In this study, we report clinical characteristics of 14 VAPP patients and virologic properties of their iVDPV isolates. All patients were identified in Iran during 1995–2014. We sought to determine relationships between clinical features of underlying immunodeficiency disorders, VDPV manifestations, and patient outcomes.

Patients in Iran with acute flaccid paralysis (AFP) were tested for shedding of polioviruses in stool specimens during national AFP surveillance up to the end of 2014. Stool specimens were collected at the earliest convenience within 48 hours after onset of paralysis. Fecal shedding of vaccine virus strains was simultaneously investigated in close contacts of patients.

Specimens were processed at the Iranian National Polio Laboratory (Tehran, Iran). Virus isolation, serotype identification, and intratypic differentiation were performed by using the World Health Organization protocol for poliovirus detection in stool specimens (14). Screening for VDPV shedding was performed by using an ELISA, reverse transcription PCR, or real-time reverse transcription PCR. Genome sequencing of isolates was conducted at the Centers for Disease Control and Prevention (CDC; Atlanta, GA, USA). The VP1 divergence rate was defined as the value (percentage) of VP1 divergence divided by years of virus replication from the original Sabin strain. To calculate this rate, we considered the interval between first OPV administration and date of the most recent isolation of iVDPVs as the duration of virus replication.

Clinical and laboratory data were obtained from the national primary immunodeficiency registry and AFP surveillance database (15,16). Patients with undetermined immune status were analyzed to determine serum levels of immunoglobulins and absolute leukocyte and different lymphocyte subpopulation counts. Immunity disorders were defined according to the Expert Committee of the International Union of Immunological Societies (17). Agammaglobulinemia, hypogammaglobulinemia, and mu heavy chain disease were considered humoral immunodeficiencies. Combined immunodeficiency (CID) included severe combined immunodeficiency, major histocompatibility complex class 2 deficiency, and Nijmegen breakage syndrome.

We evaluated patients by monthly collection of stool specimens and analysis until patients cleared their infections or died of any cause. Patients were classified into separate subgroups according to types of immunodeficiency (humoral or combined) and whether virus shedding in stool specimens was cleared before death. Patients were considered as having cleared an iVDPV infection if 2 consecutive stool specimens were negative for iVDPVs.

Continuous variables are reported as means ± SD. We used the Shapiro-Wilk normality test to assess distribution of study parameters. We used the χ² test to analyze relationships between categorical parameters, the Mann-Whitney U test to compare duration of shedding and VP1 divergence rate between different subgroups, and an independent t-test to compare age of onset of VAPP and absolute VP1 divergence. p values <0.05 were considered statistically significant.

We analyzed 14 iVDPV patients identified in Iran during 1995−2014. Detailed characteristics of patients 1–7 have been described (9–13); among these patients, only patient 6 is alive, and this patient has residual paresis. Patient 7, who had AFP at 6 months of age, continued shedding iVDPV2 for 2 months after the most recent published data and then stopped shedding virus. He died at 34 months of age from pneumonia. Among the 7 newly identified case-patients, patient 12 died before he underwent specific immunologic investigations. This patient had a history of recurrent infections, and sequencing data showed quasispecies of VDPV2 from multiple lineages, which suggested immunodeficiency for this patient. This patient was identified as an iVDPV patient in CDC annual reports (18).

Serotype 2 iVDPV has been the most prevalent serotype detected in immunodeficient patients (64%), followed by serotypes 1 (21%) and 3 (15%) (18); our series showed similar findings. We analyzed 2 patients who were simultaneously shedding serotypes 1 and 2 iVDPVs. Concurrent infection with >1 VDPV serotype has been documented in 3 other patients from China and the United States (4,6,19). Paralysis in poliomyelitis is usually asymmetric (6,20). In contrast, most of our iVAPP patients had bilateral flaccid paresis. Further studies are required to differentiate manifestations of VDPV infection in primary immunodeficiency patients with poliomyelitis caused by wild-type poliovirus.

Patient 1 received only IPV; however, all his contacts had received OPV as routine national vaccination in Iran. Subsequently, he had VAPP and died of an unknown etiology. All other patients in our study had severe complications of live vaccines (BCG and OPV) given at birth. These complications were the earliest manifestations of primary immunodeficiencies in these patients, which further aggravated their course of disease. Use of screening programs to detect immunodeficiencies at birth could prevent administration of live vaccines to primary immunodeficiency patients and their contacts and minimize the burden of adverse complications.

We observed an association between type of primary immunodeficiency and frequency of clearing iVDPV infection; patients with humoral immunodeficiency cleared VDPV infection more frequently than CID patients. Patients with CID have more severe complications and shorter survival times than patients with humoral immunodeficiency (16,21). This finding could limit the ability of CID patients to clear infections by use of proper immunoglobulin therapy before death. Although this finding might be an explanation for lower rate of infection clearance in CID patients, comparable durations of virus shedding between 2 subgroups of primary immunodeficiency patients do not support this hypothesis and imply a probable contribution of other mechanisms to this observation.

Another explanation for the relationship observed in our study between type of primary immunodeficiency and rate of clearing iVDPV infection might be differences in inherent capacity of these immunodeficiency disorders to respond to polioviruses or other enterovirus infections. To our knowledge, enteral mucosal immunity against enteroviruses is a prominently antibody-mediated mechanism, and patients with major B-cell dysfunction are at increased risk for poliomyelitis (6,22). Nevertheless, our observations suggest other major contributions of cellular immunity to clearance of poliovirus infections. The index case-patient for an outbreak of poliovirus infection in Minnesota, USA, in 2005 had been given a diagnosis of severe combined immunodeficiency. She continued to shed iVDPVs while receiving immunoglobulin therapy for several months and finally stopped shedding virus after a second bone marrow transplant (23).

Similarly, regular intravenous immunoglobulin (IVIg) administration could not protect patients 9 and 10 in our study from VAPP, and neither patient cleared the infection. Shedding of iVDPV by patient 7 is also notable because he is one of the few case-patients with CID who eliminated the infection before death (23,24). He had the longest period of iVDPV shedding among our patients before clearing the infection. His duration of virus shedding was several times longer than shedding durations in each of our patients with humoral immunodeficiencies. These observations appear to support the hypothesis of a probable role for cellular immunity in clearing enteroviral infections. Our patients were generally treated with monthly administration of IVIg. However, different IVIg preparations, doses, and settings in immunoglobulin replacement therapy might contribute to efficacy of treatment. Thereafter, to assess the independent role of immune status in clearance of infection, a comprehensive profile of potential confounding factors and therapeutic conditions, which might contribute to capability of virus clearance, is needed. This profile is missing in our study. Additional studies with comprehensive and appropriate adjustments for these factors are required to assess the immunologic response against enteroviral infections in patients with various types of primary immunodeficiencies.

A search for VDPV infection in contacts of our patients yielded negative results, and only Sabin-like polioviruses were isolated from few contacts, which indicates high vaccination coverage in Iran. Spread of iVDPVs to unimmunized contacts in the United States (6) and Morocco (25) showed potential risk for iVDPV transmission in populations with low immunity, which can lead to an outbreak. Maintaining high vaccination coverage reduces this risk.

The overall rate of accumulative nucleotide substitution in the genome of polioviruses is believed to be ≈1%/year (6,26). Some investigators consider a range of 1%–2%/year as the overall rate of nucleotide variation in polioviruses (27,28). Most of our VDPV isolates showed a divergence rate of 1%–2%/year. However, we observed unexpectedly high divergence rates beyond this range in some of our isolates. We propose 3 possible explanations for this difference.

First, for isolates with high divergence rates, the patient could have been infected with an already divergent virus, rather than the original vaccine strain. Thus, the virus had already replicated in a previous host before transmission to the patient. However, lack of any report of circulating VDPV from Iran does not support this explanation. High vaccination coverage and herd immunity against polioviruses in Iran would prevent any circulation of VDPVs. Negative results for VDPV screening in contacts of patients also does not support this explanation.

Second, it has been frequently observed that shortly after OPV administration, the initial rate of capsid region gene evolution is high in Sabin strains, which leads to VDPV emergence (28,29). A high proportion of these changes are nonsynonymous substitutions in the genomic region, which codes for structural proteins. However, the synonymous mutations, mostly in nonstructural regions, occur at the expected rates. Changes in specific antigenic sites might increase additional nonsynonymous substitutions (29). These changes could be caused by effects of direct selection against attenuating mutations in the early stages of VDPV emergence. Selected mutations might carry unselected hitchhiker mutations in these early stages, which would contribute further to a higher initial evolution rate. After these initial selection events, VDPVs appear to evolve at rates closer to the unselected steady-state rate of ≈1%–2%/year. This alternative view is supported by only limited data from more prolonged infections because dates of initiation of most VDPV infections can only be estimated from clinical records but is consistent with available evidence (27).

Third, another explanation for higher rates of divergence in some of our isolates could be presence of mixed lineages or quasispecies of VDPVs in stool specimens from corresponding patients. This explanation is consistent with previous evidence, which indicated shedding of mixed-lineage VDPVs by patients with primary immunodeficiency (27,28). Earlier detection of poliovirus shedding in our patients, possibly after exposure to OPV strains, could have provided precise information about timeline of VP1 divergence in our isolates. This detection could be another advantage of early screening for primary immunodeficiencies and VDPV shedding.

Eradication of wild-type poliovirus serotypes 2 and 3 indicates that eradication of all 3 serotypes can be expected in the near future. However, emergence of VDPVs still remains a threat to global eradication of poliomyelitis (2). A review of reports published by CDC showed that ≈100 primary immunodeficiency patients with VDPV shedding have been documented (1,4,18,19,30–32). By the end of 2014, Iran had reported the highest incidence of iVDPV infections (14 case-patients with AFP), followed by the United States (11 case-patients), and the United Kingdom, Egypt, and China (each with 6 case-patients). Similar to the United Kingdom and other industrialized countries, which reported few VAPP cases after switching to vaccination with IPV, only 3 cases occurred in the Unites States after OPV cessation in 2000 (3,23,33). In contrast, the number of iVDPV cases reported from developing and middle-income countries has increased in recent years because of increased extended VDPV surveillance in these countries that used OPV (18,34).

More than 20 outbreaks of circulating VDPV infection have been documented, mostly in countries with low OPV coverage (9). These outbreaks occurred while no cases of infection with iVDPV were reported from some of these countries, such as Nigeria, Democratic Republic of the Congo, Somalia, and Niger, where VDVPs have circulated for a long time (4,9). To explain this observation, one should consider that patients with primary immunodeficiency do not have a chance for long survival in low-income communities because of lack of proper supportive therapies (35,36). Thus, even if they are infected with OPV strains, they may die of severe consequences of primary immunodeficiency before emergence of VDPV through prolonged enteral replication.

Environmental isolation of highly divergent VDPVs without an obvious source of shedding has been reported in different countries (37–39). In Finland, although only IPV vaccine has been administered since 1985, neurovirulent and highly divergent VDPVs have been recurrently detected in sewage since 2008. Epidemiologic data and virologic characteristics suggest that these VDPVs might have originated from chronically infected patients with immunodeficiencies; however, the source of shedding remains to be identified (38). Such unidentified chronic shedding of VDPV might reintroduce neurovirulent viruses into the population and initiate outbreaks in the posteradication era if population immunity to poliovirus is not maintained.

Figure

Figure. Poliomyelitis cases in Iran, 1995–2014. WPV, wild-type polioviruses; VDPV, vaccine-derived polioviruses.

High coverage of OPV vaccination disrupted wild-type poliovirus transmission in Iran in the late 20th century, and the last case of wild-type poliomyelitis was detected in 2000 through sensitive AFP surveillance (Figure). Iran has borders with 2 polio-endemic countries (Afghanistan and Pakistan), which necessitates maintaining high levels of serum and mucosal immunity in the population. Although no wild-type poliomyelitis case has been reported from Iran for many years, vaccine-associated paralysis occurs at a rate of a few cases every year and increases illnesses in immunodeficient patients.

Routine administration of trivalent OPV would inevitably lead to emergence of VDPVs and new VAPP cases. Conversely, vaccination with IPV might not stop virus transmission because of inefficient induction of mucosal immunity (40). However, vaccination with IPV might be too expensive for some low-income countries. Accordingly, the switch to a combination of bivalent OPV (serotypes 1 and 3) and IPV vaccination was implemented worldwide in April 2016 (2). We believe that immediate achievement of maximum immunization coverage is another vital point to be addressed in this strategy. Moreover, attempts to develop safer and more efficient poliovirus vaccines and polio antiviral drugs are warranted.

This study highlights the need for efficient iVDPV surveillance and the crucial role of patient registry. Vaccine-derived polioviruses not only endanger countries using OPV but also remain a threat to industrialized countries and the global polio eradication program. Regarding the association of immunodeficiency with this public health issue, we advocate establishment of sensitive clinical and laboratory surveillance to screen primary immunodeficiency patients for shedding of polioviruses.

Eradication might not be accomplished without robust registry and tracking capacities in every country to closely monitor the persistence of polioviruses in the environment and persons at high risk for infection. Achievement of this goal requires providing public health infrastructures, which is missing in some developing countries. In addition, global vigilance of final steps of the eradication program should be increased to attain maximum public contributions. Finally, worldwide IPV immunization should continue with high levels of coverage long after OPV cessation until reliable data indicate worldwide elimination of any poliovirus.