Saturday, 18 June 2016

The elephant squeaked...

An interesting new paper came out from Pacheco and colleagues from Colombia's national institute of health (INS or Instituto Nacional de Salud) this week. The INS team have written up their Zika virus (ZIKV) data spanning from 9th August 2015 to 2nd April 2016.[1] ZIKV data were not being regularly collected prior to then. 

For me, the main new outcome from this report is that there seem to be no adverse outcomes to babies born to mothers who were infected with symptoms suggestive of ZIKV infection, in the third trimester of their pregnancy. Phew. 

I'm not surprised at the low number of ZIKV-related microcephaly diagnoses observed, because I've been following the INS's weekly reports on this which make that issue very clear. I talk a little more about that issue below.

There were quite a few other bits and pieces in this publication. I've tried to capture some of the ones I found interesting in the list below:

  1. The time taken for health care centre ZIKV disease (ZVD) data to be reported by the INS is approximately 1.5 weeks
  2. Elevations 2,000m above seal level are not considered a risk because they are above  traditional ZIKV-mosquito habitats, but they may still harbour human cases that have travelled from lower elevations and these may be under-counted
  3. In 2010, half (52%) of pregnancies in Colombia were unintended and condoms were used in the same proportion of sexual encounters reported by women. Any advice aimed at reducing risk of sexual transmission of ZIKV - or other sexually transmitted infectins - will need to innovate to get that message across 
  4. Two-thirds (67%) of suspected ZVD cases were reported in females. Incidence per 100,000 population was similar in children but higher in females. This may reflect more testing of, and concern among, women of child-bearing age
  5. Most pregnancies with symptoms suggestive of ZVD were ongoing when this report came out. This supports the thinking that it is still too early to say that Colombia will not have the same ZVD-related congenital disease problems among its pregnant women that Brazil has claimedIn a subset of 1,850 pregnant women who delivered babies, 532 (29%; 16% of pregnancies were not ongoing-why was not made clear) experienced their symptoms in the first trimester, 702 (38%; 29% not ongoing) in the second and 616 (33%; 82% born at term with normal weigh; 2% at term but low weight, 8% preterm, 1% were perinatal deaths and 7% are ongoing) in the third.
  6. Lanciotti primers were used for RT-PCR of serum samples, but they look to have been updated in this study. The names differ-1087 instead of 1086 in the cited publication by Lanciotti et al.[2], 1163 instead of 1162c, 1108-FAM instead of 1107-FAM.
    It would be good to see what sequences were actually used here. 
    Most RT-PCR testing (60% of 3,384 samples tested during this period) was on samples from pregnant women (see No. 4) and no testing of urine was discussed which is a shame because urine is reportedly a better sample for RT-PCR because the detection window can be extended beyond that of using serum alone
  7. No antibody testing was available - that means a lot of missed opportunities to confirm suspect ZVD diagnoses
  8. The INS mandates reporting of all symptomatic cases. I had a little hope that perhaps Colombia, because they have been good at reporting throughout this event, might have also looked at whether the "80% asymptomatic" figure from previous outbreaks still holds today. Okay - it was a teensy hope. It will need a specific study
  9. Fever was an integral part of the INS case definition - but fever is absent in 20% to 72% of cases.[1,3] This could mean a lot of ZVD cases were not included in the analysis - with an unknown impact on linkage to microcephaly counts from Colombia
  10. As STATNEWS reported earlier[4], the World Health Organization has indicated that Colombia is not reporting on aborted foetuses or miscarriages that might have been related to ZVD. The impact of this omission is unclear and it would be great to know more about the issues and concerns here
  11. During this period 4 infants (born between weeks 37 and 39) were reported to have microcephaly and confirmed ZIKV infection; 1 had abnormal brain findings and 3 abnormal hearing evaluations and other findings were listed as well. However, none of the 4 mothers reported symptoms of ZVD during pregnancy. Colombia is capturing newborn issues other than head size, at least following that initial diagnosis of microcephaly being made 
  12. STORCH (syphilis, toxoplasmosis, other agents [which other?], rubella, cytomegalovirus, and herpes) screens, karyotype analysis and ZIKV virus testing of the subset of 1,850 pregnant women with suspected ZVD were the only other tests described.
    Investigations into teratogenic or toxic causes of microcephaly [6] not happening - at least based on the contents of this report
  13. Among 239 ZIKV-negative samples (another 316 were ZIKV positive-all collected within 7 days of symptom onset - 8 (3%) were positive for dengue virus and 23 (10%) for chikungunya virus

The authors conclude that women with symptoms suggesting ZVD in the 1st trimester, may start to deliver affected babies soon after the early April cut-off date that this report covers. It's now the second half of June though and we have only had 6 cases reported by Colombia.[5] This may all be because of those reporting differences and timing issues discussed above and elsewhere on VDU. We await some more discussion about precisely how that would be the case.

We should also remember that clinically suspected diagnoses in this neck of the woods can be fraught with difficulty because fever+rash can be caused by a very wide range of things - including other mosquito-borne viruses known to be co-circulating. Take those totals with a big grain of NaCl and look laboratory confirmed samples for trends. This is also bias because a lot of the Colombian laboratory focus has been on pregnant women. 

Click to enlarge.
We really know very little about the incidence of ZIKV in the general population or about the proportion of confirmed ZIKV infections that have little or no disease or about the rate of microcephaly or other congenital disease outcomes from infection. These would be very useful numbers to have to help us understand different risks and they rely on an accurate denominator. By 'accurate' I mean one based on actual testing results, not one based on guesstimates and models that lean on "suspected" data. This knowledge gap equally applies across all of the epidemic countries though, not just to Colombia.

For now, the elephant in the room continues to be the subject of insufficient conversation.

References...


  1. http://www.nejm.org/doi/pdf/10.1056/NEJMoa1604037
  2. http://wwwnc.cdc.gov/eid/article/14/8/08-0287_article
  3. http://www.nejm.org/doi/pdf/10.1056/NEJMoa1602412
  4. https://www.statnews.com/2016/06/14/zika-olympics-who/
  5. http://virologydownunder.blogspot.com.au/2016/06/colombia-zika-virus-report_12.html
  6. http://www.neurology.org/content/73/11/887.full.pdf+html