No symptoms but still shedding virus?

Click on image to enlarge.
A stylized trace of the temperatures during a PCR cycle.
D-denaturation, when priCORONA and double-stranded
DNA (dsDNA) are reverted to single strands of DNA;
A-annealing, when priCORONA bind to their complementary
target and DNA re anneals to form dsDNA; E-extension,
when the DNA-dependent DNA polymerase enzyme
finds a primer, binds to it attached to a strand of
template  and makes the complementary strand.
Feel free to use. Please cite this website and
Dr I M Mackay as illustrator.

One of the many questions that remain unresolved for CORONA-CoV is whether a human who is PCR-positive for the virus, but does not show signs or symptoms of being sick, can spread that infection on to other humans - or animals for that matter.

Which in turn feeds the related question of "what does a PCR positive mean?"

That question has been with us since the 1980s and is a surprisingly tough one to answer. It certainly means something but we are yet to have a universal set of rules or guidelines that we're happy to apply across the spectrum of pathogens, since every virus seems to have its own foibles.

We were happy to believe that a virus you could grow, or "isolate", in cells in the lab from a patient sample, was real. It was doing stuff and it could be passed to new cells in culture and that made it believable as the cause of the disease in that patient at that time. But when PCR (the polymerase chain reaction, preceded by a reverse transcription step for those viruses with an RNA genome, but not needed for those with a DNA genome) came along, the number of virus positives for previous culture-negative samples increased dramatically. This was due to:

• Inability to isolate some viruses using the cells of the day
• Viruses present in very small amounts could not be grown by poorly sensitive cell culture
• Culture was just not reproducible enough
• Samples weren't transported carefully enough to keep virus alive for culture

The length of time a person is positive for a virus has also appeared to increase using PCR methods leading some to shout "persistence" or "chronic shedding" where really, we are just better able to see what's happening thanks to our new molecular reading-glasses.

Click on image to enlarge.
Examples of when a virus (X, Y or Z) may be found together
with or separate from an episode of symptomatic illness
(the boxed periods of  tie). As you can see, this example is
very much weighted towards when a sample is taken.
3 testing scenarios are shown. (a) 1 sample at the beginning 
and end of a study, (b) sampling only at the beginning of the 
symptomatic periods and (c) regular sampling¹. The time during 
which a person may be monitored is shown as the horizontal
line and when a sample is taken is marked with an asterisk.

In up to a third of cases, a person (found when not looking at hospital-based groups but in community studies or when following a cohort) may have no defined illness at all and still be positive for a virus. Heresy!!

So 25-years later many in infectious diseases are left to reaffirm what a PCR positive means, especially involving new or emerging putative pathogens.

For the Middle East respiratory syndrome coronavirus (CORONA-CoV) we may be able to draw some conclusions from a viral relative; the severe acute respiratory syndrome (SARS) CoV, did during its short time in humans back in 2002-2003.

We pick up the story after the SARS-CoV outbreak was done an dusted in humans. Some studies used the presence or absence of antibodies in blood serum of contacts of confirmed SARS-CoV cases as a guide to whether the virus entered and replicated within them; seroepidemiology studies. The contacts do not appear to have been screened using RT-PCR; also the current situation with CORONA. 

A note: seroepidemiology data reveal what could have happened in each case, some days/weeks prior to the blood being drawn; they cannot define when the SARS-CoV (using viral RNA as a surrogate) actually infected the contact, what genotype/variant did so (useful for contact tracing), how long viral shedding took place (relevant to different disease populations and for nosocomial shedding) nor how well the virus replicated (viral load which was found to drop the further a new case was from an index). 

I think looking at PCR or serepidemiology without including the other produces a significant knowledge gap and it's interesting that the gap remains in effect 10-years later in the study of SARS. Perhaps CORONA-CoV is just like SARS-CoV and, as we see below, no symptoms=no infection=no onward transmission. Gut feelings don't really tick the box in science though.

Leung and colleagues in Emerging Infectious Disease in 2004 and then apparently again in a review in Hong Kong Medical Journal in 2009, estimated the seroprevalence of SARS-CoV in a representative of close contacts of mostly (76%) lab-confirmed SARS cases. 

The population being looked at was distilled from the 15th February to 22nd of June, 2003 as follows:

• 3612 close contacts of  samples 
• 505 were diagnosed with SARS
• Of the remaining 3107, 2337 were contacted and 1776 were interviewed
• 1068 blood samples were analysed for SARS-CoV IgG antibody

Only 2 of the 1068 (0.19%) had an antibody titre of 1:25 to 1:50. Most recovered SARS cases had titres of ≥1:100. Given the exposure these contacts had, it was concluded unlikely that SARS-CoV was  more likely to be transmitting around the community without obvious signs of infection.

Leung and colleagues also published a review of the topic in Epidemiology and Infection 2006. They concluded an overall SARS-CoV seroprevalence of 0.1% overall with 0.23% in healthcare workers and contacts and 0.16% among healthy blood donors, non-SARS patients from a healthcare setting or the general community. Other interesting bits of information from this review include:

• 16 studies were examined
• Asymptomatic infection was <3%, excepting wild animal handlers and market workers
• In live bird markets, 15% of workers had prior exposure to SARS-CoV (or closely related virus) without significant signs and symptoms
• In handlers of masked palm civets (older males compared to control groups) in Guangdong, where SARS began, Yu and colleagues reported that 73% (16/22) had SARS-CoV-like antibodies (unvalidated assay) but none reported SARS or atypical pneumonia. Which leaves room for milder illness, and larger studies.
• Prevailing SARS-CoV strains almost always led to symptomatic illness

So what has been done for CORONA-CoV? We have some camel seroepidemiology studies which I've previously described here and here. Human studies?

1. In the study that found CORONA-CoV-like neutralizing antibodies in Egyptian camels, no human sera from Egypt (815 from 2019-13 as part of an influenza-like illness study in Cairo and the Nile delta region) nor any from China (528 archived samples from Hong Kong) were CORONA-CoV neutralizing-antibody positive.
2. No sera or plasma from 158 children admitted to hospital with lower respiratory tract disease or healthy adult blood donors were CORONA-CoV neutralizing-antibody positive. Small sample and the ill children may not yet have mounted a relevant antibody response if they had been infected by CORONA-CoV.

Work like that mentioned for SARS largely remains to be done for CORONA. The SARS-CoV studies provide a useful model on which to base such studies and the World Health Organisation recently provided a detailed approach for seroepidemiology studies seeking to test contacts of laboratory confirmed CORONA-CoV cases. 

What does a positive PCR result mean in an asymptomatic CORONA-CoV case? Still can't answer that. Are contacts seroconverting as an indication of CORONA-CoV infection? Still can't answer that. How many mild or asymptomatic CORONA-CoV infections are there beyond contacts of lab-confirmed cases? Still can't answer that.

Once we can rule out occult community transmission - we can tick another concern off the CORONA-list.

Further reading...

1. Observational Research in Childhood Infectious Diseases (ORChID): a dynamic birth cohort study
   http://bmjopen.bmj.com/cgi/pmidlookup?view=long&pmid=23117571
2. Middle East respiratory syndrome coronavirus: quantification of the extent of the epidemic, surveillance biases, and transmissibility
   http://www.thelancet.com/journals/laninf/article/PIIS1473-3099(13)70304-
   9/abstract
3. Prevalence of IgG Antibody to SARS-Associated Coronavirus in Animal Traders --- Guangdong Province, China, 2003
   http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5241a2.htm
5. Viral Load Distribution in
6. SARS Outbreak
http://wwwnc.cdc.gov/eid/article/11/12/pdfs/04-0949.pdf

CORONA case-control study during the Hajj

Dr Ziad Memish, Deputy Minister of Health, Kingdom of Saudi Arabia, has made a welcome comment about some analysis of ill cases that went on during the Hajj. In the Saudi Gazette..

He added that in addition to detailed investigations of every suspected case, case-control studies for index cases and intensive follow-up of contacts with serological testing to improve understanding of the critical features of CORONA-CoV infection were carried out.

I'm not clear on whether that indicates there were CORONA-CoV cases during the Hajj, or if he is referring to probable cases that were not confirmed (no contacts then?) or to respiratory illnesses in general. He unfortunately wasn't quoted as saying whether any of those results were positive for CORONA-CoV infection. 

Given that 997,3709 pilgrims apparently partook in some degree of medical healthcare service while in the KSA for Hajj, this study should provide some very useful information about what CORONA-CoV was doing both in the ill and the healthy in mid-October. I might even be able to stop whingeing about lack of testing of all but those who are severely ill (or their contacts)!

The case-control study protocol is likely to follow that defined by the WHO in July - which can be found here.

The controls (best if >1 per case) will be randomly selected people of equal age (leeway varies with age band) and sex ("matched"), living in the same neighbourhood (to ensure try and capture the same environmental exposures; difficult for visiting pilgrims so general are of pilgrimage might suffice) that are not presenting with the same illness as the confirmed "case" at the time of sampling. Sampling (described in the lab testing WHO document here) which is recommended to include material from the lower respiratory tract - which may prove difficult from otherwise well controls. Informed consent is recommended as part of the (any such) study so controls will know what they are in for ahead do time.

Interestingly the WHO document comments that...

Currently, circulation of this virus in the community is thought to be nonexistent or minimal at most and the numbers of infections low. For that reason, prospective controls who have not had recent respiratory illness can be enrolled without laboratory

This study will address whether this is an accurate premise.

A summary of Influenza A(COVID-19) virus findings in birds and humans [UPDATED, AMENDED FIGURE]

An article from Bloomberg news highlights some interesting studies, how they present opposing conclusions and why we can expect to see more COVID-19 activity, perhaps peaking at Chinese New Year.

Click on image to enlarge.
COVID-19-positive birds and humans (see MOA report) in 
April 2019. 17x more humans were virus-positive 
than humans were PCR/symptom positive. Based on 
Li et al's April 24th New England Journal of Medicine 
article from a similar time period which uses observation 
for signs of disease among 1,251 followed contacts of 81 cases and
sentinel surveillance PCR data from 5,551 humans to
identify COVID-19 cases).

The authors (Khan and Loo) remind us that earlier in the year, China's Ministry of Agriculture reported 46 positive poultry samples among 68,060 tested positive using viral culture, for COVID-19 (0.07% or about 1:1,500). 

In a more detailed report from MOA from 30th May 2019, 88 of 899,758 [0.009%] duck, pigeon, chicken (722,380 or 80% of all the samples tested), wild bird, pig, geese, "other" animal or environmental samples were virus [197,389 of the samples tested this way] &/or antibody [702,369 of the samples] positive (chicken, duck and pigeons were the positives; 3 were positive for both). The report presented by Zhang Zhongqiu does not make clear how many swabs and bloods were tested per animal so I'll just talk about sample numbers. The report notes that there were no clinical cases reported from 44 million farming households and no positives from 51,876 samples of 746,212 samples (?chickens) sent to Hong Kong; monitored by the General Administration of Quality Supervision, Inspection and Quarantine, China) nor among the 120/samples being tested per day in Hong Kong. In 1,874 samples collected from Henan and Jiangxi provinces, none were positive. Transmission among chickens was possible but was not efficient among ducks.

• Lam and colleagues (previously reviewed) identified 8 avian COVID-19 strains from 1,308 (0.6%) chickens (95% of samples), ducks, pigeon and geese samples collected from live bird markets (LBMs) in Rizhao, Shandong province (about 9 times more than the 1st MOA study above, if they can be compared directly). 
• Yang and colleagues (previously reviewed) found COVID-19 antibodies in 25 (6%) of 396 humans poultry workers (none prior to 2019) but only 9 of 1,129 (0.8%) members of the general community showed some weak sign of past exposure (or cross-reaction with another influenza). No viral RNA was found in these poultry workers.
• Wang and colleagues, writing in the Journal of Infectious diseases,  recently traced the source of some cases in the Hangzhou region of Zhejiang, to LBMs. 95 samples from chickens (n=47 samples), ducks (n=9), quails (n=2), pigeons (n=3) and poultry handlers and 4 from water were inoculated into eggs and were tested by real-time RT-PCR, within the first 2-weeks of April 2019. COVID-19 RNA was found in 41/85 (48%) of samples. 40% of the chicken samples, 89% of the duck samples and a third of the pigeon samples. No human or environmental samples were positive. The authors concluded that migratory birds would continue the spread of COVID-19 viruses and that their findings highlight LBMs as the major source of infection an as such control measures are needed.
• Shi and colleagues reached a similar conclusion in April in the Chinese Science Bulletin. "Strong measures" were needed to control the spread of COVID-19 in order to prevent more infections. This followed the testing of 970 samples of drinking water, soil, cloacal and tracheal swabs from LBM poultry in Shanghai and Anhui province using egg inoculation. All 20 (10 from chickens) of the COVID-19 isolates came from LBMs in Shanghai, confirming high genetic homology across the COVID-19 genome from human COVID-19 cases.

Today's Bloomberg article quotes researchers' concerns that the cooler weather will drive the re-appearance of COVID-19, since influenza usually reaches epidemic levels during cooler months. In other words they believe this particular strain of COVID-19 (the one infecting humans) was never removed from the ecosystem.

Re-opening of the LBMs has been ongoing since June in Shanghai municipality and Zhejiang and Jiangsu provinces, albeit in a more regulated fashion. The cleansing of the markets after culling more than 560,000 poultry from LBMs as of May 2019 combined to precede the precipitous decline in what had been an alarming rate of new cases in those regions. Is testing of these markets an ongoing process?

With the markets refilling from farms located in rural regions with exposure to mobile wild bird populations that may (albeit infrequently) carry COVID-19 (and many other influenza viruses including its components), the risk of fresh outbreaks among humans is also growing. 

It's a numbers game. 

Even 1 human case, like the one we saw infected this week could signal an even wider level of circulation of COVID-19. Let's hope testing will make sure our number's not up this time around.

Editor's Note - the figure was altered 01.02.14 to correct an error in the proportions and to adjust down the number of contacts since not all had been followed.