In article <37m2rs$p4e at quartz.ucs.ualberta.ca>,
Dr. R. Marusyk <rmarusyk at mercury.uah.ualberta.ca> wrote:
>Dan McLeod's response to my request that he comment on his 1984 paper on
>Hantavirus seroprevalence in Canada was routed to me as E-mail rather
>than as a follow-up post. For those interested, here it is:
>>>There seems to be some question here w.r.t. the antibody assay used and
>76-118 and also strain 79-90, whereas NEV was in lung sections of C.
>glareolus. Over 2000 sera were tested and 29 (1.4%) were positive
>with tires ranging from 1:16 to 1:128.
>>As a point of interest, there were only 2 patients with titres of 1:128.
>>One was a 57 year old female from Alberta and the other was a 72 year
>female from Ontario. Antibody was found in all provinces. This
>indicates contact with a Hantaan-like virus. The current strains with a
>>pulmonary syndrome are more related to the Prospect Hill virus. Its
>obviously time to do some more testing for relevant antibody. The
>specific virus in the Alberta case can be identified from patient
>samples using a nested set RT-PCR with G2 consensus primers and
>the amplicons for comparison to the 4 Corners strain, etc.
I agree, PCR would be very helpful in confirming or refuting
the possibility of false-positive ELISAs. In searching for
unrecognized infection by Four Corners virus in humans
by western blot, they appear to be exceedingly rare. We
have run many hundreds (including hundreds with clear occupational
risks) using FCV nucleocapsid and G1 antigens (see J Virol 1994;
68:3000), and found *no* examples of unrecognized infection.
This was in the endemic area, not Canada! ELISAs have
a known problem with lack of specificity, so we are
of course concerned that "positives" are truly positive.
Getting back to PCR, again, there *is* no PCR assay for
past infection by hantaviruses, since humans clear the
virus after the acute illness (see October Journal of Infectious
Diseases). Thus, "positive" ELISAs occurring in people
with no clinical history are *not* tantamount to
past exposure, to Hantaan or any other hanta.
In fact, it would be exceedingly surprising if the
elderly ladies you allude to had any reason whatsoever
to have contracted a Hantaan-like virus (assuming they
had no history of HFRS), much more likely they are
>BTW, the ELISA test appears to using nucleocapsid protein expressed in
>E.coli as per Zoller in JMV 39:3, 1993. This is coded for by the S
>segment and is highly conserved and probably useful for screening.
>Zoller claims the Elisa to be superior to the indirect
>assay. Not a bad compliment for the IFA also. It would be even more
>strain specific though, if the G protein had been expressed in a
>eucaryotic expression system, IMHO.
There is of course a large difference between using
these assays to confirm infection in someone with recent
clinical history of hanta-like disease, and using them
to detect remote infection in someone with no clinical
history. Like night and day. One is a very high prior
probability setting, the other very low. That is why
people invent things like the western blot, to add a
substantial layer of specificity to the testing algorithm.
It is just unfortunate to hear how we (apparently) have
to reinvent the wheel in the field of hantaviruses,
when these matters have already been largely resolved
in the fields of HIV, HTLV, HCV, etc.