278
T. BERGAN
contracting Shigella strains during the study period. Mixed infections,
though rare, may lead to confusing results.
The question may be elucidated in two ways: (1) serial isolation from the
same individuals, and (2) comparison of strains from several patients who
are clearly involved in the same epidemic.
Hammarstrom (1949) pursued both approaches. Among 145 1 strains
from 987 patients, only 0.9% of the isolates (ten persons-1%) reflected
instability. Rare changes have also been observed by ZieschC and Rische
(1973) and AldovP and Sucha (1973).
Grunow (196%) and ZieschC and Rische (1973) among others have
observed changes in serial isolates from the same persons. In vivo, it is
difficult to say why more than one type develops and to determine whether
it is due to reinfection, mixed initial infection, or spontaneous shift in
phage type. Extrachromosomal DNA may lead to phage restriction,
modification, or development of new phage receptors (Grunow, 1965e).
D. Typing method and reproducibility
Phage typing procedures must be carefully standardised. Pattern variation may be due to trivial technical evaluations. It is often preferable
to type strains which are to be compared simultaneously. Thereby, the
influence of slight differences in phage suspension densities and typing
medium are avoided.
With meticulous standardisation of technique, phage typing of Shigella
is sufficiently reproducible to be of considerable epidemiological value.
VIII. EPIDEMIOLOGICAL USE
Typing must be assessed on achievements in elucidating epidemics.
One relevant parameter of usefulness is the percentage of typable strains.
Aldovi and Sucha (1973) with the Hammarstrom phages obtained 17%
non-typable S. sonnei. High numbers of apparently untypable strains result
if the strains are not clearly in phase 11, i.e. only R-form colonies (not
phase I + S-form) are to be selected, Passage on endo-agar up to five times
may be necessary to obtain proper colonies (Grunow, 1965d). With such
precautions, bona Jide untypable strains are extremely rare as evidenced
from the number of reports with 100% typability.
A clinically relevant question is whether there is any correlation between
phage type and either antibiotic sensitivity or pathogenicity. There have
been reports that certain lysotypes have particular antibiotic sensitivity
patterns or biotypes but this applies only within restricted geographical
areas (Szturm-Rubinsten et al., 1974). Accordingly, such a relationship is
probably more apparent than real, due to local dominance of one or a few
T. BERGAN
contracting Shigella strains during the study period. Mixed infections,
though rare, may lead to confusing results.
The question may be elucidated in two ways: (1) serial isolation from the
same individuals, and (2) comparison of strains from several patients who
are clearly involved in the same epidemic.
Hammarstrom (1949) pursued both approaches. Among 145 1 strains
from 987 patients, only 0.9% of the isolates (ten persons-1%) reflected
instability. Rare changes have also been observed by ZieschC and Rische
(1973) and AldovP and Sucha (1973).
Grunow (196%) and ZieschC and Rische (1973) among others have
observed changes in serial isolates from the same persons. In vivo, it is
difficult to say why more than one type develops and to determine whether
it is due to reinfection, mixed initial infection, or spontaneous shift in
phage type. Extrachromosomal DNA may lead to phage restriction,
modification, or development of new phage receptors (Grunow, 1965e).
D. Typing method and reproducibility
Phage typing procedures must be carefully standardised. Pattern variation may be due to trivial technical evaluations. It is often preferable
to type strains which are to be compared simultaneously. Thereby, the
influence of slight differences in phage suspension densities and typing
medium are avoided.
With meticulous standardisation of technique, phage typing of Shigella
is sufficiently reproducible to be of considerable epidemiological value.
VIII. EPIDEMIOLOGICAL USE
Typing must be assessed on achievements in elucidating epidemics.
One relevant parameter of usefulness is the percentage of typable strains.
Aldovi and Sucha (1973) with the Hammarstrom phages obtained 17%
non-typable S. sonnei. High numbers of apparently untypable strains result
if the strains are not clearly in phase 11, i.e. only R-form colonies (not
phase I + S-form) are to be selected, Passage on endo-agar up to five times
may be necessary to obtain proper colonies (Grunow, 1965d). With such
precautions, bona Jide untypable strains are extremely rare as evidenced
from the number of reports with 100% typability.
A clinically relevant question is whether there is any correlation between
phage type and either antibiotic sensitivity or pathogenicity. There have
been reports that certain lysotypes have particular antibiotic sensitivity
patterns or biotypes but this applies only within restricted geographical
areas (Szturm-Rubinsten et al., 1974). Accordingly, such a relationship is
probably more apparent than real, due to local dominance of one or a few