SUPPLEMENT ARTICLE. S62 JID 2014:210 (Suppl 1) Alleman et al

Transcription

1 SUPPLEMENT ARTICLE Factors Contributing to Outbreaks of Wild Poliovirus Type 1 Infection Involving Persons Aged 15 Years in the Democratic Republic of the Congo, , Informed by a Pre-Outbreak Poliovirus Immunity Assessment Mary M. Alleman, 1 Kathleen A. Wannemuehler, 1 William C. Weldon, 2 Jean Pierre Kabuayi, 3 Felly Ekofo, 3 Samuel Edidi, 3 Audry Mulumba, 4 Albert Mbule, 5 Renée N. Ntumbannji, 5 Tiekoura Coulibaly, 5 Nadine Abiola, 6 Minlangu Mpingulu, 6 Kassim Sidibe, 6 and M. Steven Oberste 2 1 Global Immunization Division, and 2 Division of Viral Diseases, Centers for Disease Control and Prevention, Atlanta, Georgia; 3 Programme National de Lutte contre les IST/SIDA, 4 Expanded Programme on Immunization, Ministry of Public Health, 5 Immunization, Vaccines, and Emergencies Cluster, World Health Organization, and 6 Division of Global HIV/AIDS, Centers for Disease Control and Prevention, Kinshasa, Democratic Republic of the Congo Background. The Democratic Republic of the Congo (DRC) experienced atypical outbreaks of wild poliovirus type 1 (WPV1) infection during in that they affected persons aged 15 years in 4 (Bandundu, Bas Congo, Kasaï Occidental, and Kinshasa provinces) of the 6 provinces with outbreaks. Methods. Analyses of cases of WPV1 infection with onset during by province, age, polio vaccination status, and sex were conducted. The prevalence of antibodies to poliovirus (PV) types 1, 2, and 3 was assessed in sera collected before the outbreaks from women attending antenatal clinics in 3 of the 4 above-mentioned provinces. Results. Of 193 cases of WPV1 infection during , 32 (17%) occurred in individuals aged 15 years. Of these 32 cases, 31 (97%) occurred in individuals aged years; 9 (28%) were notified in Bandundu, 17 (53%) were notified in Kinshasa, and 22 (69%) had an unknown polio vaccination status. In the seroprevalence assessment, PV type 1 and 3 seroprevalence was lower among women aged years in Bandundu and Kinshasa, compared with those in Kasaï Occidental. Seropositivity to PVs was associated with increasing age, more pregnancies, and a younger age at first pregnancy. Conclusions. This spatiotemporal analysis strongly suggests that the outbreaks of WPV1 infection affecting young adults were caused by a PV type 1 immunity gap in Kinshasa and Bandundu due to insufficient exposure to PV type 1 through natural infection or vaccination. Poliovirus immunity gaps in this age group likely persist in DRC. Keywords. polio eradication; polio seroprevalence; neutralizing antibodies; Democratic Republic of the Congo; Africa; wild poliovirus outbreaks; polio in adults. Prior to 2001, wild poliovirus (WPV) transmission was endemic in the Democratic Republic of the Congo (DRC) [1 15]. Case-based acute flaccid paralysis Correspondence: Mary M. Alleman, PhD, Global Immunization Division, Centers for Disease Control and Prevention, 1600 Clifton Rd, NE, MS A-04, Atlanta, GA (mea4@cdc.gov). The Journal of Infectious Diseases 2014;210(S1):S62 73 Published by Oxford University Press on behalf of the Infectious Diseases Society of America This work is written by (a) US Government employee(s) and is in the public domain in the US. DOI: /infdis/jiu282 (AFP) surveillance for polio, in which each suspected polio case is investigated with collection and analysis of stool specimens, was not established nationwide in DRC until 1999 or later; however, historical reports of WPV transmission, cases of infection, and outbreaks of infection are available [1 16]. In those reports, most cases of WPV infection in which the age of the infected individual was documented occurred in individuals <5 years old; documentation of cases in individuals aged 15 years existed but was rare [4, 6, 7, 9,14]. WPV transmission persisted in DRC despite the inclusion of 3 S62 JID 2014:210 (Suppl 1) Alleman et al

2 doses of trivalent oral polio vaccine (topv) in the country s Expanded Programme on Immunization (EPI), which was initiated in 1977 (Figure 1) [1 12, 17 19]. National-level World Health Organization (WHO)/United Nations Children s Fund (UNICEF) coverage estimates for the third dose of topv (OPV3) during ranged from 15% in 1980 to 41% in 1988; other reports suggest that OPV3 coverage was poor and not uniform geographically across the DRC during these years (Figure 1)[1 7, 9, 11, 12, 17 20]. OPV3 coverage estimates remained <50% until 2004, when coverage began increasing, reaching 78% by 2011 [20]. In line with strategies of the Global Polio Eradication Initiative (GPEI), supplemental immunization activities (SIAs) with topv, in the form of subnational immunization days, began in 1996 in DRC (Figure 1) [1 3, 5, 11, 12, 21]. The first national immunization day (NID) was conducted in 1999, followed by others in ; after those years, NIDs were not conducted until 2011 [2 4, 12, 21 23]. These SIAs targeted children aged <5 years with topv through 2005, after which other poliovirus-containing vaccines became available and were used, and total populations were vaccinated (in 2011 and later) according to the particular epidemiological situation [2, 3, 21 23]. After the initiation of SIAs, there was a period ( ) with no confirmed cases of WPV infection in DRC, and the interruption of WPV transmission was assumed (Figure 1) [4, 23 32]. DRC s last case of WPV type 2 (WPV2) infection was documented in 1997 [4]. A series of WPV importations from Angola during led to outbreaks of WPV type 1 (WPV1) and WPV type 3 (WPV3) infections that lasted through 2011 [31 44]. In combined, there were 58 and 4 cases of WPV1 and WPV3 infection, respectively [45]. Of these 62 cases, 100% occurred in individuals aged <15 years. The last case of WPV3 Figure 1. Historical time line of implementation of polio eradication strategies in the Democratic Republic of the Congo, with national-level oral polio vaccine 3 (OPV3) coverage estimates (OPV3 WHO/UNICEF national-level coverage estimates are not available for years prior to 1980) and indication of birth years relative to the collection of blood ( ) used in the poliovirus sero-prevalence assessment and to the occurrence of wild poliovirus type 1 (WPV1) cases aged years during the WPV1 outbreaks. Abbreviations: AFP, acute flaccid paralysis; DRC, Democratic Republic of the Congo; EPI, Expanded Programme on Immunization; NID, national immunization day; WPV2, wild poliovirus type 2; WPV3, wild poliovirus type 3; WHO, World Health Organization; UNICEF, United Nations Children s Fund; y, year. Factors Contributing to Outbreaks of WPV1 Infection in DRC JID 2014:210 (Suppl 1) S63

3 infection in the DRC had paralysis onset in June 2009 [45]. During the outbreaks of WPV1 infection, a higher than usual percentage of cases was notified among persons aged 15 years in Bandundu, Bas Congo, Kasaï Occidental, and Kinshasa provinces [46, 47]. To better understand the outbreaks and the susceptibility to WPV infection among individuals aged 15 years, the outbreak was analyzed, and a polio seroprevalence assessment was undertaken in Bandundu, Kasaï Occidental, and Kinshasa provinces. The results are presented in this report. METHODS Review of Epidemiological Data AFP cases reported to DRC s Ministry of Public Health and WHO-DRC with symptom onset during 1 January December 2011 were reviewed for WPV and compatible cases by geographic location, age, sex, and polio vaccination status. AFP cases were classified per definitions in the WHO African Region AFP surveillance guidelines [16]. Poliovirus Seroprevalence Assessments Titers of neutralizing antibodies against poliovirus (PV) types 1 (PV1), 2 (PV2), and 3 (PV3) were measured in sera prepared from venous blood collected during from 5942 pregnant women aged years from 7 different antenatal clinic (ANC) sites in Bandundu, Kasaï Occidental, and Kinshasa provinces (Figure 2) [48, 49]. The Binza Meteo, Boyambi, and Kingasani sites (urban sites in DRC s capital, Kinshasa) in Kinshasa province and the Vanga (rural site) and Kikwit (Kikwit city, urban site) sites in central Bandundu province were chosen because they were geographically near the location of WPV1 outbreaks in which a higher than usual percentage of cases occurred in persons aged 15 years. The remaining 2 sites, Mikalayi (rural site) and Tshikapa (Tshikapa city, urban site) both in Kasaï Occidental province, were Figure 2. Cases of wild poliovirus type 1 (WPV1) infection, and approximate location of antenatal clinic (ANC) sites where blood specimens were collected from persons in the seroprevalence assessment population, by zone de santé (health zone) and province, Democratic Republic of the Congo. One symbol represents 1 case of WPV1 infection. Symbols are drawn at random within the zones de santé. Boundaries of zones de santé are not shown except in the Kinshasa province inset. The geographic locations of the ANC sites are approximate and based on the zones de santé in which they are located. S64 JID 2014:210 (Suppl 1) Alleman et al

4 geographically near outbreaks of WPV1 infection in which there were cases in the older age group ( 15 years of age), but the percentage of cases in the older age group was relatively low compared to Kinshasa and Bandundu provinces. The seroprevalence assessment protocol was finalized prior to the onset of most WPV1 cases in persons aged 15 years in Bas Congo; thus, this province was not included. The sera were from 6615 original blood samples collected in the 7 ANC sites as part of nationwide sentinel site human immunodeficiency virus (HIV) seroprevalence surveys [48, 49]. From the 6615 samples, those for which the DRC National HIV Program laboratory had sufficient volume of sera (approximately 500 μl) were used for the PV antibody assessments; the 5942 sera analyzed for this report represent 90% of the 6615 samples. Delinked demographic attributes were available for each woman from whom blood was collected; only those attributes considered relevant to PV immunity were analyzed for associations with PV seroprevalence [48, 49]. Sera were shipped from Kinshasa to the Centers for Disease Control and Prevention (CDC-Atlanta) via air transport on dry ice and were thereafter stored at 20 C until antibody assessments were performed at CDC-Atlanta s Poliovirus and Picornavirus Laboratory according to previously described methods [50, 51]. Sera with 50% end point titers of 1:8 were considered seropositive [50, 51]. Data Analyses Seroprevalence for each PV type was calculated by dividing the number of seropositive sera in the stratum of interest by the total number of sera in that stratum. The seroprevalence estimates represent seroprevalence within the ANC assessment population only and are not representative of any DRC populations at large. Data analyses were conducted using SAS, version 9.2 (SAS Institute, Cary, NC); Epi Info, version 7 (CDC-Atlanta, Atlanta, GA); and Excel, version 2010 (Microsoft, Redmond, WA). Figure 2 was created using ArcGIS, version 10.1 (Environmental Systems Research Institute, Redlands, CA). The Cochran-Armitage test for trend was used to test for linear trends in PV seroprevalence across substrata of increasing age, overall and by ANC site. The Cochran-Mantel-Haenszel test was used to assess the association between demographic attributes and PV seroprevalence while controlling for age or province, as indicated in the Results section. The test of binomial proportions was used to determine whether sex percentages among cases of WPV1 infection were different than 50%; the ratio of males to females is approximately 1:1 in DRC in the age groups of interest in this analysis [52]. Tests were considered statistically significant at P values of <.05. Ethical Approval The Human Subjects Research Coordinator at the CDC-Atlanta National Center for Immunization and Respiratory Diseases reviewed the seroprevalence assessment protocol. It was determinedtobeaprogramevaluationinresponsetothe outbreaks of WPV1 infection and was therefore exempt from requiring institutional review board approval. The protocol was reviewed and approved by DRC s Ministry of Public Health. RESULTS In 2010 and 2011, there were outbreaks of 100 (in 5 provinces) and 93 (in 6 provinces) cases of WPV1 infection, respectively, in DRC (Table 1 and Figure 2). Combining the 2 years, there were 32 cases among individuals aged 15 years in Bandundu, Bas Congo, Kasaï Occidental, and Kinshasa provinces; more specifically, 21%, 16%, 3%, and 50% of all cases with known age in these provinces, respectively, were in this older age group [46].Theagerangeofthe32caseswas16 31 years, with 31 of 32 (97%) aged years; no deaths among these cases were documented. In 2011 alone, 29% of 93 cases of WPV1 infection nationally were aged 15 years. Comparing the provinces, Kinshasa had the highest annual percentage of cases of WPV1 infection in this older age group in 2011 (17 of 33 total cases [52%]; Table 1). Sixteen (94%) of the 17 cases were aged years (Figure 3A). All 17 presented in zones de santé (ZS; health zones) within Kinshasa city (Figure 2). Nine reported an unknown OPV vaccination status; of the 8 reporting vaccination, 75% had received 3 OPV doses. In addition to the confirmed cases of WPV1 infection in Kinshasa, there were 16 polio-compatible cases in , 7 (44%) of which involved individuals aged 15 years. Of these 7, 5 (71.4%) were aged years (data not shown). In Bandundu, among the 9 cases of WPV1 infection aged 15 years in , 100% were aged years (Table 1 and Figure 3B). Seven of 9 presented in ZS in the general vicinity of the Vanga and Kikwit ANC sites in central Bandundu, and 1 of the 9 had a known vaccination status reporting 0 OPV doses (Table 1 and Figure 2). In Bas Congo, the 4 cases of WPV1 infection aged 15 years in 2011 were all years of age (data not shown); 1 of the 4 had a known vaccination status (0 OPV doses reported), and the individuals presented in ZS scattered throughout the province (Table 1 and Figure 2). Two of 67 cases of WPV1 infection in Kasaï Occidental were aged 15 years. The adult cases involved individuals aged 18 (2010) and 25 years (2011), both had unknown vaccination status, with one presenting in a ZS in the same region of the province as the Mikalayi ANC site and the other presenting in a ZS relatively far from the Mikalayi and Tshikapa ANC sites (Table 1, Figure 2, and Figure 3C). At the national level, in , among cases of WPV1 infection aged <15 years and 15 years, 48% and 66% were males, respectively (Table 1). These percentages were not statistically significantly different from the expected percentage of males in these age groups in the population (ie, 50%) [52]. Factors Contributing to Outbreaks of WPV1 Infection in DRC JID 2014:210 (Suppl 1) S65

5 Table 1. Characteristics of Cases of Wild Poliovirus Type 1 (WPV1) Infection Nationally and by Province, Age Category, Oral Polio Vaccination Status, and Sex, , Democratic Republic of the Congo (DRC) Province Bas Kasaï Characteristic, by Year(s) National Bandundu Congo Occidental Katanga Kinshasa Maniema 2010 Cases, no. 100 a 23 a Cases in persons aged <15 y, no NA Cases in persons aged 15 y, no NA Cases in persons with unknown age, no. 2 a 2 a NA NA NA NA NA Cases in persons aged 15 y, % NA 2011 Cases, no Cases in persons aged <15 y, no Cases in persons aged 15 y, no Cases in persons aged 15 y, % combined Cases in persons aged 15 y, % 17 a 21 a Among cases in persons with age 15 y, those with NA 47 NA known vaccination status, % Among cases in persons with age 15 y and with known NA NA 75 NA vaccination status, those reported to have received 3 OPV doses, % Cases in persons aged <15 y who were male, % 48 b Cases in persons aged 15 y who were male, % 66 b NA 65 NA In , there were no cases of wild poliovirus infection in the DRC s remaining 5 provinces (ie, Equateur, Kasaï Oriental, Province Orientale, Nord Kivu, and Sud Kivu). Abbreviations: NA, not applicable; OPV, oral polio vaccine. a Two of the 23 cases from Bandundu had missing data on age. Of the 2, 1 was a female, and the other s sex was unknown. b P >.05, for the test of binomial proportions that the percentage is different than 50%. The percentages of males among cases in these age categories by province were too few to determine whether sex-specific percentages were statistically significantly different than 50% (Table 1 and Figures 3A C). Sera from blood specimens collected during from pregnant women attending ANC sites in Bandundu, Kasaï Occidental, and Kinshasa provinces were assessed for neutralizing antibodies against PV1, PV2, and PV3 (Figures 2 and 3). Selected demographic attributes of these women are described in Table 2. Of note, >99% of the women from each of the 3 sites in Kinshasa province declared residence in towns, consistent with the location of those ANC sites within Kinshasa city. For 6 of 7 sites, the highest percentage of women was in the years of age category at the time of blood collection; in Binza Meteo, the highest percentage was in the next highest age category. In all sites, 75% of women had their first pregnancy prior to age 24 years, and in all sites except those in Binza Meteo and Boyambi, 50% of women had been pregnant 3 times. HIV seroprevalence ranged from 0.8% to 3.2% in the various sites. The prevalence of neutralizing antibodies to PV1, PV2, and PV3 among the overall assessment population by age category is provided in Table 3; 73.8% of the women were seropositive for all 3 PV types (data not shown). For each PV type, there was a statistically significant increase in seroprevalence with increasing category of age. PV1 seroprevalence for all age categories was >95% in Tshikapa; >85% in Kingasani, Mikalayi, and Vanga; >75% in Boyambi and Kikwit; and >65% in Binza Meteo (Figures 3A C). In all sites, by age 30 years, PV1 seroprevalence was 94%, and, excepting Tshikapa, there was a statistically significant increase in seroprevalence with increasing age category (P <.05 for all sites except Tshikapa, where P =.12; Figures 3A C). PV2 seroprevalence was >90% for all age categories in all sites except Boyambi and Binza Meteo (Figures 3D F ). In all sites except Boyambi and Binza Meteo, by 30 years of age, seroprevalence was 96%, and in all sites except Boyambi and Kikwit, there was a statistically significant increase in seroprevalence with increasing age category (P <.05 for all sites except Boyambi, where P =.39, and Kikwit, where P =.06). PV3 seroprevalence was >70% for all age categories in all sites in Bandundu and Kasaï Occidental and in the Kingasani site S66 JID 2014:210 (Suppl 1) Alleman et al

6 Figure 3. Cases of wild poliovirus type 1 (WPV1) infection, , by age, sex, and province, and seroprevalence to polioviruses type 1 (PV1), 2 (PV2), and 3 (PV3) in the seroprevalence assessment population, by age and antenatal clinic site in Kinshasa (A, D, and G), Bandundu (B, E, and H ), and Kasaï Occidental (C, F, and I) provinces, Democratic Republic of the Congo. In all graphs, the secondary y-axis indicating seroprevalence is truncated and starts at 50%. Horizontal gridlines are associated with the secondary axis (seroprevalence values) and not the primary axis (cases of WPV1 infection). (Figures 3G I). In Boyambi, seroprevalence was 70% for all age categories except for years, and in Binza Meteo, it was 59% for all age categories. In Mikalayi and Vanga, by age 30 years, prevalence was 94%, whereas in Binza Meteo and Boyambi, by age 30 years, the seroprevalence was <80%. For all sites, there was a statistically significant increase in seroprevalence with increasing age category (P <.05). The associations of demographic attributes with seroprevalence to PV1, PV2, and PV3 are presented in Table 3. Inthe assessment population overall, after controlling for age at the time of blood collection, there was a statistically significant association between being of younger age during the first pregnancy and being seropositive for PV2 and PV3. Controlling for province, there was a statistically significant association between a greater number of pregnancies and a higher seroprevalence for all PV types in both the year and year age groups. There was no statistically significant association between seroprevalence for HIV and seroprevalence for any PV type. DISCUSSION In , DRC had historically atypical outbreaks of WPV1 infection in that they affected numerous persons aged 15 years in 4 (Bandundu, Bas Congo, Kasaï Occidental, and Kinshasa) of the 6 provinces with outbreaks. Prior to these 2 years, nearly all documented WPV cases had been in children <15 years of age [4, 6, 7, 9, 14]. Of the 4 provinces in the 2 years combined, Kinshasa had the highest percentage of cases of WPV1 infection in this older age group, followed by Bandundu, Bas Congo, and Kasaï Occidental. The analysis of poliocompatible cases from suggests that Kinshasa s outbreak might have been of greater magnitude and affected a larger number of persons aged 15 years than previously thought [46]. DRC s response to the outbreaks included numerous SIAs conducted during with monovalent oral polio vaccine targeting PV1 (mopv1), bivalent OPV targeting PV1 and PV3 (bopv1,3), and topv [21]. The campaigns generally targeted children aged <5 years; however, Factors Contributing to Outbreaks of WPV1 Infection in DRC JID 2014:210 (Suppl 1) S67

7 Table 2. Demographic Characteristics of the Poliovirus Seroprevalence Assessment Population, Overall and by 7 Antenatal Clinic Sites, in Bandundu, Kasaï Occidental, and Kinshasa Provinces, Democratic Republic of the Congo Bandundu Province, No. (%) Kasaï Occidental Province, No. (%) Kinshasa Province, No. (%) Characteristic Overall, No. (%) (n = 5942) Kikwit (n = 954) Vanga (n = 936) Mikalayi (n = 755) Tshikapa (n = 682) Binza Meteo (n = 805) Boyambi (n = 835) Kingasani (n = 975) Residence Town 4110 (69.2) 830 (87.0) 8 (0.9) 8 (1.1) 660 (96.8) 802 (99.6) 833 (99.8) 969 (99.4) Village 1830 (30.8) 124 (13.0) 928 (99.2) 747 (98.9) 22 (3.2) 3 (0.4) 0 (0.0) 6 (0.6) Unknown 2 (0.03) 2 (0.2) Age at time of blood collection, y (16.3) 154 (16.1) 128 (13.7) 170 (22.5) 164 (24.1) 99 (12.3) 120 (14.4) 131 (13.4) (27.6) 239 (25.1) 264 (28.2) 191 (25.3) 170 (24.9) 220 (27.3) 282 (33.8) 274 (28.1) (25.5) 232 (24.3) 251 (26.8) 164 (21.7) 155 (22.7) 222 (27.6) 229 (27.4) 263 (27.0) (18.0) 197 (20.7) 162 (17.3) 121 (16.0) 110 (16.1) 150 (18.6) 132 (15.8) 197 (20.2) (12.6) 132 (13.8) 131 (14.0) 109 (14.4) 83 (12.2) 114 (14.2) 72 (8.6) 110 (11.3) Age at first pregnancy, y (33.4) 225 (23.6) 312 (33.3) 490 (64.9) 375 (55.0) 162 (20.1) 185 (22.2) 236 (24.2) (52.8) 563 (59.0) 538 (57.5) 254 (33.6) 265 (38.9) 440 (54.7) 515 (61.7) 565 (58.0) (13.8) 166 (17.4) 86 (9.2) 11 (1.5) 42 (6.2) 203 (25.2) 135 (16.2) 174 (17.9) No. of pregnancies, including current pregnancy (25.7) 272 (28.5) 182 (19.4) 145 (19.2) 146 (21.4) 244 (30.3) 249 (29.8) 291 (29.9) (20.8) 206 (21.6) 185 (19.8) 105 (13.9) 100 (14.7) 189 (23.5) 253 (30.3) 197 (20.2) (53.5) 476 (49.9) 569 (60.8) 505 (66.9) 436 (63.9) 372 (46.2) 333 (39.9) 487 (50.0) HIV serologic test result Positive 113 (1.9) 8 (0.8) 22 (2.4) 9 (1.2) 19 (2.8) 13 (1.6) 11 (1.3) 31 (3.2) Negative 5829 (98.1) 946 (99.2) 914 (97.7) 746 (98.8) 663 (97.2) 792 (98.4) 824 (98.7) 944 (96.8) Abbreviation: HIV, human immunodeficiency virus. certain SIAs targeted all ages in relevant areas of Bandundu, Bas Congo, and Kinshasa [21, 46]. No cases of WPV infection were reported from Bandundu, Bas Congo, Kasaï Occidental, or Kinshasa after the third quarter of 2011; the last case of WPV infection (due to WPV1) in the DRC had paralysis onset in December 2011 (Maniema province) [43, 44, 46, 53]. The epidemiological data from the outbreak suggest that among those aged 15 years, the age group most susceptible to WPV1 infection were young adults (ie, those aged years). In 3 of the 4 provinces (Bandundu, Bas Congo, and Kasaï Occidental), 100% of cases aged 15 years were in persons years of age; in Kinshasa, it was 94%. Overall, of the 32 infected persons aged 15 years in , only 1 was >29 years old (age, 31 years). Outbreaks of WPV1 infection affecting adults have been reported recently in other countries, such as Albania, China, Namibia, the Republic of the Congo (ROC), and Tajikistan [54 60]. The analyses of infected persons aged 15 years from Namibia, the ROC, and Tajikistan yielded results similar to those of DRC, insofar as there was a high proportion/high incidence of cases in adults aged approximately years; in Albania and China the adults were approximately years of age [54 58, 60]. The high case-fatality ratios reported among infected young adults in Albania, Namibia, and the ROC and among WPV-infected adults in general were not documented among the cases notified in DRC [10, 54, 56 58, 61 63]. Moreover, a statistically significantly higher percentage of cases in one gender versus the other was not observed among the WPV1-infected individuals aged 15 years in DRC [62]. The investigators of the Albania, China, Namibia, and ROC outbreaks concluded that the young adults affected by those outbreaks were young children at a time when routine immunization coverage with topv was sufficiently high to reduce WPV transmission, and consequently opportunities for acquiring natural immunity, yet not high enough, leaving much of the cohort unvaccinated or incompletely vaccinated because of suboptimal routine immunization coverage coupled with age-based ineligibility for polio SIAs [54 58]. Other outbreaks, affecting adults, documented prior to these were explained similarly [10]. Persons aged years in DRC at the time of the outbreaks would have been born between approximately 1980 and 1996, a span when topv was being increasingly provided through the EPI; however, OPV3 coverage was low, only occasionally approaching 40% nationally, and the provision of S68 JID 2014:210 (Suppl 1) Alleman et al

8 Table 3. Seroprevalence to Poliovirus Type 1 (PV1), PV2, and PV3, by Demographic Characteristic, for the Assessment Population Overall and Associations Controlling for Age and Province, Democratic Republic of the Congo Characteristic PV1 Seropositivity PV2 Seropositivity PV3 Seropositivity Age at time of blood collection /966 (84.7) 883/966 (91.4) 699/966 (72.4) /1640 (89.7) 1527/1640 (93.1) 1247/1640 (76.0) /1516 (93.8) 1445/1516 (95.3) 1247/1516 (82.3) /1069 (96.6) 1028/1069 (96.2) 917/1069 (85.8) /751 (97.6) 732/851 (97.5) 670/751 (89.2) Overall 5477/5942 (92.2) 5615/5942 (94.5) 4780/5942 (80.4) Cochran-Armitage trend test (for age) statistic (P) (<.0001) 6.52 (<.0001) (<.0001) Age at first pregnancy, y /1985 (92.5) 1892/1985 (95.3) 1637/1985 (82.5) /3140 (91.4) 2968/3140 (94.5) 2511/3140 (80.0) /817 (94.4) 755/817 (92.4) 632/817 (77.4) CMH statistic (P), controlling for 2 categories of age at time 4.54 (.10) (<.0001) (<.0001) of blood collection a No. of pregnancies, including current pregnancy, by age y /1304 (85.2) 1179/1304 (90.4) 917/1304 (70.3) 2 675/763 (88.5) 724/763 (94.9) 585/763 (76.7) 3 503/539 (93.3) 507/539 (94.1) 444/539 (82.4) CMH statistic (P), controlling for province (.0002) (.001) (.0001) y 1 201/218 (92.2) 193/218 (88.5) 156/218 (71.6) 2 419/451 (92.9) 419/451 (92.9) 342/451 (75.8) /1916 (95.8) 1861/1916 (97.1) 1666/1916 (87.0) CMH statistic (P), controlling for province 6.95 (.031) (<.0001) (<.0001) HIV serologic test result Positive 108/113 (95.6) 106/113 (93.8) 91/113 (80.5) Negative 5369/5829 (92.1) 5509/5829 (94.5) 4689/5829 (80.4) CMH statistic (P), controlling for 2 categories of age at time of blood collection a 1.61 (.20) 0.15 (.69) (.93) Data are no. of women seropositive for the specified PV type/total no. in stratum (%), unless otherwise indicated. Trends and associations are considered statistically significant at a P value of <.05. Abbreviations: CMH, Cochran-Mantel-Haenszel; HIV, human immunodeficiency virus. a Ages y and 25 y. routine health services was not uniform geographically across DRC (Figure 1)[1, 4 7, 9, 11, 12, 17 20]. While WPV transmission remained endemic during those years and outbreaks were documented, it is possible that the increasing use of topv, despite low coverage, had the impact of gradually reducing transmission, leading to fewer opportunities for immunity through natural infection [1, 6 12]. The only additional opportunity for immunization for the majority of these persons would have been through exposure to topv excreted by vaccinated contacts [63]. Only the very youngest persons in this cohort (those born during 1991 or later) would have been eligible for polio subnational immunization days and NIDs and only until they were aged 5 years (Figure 1) [1 3, 5, 11, 12, 21]. It is thus plausible, based on the country s history of WPV transmission and polio vaccination, that the young adult population in certain of DRC s provinces was susceptible to WPV1 infection in because of lack of exposure to WPV transmission combined with insufficient exposure to OPV. The results of the PV seroprevalence assessment in Bandundu, Kasaï Occidental, and Kinshasa support this hypothesis and suggest that susceptibility might have differed geographically. More specifically, in the assessment population, in the Binza Meteo and Boyambi sites in Kinshasa city and in the Kikwit site in Kikwit city, PV1 seroprevalence was <80% in year-olds and <90% in year olds, notably lower than in the same age groups in the other 4 sites. Of the 4, of note are Factors Contributing to Outbreaks of WPV1 Infection in DRC JID 2014:210 (Suppl 1) S69

9 the 2 sites in Kasaï Occidental where the PV1 seroprevalence was nearly 90% in all age groups. The spatiotemporal overlap of low PV1 seroprevalence in the young adult women attending ANC clinics and numerous cases of WPV1 infection in persons in the same age range strongly suggests the presence of a subpopulation of young adults susceptible to WPV1 infection in Kinshasa and Bandundu provinces at the time of the outbreaks. Similarly, a spatiotemporal relationship between low polio vaccination coverage and high numbers of cases of WPV1 infection in the young adult population was observed in Pointe Noire, ROC, when a cross-sectional vaccination survey was conducted among young adults just after that outbreak [58]. In contrast are findings from Kasaï Occidental, where there was a spatiotemporal overlap of high PV1 seroprevalence among young adult women attending ANC clinics and few cases of WPV1 infection among young adults. Unfortunately, the polio vaccination status of 69% of the cases of WPV1 aged 15 years in the DRC s outbreaks was unknown, preventing analysis of individual susceptibility; this is a limitation of the analysis. In developing countries, it is estimated that immunity to PV1 needs to be at least 97% to prevent outbreaks [6, 63]. The seroprevalence assessment also provides an indication of PV1 immunity in women attending ANC clinics aged 30 years and of PV2 and PV3 immunity in all age groups. For the older women, born between approximately 1960 and 1978, seroprevalence to PV1 and PV3 was, with few exceptions, significantly higher statistically than in the younger women. Since only the youngest among the women aged 30 years (those born 1977 or later) would have been eligible for topv as infants through the EPI, the vast majority of these older women obtained PV immunity through natural infection in childhood, when PV was highly endemic in DRC, and/or through up to 30 years of contact exposure to OPV, boosted once polio SIAs were initiated (Figure 1) [1 5, 11, 12, 17, 21, 63]. The concept of obtaining PV immunity either through asymptomatic exposure to WPV or through OPV contact exposure is supported by the findings from this seroprevalence assessment in that, for all 3 PVs, seropositivity was associated with having had a greater number of pregnancies (a potential proxy for having more living children and, thus, greater doses and frequency of contact exposure) and that, for PV2 and PV3, seropositivity was associated with younger age at first pregnancy (a potential proxy for a longer duration of contact exposure) [63]. PV2 seroprevalence was uniformly high across all ages and most ANC sites (excepting Binza Meteo and Boyambi), consistent with exposure to WPV2 by older women and with high transmissibility and the high rate of seroconversion after exposure to PV2-containing OPV, even in the context of DRC s low vaccination coverage [63, 64]. Thus, there can be high PV2 seropositivity in a population simultaneous with lower seroprevalence to PV1 and PV3 [63]. Relative to PV1 and PV2, seroprevalence to PV3 was lower across most age groups in the assessment population; particularly noteworthy was the lowest PV3 seroprevalence in the 3 Kinshasa ANC sites and the Kikwit ANC site. The PV3 seroprevalence results reflect the lower immunogenicity of PV3 and could be used as a proxy indicator of exposure in the assessment population to topv, mopv3, or WPV3 [21, 45, 60, 63]. The assessment showed no differences in PV seroprevalence in women with HIV-positive results of serological tests, compared with women with HIV-negative results. There are few publications describing PV seroprevalence in HIV-seropositive African adults. A recent polio seroprevalence assessment in women attending ANC clinics during 2010 in Namibia, where the HIV prevalence among adults is higher than that in DRC, reports that among HIV-seropositive women there was lower seroprevalence for PV1 and PV3 [65]. Additional studies are needed in countries with high HIV seroprevalence to fully assess the relationship and to inform the GPEI. This analysis has a number of limitations. First, it is possible that in years prior to the outbreak, although cases of WPV infection occurred in persons aged 15 years, they were less likely to be identified, notified, and documented than cases in younger persons. It is only since 2011 that the DRC s AFP surveillance case definition was expanded from its focus on persons <15 years of age to specifically mention clinical signs in adults [66]. However it should be noted that the previous AFP case definition had always accounted for persons of any age in which a clinician suspected polio and that cases in persons aged 15 years, although few, had been documented in 2000, indicating a certain level of awareness that polio could occur in this older age group [4, 16]. Further, community-based active WPV case search conducted just after a 1995 outbreak of nearly 500 cases of WPV infection in the city of Mbuji Mayi revealed no adult cases [9]. Second, it is assumed that during the outbreaks, AFP surveillance was being conducted with equal rigor across all geographic areas of DRC and that the cases of WPV1 infection notified were a reflection of the true trends of cases in terms of absolute numbers, age, sex, and other relevant epidemiologic factors. Finally, the sera used in the seroprevalence assessment were from pregnant women voluntarily seeking antenatal care in certain geographic areas and were not representative of DRC s population at large; moreover, the women might have been more likely than males to be seropositive for PV because of greater contact with children excreting WPV or OPV [63]. Unfortunately, sera from adult males from this assessment period were not available. Data regarding the lifetime residential history of the women in the seroprevalence assessment would have been useful for more in-depth analyses of the lower PV seroprevalence in Kinshasa and Bandundu relative to that in Kasaï Occidental. In summary, this report describes a rare opportunity to obtain an indication of PV immunity in a large young adult S70 JID 2014:210 (Suppl 1) Alleman et al

10 population prior to outbreaks of WPV infection affecting young adults in the same geographic area. The results strongly suggest that the outbreaks of WPV1 infection affecting young adults in Kinshasa and Bandundu provinces could be explained by a PV1 immunity gap caused by lack of exposure to WPV and suboptimal exposure to OPV. The outbreak and the consequent vaccination response with mopv1, bopv1,3, and topv targeting all ages in certain areas might have narrowed the gap. However, susceptibility likely remains in young adults in DRC s other provinces. Except under exceptional circumstances, adults are rarely targeted for vaccination through polio eradication activities. Thus, the next best strategy to prevent future outbreaks in any age group is to attain and maintain the highest polio vaccination coverage possible among children, which will simultaneously boost immunity in contacts when OPV is used. While OPV3 coverage has improved in DRC during recent years and SIAs have been prevalent, not all children receive the minimum required doses. Thus, until OPV3 coverage is consistently sufficiently high across DRC, SIAs will be needed to build and maintain population immunity and to reduce the risk of widespread PV circulation should an introduction occur. A Demographic and Health Survey has just been conducted in DRC, during which blood specimens were obtained from a representative sample of children aged 6 months to 5 years from all provinces for PV neutralizing antibody assessments. The results of this survey will be extremely informative for guiding future polio eradication strategies in DRC. Moreover, the country should continue to include adults in AFP surveillance. These should be considerations for all countries with a history of polio epidemiology and vaccination similar to that of DRC. Notes Acknowledgments. We thank Luca Flamigni and Hypolite Sadiki, formerly, and Rogers Ngalamulume, currently, of the Division of Global HIV/ AIDS, CDC (Kinshasa, DRC), for facilitating the collaboration with DRC s Programme National de Lutte contre les IST/SIDA (PNLS); the technical staff of the PNLS laboratory who prepared the serum samples used in the seroprevalence assessment and shipped them to CDC-Atlanta; various staff of the Poliovirus and Picornavirus Laboratory at CDC-Atlanta, specifically, Michael McDonough, for assistance with the formalities for shipping the serum samples from Kinshasa to Atlanta, and Deborah Moore, Sharla McDonald, Yiting Zhang, and William Hendley, for working determinedly to perform the neutralizing antibody assessments in a short period; Brian Kaplan and Gina Marie Perleoni of the Geospatial Research Analysis and Services Program at the Agency for Toxic Substances and Disease Registry of CDC-Atlanta, for preparing the map in Figure 2; Kim Porter and Kristin Brown of CDC-Atlanta s Global Immunization Division, for assisting with the cleaning and merging of the various databases that were used in the analysis of the seroprevalence assessment; Steve Cochi, Allen Craig, Olen Kew, Melinda Mailhot, Mark Pallansch, Steve Wassilak, and Karen Wilkins, of CDC-Atlanta, and various staff of the DRC s Expanded Programme on Immunization, for invaluable comments and suggestions about the seroprevalence assessment protocol, analyses, interpretation of results, and the original version of the manuscript; and all staff of the DRC s Ministry of Public Health who work in HIV/AIDS surveillance and toward polio eradication, for their tireless efforts, often under difficult working conditions. Disclaimer. The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention. Financial support. This work was supported by the Ministry of Public Health, Democratic Republic of the Congo (to J. P. K., F. E., S. E., and A. M.); the World Health Organization, Democratic Republic of the Congo (to A. M., R. N. N., and T. C.); and the Centers for Disease Control and Prevention (to M. M. A., K. A. W., W. C. W., N. A., M. M., K. S., and M. S. O.). Supplement sponsorship. This article is part of a supplement entitled The Final Phase of Polio Eradication and Endgame Strategies for the Post-Eradication Era, which was sponsored by the Centers for Disease Control and Prevention. Potential conflicts of interest. All authors: No reported conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed. References 1. Centers for Disease Control and Prevention. Progress toward poliomyelitis eradication Africa, MMWR Morb Mortal Wkly Rep 1997; 46: Centers for Disease Control and Prevention. Progress toward poliomyelitis eradication Democratic Republic of Congo, MMWR Morb Mortal Wkly Rep 2000; 49: World Health Organization. Progress towards poliomyelitis eradication, Democratic Republic of the Congo, Wkly Epidemiol Rec 2000; 75: Kabue JP, Mushiya F, Pukuta E, et al. Surveillance virologique des paralysies flasques aiguës en République Démocratique du Congo (RDC) [Acute flaccid paralysis virologic surveillance in the Democratic Republic of the Congo ]. Med Trop 2004; 64: Centers for Disease Control and Prevention. Progress toward poliomyelitis eradication African Region, MMWR Morb Mortal Wkly Rep 1998; 47: Patriarca PA, Sutter RW, Oostvogel PM. Outbreaks of paralytic poliomyelitis, J Infect Dis 1997; 175(suppl 1):S World Health Organization. Expanded Programme on Immunization (EPI), poliomyelitis surveillance. Wkly Epidemiol Rec 1991; 66: World Health Organization. Expanded Programme on Immunization (EPI), Poliomyelitis in Wkly Epidemiol Rec 1993; 68: Lambert ML, Doussantousse S, Onadikondo L, et al. Poliomyelitis outbreak in Zaire. Lancet 1995; 346: Crawford C. Epidemics of poliomyelitis in Africa since Clin Infect Dis 2009; 48: Okwo-Bele JM, Lobanov A, Biellik RJ, et al. Overview of poliomyelitis in the African region and current regional plan of action. J Infect Dis 1997; 175(suppl 1):S Tangermann RH, Hull HF, Jafari H, et al. Eradication of poliomyelitis in countries affected by conflict. Bull World Health Organ 2000; 78: Barski G, Lépine P, Cornefert F. Recherche des anticorps neutralisants de la poliomyélite chez les Africains (noirs et pygmées) du Congo Belge [Neutralizing antibodies against poliomyelitis in pygmy and black Africans in the Belgian Congo]. Bull World Health Organ 1956; 14: De Moor J. Evolution de la poliomyélite à Léopoldville de 1951 à 1963 [Occurrence of poliomyelitis in Leopoldville, ]. Ann Soc Belg Med Trop 1965; 45: Pattyn SR, Delville JP, DeBont AF. Etude des anticorps antipoliomyélitiques dans la population congolaise de Léopoldville [Study of antipoliomyelitis antibodies in the Congolese population in Leopoldville]. Ann Soc Belg Med Trop 1957; 37: World Health Organization Regional Office for Africa. Acute flaccid paralysis surveillance field guide, January World Health Organization Regional Office for Africa. 17. ma-disu M. Role of nongovernmental agencies in vaccine delivery. Rev Infect Dis 1989; 11(suppl 3):S Factors Contributing to Outbreaks of WPV1 Infection in DRC JID 2014:210 (Suppl 1) S71

11 18. Mashako LMN, Kapongo CN, Nsibu CN, et al. Evaluation de la couverture vaccinale des enfants de moins de 2 ans à Kinshasa (Zaïre) [Evaluation of vaccination coverage among infants aged less than 2 years in Kinshasa (Zaïre)]. Arch Fr Pediatr 1992; 49: Tonglet R, Soron gane M, Lembo M, et al. Evaluation of immunization coverage at local level. World Health Forum 1993; 14: World Health Organization. WHO/UNICEF estimates of national immunization coverage, WHO-UNICEF estimates of POL3 coverage. timeseries/tscoveragepol3.html. Accessed 23 March World Health Organization, Democratic Republic of the Congo Country Office. Database of supplementary immunization activities for polio, January 1996 April World Health Organization, Democratic Republic of the Congo Country Office. 22. World Health Organization, Democratic Republic of the Congo Country Office. Activités d éradication de la poliomyélite, République Démocratique du Congo, 20 mai 2011 (Polio Eradication Activities, the Democratic Republic of the Congo, 20 May 2011). World Health Organization, Democratic Republic of the Congo Country Office. 23. Centers for Disease Control and Prevention. Progress toward poliomyelitis eradication Angola, Democratic Republic of Congo, Ethiopia, and Nigeria, January 2000 July MMWR Morb Mortal Wkly Rep 2001; 50: Centers for Disease Control and Prevention. Progress toward global eradication of poliomyelitis, MMWR Morb Mortal Wkly Rep 2002; 51: Centers for Disease Control and Prevention. Progress toward global eradication of poliomyelitis, MMWR Morb Mortal Wkly Rep 2003; 52: Centers for Disease Control and Prevention. Progress toward poliomyelitis eradication Angola and the Democratic Republic of Congo, January 2002 June MMWR Morb Mortal Wkly Rep 2003; 52: Centers for Disease Control and Prevention. Progress toward global eradication of poliomyelitis, January 2003 April MMWR Morb Mortal Wkly Rep 2004; 53: Centers for Disease Control and Prevention. Progress toward interruption of wild poliovirus transmission worldwide, January 2004 March MMWR Morb Mortal Wkly Rep 2005; 54: Centers for Disease Control and Prevention. Conclusions and recommendations of the Advisory Committee on Poliomyelitis Eradication Geneva Switzerland, October MMWR Morb, Mortal Wkly Rep 2005; 54: Centers for Disease Control and Prevention. Progress toward interruption of wild poliovirus transmission worldwide, January 2005 March MMWR Morb Mortal Wkly Rep 2006; 55: Centers for Disease Control and Prevention. Progress toward interruption of wild poliovirus transmission worldwide, January 2007 April MMWR Morb Mortal Wkly Rep 2008; 57: Centers for Disease Control and Prevention. Wild poliovirus type 1 and type 3 importations 15 countries, Africa MMWR Morb Mortal Wkly Rep 2009; 58: Centers for Disease Control and Prevention. Progress toward interruption of wild poliovirus transmission worldwide, January 2006 May MMWR Morb Mortal Wkly Rep 2007; 56: Centers for Disease Control and Prevention. Progress toward interruption of wild poliovirus transmission worldwide, MMWR Morb Mortal Wkly Rep 2009; 58: Centers for Disease Control and Prevention. Progress toward interruption of wild poliovirus transmission worldwide, MMWR Morb Mortal Wkly Rep 2010; 59: Centers for Disease Control and Prevention. Progress toward interrupting wild poliovirus circulation in countries with reestablished transmission Africa, MMWR Morb Mortal Wkly Rep 2011; 60: Centers for Disease Control and Prevention. Tracking progress toward global polio eradication worldwide, MMWR Morb Mortal Wkly Rep 2011; 60: Centers for Disease Control and Prevention. Progress toward interruption of wild poliovirus transmission worldwide, January 2010 March MMWR Morb Mortal Wkly Rep 2011; 60: Centers for Disease Control and Prevention. Progress toward global polio eradication Africa, MMWR Morb Mortal Wkly Rep 2012; 61: Centers for Disease Control and Prevention. Tracking progress toward global polio eradication, MMWR Morb Mortal Wkly Rep 2012; 61: Centers for Disease Control and Prevention. Progress toward interruption of wild poliovirus transmission worldwide, January 2011 March MMWR Morb Mortal Wkly Rep 2012; 61: Centers for Disease Control and Prevention. Progress toward eradication of polio worldwide, January 2011 March MMWR Morb Mortal Wkly Rep 2013; 62: Global Polio Eradication Initiative. List of wild polio virus by country, wild polio virus Document/Data&Monitoring/Wild_poliovirus_list_2008_2013_16Jul. pdf. Accessed 23 March World Health Organization, Democratic Republic of the Congo Country Office. Activités d éradication de la poliomyélite (IEP), République Démocratique du Congo, 2 août 2013 [Polio eradication activities (GPEI), the Democratic Republic of the Congo, 2 August 2013]. World Health Organization, Democratic Republic of the Congo Country Office. 45. World Health Organization, Democratic Republic of the Congo Country Office. AFP surveillance databases for World Health Organization, Democratic Republic of the Congo Country Office. 46. Nsambu MN, Bazira L, Coulibaly T, et al. Investigation et riposte à une épidémie de poliovirus sauvage à Kinshasa [Investigation of and response to a wild poliovirus outbreak in Kinshasa]. Pan Afr Med J 2013; 15: World Health Organization, Democratic Republic of the Congo Country Office. Activités d éradication de la poliomyélite, République Démocratique du Congo, Semaine 43, 28 octobre 2011 [Polio eradication activities, the Democratic Republic of the Congo, week 43, 28 October 2011]. World Health Organization, Democratic Republic of the Congo Country Office. 48. Ministry of Public Health-Democratic Republic of the Congo. Rapport épidémiologique de surveillance du VIH chez les femmes enceintes fréquentant les structures de CPN 2008 [Epidemiological report of HIV surveillance in pregnant women attending prenatal consultations 2008]. National HIV/AIDS Program, Ministry of Public Health-Democratic Republic of the Congo. 49. Ministry of Public Health-Democratic Republic of the Congo. Rapport épidémiologique de surveillance du VIH/SIDA chez les femmes enceintes fréquentant les structures de CPN 2009 [Epidemiological report of HIV/AIDS surveillance in pregnant women attending prenatal consultations 2009]. National HIV/AIDS Program, Ministry of Public Health-Democratic Republic of the Congo. 50. WHO Collaborative Study Group on Oral and Inactivated Poliovirus Vaccines. Combined immunization of infants with oral and inactivated poliovirus vaccines: results of a randomized trial in The Gambia, Oman, and Thailand. J Infect Dis 1997; 175:S Expanded Progamme on Immunization. Report of a WHO informal consultation on polio neutralizing antibody assays. Nashville, 5 6 December Geneva: World Health Organization, Index Mundi. Congo, Democratic Republic of the, Demographics Profile congo/demographics_profile.html. Accessed 23 March World Health Organization, Democratic Republic of the Congo Country Office. Activités d éradication de la poliomyélite, République Démocratique du Congo, Semaine 52, 30 décembre 2011 [Polio eradication activities, the Democratic Republic of the Congo, week 52, 30 December 2011). World Health Organization, Democratic Republic of the Congo Country Office. 54. Prevots DR, Ciofi degli Atti ML, Sallabanda A, et al. Outbreak of paralytic poliomyelitis in Albania, 1996: High attack rate among adults and S72 JID 2014:210 (Suppl 1) Alleman et al