Rapid Risk Assessment: the risk of Zika virus to Canadians (first update)

Last reviewed: July 22 2016
Last updated: July 22 2016

What’s New?

• There are no major updates in this version.
• Several isolated instances of unusual transmission have now been documented; an asymptomatic sexual transmission (likely male-to-female), a likely female-to-male transmission, and a person-to-person transmission without sexual contact. All of these are thought to be rare modes of transmission, requiring particular circumstances to be realized, but these events demonstrate that they are possible.

Summary

• An epidemic of mosquito-borne Zika virus (ZIKV) is occurring in many parts of the Americas and several other tropical/sub-tropical areas. The virus causes no or relatively mild illness in most adults. However, it also can cause more severe outcomes like congenital abnormalities in the fetus and Guillain Barré Syndrome (GBS) in infected people.
• Risk of establishment of ZIKV in Canada is considered negligible, because the species of mosquitoes associated with transmission are not found here (Very Low likelihood, with high confidence).
• More than 160 travel-related ZIKV infections have been reported among Canadian travellers (of several million travellers to affected regions, and several thousand tested samples). For the individual, it is estimated that travel to an affected area is associated with a medium chance of infection with ZIKV (Medium likelihood, with high confidence).
• Sexual transmission, usually from symptomatic male travellers to a sexual partner (female or male) who has not travelled, has been reported. As relatively few travellers will be infected with ZIKV, the likelihood of transmission via this route is low (medium confidence). This changes if a man does become infected and is symptomatic with ZIKV, the likelihood of transmission to his sexual partner is assessed as Medium (low confidence).
• For most infected travellers, ZIKV will have little or no health impact (Low impact, with medium confidence). However, severe outcomes (e.g., GBS) might occur in some affected individuals (High impact, high confidence).
• Based on recent evidence, we assess that there could be Very High impact (with high confidence) to the unborn children of women who become infected with ZIKV while pregnant.
• Canadian recommendations for the prevention and management of ZIKV-disease have been developed by the Committee to Advise on Tropical Medicine and Travel.

1. Date of initial assessment
   28 January 2016
2. Date of last update
   July 22 2016
3. Risk identification
   Individual and public health risk to Canadians associated with the 2015/16 ZIKV outbreak in the Americas.
4. Prepared by
   Public Health Agency of Canada, with expert consultation from the Department of National Defence, Force Health Protection.
5. Key information sources
   See Reference list
6. Review & update
   This assessment will be revised as required, informed by a continuous review of outbreak data and the scientific literature. Triggers for a revision include substantial change in epidemiology, new scientific developments, or increased rates of severe outcomes.

Document information

In this Rapid Risk Assessment (RRA) document we assess the risk to Canadians posed by the ongoing Zika virus (ZIKV) epidemic. We separately consider the likelihood of infection with ZIKV and the impact of infection. We consider the likelihood of infection for (i) Canadians remaining in Canada (by assessing the likelihood of local mosquito-associated or sexual transmission), and (ii) Canadians travelling to countries with ongoing ZIKV activity. We then consider the impact of ZIKV infection. The terms “likelihood”, “impact”, and “confidence” are used in a specific sense, and are defined in Appendix 1.

Our assessments will change as more information becomes available, and this document will be updated accordingly.

Disease background information

Infectious agent

Zika virus disease is a mosquito-associated flaviviral disease caused by Zika virus (ZIKV). It is related to other Flaviviridae, including Japanese Encephalitis, West Nile, Yellow Fever, St. Louis Encephalitis, and Dengue viruses.

Transmission modes

The natural cycle of ZIKV involves mosquito vectors and vertebrate hosts, probably African primates^{Reference 1 Reference 2 Reference 3}. There is limited documentation of non-primate reservoirs of ZIKV; serological evidence has been demonstrated in rodents^{Reference 4}, but their role (if any) in transmission is unknown. Once infected, a mosquito is thought to remain so for their lifetime^{Reference 5}. In the current outbreak, the vertebrate hosts are humans. Aside from mosquitos, blood transfusion-associated^{Reference 6 Reference 7 Reference 8} and sexual transmission have been documented. The latter includes several instances where symptomatic male travellers infected their partner (female or male), who had not travelled^{Reference 9 Reference 10 Reference 11 Reference 12}, including a Canadian ^{Reference 13}.

A significant concern with the current ZIKV outbreak is the potential for vertical transmission from mother to infant which can cause microcephaly and other congenital abnormalities ^{Reference 14}.

Vector

Aedes aegypti is the primary vector of ZIKV (as well as Dengue and Yellow fever virus). This species is largely restricted to tropical and subtropical regions, though temperate populations may occur in isolated refuges^{Reference 15}. Ae. albopictus has been implicated as a vector, though its role in the current outbreak is uncertain. This species is widely distributed outside the tropics^{Reference 16}, but is not known to be established in Canada^{Reference a}.

Epidemiology

Although it has long been known that ZIKV can cause human disease ^{Reference 3 Reference 17}, it was perceived as a low-impact and obscure disease. However, in 2007, the first major outbreak outside of these areas was reported on the island of Yap (Micronesia) in the southwestern Pacific Ocean ^{Reference 18}. Between 2013 and 2015, additional outbreaks occurred on islands and archipelagos in the Pacific region including a large outbreak in French Polynesia^{Reference 19 Reference 20}. More recently, the virus has caused widespread outbreaks across Central and South America, Mexico, and the Caribbean, with some activity also occurring in Asia, , Africa (Cape Verde), and the South Pacific.

Clinical presentation

Asymptomatic infections appear to be the norm, with only one in four or five people (20-25%) infected with ZIKV developing clinical symptoms ^{Reference 25 Reference 29}. Disease is generally relatively mild, with symptoms that include: low-grade fever (usually <38.5°C), transient arthritis/arthralgia with possible joint swelling mainly in the smaller joints of the hands and feet, maculo-papular rash often spreading from the face to the body, conjunctival hyperaemia or bilateral non-purulent conjunctivitis, and general non-specific symptoms such as myalgia, asthenia, and headaches ^{Reference 25}. Less commonly, infection with ZIKV is now known as a cause of serious complications such as Guillain Barré Syndrome (GBS)^{Reference b Reference 22 Reference 23 Reference 24} and other neurologic complications^{Reference 25 Reference 26}. There have been a handful of reports of ZIKV-associated deaths in adults^{Reference 27}.

The incubation period (i.e. the amount of time between exposure and symptom onset) is thought to range from 3 to 12 days ^{Reference 28}. The disease symptoms are usually mild and last for 2 to 7 days. Most people recover fully without severe complications, and hospitalization rates are low ^{Reference 18}. Viremia (the period when ZIKV is measurably present in the blood) has been estimated as lasting 3 to 5 days following symptom onset ^{Reference 29 Reference 30}, although viral RNA has been detected in saliva ^{Reference 31} or urine ^{Reference 32} more than a week after clearance from the blood. Infection may go unrecognized or be misdiagnosed as dengue, chikungunya, or other viral infections causing fever and rash.

Event background information

Between 2013 and 2015, several significant ZIKV outbreaks occurred on islands and archipelagos from the Pacific region including a large outbreak in French Polynesia ^{Reference 19 Reference 20}. In 2015, ZIKV emerged in South America with widespread outbreaks reported in Brazil and Colombia. Mosquito-borne cases have been reported to WHO by many countries, mostly in the Americas. Information related to the ZIKV outbreak continues to evolve. For a summary of the most up-to-date information related to the ZIKV outbreak, readers are referred to the World Health Organization situation report.

Risk assessment

In this section, we consider i) the likelihood of infection with ZIKV for Canadians who have not travelled to an area where ZIKV is actively circulating, ii) the potential for local transmission (i.e., in Canada) of ZIKV, and iii) the likelihood of infection with ZIKV for travelling Canadians. We then separately consider the impact for individuals who become infected, first in the general case and then specifically for the unborn child of a woman infected while pregnant.

Likelihood of infection with ZIKV via blood and sexual transmission for Canadians who have not travelled

Human-to-human transmission of ZIKV can occur through several routes, including sexually^{Reference 11} and via contaminated blood^{Reference 7}. Individual Canadians who do not travel to areas of ongoing ZIKV outbreaks could be exposed to ZIKV in Canada via these routes.

Because ZIKV does not persist for long in the blood (typically 3-5 days following symptom onset)^{Reference 29 Reference 30}, there is a narrow window for blood-based transmission. Given that Canadian blood safety agencies have enacted a 21-day donor deferral period following travel to ZIKV outbreak countries, the likelihood of transmission via this route is negligible (Very Low Likelihood with high confidence).

Several cases of sexual transmission from a symptomatic male to his sexual partner have been documented^{Reference 11}, including in a Canadian traveller. Male-to-male transmission has also occurred^{Reference 12}. It is thought that viable virus might persist in semen for at least several weeks after it has been cleared from the blood^{Reference 7 Reference 9 Reference 10} .

Although rare, cases of sexual transmission not involving a symptomatic male have also been observed. One instance of likely asymptomatic transmission has been reported ^{Reference 35}, in which both partners travelled to a ZIKV risk area but the timing of infection is more compatible with sexual transmission from the (asymptomatic) male to his female partner than with mosquito-acquired infection. Additionally, a case likely to be female-to-male transmission^{Reference 69} has been reported. An unprotected sexual episode between a returned US female traveller and her male partner (who did not travel) seems to have coincided with peak viremia in the female (i.e., immediately preceding symptom onset), leading the male to develop symptomatic ZIKV illness.

In addition, a case of likely person-to-person transmission without sexual contact has been reported. This seems to be an isolated event involving (i) very close, regular contact with (ii) an infected person who had developed an unusually high viral load. An ongoing investigation by US CDC should provide additional information on this unusual case.

We expect that transmission from symptomatic males to their partners (i.e. via semen) will remain the dominant form of sexual transmission, but note that these examples demonstrate the possibility of rare, alternate transmission pathways under certain circumstances.

In summary, because the absolute probability of infection with ZIKV is considered low^{Reference c}, the likelihood of transmission via any form of sexual contact must be lower still (Low likelihood, medium confidence). However, for a symptomatic man infected with ZIKV, the likelihood of transmission to his sexual partner is assessed as Medium (low confidence).

Likelihood of local transmission in Canada

Local transmission would require the presence of a species of mosquito that can be infected with and transmit to human hosts ZIKV. Ae. aegypti is thought to be the principal vector in the current outbreak, with Ae. albopictus possibly playing a role. The distribution of Ae. aegypti is largely limited to tropical and subtropical areas and, while Ae. albopictus can occur in temperate regions, it also is limited by climate^{Reference 16 Reference 36 Reference 37}.

The likelihood that Ae. aegypti will become established in any area of Canada under current climatic conditions is estimated as Very Low (high confidence). This assessment is supported by a recent study^{Reference d Reference 23}.

A potential secondary vector, Ae. albopictus, occurs in the United States, including the southern parts of some Eastern and upper Midwest states that border Canada. The likelihood that this species will become established in any area of Canada under current climatic conditions is considered Low (medium confidence^{Reference e}). This assessment is based on recent work ^{Reference 23 Reference 39 Reference 40} that estimated the current probability of this species occurring in most of Canada as at, or approaching, zero. One study did report that small areas in southern Ontario, southern Nova Scotia, as well as southern coastal British Columbia had low to moderate climatic suitability for this species^{Reference 40}. Additionally, a recent report from the National Public Health Institute of Quebec reached a similar conclusion, identifying a low probability of potential establishment of Ae. albopictus in Southern Quebec under some future scenarios .

In summary, self-sustaining populations of the mosquitoes associated with the current ZIKV outbreak are not thought to occur in Canada. Based on this assumption, (the absence of Ae. aegypti or absence of Ae. albopictus from all but a very few limited locations in Canada), there is Very Low likelihood (with high confidence) of epidemic spread and subsequent endemic establishment of ZIKV in Canada.

It is not known if mosquitoes that occur in Canada are capable of transmitting ZIKV nor, if they are, whether they could do so in their natural environment. However, the absence of reports of autochthonous transmission of ZIKV or other arboviruses associated with warm climates (e.g., dengue viruses, chikungunya virus) in Canada or similar climatic regions suggests that, even if native species could transmit ZIKV, they are very unlikely to support anything other than sporadic transmission (Very Low likelihood, high confidence).

Assumptions for assessments of likelihood of local transmission

We have made a number of assumptions in assessing the likelihood of local transmission:

1. Populations of Ae. aegypti, are not established in Canada. This assumption is supported by a robust assessment of the distribution of this vector species^{Reference 16}.
2. Populations of Ae. albopictus are not established in Canada, or occur with a geographic distribution too limited to support sustained endemic transmission of ZIKV. This assumption is supported by robust assessments of the distribution^{Reference 17} and climatic suitability of Canada^{Reference 37} for this species as well as its absence (except for rare individual specimens in southeastern Canada) in recent mosquito surveillance for West Nile virus in Canada.
3. Other mosquito species occurring in Canada are not capable of maintaining cycles of Zika virus transmission. This assumption is based upon the absence of records of ZIKV from northern temperate regions of the world ^{Reference 38 Reference 39 Reference 40}.
4. There are temperature limitations to transmission of ZIKV (e.g., effects of temperature on the extrinsic incubation period in mosquitoes) that limit endemic transmission to tropical and sub-tropical areas. This assumption is based on similar limits on transmission of dengue^{Reference 38 Reference 39 Reference 40 Reference 41 Reference 42 Reference 43}.
5. There are no alternative (i.e. non-mosquito-borne) routes of ZIKV transmission, including sexual transmission^{Reference f}, capable of maintaining epidemic or endemic transmission cycles in Canada.

Likelihood of ZIKV infection for travellers

At the time of writing, many countries have reported recent, mosquito-borne ZIKV cases. Continued spread of ZIKV is expected^{Reference 41} with the result that more areas will present a risk to Canadian travellers. To date,of the approximately 7000 individuals tested by the Canadian health care sector, over 160 have yielded positive results for ZIKV infection. Thus, among Canadians who have travelled to countries with active ZIKV transmission, the likelihood of infection is Medium ^{Reference g} (with high confidence). In absolute terms, the true prevalence of ZIKV infection among Canadian travellers is likely to be substantially lower than the estimated 2% prevalence among individuals who have been tested to date.

Factors that could affect the likelihood or impact of infection

The impact of traveller- and/or itinerary-specific factors on the likelihood or impact of ZIKV infection has not been well described. However, there are a number of plausible relationships:

• Conditions at higher elevations (≥ 2,000 m) are generally not supportive of viral replication in, or survival of, Ae. aegypti populations. Correspondingly, the relative likelihood of infection with ZIKV might be substantially lower for travellers (depending on how much time they spend at altitude compared to lower elevations) to such areas.
• All else held equal, the likelihood of infection is higher in countries/areas that are reporting high levels of ZIKV activity compared to those that are not.
• The likelihood of infection is likely lower for shorter travel durations and/or when staying in protected environments (e.g., well screened and air-conditioned accommodations, transiting through an airport in a risk area, staying in an area where robust mosquito abatement practices are in use). This might also apply to situations where the traveller is staying in an isolated location, i.e., where there are relatively few residents who might support sustained transmission.
• Travellers with co-morbidities (often, these are associated with age) might be at increased risk for more serious ZIKV-associated outcomes.

Impact of ZIKV infection

For most travellers, ZIKV will have little or no health impact (Low impact, with medium confidence). This is because infection will be relatively infrequent, and, of those infected, only about 20-25% will develop illness, most of it mild^{Reference 18Reference 21} .

However, infection with ZIKV is a cause of rare but severe outcomes such as GBS^{Reference 44}. Indeed, cases of GBS and other neurological syndromes^{Reference 25Reference 26} have been reported in many countries or territories affected by the current outbreak^{Reference 24}. Perhaps the most compelling evidence related to infection with ZIKV and GBS comes from a retrospective case-control study done with patients from French Polynesia ZIKV outbreak during 2013-2104^{Reference 23}. Among the 42 GBS cases, all but one (98%) showed serologic evidence (anti-ZIKV IgM and IgG) of ZIKV infection, compared to 35 of 98 (36%) matched controls (Odds Ratio 59.7, 95% Confidence Interval 10.4 to ∞). Based on an estimated ZIKV attack rate of 66% among the population of French Polynesia the authors estimate the proportion of post-ZIKV infection GBS as approximately 1 GBS case per 4,000 ZIKV infections. This estimate would put ZIKV among the higher-risk pathogens of those known to cause GBS, equal to or less than the risk posed by Campylobacter jejuni (1 to 2.6 GBS cases per 4,000 infections^{Reference 45}), and less than the risk posed by cytomegalovirus (2.4 to 8.8 GBS cases per 4,000 infections^{Reference 46}).

In summary, for most infected travellers, ZIKV will have little or no health impact (Low impact, with medium confidence). Severe outcomes (e.g., GBS) might occur in some affected individuals (High impact, high confidence).

Impact of infection on unborn children of women who acquire ZIKV

There is scientific consensus that ZIKV is a cause of infant microcephaly or other severe brain anomalies in children born to mothers infected during pregnancy^{Reference 14Reference 24} . Evidence supporting this conclusion is well described in a recent publication^{Reference 14}, and thus we estimate a Very High impact (with high confidence) on the unborn child of a woman who acquires ZIKV while pregnant. However, the likelihood that prenatal ZIKV infection will cause fetal harms remains uncertain. There is some evidence that this might not be an infrequent occurrence. For example, during the French Polynesia outbreak, it was estimated that congenital abnormalities occurred in the developing fetus of approximately 1% of the time in mothers infected during the first trimester^{Reference 47}. Another study, using data from Habia, Brazil, estimated the probability of microcephaly following first trimester infection from 0.88% (95% Credible Interval 0.8 to 0.97) to 13.2% (95% Credible Interval 12.0 to 14.4), depending on assumptions of true ZIKV infection rate and the degree of microcephaly over reporting^{Reference 48}.

In conclusion, ZIKV is a cause of infant microcephaly or other severe brain anomalies. Substantial effort continues to be directed towards clarifying the effect of infection timing on fetal development, and determining if there are other contributing risk factors (e.g., nutritional status, or other environmental factors)^{Reference 49}.

Biosafety information for laboratory workers

ZIKV is classified as a Risk Group 2 human pathogen, since this virus presents a moderate individual risk and a low community risk. Laboratory work involving ZIKV requires Containment Level 2 physical and operational requirements as outlined in the Canadian Biosafety Standard. The impact on the unborn children of pregnant women working in the laboratory with ZIKV could be Very High (with high confidence; see Impact section), if she became infected.

The main risks for laboratory acquired infection include accidental auto-inoculation and exposure to specimens from viremic patients, experimentally-infected animals, or ZIKV cultures. Local risk assessments should take into consideration the activities taking place in the laboratory to determine if risk reduction strategies in place are sufficient. Pregnant women or women trying to get pregnant should be made aware of the potential risks for the fetus and take extra care when working in a laboratory where ZIKV is propagated or handled.

Conclusions and Risk mitigation

In summary, we conclude that the overall risk of infection to Canadians (while in Canada) is very low. This is driven by the Very Low likelihood of exposure in Canada through mosquito-based transmission. An exception is transmission from symptomatic males to their sexual partner (Medium likelihood, with low confidence). The overall likelihood of infection for a Canadian travelling to an outbreak area is Low (medium confidence), and the impact for most of those infected will be Low (medium confidence). However, for some of those who are infected, impact may be High (high confidence) specifically because of the association with severe outcomes such as GBS. Although pregnant women or those who may soon become pregnant are assumed to be equally likely to be infected as are other travellers, the impact of such infection could be, for the unborn child, Very High (high confidence).

Persons considering travel should consult the Agency’s Travel Health Notice on ZIKV, and Canadian recommendations for the prevention and management of ZIKV-disease developed by the Committee to Advise on Tropical Medicine and Travel.

What can be anticipated in the next 6 months

The Agency continues to closely monitor the situation. Changes in scientific information are continually monitored, and this review will be updated on an as-needed basis.

Footnotes

References

Appendix 1 – Definition of technical terms