Recent scientific findings, based on literature reviewed after the seventh update of the ECDC Rapid Risk Assessment on Zika virus infection (12 July 2016)

Recent scientific findings, based on literature reviewed after the seventh update of the ECDC Rapid Risk Assessment on Zika virus infection (12 July 2016)

This section presents relevant scientific literature and guidance on Zika virus research published between the rapid risk assessment of 12 July 2016 and 19 August 2016. This summary aims to outline the main findings with regard to Zika virus research.

Pathogenesis

Recent research findings provided new insights on the pathogenesis of neurological damage caused by the Zika virus through the destruction of progenitor cells.

A detailed investigation of a miscarriage associated with Zika virus infection showed the presence of the virus in amniotic fluid and in foetal and placental tissue [1]. Using specific immunohistochemical methods, researchers could detect Zika virus RNA in the placental amniotic epithelium and foetal mesenchymal cells. No brain tissue was available for testing but other foetal tissues tested negative for Zika virus. This demonstrates that Zika virus can replicate in vivo in pluripotent amniotic stem cells involved in embryogenesis. It also supports the viruses’ role in miscarriages reported in previous epidemiological reports [2,3].

In pregnant mice, contemporary Zika virus strains can provoke placental infection and injury, spread to the foetal brain and induce the destruction of neuronal progenitor cells. Cellular biology studies using human cells corroborate the studies in pregnant mice. Overall, results demonstrate that Zika virus is neurotropic and cytotoxic for several foetal neural cells and interferes with cerebral embryogenesis, causing microcephaly and other neurological abnormalities. These findings conceivably support the role of Zika virus in neurological disorders such as microcephaly and other congenital malformations that have been observed in human neonates [4].

Zika virus may also infect neural progenitors still present in adults [5]. In a preliminary study on mice sensitive to Zika virus (Interferon-3, -5, and -7 triple-knockout mice), infected adult mice showed evidence of cell death and reduced generation of new neurons in specific regions. A similar alteration of neural cell proliferation may be possible in individuals with weak immune systems [5].

Microcephaly and congenital central nervous system malformations

In March 2016, WHO reported that there was ‘strong scientific consensus that Zika virus is a cause of GBS [Guillain–Barré Syndrome], microcephaly and other neurological disorders’ [6].

Other data from scientific investigations add to the evidence that the emerging Zika virus strain from the Asian lineage can cause transplacental infection and congenital central nervous system malformations in the developing brain [7,8].

Severe congenital malformations, such as microcephaly, are severe forms of the Zika congenital syndrome, which might also include other congenital impairments and adverse outcomes of pregnancy. A recent retrospective study in Pernambuco state, Brazil, investigated seven infants with arthrogryposis associated with congenital infection, presumably by Zika virus [9]. Laboratory tests for toxoplasmosis, syphilis, rubella and infection with HIV and cytomegalovirus were negative. IgM serology for Zika virus in the cerebrospinal fluid was positive in two of the children (the five others were not tested). Microcephaly was found in six cases, but all cases presented calcifications, predominantly in the cortex and subcortical white matter, with abnormalities of cortical development and brainstem and cerebellar atrophy. The authors propose to add arthrogryposis as a possible manifestation of the congenital Zika syndrome.

Meaney-Delman et al. reported a longitudinal investigation of five Zika infection cases (four symptomatic and one asymptomatic) during pregnancy from the US Zika Pregnancy Registry [10]. One case had onset of Zika infection symptoms at week 12 of gestational age. A subsequent RT-PCR test for Zika virus in the amniotic fluid was positive (week 16), and the pregnancy was terminated at week 18. In the same study, three pregnant women (with symptomatic infection at weeks 18, 19 and 20 of pregnancy, respectively) experienced no adverse effects: two gave birth to apparently healthy children, while for the third pregnancy – still ongoing at the time the study was published – obstetrical monitoring of the foetus did not detect signs of congenital impairments. The fifth case, an asymptomatic case, was exposed to Zika virus between weeks 1 and 20 of gestational age. The Zika diagnosis was based on a positive RT-PCR test for Zika virus in both blood and urine at week 24. The infant was born apparently healthy. This case report is in line with previous findings which support that the first trimester of pregnancy is a high-risk period for severe congenital impairments.

The results of several other longitudinal studies with follow-ups of newborns who were exposed during the second trimester are not yet available [4].

Clinical features and sequelae

Zika is considered a mild infection, but there are case reports on potential Zika-related deaths. Three fatal cases, potentially Zika-associated, were described in Suriname [11]. For all three cases, Zika viral infection was confirmed by RT-PCR; chikungunya, dengue, HIV and leptospirosis could be ruled out for these three patients who also had co-morbidities. After unspecific symptoms, such as vomiting and watery diarrhoea without rash, patients rapidly deteriorated and went into shock, requiring intensive supportive care. A post-mortem analysis could not be conducted. It remains to be determined whether the direct involvement of Zika virus caused acute myocarditis and triggered a multi-organ failure after bacterial co-infections.

Previous reports from Colombia of four fatal cases with positive RT-PCR for Zika virus after atypical presentation and acute clinical deterioration present similarities with the cases reported in Suriname [12]. Although further comprehensive investigations are needed, clinicians should be aware of atypical presentations of Zika virus infection.

Several news media reported a case of sensorimotor polyneuropathy potentially associated with Zika virus infection in a 62-year-old man from Tegucigalpa, Honduras [13].

Several co-infections with other arboviruses such as dengue and/or chikungunya were also reported, all without a significant increase in morbidity [14,15].

Epidemiology

An up-to-date review about the epidemiology, natural history, and perspective on the public health impacts of ZIKV infection was published by Lessler et al. [16].

In France, a descriptive analysis of Zika virus cases from first January to 15 July 2016 reported 617 travel-related cases [17]. In addition, eight cases were associated with infected travellers who passed on the virus to their sexual partners.

After their return to mainland France, 185 of the 617 travel-related cases stayed in a ‘receptive area’, i.e. an area with an established Aedes albopictus mosquito population and high mosquito activity between May and November; 84% of these cases were viraemic, which could – according to the authors – lead to local vector-borne transmission unless appropriate vector control measures are taken.

A modelling study estimated the basic reproduction number (R0) of Zika virus for areas in the EU with Aedes mosquito populations (Ae. aegypti in Madeira, Portugal, and Ae. albopictus in continental Europe) [42]. Using a temperature-driven vectorial capacity model and assuming that the European population of Ae. albopictus has competence for the transmission of Zika virus in natural settings, potential areas for autochthonous transmission included Italy, southern France, the southern and eastern coasts of Spain, the western regions of the Balkans, and southern and northern Greece. Potential transmission rates were predicted to follow a marked seasonal pattern, with an increase in July (R0 values mainly ranging from 2 to 3), reaching a peak in August (R0 values of 3 to 4); R0 values would decrease in September and fall below one in October.

Transmission

Physical properties and inactivation

Environmental studies showed that dried Zika virus droplets remained infectious for more than three days [18]. The virus loses its infectivity at ≥60 °C and is destroyed by classical disinfectants and inactivation methods.

Mosquito transmission

Jupille et al. published the results of a study on vector competence of several European Aedes mosquitoes which was mentioned in an earlier ECDC risk assessment ( Zika virus disease epidemic, sixth update – 20 May 2016 ) [19]. In this study, two populations of Ae. aegypti from the island of Madeira and two populations of Ae. albopictus from France (near Nice) were challenged with Zika virus Asian genotype (isolated in New Caledonia, 2014) [20]. Zika virus dissemination was higher for Ae. aegypti. The virus was detected earlier in the saliva of Ae. aegypti (on day 3) then in Ae. albopictus (day 9). Both mosquito species showed somewhat similar transmission efficiencies (around 5%) between days 9 and 14 post-infection. Transmission rates are comparable to those previously found by the same research team using Ae. aegypti from Brazil (Rio de Janeiro) and Ae. albopictus from the USA (Florida) [21]. In a study by Aliota et al., Ae. aegypti and Ae. albopictus from the US showed no significant difference in transmission rate (around 22%) at day 14 post-infection on a limited number of specimens [22].

Zika virus was isolated from several unengorged Ae. aegypti batches collected around households of suspected Zika virus patients in December 2015 in Chiapas state, Mexico [23]. The 95% confidence interval of the minimal infection rate was estimated between 52 and 172 per 1000 mosquitoes. In addition, none of the unengorged Culex quinquefasciatus mosquitoes collected at the same time were positive by RT-PCR.

Brazilian scientists claimed that they detected Zika virus in salivary glands of Cx. quinquefasciatus mosquitoes on the third day of artificial feeding [24]. These unpublished results need to be confirmed in order to assess the vectorial capacity of this species [25].

A recent study found that Cx. pipiens mosquitoes in the US were not capable of transmitting the Zika virus [22].

Transmission

Observed transmission modes include vector-borne transmission, intra-uterine vertical transmission, perinatal and sexual routes (see below). There is also a risk of transmission through blood transfusions [26,27]. Further investigations are still required to describe the viral kinetics in various bodily fluids in order to adapt prevention and control measures.

Transmission by sexual contact

The first female-to-male sexual transmission of Zika virus was reported in July 2016 [28]. Relevant observations with regard to sexual transmission of Zika virus can be outlined as follows:

• Sexual transmission of Zika virus has been reported from males with symptomatic infections to both female and male sexual partners and, in one instance, from female to male [28,29]. Zika virus might be transmitted to sex partners of infected women through exposure to vaginal secretions or menstrual blood [30].
• The longest interval reported so far between onset of symptoms in a man and symptoms in his female partner was 44 days, which corresponds to a sexual transmission occurring between 34 and 41 days after the index case [31].
• Freour et al. reported a probable case of asymptomatic sexual transmission (male to female) in two asymptomatic people which likely occurred between 21 to 36 days after returning from an area with ongoing Zika virus outbreak [32].
• Replicative Zika virus particles in semen were detected twice, at 21 and 24 days after onset of symptoms [33,34]. In several studies, the virus could not be isolated.
• Zika viral RNA in semen was detected in more than 10 instances between day 14 and day 100 after onset of symptoms (personal communication: Emma Aarons and Daniel Bailey, Public Health England, and [32-41]). In two recent publications, however, the range of detection of Zika viral RNA in semen was given at between 14 and 188 days after onset of symptoms [35,42].
• Studies with longitudinal follow-up laboratory testing of Zika viral RNA kinetics in semen are scarce [37,39,42]. Reusken et al. documented a progressive decrease of the viral load in semen down to undetectable levels at day 62 after infection [37]. A gradual decrease of Zika RNA viral load in two patients was also observed by Huits at al.: Zika virus was undetectable after 56 and 68 days post-onset of symptoms, respectively. Tests for Zika virus RNA detection in semen were negative for two other patients during follow-up [39]. Another case report showed Zika virus RNA in semen in samples taken at days 91, 134 and 188 after onset of symptoms [42]. Conducted by researchers of the University of Padua, Italy, an in-depth longitudinal follow-up of a 40-year-old man with 40 samples (taken between day 5 and 181 after onset of symptoms) demonstrated a decreasing but persistent shedding of viral nucleic acid in semen associated with the cellular component of semen; the Zika virus RNA was undetectable in seminal plasma up to at least 181 days [43].
• The first case of female-to-male transmission was reported in New York in July 2016 [28]. A woman reported a single event of condomless vaginal intercourse with a male partner the day she returned from travel to an area with ongoing Zika virus transmission [28]. The sexual intercourse took place the day before typical Zika virus infection symptoms appeared, which supports the assumption that the woman was viraemic. The man confirmed that he had not travelled outside the United States during the year before his illness, had not engaged in oral or anal intercourse with his partner, that he had no other recent sexual partners, and reported no mosquito bites within the week preceding his illness. The event supports female-to-male Zika virus transmission through unprotected sexual intercourse.
• Recent studies of nonhuman primates found Zika virus RNA in the vaginal fluid and cervicovaginal secretions after laboratory Zika virus infection [44,45]. A recent report describes the detection of Zika virus RNA from a woman’s cervical mucus, genital swabs, and endocervical swabs at day 3 after symptoms and up to 11 days after onset of symptoms in cervical mucus [46], which substantiates the possibility of female-to-male Zika virus transmission.
• Isolation of virus from saliva was reported twice: on day 6 post-onset of illness in one case, and during the acute phase of illness symptoms in another case (sampling date after onset of symptoms not provided) [43,47].

Zika virus genome has been detected in saliva on numerous instances during and following the acute phase of the disease. Two recent observations reported the prolonged detection of Zika virus RNA in saliva, along with high titres of Zika-neutralising antibodies at days 47 and 91 after onset of symptoms [42,43].

• At population level, the results of two modelling studies estimating the reproduction number of Zika virus support both mosquito-control and personal biting protection as well as safe sexual practices to reduce Zika virus incidence during outbreaks [48,49]. Both studies highlighted that sexual transmission can both increase the risk of infection and the size of an epidemic, but that sexual transmission alone cannot cause an outbreak.

Diagnostics

WHO regularly publishes a list of in vitro diagnostic assays (NAT, rapid diagnostic tests and ELISA) for Zika virus accepted for procurement through Emergency Use Assessment and Listing procedures [50].

Table 1. Time of detection of Zika virus in human samples from symptomatic cases (pregnant women excluded)

NA = not available. Details on sample preparation and laboratory methods for detection available in original publications.

Zika virus RNA was detected in asymptomatic individuals on four occasions: i) in a person returning from Venezuela to China among a family cluster of symptomatic cases (positive sample of serum, urine, and saliva) [52], and ii) in three individuals during a follow-up of a military community in Suriname exposed to Zika virus over a two-week period (positive urine samples in two cases and positive blood sample in one case) [54].

In non-pregnant women, Zika virus viraemia usually lasts less than a week [61]; in connection with the investigation of five pregnant women enrolled in the US Zika Pregnancy Registry, prolonged duration of detection of Zika virus RNA in serum was reported [10]. Prolonged detection of Zika virus RNA by RT-PCR in serum was reported up to 46 days after onset of symptoms for the symptomatic case and up to 53 days after exposure for asymptomatic pregnant women. Foetal abnormalities and evidence of Zika infection (through histopathology and molecular diagnostics) were found in one of the symptomatic pregnancies, which was then terminated at 18 weeks of gestational age.

These new results corroborate previous case reports of a foetal Zika virus infection where Zika virus RNA in the serum was positive at 15 days [62], 21 days [1], 4 to 10 weeks after onset of symptoms [63], and two months after the acute phase of the disease [64].

Preliminary results from the Colombian pregnant women register showed that 326 of 1 850 pregnant women with complete results were positive for Zika virus RNA in serum, and that initial positive results of Zika RT-PCR mostly ranged from 0 to 15 days after onset of symptoms [65]. Results of longitudinal follow-ups have not yet been published. The underlying mechanism of prolonged presence of viral RNA in the serum of pregnant women could be related to delayed maternal immune response or foetal infection, but the clinical significance and prevalence of prolonged viraemia needs to be explored further.

Prevention and vaccine development

A review of safety, efficacy, and effectiveness of personal protective measures for preventing Aedes spp. mosquito bites and/or related arbovirus infections was submitted to the Bulletin of the World Health Organization ( Zika open repository). The review provides an updated overview of safe and potentially effective measures for preventing Zika virus infection during pregnancy (insecticide-treated nets, mosquito repellents, etc.) [66].

Dawes et al. conducted an in-depth review on vaccine development for Zika virus based on vaccine development strategies used for other flavivirus vaccines [67]. Dowd et al. showed that infection with a single Zika virus strain elicited broadly neutralising antibodies for South American, Asian, and early African Zika virus strains. These results provide additional evidence that the choice of the Zika virus strain for vaccine or for serological diagnostics development may not be a critical parameter [68].

Recent studies have shown the protective efficacy of multiple vaccine platforms against Zika virus in rhesus monkeys and support the possibility of a rapid clinical development of Zika virus vaccine for humans [45].

Phase 1 trials of two Zika virus DNA vaccine candidates have begun in the USA and Canada to evaluate safety, tolerability and immunogenicity in healthy adult volunteers [69].

Vector control

Several new approaches to control Ae. aegypti mosquitoes are currently undergoing field trials. An overview of new tools to decimate Ae. aegypti mosquito populations – namely Wolbachia-infected mosquitoes and genetically modified male mosquitoes – was published in the Bulletin of the World Health Organisation [70,71]. A recent study showed that Wolbachia-infected Ae. aegypti mosquitoes showed a reduced vector competence for Zika virus [70]. In Brazil, Wolbachia-infected mosquitoes were released in early August 2016. Field trials with genetically modified mosquitoes started in April 2015 in Piracicaba, a city located in the Brazilian state of São Paulo [70].