Serological diagnosis in acute leptospirosis.
The need for non-bacteriological methods for diagnosis increases with time after onset, as the likelihood
of detecting leptospires decreases; the diagnosis is frequently not considered during the early stages when
leptospires can be detected. In addition, culture is slow and often uncertain. Traditionally, serology has been
used to provide evidence of current or recent leptospirosis. IgM agglutinating antibodies usually appear within
3-10 days; fetuses, stillborn or neonatal victims have IgM antibodies in cord blood and in the placenta.
The main methods used are the microscopic agglutination test (MAT), and enzyme-linked immunoassay
(EIA, ELISA), with a variety of antigens. Much more rarely, in very few places, micro-capsule agglutination,
immune hemagglutination or hemolysis of antigen-sensitized erythrocytes, and complement fixation are used.
The specimens are samples of serum obtained from blood, taken aseptically and sent to the laboratory in sterile
containers, without anticoagulant. At autopsy, blood clot, or serum which has separated, may be taken.
Alternatively a square or circle of sterile filter paper which will absorb a standard volume of 0.1 mL of blood
can be used to soak up capillary blood from a finger or ear prick. The specimen is dried, labeled for
identification and sent to a laboratory, where the serum dried on the paper is eluted and used for titration. At all
times, standard biosafety precautions for handling blood and body fluids should be taken.
In general, blood for serology should be taken as soon as possible in the illness. A second specimen
should be taken 5-7 days later, and repeated at similar intervals if necessary. The test may be negative in the
early stages, but the second specimen may be positive or show a rise in titer compared with the first. A third or
later successive specimen usually shows a rise in titer to a peak in 2-3 weeks. Sometimes antibodies do not
appear until 3-4 weeks after infection; delays of months have been recorded. There are views that early
antibiotic treatment can eliminate leptospires so quickly that antibodies cannot be detected later; in such cases
diagnosis depends on successful culture from blood or other specimens. In the latter study. it is probable that
infections with both serovars pomona and hardjo were epidemic, but only pomona was recognized, by serology,
at the time. Some successfully treated patients who did not respond serologically to pomona antigens had titers
to serovar medanensis, which cross-reacts a little with hardjo; response to hardjo was not tested.
Antibody levels drop over weeks or months, but may persist for 2-10 years at least in humans, and for
comparable periods at low levels in animals, where they may persist throughout the life of the animal. In
endemic areas, where many people or animals may have had leptospirosis, the interpretation of a single low titer
may be impossible; a rising titre in successive specimens is then mandatory for serological diagnosis. Where
there are antibodies to several serovars detected in early sera, the highest titer may not be against the infecting
serovar. The highest titer in later sera tends to be more serovar-specific. MAT involves the reaction of any class
or classes of immunoglobulin; the class cannot be determined from the routine MAT test. IgM antibodies
produced early in infection can detected with specific anti-IgM ELISA. The early IgM antibodies react broadly
against a variety of serovars more than do the later IgG. However, some people produce little or no IgG. Thus it
may not be possible to recognize the infecting serovar from the results of serological tests.
In some laboratories serum for MAT is serially diluted from an initial dilution of 1 in 100, to avoid low-5
level cross reactions. This procedure too insensitive to detect lower titer antibodies at a very early stage of
infection, when they have most clinical significance. MAT tests for diagnosis should start at a titer of 1 in 50 or
less. The maximum titer varies, depending largely on the serovar. Titers to 20-30,000 may follow infections
with Icterohaemorrhagiae serogroup leptospires while serovar hardjo infections produce much lower levels,
typically 1600-3000, in people or animals, corresponding to the relatively lower levels seen after immunization
with hardjo. Low level early responses are thus especially significant with hardjo infections, in which maximum
titers of 100 have been recorded in patients whose diagnosis of leptospirosis was proved by blood culture.
The MAT is specific for the infecting serovar or closely antigenically related serovars. Thus the likely
infecting serovar must be known, or a battery of serovars used to ensure that there is a good chance of detecting
antileptospiral antibodies. Most laboratories use at least 6, up to 15 serovars, representative of the locally
common serogroups. Some reference laboratories use representative serovars of all serogroups. In the early
diagnosis of clinical leptospirosis in humans it is vital to know whether the patient has leptospirosis rather than
what sort of leptospirosis. The MAT has the disadvantage that it is tedious and wasteful to test against a large
battery of serovars to ensure that the infecting serovar is included. Less serovar-specific agglutination tests used
for diagnosis and epidemiological screening include agglutination of L. biflexa serovar patoc strain Patoc I,
which was found empirically to be agglutinated by sera from patients and animals convalescent from infection
with many serovars. In the author's laboratory, where infections are almost exclusively hardjo, pomona and
tarassovi forms of leptospirosis, there was no significant difference between the proportion of "positive"
reactions with the Patoc antigen in leptospirosis and non-leptospirosis patients. Other less tedious and less
specific tests are a macroscopic agglutination reaction, as a slide test using pools of formalin-fixed antigens or
heated cultures of strain Patoc I; immune hemagglutination and complement fixation; and microcapsule
agglutination. They are all less sensitive than MAT, and are thus less useful for early clinical diagnosis. ELISA
tests using broadly reactive antigens derived from boiled leptospires have not been used routinely because of
poor correlation with MAT. ELISA using sonicated leptospires or LPS correlate better in human or animal
leptospirosis and are more sensitive. IgM ELISA is more promising for detection of early leptospirosis but
further evaluations are needed. Immunodot ELISA was comparable with Patoc agglutination for sensitivity,
while a similar "line blot" immunoassay was comparable with Patoc agglutination and more sensitive than MAT
starting at 1 in 100 dilution. A further adaptation of Elisa-IgM consists in a dipstick assay using a classical heatextract
of L. biflexa as antigen. Its main appeal is its ease of use, compatible with field conditions encountered
where there are few medical resources. This type of test is essentially useful only for screening as it is limited
by the same constraints that apply to conventional ELISA methods. In epidemiological studies the data provided
by this type of method need to be validated using MAT in parallel as a reference method.
A general recommended standard is that, in an endemic area, a serological diagnosis of leptospirosis is
confirmed in human patients or animals with a compatible clinical illness, where the microscopic agglutination
titre (MAT) in sera taken 5-10 days apart rises from less than the starting dilution of the test (usually 1:50, or
1:100), to more than 100, or 4-fold or more, or is initially more than 400. In a non-endemic area, a single titre of
50 or more, with a clinically compatible illness, indicates likely leptospirosis; the titre will usually rise 4-fold or
more in a second specimen a few days later. However, a single titre of 400 or more can lead to only
presumptive diagnosis, indicating recent contact with leptospires which are not necessarily the cause of the
current illness. The titre to some serovars is consistently higher than to others. Consistent criteria are required
for the definition of presumptive and confirmed cases using serological diagnosis. These two categories need to
be separated, otherwise diagnosis is subjective and epidemiological patterns can not be validly compared.