Saturday, 5 March 2016

Zika virus and brain injury among foetuses of 88 women with rash in Rio...[UPDATED]

In what may be looked back upon as one of the first and most important papers in the Zika virus (ZIKV)/microcephaly story to yield helpful data, the New England Journal of Medicine has just released a preliminary report by Brasil and colleagues describing an ongoing study of 88 women from Rio de Janeiro, Brazil, enrolled because they had developed a rash in the previous 5 days.[1] No women had any indication of foetal malformations before enrolment, the mothers were healthy and reported use of any medications.

There are a number of interesting findings from this study...
  1. Referral for an ultrasound was made before 20 weeks of gestation, between 20 and 30 weeks of gestation, and after 30 weeks of gestation
  2. Immunity to rubella virus, cytomegalovirus was identified
  3. 16/88 women were not infected by ZIKV after testing blood, urine or both using a previously described real-time RT-PCR [2] and all had normal foetal ultrasound findings
  4. 72/88 women were infected (82% of the women with fever/rash) by ZIKV according to RT-PCR results in sera (60 women; 26 only in serum), urine (46 women; 12 only in urine) or both (34 women)
    • 55/72 women were infected in the 2nd or 3rd trimester (gestation week 14 and above) but infections occurred between 6 and 35 weeks in this cohort
    • 2 (2% of 88 enrolled women) ZIKV positive women miscarried in the 1st trimester
    • 28/70 declined prenatal ultrasound examinations
    • 42/70 had foetal ultrasound
      • 12/42 had abnormal findings (29% although since 28 women declined ultrasounds, this may not be a useful % to extrapolate from) 
        • 5/12 foetuses had growth restriction in utero
        • 4/12 had cerebral calcifications noted, with maternal ZIKV infection as late as 2nd trimester
        • 2/12 had central nervous system alterations
        • 4/12 had abnormal arterial blood flow in the cerebral or umbilical arteries
        • Foetal deaths were identified in two mums who had been infected by ZIKV at 25 weeks and 32 week of gestation - third trimester,
  5. The foetal death rate in this cohort was 4.8% (2/42 who had ultrasound studies) compared to 2.5% in a cohort of 662 HIV-uninfected women in Rio de Janeiro. However, the rate in the current study Rises to 5.5% if looking at the number of deaths among all the ZIKV infected women (4/72) which includes the 2 miscarriages and 2 foetal deaths identified by ultrasound [UPDATE #1]
  6. A maculopapular rash was seen more often in ZIKV infected women than in ZIKV uninfected women (P=0.02) as was conjunctival injection (P=0.002) and arthralgia (P=0.16) and respiratory findings were rare (7%)
  7. Fever was only present in 28% of the women-not a good measure of infection then
  8. Dengue virus antibodies were present in 77/88 women (88%) but, depsite startign as a study of denguelike illness, no Dengue virus RT-PCR was conducted and no analyses for Chikungunya virus were conducted
From #4 we see instances of infection of the mum of a foetus with abnormal findings, happening in the third trimester - later than some had been assuming.

We can also see from #4 why the Brazilian Ministry of Health might simply believe that all pregnant women with traditional ZIKV disease symptoms are infected by ZIKV-82% of women in this study were at some point. However, laboratory testing to confirm belief is very important in the early days of attempting to associate a new disease to an infection. RT-PCR is also not widely available in Brazil. But as we see from #7, reliance on some traditional symptoms of ZIKV disease should not occur.

There is not yet understanding of precisely *how* ZIKV might be doing this damage, what other viruses may be in play or whether environmental factors or foods, creams or behaviours may also be a culprit alone or in combination. This study of women who already had a rash is more useful than the previous studies looking only at those mums who were infected and delivered babies with microcephaly diagnoses and/or other serious injuries. The study does not look at the proportion of all pregnant women in Rio de Janeiro though - including those who had no signs of disease.

The absence of any ultrasound abnormalities of the foetuses of women who were not RT-PCR positive for ZIKV is very interesting. If we do use the 29% figure from #4, then we would expect to see 5 with abnormalities. But we do not see any. However, if 1.4% to 4.8% of babies are stillborn, then 16 women is too small a number to support that ZIKV is the cause. That fraction equates to 0.22 to 0.77 babies and since babies don't come in fractions, we cannot really say that this severe outcome is not happening because of something other than ZIKV infection. 

Bigger studies are needed and its hard to conclude too much more than what has been observed.

This study is ongoing and all the surviving infants will be followed. It really is great to see some solid human data at last.

  1. Zika Virus Infection in Pregnant Women in Rio de Janeiro — Preliminary Report
  2. Genetic and Serologic Properties of Zika Virus Associated with an Epidemic, Yap State, Micronesia, 2007
  1. Thanks to Jody Lanard for pointing out that I had misinterpreted the paper - there were 4, not 2, foetal deaths in total. Now corrected.
  2. Corrected some typos and made note the DENV RT-PCR was not done
  3. Thanks to Judy Stone for correction on number of foetuses with growth restriction (5/12 not 5/1)