Challenges in diagnosis and management of chikungunya and Zika virus infections

Arthropod-borne viruses (arboviruses) are viruses that can be transmitted between animal or human hosts by arthropod vectors, such as mosquitoes, ticks and sandflies. Transmission of virus occurs when the arthropod bites the host to feed on his blood. An infectious bite may then lead to a wide range of symptoms, from inapparent (asymptomatic) to severe disease and death. Approximately 150 virus species are known to cause human disease, and some of them have threatened human health for centuries. Others, like chikungunya virus (CHIKV) and Zika virus (ZIKV), have caused outbreaks on a global scale over the past decades. CHIKV and ZIKV belong to different families of viruses (Togaviridae and Flaviviridae, respectively), but they are both primarily transmitted by Aedes species mosquitoes, that are widely distributed across the continents. The emergence of CHIKV and ZIKV into new territories has emphasized that significant gaps exist in our knowledge of the natural history and clinical epidemiology of arboviral illnesses. As explained in the Introduction, these gaps present many challenges to the diagnosis and management of CHIKV and ZIKV infections. People who are infected or at risk of infection, clinicians, public health authorities and policy makers could benefit from multidisciplinary scientific efforts to advance our understanding of the risk factors, clinical presentation and diagnosis of Aedes-borne infections. The studies that were conducted in the context of this PhD thesis aim to contribute to these efforts. CHIKV was first introduced in the Caribbean and the New World in December 2013. It spread rapidly, and from October 2014 to March 2015, the outbreak affected the island Aruba in the Dutch Caribbean. The diagnostic capacity for CHIKV infections on the island at that time was limited to CHIKV-specific antibody detection by an immunofluorescence assay at the Landslaboratorium Aruba. At the Institute of Tropical Medicine in Antwerp (ITM), we retrospectively tested the sera of 498 patients who were referred to the Landslaboratorium for CHIKV diagnostics by their primary care physicians during the outbreak (Chapter 1). Two-hundred sixty-nine CHIKV cases were identified, 210 by antibody detection and in addition 59 (28%) cases using real-time reverse transcription polymerase chain reaction (RT-PCR). This finding highlights the substantial risk of under-diagnosis in field settings where RT-PCR is not available and follow-up samples are not easily obtained. Hundred seventy-one of 248 patients who were eligible for interview were contacted by telephone. We found arthralgia, fever and skin rash to be the dominant acute phase symptoms. Twenty-six percent of cases suffered from persisting joint pains that lasted longer than one year. Persistence of joint pains was predicted by female gender of the patient (odds ratio 5.9, 95% confidence interval (CI) [2.1-19.6]), the pattern and number of joints involved in the acute phase of infection (odds ratio 7.4, 95% CI [2.7-23.3]), and viremia beyond 7 days of symptom onset (odds ratio 6.4, 95% CI [1.4-34.1]). The considerable burden of long-lasting sequelae of CHIKV infection and the poor performance of antibody detection-based assays in the acute phase, emphasize the need for improved diagnostics. In Chapter 2, we evaluated a prototype immunochromatographic rapid diagnostic test that uses mouse antibodies to target the envelope protein E1 of CHIKV. When evaluated in a panel of clinical samples from returning travellers that contained ECSA genotype CHIKV (different lineages), the test had fair diagnostic sensitivity (88.9%, 95% CI [56.5 - 98.0]), but the sensitivity for samples from Aruba, containing Asian genotype CHIKV (33.3%, 95% CI [19.2-51.2]) was low. The overall specificity (83.1%, 95% CI [71.5-90.5]) of the test against sera from patients with other febrile conditions or sera that contained other alphaviruses or flaviviruses was poor. We suggest further development of a rapid test for CHIKV requires the use of antibodies that react across CHIKV genotypes and that the performance of such a new assay should be evaluated against different CHIKV genotypes. ZIKV, a member of the family Flaviviridae, was considered a cause of mild illness or, in 80% of cases, asymptomatic infection. However, its emergence in French Polynesia (2013) and the Americas (2015) has unveiled associations of ZIKV infection with neurological disease and, when pregnant women are infected, with microcephaly and other birth defects. In addition, the notion that this arbovirus can be transmitted from person to person by sexual intercourse, has caused great concern among scientists, health professionals and the general public. Precautionary advice was given to travellers from non-endemic areas, who consulted travel clinics for risk assessment prior to a journey into areas affected by the outbreak. This advice included travel restrictions for pregnant women. Upon return after travel, clinicians and laboratories were overwhelmed by the demand for diagnostic evaluations for ZIKV. The evidence base for diagnosis and management of ZIKV infection, particularly in regard to family planning, was small. In a prospective cohort study among 55 adult participants who travelled to areas with epidemic vector-borne transmission in the Americas in 2016 (Chapter 3), we observed 9 cases of ZIKV infection. The ZIKV incidence rate was 17.0% (95% CI [7.8 - 32.2]) per month of travel, and during the outbreak it ranked second only to travellers' diarrhea among travel-associated health hazards. Symptomatic infection presented as an exanthematous, rather than a febrile illness. Only one of 9 ZIKV-cases (11.1%) was asymptomatic, suggesting that asymptomatic ZIKV infection in travellers is much lower than previously reported in studies from endemic areas. The majority of reports of sexual transmission of ZIKV involved male-to-female sexual intercourse. The virus had been detected in high loads, and it was isolated from semen samples for up to 69 days after the onset of symptoms. The potential for sexual transmission of ZIKV seems therefore closely associated to viral persistence in semen, which we studied in symptomatic returning travellers with confirmed ZIKV infection (Chapter 4). ZIKV RNA was detected by RT-PCR in the semen of nine out of 15 participants (60%). It remained detectable for 83 days after symptom onset (95% CI [57–108]), and the longest duration of viral shedding in semen recorded in our cohort was 144 days after symptom onset. ZIKV was successfully isolated from one sample only, but as long as viral RNA can be detected in semen, we would not exclude the potential for sexual transmission of ZIKV. In the semen samples of 11 participants whose semen was microscopically analyzed, we found leukocytes (n=11), red blood cells (n=10) and oligospermia (n=6). These abnormalities occurred irrespective of Zika virus detection in semen and may indicate some degree of tissue damage to the male reproductive tract. At the start of the outbreak in the Americas, validated diagnostic assays for ZIKV infection were in scarce supply. At ITM, symptomatic travellers were tested with a ZIKV-specific RT-PCR on serum samples upon presentation within 7 days post symptom onset (DPSO) and on urine within 14 DPSO and an anti-ZIKV Immunoglobulin (Ig)M and IgG Enzyme Linked Immunosorbent Assay (ELISA). All positive or equivocal anti-ZIKV IgM/IgG results were considered diagnostic only when confirmed by ZIKV virus neutralization testing (VNT). Asymptomatic travellers were tested using ELISA only, preferably from 20 days after the last exposure. In Chapter 5 we performed a cross-sectional cohort analysis to evaluate our approach to the diagnosis of Zika infection in non-pregnant travellers. During a 12 month period, we evaluated 462 travellers. ZIKV infection was confirmed in 49, and was frequent in symptomatic cases (46/227, 20.3%), but not in asymptomatic persons (3/235, 1.3%). Asymptomatic travellers had similar baseline characteristics, but they had reproductive concerns more often (75.8% vs. 24.2%). Rash (positive likelihood ratio (LRP) 5.6) and conjunctivitis (LRP 10.8) predicted ZIKV infection. The post-test probability of a negative ELISA result at 20-25 days was below 0.1%. We consider negative ELISA results at 20-25 days after exposure a safe strategy to rule out ZIKV infection. Testing for ZIKV-specific antibodies within this timeframe could be particularly valuable in the management of returning travellers who wish to conceive.status: publishe