The present paper reports the results of routine testing of WNV RNA by RT-PCR for the diagnosis of patients with probable WNV infection, enrolled during the 2018 Tunisia outbreak. Confirmation of WNV infection by detection of WNV RNA in urine samples of symptomatic patients was possible for 50,5% of probable cases, while WNV RNA detection in CSF was much less frequent (2,8%). Patients were mainly men 64,6% and incidence of confirmed cases increases with age over than 60-year-old (56,3%, p < 0,001). Similar age and gender distribution of WNV neuro-invasive disease was reported (Jean et al., 2007; Koch et al., 2021; Mada et al., 2020).
Urine is, nowadays, becoming an attractive alternative sample for the detection of several pathogens that may infect humans, since it may be obtained easily, quickly, with high volume, in serial samples, and with no exposure to potential infectious risk for the health care staff comparing to veinopuncture. Urine samples are used for the detection of Cytomegalovirus genome for newborns with congenital cytomegalo-infection, of Zika Virus genome for patients having neuro-invasive disease in endemic regions, of Mycobacterium tuberculosis genome for individuals with active pulmonary tuberculosis, of Chlamydia trachomatis genome for persons screened for sexually transmitted infections and of Polyomaviruses genome for renal post-transplanted patient (Jorge et al., 2020; Marangu et al., 2015; Meyer, 2016; Nicloux et al., 2020; Xu et al., 2021). WNV RNA was detected in urine samples of an infected patient with encephalitis for the first time in 2005, 8 days after symptom onset (Tonry et al., 2005). Subsequent studies reported that WNV RNA is detected more frequently, with higher viral load and for longer period in patients’ urine samples than in plasma or CSF (Barzon et al., 2013a; Magurano et al., 2012; Niedrig et al., 2018; Pacenti et al., 2020; Papa et al., 2014). For that reasons, it was proposed in recent international guidelines to include the use of urine samples as suitable alternative diagnostic sample comparing to serum and CSF ("Piano Nazionale integrato di sorveglianza e risposta al virus della West Nile – 2016";
"Mission COREB nationale. Infection à West-Nile virus (VWN) : repérer et prendre en charge un patient suspect."). However, published reports on the results of WNV detection in urine samples as part of routine diagnostic are still scarce.
Detection of WNV RNA in urine samples by RT-PCR was introduced for the first time in Tunisia during the 2018 outbreak; that’s why, not all probable cases were sampled for urine samples. Regardless, we found that it allows confirmation of 50,5% of probable cases (48/95). Previous studies such one conducted in northeastern Italy in 2012, one in Greece in 2013 and another in Hungary in 2014, revealed positivity rates in urine samples of 43,8% (14/32), 40% (14/35) and 27,3% (3/11), respectively (Barzon et al., 2013a; Papa et al., 2014; Nagy et al., 2016). We also found that WNV RNA detection was more effective in urine samples than in CSF which confirms previous findings (Barzon et al., 2015, 2013a, 2012; Pacenti et al., 2020). While 50,5% of urine samples from our patient were positive, only one CSF, sampled on day 3 after symptom onset, was positive. These findings confirm how fleeting virorachia is and underlies the superiority of detecting virus RNA in urine samples.
Barzon et al. Pacenti et al. found that viral shedding in urine samples starts within the first week after symptom onset then lasts up to 30 days (Barzon et al., 2012; Pacenti et al., 2020). In our series, WNV RNA could be detected in urine samples from 1 up to 41 days after symptom onset. This extends further the possibility to confirm WNV infection even late. No relationship was observed between the time after symptom onset and the WNV RNA load in urine samples. Otherwise, viruria was detectable even for years after primo infection suggesting chronic renal infection (Baty et al., 2012; Murray et al., 2010). For these reasons, RT-PCR positive results should be interpreted in conjunction with clinical and serological results.
In this series, WNV RNA was not detected in 49,5% patients (n = 47) having positive IgM and IgG. Negative results should be interpreted by caution for many reasons. First, as it was experimentally reported by Tonry et al. that WNV RNA excretion is intermittent (Tonry et al., 2005), one unique negative urine sample should to exclude WNV infection, theoretically. In this series, viral excretion kinetic in urine samples could not be studied as serial samples were not systematically obtained. Second, positive serology results are not always reliable. IgM may cross react with other flaviviruses such as arboviruses like dengue virus, Zika virus and Japanese encephalitis virus, leading to false positive results (Barzon et al., 2015; Khan et al., 2018). In fact, several commercial kits are available nowadays, targeting antigens from recombinant proteins, virus-like particles or lysates of infected cells (Barzon et al., 2015). It was suggested that sequence similarity of structural proteins of flaviviruses, mainly of the immunodominant envelope protein, may lead to cross reacting antibodies (Mansfield et al., 2011). Thus, it is recommended to exclude the presence of other flaviviruses antibodies, especially those co-circulating in the same geographic area (“ National Center for Emerging and Zoonotic Infectious Diseases, Centers for Disease Control and Prevention. West Nile virus in the United States: guidelines for surveillance, prevention, and control. 4th revision. ” ). In addition, many studies reported that IgM may persist until 30 to 90 days after primary infection and even longer in safe blood donors, until 200 days (Prince et al., 2005). Thus, their presence in the absence of any other confirmation by RT-PCR may indicate a past infection. The rest of probable cases requires absolute confirmation by seroneutralisation, which will be the only laboratory method to confirm WNV IgM positivity in the absence of virus RNA detection by RT-PCR(Barzon et al., 2015).