Læknablaðið - 01.12.1978, Qupperneq 66

200
LÆKNABLAÐIÐ
the low rates are very interesting re-
presenting as they do communities which
are somehow protected from the environ-
mental carcinogens which produce cancers
elsewhere. Summation suggests that there
could be communities with fantasticaliy
high rates for human cancer and others in
which cancer is virtually unknown.
If you wish to study the clinical aspects,
pathology and treatment of a particular
cancer, then choose an area where that
cancer commonly occurs. But in the hunt
for causative factors, the uncommon indi-
vidual who developes a cancer rare in his
or her community may be a more reward-
ing subject for study. As attention is
focused on specific individual cancers, then
the inadequacy of national and inter-
national statistics become the more obvious,
especially if based on mortality. The focus
must be on living individuals with a speci-
fic cancer. This is indeed the newer epi-
demiology of cancer on which I hope to
dwell. We have seen its success in the case
of hepatocellular carcinomas in which a
number of specific causative factors have
now been idendified. It has achieved a
great triumph with the incrimination1617 6
of di-ethyl-stilbestrol with vaginal adeno-
carcinoma in young women, the offspring
of women to whom in the pregnancy that
produced the victim the chemical had been
administered. The sequence of events in
this discovery is worth noting, in routine
histopathology a small sequence of distinc-
tive cancer is seen in a group of patients
in whom this cancer is rarely found,
followed by detailed studies, not only of
the victims themselves, but of the families,
especially the mothers leading to the identi-
fication of the causal factor, and one
which it should be noted can be avoided,
leading to a decline in the incidence of this
cancer without regard to the patho-
genesis.18 The investigation began where
today such investigators should start, in a
routine auxilliary diagnostic department
when the patient is alive and families are
cooperative.
So let us with this in mind turn to H.D.
The classic study is that of Dr. Martin
Uddstromer in Sweden.47 Recently, we
were pleased to receive a letter from this
pioneer in H.D. epidemiology who in this
study remarked on the surgeon described
by Sisto44 who cut his hand when opening
an H.D. gland and died of H.D. very
rapidly. This influenced Uddstromer to
feel that H.D. might be an infectious dis-
ease. Later in a brilliant series of studies,
MacMahon-1'30 drew attention to the bi-
modal nature of the age incidence curve
of H.D. in the US., true as Clemmesen3
showed of many countries, but seemingly
not of Japan where the young adult peak
appeared to be missing. In the countries
cited there was no peak in childhood but
MacMahon was aware that such peak could
occur. He then made the stimulating and
challenging suggestion that perhaps what
we identify as H.D. subsumes two diffe-
rent entities, an infection in the young, a
cancer in the old.
Meanwhile, various long term studies of
mortality and survival in H.D. had re-
emphasised Rosenthal’s conclusions42 as to
the relationship between survival and the
lymphocyte population and Lukes27 had
created the system of subtypes of histo-
logy, replacing the Jackson-Parker system
which was subsequently modified into the
Rye system28. This includes four subtypes,
two of good prognosis — the lymphocyte
predominant and the nodular sclerosing,
and two of poor prognosis, the mixed cellu-
larity and the lymphocyte depleted type.
Both these it should be noted reflect pro-
gnosis in natural survivorship and after
radiotherapy but not with chemotherapy,
this latter point being evident from the
work of Ziegler et alB0 in Uganda where
it was found that the Rye type distributions
both of children2 and adults35 differed
widely from that of the U.S. with a pre-
ponderance of the unfavourable types in
Africa, and also in S. America.5
Meanwhile Miller,32 in the age incidence
curve of cancer mortality in the U.S. had
noted that the male predominance, seen in
all countries and marked in the U.S., was
least evident in the U.S. in the later years
of childhood when the rates for boys and
girls were rising rapidly. With an average
survival then of 2—7 years, this pointed