Prevalence of Zika virus in blood donations: a systematic review and meta-analysis

As a zoonotic pathogens and reemerging arbovirus, Zika virus (ZIKV) has caught extensive attention since its outbreak in Brazil 2015 [1]. ZIKV belongs to the genus Flavivirus, family Flaviviridae. It shares many biological and molecular characteristics in symptoms, genome and pathogenicity with other family members such as dengue virus (DENV), yellow fever virus (YFV), West Nile virus (WNV) and Japanese encephalitis virus (JEV) [2]. Approximately 80% of Zika infections are asymptomatic [3, 4], the remainder usually develop non-specific clinical symptoms such as mild fever, rash, arthralgia, myalgia, conjunctivitis and headache [5]. Nevertheless, sometimes ZIKV infection can lead to much severer neurological complications, such as Guillain-Barre Syndrome in adults [6, 7] and microcephaly in neonates [8].

ZIKV is transmitted to humans mainly through Aedes mosquitoes bite, but it can also be spread by sexual activities, maternal-fetal pathway, physical contact and blood transfusion [3, 9]. ZIKV RNA has been detected in blood, semen, saliva, urine and other biofluids samples, but the viremia period of ZIKV is not definite. It was estimated that the median incubation period for the infection was 5.9 days and the virus was detectable in blood for 9.9 days [10]. Notably, ZIKV RNA can persist in serum up to 54 days in some cases [11]. So, asymptomatic infectors in viremia period can be a nonnegligible source of ZIKV transmission and largely increase the risk that blood donations were already contaminated before collection.

WHO [12] and FDA [13] have instituted multiple interventions to reduce the risk of transfusion-transmission (TT) ZIKV, such as deferral of blood collection in endemic areas and donors returned from Zika-endemic countries within 28 days; import of blood from low-risk regions; screening blood donations by ZIKV NAT (Nucleic Acid Tests) and/or treating them with PRT (Pathogen Reduction/inactivation Technology). On March 30, 2016, the first ZIKV NAT assay for blood screening was permitted by FDA for emergency use, enabling the blood collection in Puerto Rico resumed [14]. However, there are opposite attitudes towards mandating routine screening of ZIKV in blood collections because it is time-consuming and costly. Some consider that the TT risk of Zika in their country is low [15, 16], and there is no urgent need to introduce universal screening of donated blood for ZIKV [17]. On the other hand, some studies mentioned that present measures were not effective enough to prevent transfusion of ZIKV RNA-reactive blood products, and ZIKV NAT should be used [18, 19]. Their attitudes varied greatly depend on the prevalence rate of ZIKV considered as low or high by different public health officials in different areas, but the cut-off value hasn’t been revealed. ZIKV TT risk modeling is not yet available to date due to the absence of critical parameters [20] and US FDA Zika guidance remained to be evaluated by a formal risk assessment and stakeholder consultation [21]. Many factors can influence the interventions to reducing TT risk of ZIKV, including social, economic and ethical factors, but the foremost factor is the infection rate of ZIKV in blood components, which should be a topic for further discussions. In this study, we performed the first systematic review and meta-analysis to assess the prevalence of ZIKV in blood donors or blood donations and find out potential risk factors. The result will provide valuable reference information for strategy making to ensure blood supply safety.

Flavivirus transmission through blood transfusion has been recorded for Dengue virus [33], West Nile virus [34], Yellow Fever vaccine virus [35] and Zika virus [9, 36]. The major challenges in preventing transfusion-associated transmission of ZIKV are the high rate of asymptomatic infections and the high proportion of infected people in endemic-areas [37]. Asymptomatic infectors have long been a challenging aspect in the control and prevention of infectious disease, for their ability to spread virus and high mobility without surveillance. Asymptomatic blood donors during viremia period can pose a great threat to blood supply system with viremia of Zika virus reaching to a high level exceeding 10⁸copies/mL [27]. Therefore, an extensive investigation of ZIKV prevalence in blood donations is quite essential to provide information for interventions implemented to ensure the safety of blood transfusion.

Through this meta-analysis, we concluded the overall prevalence of ZIKV in blood donations was 1.02%, which is higher than DENV RNA (around 0.19% [38]~ 0.4% [39]) and CHIKV RNA (0.36–0.42% [40]) detected in blood donors, and much higher than the rates of window period infections detected by NAT screening for HIV, HCV and HBV over the past two decades [41]. AABB (formerly the American Association of Blood Banks) have categorized dengue as high priority agents in terms of threats to US transfusion recipient safety [42]. Zika virus should also be classified as high-risk agent according to the criteria and even severer than dengue for its possible permanent neurological damage. Routine nucleic acid testing for WNV was performed for all US donations in 2003 and prevented thousands of WNV transfusion-transmissions, more than 27 million donations were tested during transmission periods with 1576 WNV RNA-positive donations identified [43], which is much lower than the rate of ZIKV prevalence in blood donations we summarized here. The virus load of ZIKV in blood donations in included studies ranged from 10²~10⁷ copies/ml, even though the dose required to cause infection in a recipient is unknown due to few recipients were investigated and the difficulty to confirm a TT case excluding vectorial transmission, the virus load is much higher than WNV screened in blood donations [43]. Thus, since routine blood screening test was already implemented for WNV, ZIKV should also be taken into consideration, especially in ZIKV-endemic areas.

The prevalence of anti-ZIKV antibodies (1.61%) in blood donation was almost twice as much as ZIKV RNA prevalence (0.85%), this may due to antibodies can persist longer in blood than nucleic acid, thus are more likely to be detected. But the existence of antibodies doesn’t mean the blood is contagious, for IgG antibody only indicate remote infections while IgM represent recent infections. On the contrary, a mere absence of ZIKV nucleic acid in blood samples does not absolutely rule out the existence of virus. So, both ZIKV RNA and antibodies in blood donations implied risks in blood supply system. Here we calculated the overall pooled prevalence of ZIKV including both ZIKV RNA and antibody, and then computed separately within each method. The 3 included antibodies-detected studies all tested for anti-ZIKV IgG antibodies, only 1 study (p5) tested both IgM and IgG and merely 3 cases were reactive with IgM in all included samples. Additionally, antibody detection assay may be compromised by cross-reacting antibodies and lead to false-positive results [44]. So, antibody-test may not be an appropriate screening assay for blood inspection. But there was no statistically significant difference between two detection methods, which may be due to the limited amount of studies in antibody detection. The prevalence of ZIKV in blood donations during epidemic period (1.37%) was more than two times higher than in non-epidemic period (0.61%). Blood donations collected during non-epidemic period were likely get infected because of the donors may have travel history to ZIKV endemic areas, and get infected without notice before donation. Another possibility that ZIKV has already transmitted into the study areas where are outside of the active-transmission regions or at risk of virus invading. There are four studies conducted in non-epidemic period, and three of them were detected ZIKV antibodies. So, it’s also a possibility that false-positive results were made because of antibody cross-reacting with other flavivirus. The result shows that South America has the highest prevalence of ZIKV in blood donations (2.17%), which is consistent with the geographic distribution of Zika infection [45, 46] and Brazil contributed most of the cases. Africa has a high prevalence of anti-Zika IgG (4.89%), which can be attributed to the sporadic circulation history of Zika in Africa [45].

Because of the high volume of blood needed and difficulties in blood transportation to remote areas, as well as quickly-spread and large-scale of Zika virus epidemics, it is not practical to import blood from low-risk regions to supply most endemic areas. So, safe and cost-effective technology was urgently needed to guard blood supply safety. The existing blood donor-screening ZIKV NAT assays demonstrate excellent sensitivity [47] and effectiveness. The risk of ZIKV TT infection can be reduced greatly of 29% by antibody screening and 7% by symptom-based screening [10]. Another important measure to ensure blood safety is PRT, but the licensed methods are only available for plasma and platelets now. Fortunately, many studies have shown that amustaline/glutathione treatment for red blood cells can effectively inactivate Zika [48], Dengue [49] and Chikungunya virus [50], and other methods aimed at red blood cells were under clinical trials and promising to be applied in the future [51]. Cost of screening and PRT assays is another crucial concern before carrying out a policy. Although the screening assays are high-cost, the adoption of a new system could be effective for many viruses and thus eliminating other existing tests along with their costs, so that the total costs would not be increased greatly. Blood donations NAT screening and PRT were strongly suggested in Zika virus-endemic areas. Given 39% of asymptomatic donors showed symptoms postdonation, it will be helpful to develop follow-ups after blood collection. Areas with no capacity to provide screening assays to all blood donations and non-endemic areas should strengthen general inspection measures, such as detailed questionnaire survey before donation and screening tests should focus on at-risk recipients, such as pregnant women.

There are spaces for improvements to this review: Firstly, the amount of eligible studies was not enough for deep analysis of subgroups, for some subgroups contained only 1 study and some are null, which could reduce the accuracy of results. Secondly, significant high heterogeneity and publication bias were observed. Several factors might contribute to the existence of heterogeneity, for example, socioeconomic status, sanitary conditions and most importantly, the method used to detect ZIKV infection and the diagnostic cutoffs adopted. In addition, the heterogeneity of single-rate meta-analysis is usually higher than two-group study, such as case-control study and RCTs, because the data were extracted only from one group. It is common to see that I² statistic often exceed 90% in some other meta-analysis studies [52, 53]. So I² statistic should not be the only standard to judge the reliability of a study. Based on the precisely sampling methods and detailed subgroup analyses, we believe that the pooled prevalence of ZIKV in blood donations calculated in this study are of high reference value. Thirdly, those studies that detected antibodies didn’t confirm the result with RNA tests, this caused much uncertainty in the interpretation of the result. Finally, it would be better if detailed information was provided in included studies, such as the demographic characteristics of blood donors or the types of blood products (plasma, platelet, red cell and whole blood). With attention on ZIKV increases, more research data are expected in the future to provide more insights for intensive research.

Based on this meta-analysis, the prevalence of Zika infection in blood donations is at high-risk level of 1.02%, the prevalence varied greatly in different areas. NAT blood screening tests and PRT are recommended in Zika virus-endemic regions and appropriate strategies should be made according to different conditions. More studies are needed in future to carry out more detailed researches.