NCR nt 292 (G > A) in the domain IV of the IRES was also found in other PV3 isolated from iVDPV and VAPP cases [24, 25] but its role in conferring neurovirulence has not been demonstrated. Mutations in VP1 quickly increased over time reaching 1 % genome variation in the samples (ITA C1 and C2) collected in September 2014. The strains were still Sabin-like, but the percentage of accumulated nucleotide mutations was very close to the value >1 % that defines an iVDPV.

Viruses replicating in immunodeficient patients tend to select characteristic mutations at different times during the infection, suggesting that specific selection pressures may operate in the gut over time. The amino acid substitutions at position 105 (Met > Thr) and 54 (Ala > Val) of VP1 have already been described by Martin and colleagues in a PV3 strain isolated from an immunodeficient patient [24]. Amino acid VP1 105 is located at the north rim of the canyon in the hydrophobic pocket, and plays a role for the uncoating of the virus. The substitution 54 (Ala > Val) in VP1 acts as a suppressor of the temperature sensitivity and attenuation phenotype (the capacity of growth at elevated temperatures is indeed typical of wild poliovirus strains) [42]. Other mutations that are known to be involved in growth adaptation, viral persistence and increased neurovirulence, such as sequence variation at nucleotide 2493 in VP1[43, 44], were not found in our patient.

Finally, sequencing of the 3D polymerase revealed the presence of an intertypic recombination with type 2 Sabin (recombinant PV3/PV2 Sabin) in all isolates. This is an interesting finding, even though similar recombinations have already been described in healthy vaccinees, iVDPV and VAPP cases [33, 45].

The global prevalence of immunodeficient subjects with chronic polio infection is unknown and if asymptomatic they remain undetected. This poses a risk in the context of the GPEI endgame. Therefore, the detection of chronic iVDPV excretors in all countries and the development of antivirals to eradicate chronic infections is a WHO priority. Sequence properties of circulating VDPV (cVDPV) strains are distinguishable from iVDPV strains excreted by chronic infected persons. The comparison between the mutational patterns found in iVDPV and cVDPV can predict the origin of the anonimous VDPV (aVDPV) strains isolated during the environmental surveillance. This remains crucial for epidemiological interpretation because it may alert for the presence of a chronic iVDPV excretor in the community [3]. Environmental surveillance will be implemented by the GPEI in low income countries since it proved to be very sensitive in detecting cVDPV and iVDPV strains [46]. In this sense the mutations found in our study can add knowledge to the definition of iVDPVs genetic pattern.

OPV plays a key role in the eradication of wild poliovirus and is still used in regions where wild poliovirus is not definitively eradicated or with high risk of polio outbreak. Occurrence of cVDPV outbreaks and iVDPV cases and the potential spread of highly neurovirulent poliovirus strains to the environment has recently led the WHO to state that OPV usage will globally be discontinued soon after the certification of global eradication [3, 6, 24]. The endgame plans for the GPEI actually include the synchronic replacement of the trivalent Sabin live-attenuated oral poliovaccine (tOPV) with the bivalent bOPV (PV1,3) as a first step to prevent the more frequent cVDPV type 2 outbreaks. The subsequent withdrawal of all OPV use and the maintenance of high immunization coverages with IPV will protect against imported wild polioviruses and prevent cVDPV outbreaks and new iVDPV infections [47]. As Albania did in 2014, most low income countries are incorporating at least 1 dose of IPV into routine childhood immunization schedules. The production of less expensive inactivated Salk IPV vaccine based on the Sabin strains vaccines is encouraged by the WHO for environmental safety and a large-scale use [47]. Because many high-income countries have replaced OPV with IPV, the VAPP burden is currently concentrated in lower-income countries. The planned universal introduction of IPV will also substantially decrease the global VAPP cases [2]. However, it is important that all countries maintain high quality of AFP surveillance and improve the current laboratory typing methods to better differentiate between VDPVs and wild poliovirus strains.

This is the first case of poliomyelitis and long-term excretion from an immunodeficient patient detected in Albania through the Acute Flaccid Paralysis surveillance system. Due to the high levels of immigration across the Mediterranean sea, Italy, Albania and other Mediterranean countries remain at risk of importing wild poliovirus from endemic areas as well as Sabin and neurovirulent VDPVs from countries currently using OPV. Moreover, since IPV vaccine (adopted in Italy since 2002) may not elicit a consistent mucosal immunity, silent transmission of neurovirulent poliovirus might occur through IPV-immunized individuals, favouring possible infection of unvaccinated subjects or children receiving delayed vaccination [22, 48, 49]. For these reasons an Active AFP surveillance is in force at national level as well as an environmental surveillance in seven large Italian cities with high immigration rate [45]. Immunodeficient patients have been monitored for several years in Italy and no poliovirus long-term excretors have been detected [50]. In Albania this screening has not been performed, but in the future environmental surveillance and monitoring of immunodeficient persons will be implemented.

No antivirals are currently available to interrupt poliovirus excretion in immunodeficient subjects. The capsid-binding drug pleconaril that prevents poliovirus cell-entry has been tried in a reported case by our group with successful interruption of viral excretion when combined with IVIG treatment [16]. Developing effective drugs/treatments is a WHO priority and research is ongoing in this field.

AA (or aa), amino acid; AFP, acute flaccid paralysis; CNS, central nervous system; CRP, C-reactive protein; cVDPV, circulating VDPV; ENG/EMG, electroneurography and electromyography; GPEI, global polio eradication initiative; IPV, inactivated polio vaccine (Salk); ITD, intratypic differentiation; iVDPV, immunodeficiency-associated VDPV; IVIG, polyclonal intravenous immunoglobulins; L20B, recombinant murine cell line; NRL, National Reference Laboratory for poliovirus; NT (or nt), nucleotide; OPV, oral polio vaccine (Sabin); PV, poliovirus; PV1, poliovirus type 1; PV2, poliovirus type 2; PV3, poliovirus type 3; RD, rhabdomyosarcoma cell line; RRL, Regional Reference Laboratory for poliovirus; rRT-PCR, real-time reverse transcription polymerase chain reaction; VAPP, vaccine-associated paralytic poliomyelitis; VDPV, vaccine-derived poliovirus