Multiple independent emergences of type 2 vaccine-derived polioviruses during a large outbreak in northern Nigeria

Abstract

Since 2005, a large poliomyelitis outbreak associated with type 2 circulating vaccine-derived poliovirus (cVDPV2) has occurred in northern Nigeria, where immunization coverage with trivalent oral poliovirus vaccine (tOPV) has been low. Phylogenetic analysis of P1/capsid region sequences of isolates from each of the 403 cases reported in 2005 to 2011 resolved the outbreak into 23 independent type 2 vaccine-derived poliovirus (VDPV2) emergences, at least 7 of which established circulating lineage groups. Virus from one emergence (lineage group 2005-8; 361 isolates) was estimated to have circulated for over 6 years. The population of the major cVDPV2 lineage group expanded rapidly in early 2009, fell sharply after two tOPV rounds in mid-2009, and gradually expanded again through 2011. The two major determinants of attenuation of the Sabin 2 oral poliovirus vaccine strain (A481 in the 5'-untranslated region [5'-UTR] and VP1-Ile143) had been replaced in all VDPV2 isolates; most A481 5'-UTR replacements occurred by recombination with other enteroviruses. cVDPV2 isolates representing different lineage groups had biological properties indistinguishable from those of wild polioviruses, including efficient growth in neuron-derived HEK293 cells, the capacity to cause paralytic disease in both humans and PVR-Tg21 transgenic mice, loss of the temperature-sensitive phenotype, and the capacity for sustained person-to-person transmission. We estimate from the poliomyelitis case count and the paralytic case-to-infection ratio for type 2 wild poliovirus infections that ∼700,000 cVDPV2 infections have occurred during the outbreak. The detection of multiple concurrent cVDPV2 outbreaks in northern Nigeria highlights the risks of cVDPV emergence accompanying tOPV use at low rates of coverage in developing countries.

Fig 1

Map of Nigeria showing state boundaries, population density by LandScan (http://www.ornl.gov/landscan/), and location (circles color-coded by estimated year of emergence) of first acute flaccid paralysis (AFP) cases found to be associated with each independent type 2 vaccine-derived poliovirus (VDPV2) emergence. The numbers 1 to 23 correspond to emergences listed in Table 1 according to the order of the estimated date of the initiating tOPV dose. The bold red line indicates the boundary between the northern and southern states. Abbreviations for Nigerian states (S) with VDPV2 cases (14, 15) are as follows: ADS, Adamawa; ANS, Anambra; BAS, Bauchi; BNS, Benue; BOS, Borno; JIS, Jigawa; KBS, Kebbi; KDS, Kaduna; KNS, Kano; KTS, Katsina; KGS, Kogi; LAS, Lagos; NAS, Nasarawa; NIS, Niger; PLS, Plateau; SOS, Sokoto; YBS, Yobe; ZAS, Zamfara.

Fig 2

Distribution of nucleotide substitutions in the VP1 regions of Sabin 2-related polioviruses isolated from patients with AFP from 2005 to 2011 in the northern (n = 612) (A) and southern (n = 83) (B) states of Nigeria. Each isolate is from an individual patient. The northern and southern states are identified in Fig. 1. Type 2 vaccine-derived polioviruses (VDPV2s) (black bars) are more divergent, having ≥6 nt substitutions in the VP1 region (48).

Fig 3

Maximum clade credibility subtree of P1/capsid region sequences (2,637 nt) of 62 2005-2011 isolates representing the seven well-defined cVDPV2 lineage groups (in boldface type) and 16 other isolates that signaled additional independent VDPV2 emergences (lightface type). Lineage groups and emergences are numbered by year and ordered within years by the estimated dates of the initiating OPV doses. *, VDPV2/WPV1 coinfections. Color-coding of isolate identifiers corresponds to that used in Fig. 1.

Fig 4

Timeline of independent VDPV2 emergences and spread in Nigeria from 2004 to 2011 based on the dates of VDPV2 specimen collection (specimens were collected 10 ± 6 days after onset of AFP). cVDPV2 lineage groups are indicated in boldface type. Dates of the initiating tOPV doses for the 14 emergences associated with single isolates are point estimates with indeterminate confidence intervals. Arrows indicate months of tOPV supplemental immunization activities (SIAs) (mass vaccination campaigns) in the northern states. The blue point symbolizes two superimposed points. The blue wedge in the lower left corner symbolizes the duration of type 2 vaccine-related poliovirus excretion in immunologically healthy primary-dose OPV recipients (30).

Fig 5

Maximum likelihood estimates of relative frequencies of specific base changes (10⁻²) (black arrows, transitions; gray arrows, transversions) at all codon positions within the P1/capsid regions of the 361 2005-2011 cVDPV2 isolates from the major 2005-8 lineage group during divergence from the Sabin 2 root sequence.

Fig 6

Cumulative seasonality of specimens containing VDPV2 in Nigeria, 2004 to 2011. Note that AFP cases appeared ∼10 days earlier than specimen collection and that VDPV2 exposure likely occurred 3 to 4 weeks earlier than specimen collection.

Fig 7

Estimated number of observed lineages that survived the 2004-2005 to 2010-2011 December-to-January seasonal bottlenecks. Color-coding of individual emergences and lineage groups (boldface type) does not correspond to that used in Fig. 1.

Fig 8

Bayesian skyline plot of population dynamics of all Nigerian VDPV2s, 2005 to 2011. Shaded areas represent the 95% highest posterior density (HPD) interval around the mean of the effective virus population size (N_e) estimates. Arrows indicate months of tOPV SIAs in the northern states.

Fig 9

Frequency of nucleotide substitution into the VP1 region relative to the P1/capsid region among the 2005-2011 Nigerian VDPV2 isolates.