Reemergence of Recombinant Vaccine–derived Polioviruses in Healthy Children, Madagascar
Reemergence of Recombinant Vaccine–derived Polioviruses in Healthy Children, Madagascar
复制标题
重组疫苗衍生的脊髓灰质炎病毒在马达加斯加健康儿童中重新出现
DOI：
--
复制
发表时间：
2013
影响因子：
11.8
通讯作者：
F. Delpeyroux
中科院分区：
文献类型：
--
作者：
R. Razafindratsimandresy;M. Joffret;Sendraharimanana Rabemanantsoa;Seta Andriamamonjy;J. Héraud;F. Delpeyroux
致编辑：致病性疫苗衍生脊髓灰质炎病毒（VDPVs）引起的脊髓灰质炎疫情主要是脊髓灰质炎疫苗覆盖率低的结果。低覆盖率使口服脊髓灰质炎疫苗（OPV）中的脊髓灰质炎病毒（PV）能够在人与人之间传播，并使病毒发生遗传漂移，随后逆转为神经毒性表型（1）。2001-2002年和2005年在马达加斯加南部的Toliara省报告了与2型或3型VDPV（VDPV 2或VDPV 3）相关的脊髓灰质炎暴发（2、3）。这些VDPV见于急性弛缓性麻痹（AFP）患者和与AFP患者接触的健康儿童（2，4）。这些VDPV的基因组属于几个独立的、复杂的镶嵌重组谱系，由源自疫苗脊髓灰质炎病毒和其他共循环物种C人类肠道病毒（人类EV-C）的序列组成（4，5）。在开展了快速有效的口服脊髓灰质炎疫苗接种运动之后，2001-2002年和2005年在托里亚拉省爆发的疫情得到了遏制。
自2005年以来，马达加斯加没有发现脊髓灰质炎病例。然而，口服脊髓灰质炎疫苗在图利亚拉省的覆盖率有所波动。2011年6月，为了确定VDPV是否在该省流行，我们从616名5岁以下健康居民中采集了粪便样本（马达加斯加伦理委员会协议011-MSANP/CE）。使用样本提取物培养表达人脊髓灰质炎病毒细胞受体CD 155的人RD和HEp-2c细胞以及小鼠L细胞（L20 B细胞）（6）。在616份样本中，238份在人细胞中诱导致细胞病变效应，其中20份还在L20 B细胞中诱导致细胞病变效应;后一份样本通过逆转录PCR分子检测（7）确认含有PV菌株。所有PV分离株均来源于采集样本前30天内未接受OPV的儿童。
我们对20个确认含有PV毒株的分离株的编码衣壳病毒蛋白（VP 1）的基因组区域进行了测序：3个2型PV分离株与2型OPV毒株（Sabin 2）的nt序列差异>0.5%，因此被鉴定为潜在致病性VDPV 2。其中两种VDPV 2在VP 1区（903 bp）携带大量突变：分离株MAD-2593-11（30 nt突变）和MAD-2642-11（33 nt突变）。第三个VDPV 2，MAD-2675-11，携带6个核苷酸突变。Sabin 2与MAD-2593-11（3.3%）和MAD-2642-11（3.6%）之间的核苷酸差异表明，这些VDPV 2已繁殖或传播了10.3年，Sabin 2与MAD-2675-11之间的差异表明其已繁殖或传播了10.5年。随后进行了从nt 36至nt 7，420（Sabin 2编号）的几乎完整的基因组测序（EMBL登录号Nos. {“type”：“entrez-nucleotide-range”，“attrs”：{“text”：“HF913426-HF913428”，“start_term”：“HF913426”，“end_term”：“HF913428”，“start_term_id”：“510934497”，“end_term_id”：“510934501”}}HF913426-HF913428）。
分离株MAD-2675-11显示为突变的Sabin 2 PV，但分离株MAD-2593-11和MAD-2642-11显示由源自Sabin 2和其他人EV-C的序列组成的嵌合重组基因组。这两个重组体都具有来自非PV人类EV-C的5′-非翻译区（UTR）、非结构性P2-P3区和3′-UTR。2个重组位点类似地位于这些重组基因组中的每一个中（大约在nt 760和nt 3，368处），在结构基因组P1区的2个末端。虽然这些重组体似乎是相关的，但它们在P2-P3区和3′-UTR中具有不同的非PV人EV-C序列（6.0%-20.0%nt序列差异）。评估了这些分离株与先前马达加斯加VDPV 2的系统发育关系：2011年分离株来源于2个新的独立事件（图）。2011年的VDPV在5′-UTR（在nt 481处鸟嘌呤变为腺嘌呤）和VP 1区（在密码子143处异亮氨酸变为苏氨酸）区域中丢失了主要减毒决定簇（8）。这一发现有力地表明，2011年VDPV已经恢复了一定程度的神经毒力。
图
显示疫苗衍生脊髓灰质炎病毒（VDPV）分离株序列之间遗传关系的系统发育树。这些树是基于各种亚基因组区域的核苷酸序列比对。用CLC进行多序列比对。
虽然目前没有证据表明AFP病例与这些新的VDPV有关，但在3名儿童中发现的VDPV及其遗传特征强烈表明Toliara省口服脊髓灰质炎疫苗覆盖率不足。只有27%的参与者可以获得疫苗接种证明（疫苗接种卡）。为了防止第三次脊髓灰质炎爆发，马达加斯加卫生部于2011年10月和12月以及2012年1月在Toliara省组织了挨家挨户的口服脊髓灰质炎疫苗接种运动。自那时以来，该省没有报告任何脊髓灰质炎病例，这表明VDPV的传播受到这些运动的限制或停止。
全世界大规模的口服脊髓灰质炎疫苗免疫运动大大降低了野生型PV引起的脊髓灰质炎的发生率。然而，在发展中国家消灭了脊髓灰质炎后，低脊髓灰质炎疫苗覆盖率常常会导致VDPV的出现或疾病流行国家重新引入野生型PV;这两种情况都威胁到脊髓灰质炎根除计划的成功（9，10）。据报告，在21个发展中国家，有640多例由正在传播的VDPV引起的脊髓灰质炎病例。我们已经证明，即使在没有脊髓灰质炎病例的情况下，也可以在从生活在脊髓灰质炎疫苗覆盖率波动且通常较低的地区的儿童收集的粪便样本中检测到潜在致病性循环VDPV。因此，这种抽样可以帮助确定是否应该在脊髓灰质炎可能重新出现的地区开展疫苗接种运动。在马达加斯加，疫苗接种似乎清除了新出现的VDPV，并预防了儿童脊髓灰质炎病例。
To the Editor: Poliomyelitis outbreaks caused by pathogenic vaccine-derived polioviruses (VDPVs) are primarily a result of low polio vaccine coverage. Low coverage enables interhuman circulation of polioviruses (PVs) from the oral polio vaccine (OPV), and it enables genetic drift of the viruses and their subsequent reversion to neurovirulent phenotypes (1). Polio outbreaks associated with type 2 or 3 VDPVs (VDPV2s or VDPV3s) were reported in 2001–2002 and 2005 in Toliara Province in southern Madagascar (2,3). These VDPVs were found in patients with acute flaccid paralysis (AFP) and in healthy children who were contacts of the patients with AFP (2,4). The genomes of these VDPVs belong to several independent, complex mosaic recombinant lineages composed of sequences derived from vaccine polioviruses and other co-circulating species C human enteroviruses (human EV-C) (4,5). The 2001–2002 and 2005 outbreaks in Toliara Province were stopped after rapid and efficient OPV vaccination campaigns.
No polio cases have been detected in Madagascar since 2005. However, OPV coverage fluctuates in Toliara Province. In June 2011, to determine if VDPVs were circulating in the province, we collected fecal samples from 616 healthy residents <5 years of age (Madagascar Ethics Committee agreement 011-MSANP/CE). Sample extracts were used to inoculate human RD and HEp-2c cells and mouse L cells expressing the human poliovirus cellular receptor CD155 (L20B cells) (6). Of the 616 samples, 238 induced cytopathogenic effects in human cells, of which 20 also induced cytopathogenic effects in L20B cells; the latter samples were confirmed by reverse transcription PCR molecular testing (7) to contain PV strains. All PV isolates originated from children who had not received OPV in the 30 days before samples were collected.
We sequenced the genomic regions encoding the capsid viral protein (VP1) of the 20 isolates confirmed to contain PV strains: 3 type 2 PV isolates showed >0.5% nt sequence divergence from the type 2 OPV strain (Sabin 2) and were thus identified as potentially pathogenic VDPV2s. Two of these VDPV2s carried numerous mutations in the VP1 region (903 bp): isolates MAD-2593–11 (30 nt mutations) and MAD-2642–11 (33 nt mutations). The third VDPV2, MAD-2675–11, carried 6 nt mutations. The nucleotide differences between Sabin 2 and MAD-2593–11 (3.3%) and MAD-2642–11 (3.6%) suggest that these VDPV2s had been multiplying or circulating for ≈3.0 years, and the differences between Sabin 2 and MAD-2675–11 suggest it had been multiplying or circulating for ≈0.5 year. Almost complete genomic sequencing, from nt 36 to nt 7,420 (Sabin 2 numbering), was subsequently performed (EMBL accession nos. {"type":"entrez-nucleotide-range","attrs":{"text":"HF913426-HF913428","start_term":"HF913426","end_term":"HF913428","start_term_id":"510934497","end_term_id":"510934501"}}HF913426-HF913428).
Isolate MAD-2675–11 was shown to be a mutated Sabin 2 PV, but isolates MAD-2593–11 and MAD-2642–11 displayed mosaic recombinant genomes composed of sequences derived from Sabin 2 and other human EV-Cs. Both recombinants had the 5′–untranslated region (UTR), the nonstructural P2–P3 regions, and the 3′-UTR derived from non-PV human EV-C. The 2 recombination sites were similarly located in each of these recombinant genomes (approximately at nt 760 and nt 3,368) at the 2 extremities of the structural genomic P1 region. Although these recombinants appeared to be related, they had different non-PV human EV-C sequences in the P2–P3 region and the 3′-UTR (6.0%–20.0% nt sequence difference). The phylogenetic relationship of these isolates to the previous Madagascar VDPV2s was assessed: the 2011 isolates originated from 2 novel, independent events (Figure). The 2011 VDPVs had lost the major attenuating determinants in the 5′-UTR (guanine to adenine at nt 481) and VP1 region (isoleucine to threonine at codon 143) regions (8). This finding strongly suggests that the 2011 VDPVs had regained a degree of neurovirulence.
Figure
Phylogenetic trees showing genetic relationships between sequences of vaccine-derived poliovirus (VDPV) isolates. The trees are based on nucleotide sequence alignments of various subgenomic regions. Multiple sequence alignments were performed with CLC ...
Although there is currently no evidence of AFP cases linked to these new VDPVs, their detection in 3 children and their genetic characteristics strongly suggest insufficient OPV coverage in Toliara Province. We could obtain proof of vaccination (vaccination card) for only 27% of the participants. To prevent a third poliomyelitis outbreak, the Ministry of Health of Madagascar organized house-to-house OPV vaccination campaigns within Toliara Province in October and December 2011 and in January 2012. No polio cases have been reported in the province since then, suggesting that VDPV circulation was limited or stopped by these campaigns.
Massive OPV immunization campaigns worldwide have greatly decreased the frequency of poliomyelitis caused by wild-type PVs. However, after the disease has been eliminated in developing countries, low polio vaccine coverage frequently enables the emergence of VDPVs or the reintroduction of wild-type PV from disease-endemic countries; both scenarios threaten the success of the poliomyelitis eradication program (9,10). More than 640 polio cases caused by circulating VDPVs have been reported in 21 developing countries. We have shown that even in the absence of polio cases, potentially pathogenic circulating VDPVs can be detected in fecal samples collected from children living in areas with fluctuating and often low polio vaccine coverage. Thus, such sampling can help determine whether vaccine campaigns should be implemented in areas where polio could reemerge. In Madagascar, vaccination appears to have cleared emerging VDPVs and to have prevented polio cases in children.