Clin.
exp.
Immunol.
(1984)
57,
511-519.
Critical
analysis
of
T
cell
subset
and
function
evaluation
in
patients
with
persistent
generalized
lymphadenopathy
in
groups
at
risk
for
AIDS
M.
CAVAILLE-COLL,*
A.
MESSIAHt
D.
KLATZMANN,*$
W.
ROZENBAUM,t
DOMINIQUE
LACHIVER,§
S.
KERNBAUM,T
EDITH
BRISSON,*
FRANCOISE
CHAPUIS,*
CATHERINE
BLANC,$
P.
DEBRE$
&
J.-C.
GLUCKMAN*
*
Laboratoire
d'Immunologie
Niphrologique
et
de
Transplantation,
UER
Pitwi-Salpetriere;
t
Departement
de
Sante
Publique;
$
Laboratoire
d'Immunologie
Cellulaire
et
Tissulaire,
Groupe
Hospitalier
Pitie-Salpitriere;
§
Institut
Arthur
Vernes
and
¶
Clinique
des
Maladies
Infecticuses
et
Tropicales,
H6pital
Claude
Bernard,
Paris,
France
(Acceptedfor
publication
11
April
1984)
SUMMARY
In
order
to
determine
whether
there
exists
a
distinct
immunological
profile
that
discriminates
between
healthy
homosexual
men
and
those
expressing
persistent
genera-
lized
hyperplastic
lympadenopathy
GHL
and
which
of
the
observed
abnormalities
in
GHL
would
be
most
indicative
of
patients
who
would
eventually
develop
AIDS,
we
evaluated
T
lymphocyte
subset
and
in
vitro
proliferative
responses.
The
respective
profiles
of
44
symptom
free
homosexual
men
(HG),
42
patients
with
GHL
and
30
cases
of
AIDS
were
compared
to
that
of
blood
bank
volunteers
(BBV)
and
to
each
other,
using
stepwise
discrimination
analysis.
HC,
GHL
and
AIDS
patients
demonstrated
upon
their
first
investigation
varying
degrees
of
the
same
type
of
abnormal
immunological
character-
istics.
The
percentage
of
OKT4+
cells
was
the
most
discriminant
variable
between
BBV
and
HC,
GHL
or
AIDS.
GHL
was
distinguished
from
HC
primarily
by
a
higher
percentage
of
OKT8+
cells
and
from
AIDS
by
higher
T
lymphocyte
counts,
while
the
OKT4+
cell
count
discriminated
between
AIDS
and
HC.
However,
none
of
these
major
discriminant
variables
could
classify
correctly
more
than
71%
within the
HC
and
the
two
patient
groups.
Thus,
there
exists
a
continuity
between
GHL
and
AIDS
in
the
range
of
observed
immunological
abnormalities
without
clearcut
boundaries.
This,
together
with
the
existence
of
abnormal
background
values
in
HC
will
not
permit
a
single
investigation
of
T
lymphocyte
immune
parameters
to
provide
substantial
diagnostic
and
prognostic
evidence.
Keywords
acquired
immune
deficiency
syndrome
T
lymphocytes
lymphadenopathy
homosexuals
INTRODUCTION
The
acquired
immune
deficiency
syndrome
(AIDS)
is
characterized
by
a
major
defect
in
helper
inducer
T
lymphocyte
numbers
and
functions
(Gottlieb
et
al.,
1981;
Mildvan
et
al.,
1982;
Rogers
et
al.,
1983)
that
results
in
high
suspectibility
to
opportunistic
infections.
This
syndrome
has
now
been
Correspondence:
Dr
Jean-Claude
Gluckman,
Laboratoire
d'Immunologie
Nephrologique
et
de
Transplan-
tation,
UER
Pitie-Salp&triere,
91
bd
de
l'H6pital,
Paris,
France.
5
I"
shown
to
affect
various
distinct
high
risk
groups
(Centers
for
Disease
Control
Task
Force,
1
982a,
1982b)
including
male
homosexuals
(Centers
for
Disease
Control
Task
Force,
1982c).
Besides
this
well
defined
disease,
a
variety
of
chronic
but
non-specific
symptoms
such
as
persistent
generalized
hyperplastic
lymphadenopathy
(GHL)
recurrent
fever,
diarrhoea
and
weight
loss,
have
also
recently
been
observed
with
increasing
frequency
(Centers
for
Disease
Control
Task
Force,
1982d).
Currently
it
is
not
yet
known
whether
this
syndrome
that
some
have
called
an
AIDS
related
symptom
complex,
which
occurs
among
the
same
high
risk
populations,
is
related
to
AIDS
either
as
a
precursor
and/or
as
a
milder
form,
or
even
whether
it
shares
the
same
causative
agent
(Metroka
et
al.,
1983).
Although
the
clinical
definition
of
both
syndromes
is
different,
the
overall
pattern
of
biological
abnormalities
found
in
GHL
resembles
that
seen
in
patients
with
AIDS
(Amman
et
al.,
1983),
with
imbalance
of
T
lymphocyte
subsets
and
impairment
of
their
functions.
The
purpose
of
this
study
was
to
address
the
questions
as
to
whether
there
exists
a
distinct
cellular
immunological
profile
that
discriminates
between
healthy
and
GHL
expressing
homosex-
uals
and
which
of
the
observed
abnormalities
would
be
most
indicative
of
patients
who
would
eventually
develop
AIDS,
in
light
of
the
fact
that
most
of
the
immune
parameters
are
interrelated
and
must
provide
redundant
information.
MATERIALS
AND
METHODS
Patients.
Seventy-two
individuals
were
referred
to
us
for
investigation
between
January
1982
and
October
1983.
Fourty-two
individuals
from
groups
at
risk
for
AIDS
had
multiple
extra-inguinal
hyperplastic
lymph
nodes
with
no
underlying
cause
of
disease,
persisting
for
at
least
3
months,
with
or
without
other
non-specific
constitutional
symptoms
and
were
defined
as
GHL.
Thirty
suffered
from
confirmed
AIDS
according
to
the
definition
of
the
Centers
for
Disease
Control
Task
Force
(1982c).
Among
these,
15
had
only
Kaposi's
sarcoma
(KS),
12
had
opportunistic
infections
(01)
and
three
had
both.
Since
several
samplings
were
performed
over
the
course
of
the
disease,
and
because
some
patients
initially
seen
with
one
manifestation
later
developed
another,
only
the
initial
sampling
and
its
contemporary
diagnosis
were
taken
into
account
for
this
analysis.
All
patients
included
in
this
study
have
been
followed
at
Piti6-Salpetriere
and
Claude
Bernard
Hospitals,
where
most
of
the
French
AIDS
patients
have
been
hospitalized
up
to
now.
The
main
characteristics
of
the
groups
studied
are
presented
in
Table
1.
All
patients
with
GHL,
except
for
one
female
intravenous
drug
user,
were
homosexual
men
who
spontaneously
consulted
our
patient
clinic
for
their
symptoms
and
the
fear
of
AIDS.
Control
groups.
Fourty-four
male
homosexuals,
who
had
no
history
of
lympadenopathy,
KS
or
01,
were
included
for
comparison
in
the
study.
Twenty-eight
were
selected
from
patients
attending
a
clinical
for
sexually
transmitted
diseases
(STD).
Only
healthy
individuals
or
those
presenting
with
Table
1.
Main
characteristics
of
the
groups
studied
Diagnosis
at
initial
sampling
Characteristics
of
groups
studied
GHL
AIDS
HC
BBV
No
of
cases
42
30
44
50
Mean
age:
years
(range)
31
(20-47)
34
(24-54)
30
(20-55)
36
(18-59)
%
males
98
93
100
70
%
homosexuals
98
87
100
?
%
Haitians
0
7
0
0
%
Black
Africans
0
7
0
0
%
heroin
addicts
5
0
0
0
M.
Cavaille-Coll
et
al.
512
T
cell
subset
analysis
in
AIDS
risk
patients
513
gonorrhoea
(n
=
5)
but
with
no
other
active
STD
were
included
in
this
group.
None
had
active
syphilis
as
determined
by
a
Kline
test,
quantitative
TPHA
and
FTA.
Any
individual
with
skin
erruptions
suggestive
of
active
Herpes
simplex
virus
infection
or
with
clinical
symptoms
of
viral
hepatitis
was
excluded
from
this
group.
Sixteen
were
volunteers
found
among
friends
of
AIDS
patients,
11
of
whom
were
past
or
present
sexual
contacts
of
such
patients.
Fifty
blood
bank
volunteers
(BBV)
studied
over
the
same
period
were
used
to
establish
the
normal
values
for
our
laboratory
procedures.
Women
were
not
excluded
from
this
control
group
since
their
immunological
parameters
did
not
differ
significantly
from
those
of
men
(data
not
shown).
Methods.
Peripheral
blood
mononuclear
cells
(PBMC)
were
separated
from
a
sample
of
heparinized
fresh
blood
by
centrifugation
over
Ficoll-hypaque.
T
lymphocyte
subsets
were
quantitated
by
indirect
immunofluorescence
staining
on
freshly
prepared
PBMC,
using
a
Leitz
Orthoplan
microscope
with
epifluorescence
or
by
using
a
Spectrum
III
cell
analyzer
(Ortho
Diagnostic
System,
Raritan,
New
Jersey,
USA.).
OKT3
+
(T
cells),
OKT4+
(helper
inducer
T
cells)
and
OKT8+
(suppressor/cytotoxic
T
cells)
lymphocytes
were
thus
measured
with
the
help
of
specific
monoclonal
antibodies
(Ortho
Diagnostics)
(Reinherz
&
Schlossman,
1980).
Only
the
31
GHL,
27
AIDS
patients
and
34
homosexual
controls
(HC)
who
had
a
complete
blood
count
with
differential
performed
on
the
day
of
the
first
sampling
were
included
in
the
quantitative
analysis
of
T
lymphocyte
subpopulations.
PBMC
were
cryopreserved
for
later
in
vitro
study
of
proliferative
responses
as
previously
described
(Oppenheim
&
Schecter,
1976;
Klatzmann
et
al.,
1983).
After
thawing
they
were
stimulated
for
3
days,
using
a
microculture
technique,
by
optimal
concentrations
of
PHA
(PHA-M
DiFco),
Con
A
(Miles)
or
pokeweed
mitogen
(PWM,
GIBCO)
or
for
6
days
by
a
pool
of
20
Gy
irradiated
lymphocytes
from
10
different
donors
(MLR).
Proliferative
responses
were
measured
by
the
incorporation
of
3H-thymidine
at
the
end
of
the
culture
period.
The
logio
mean
ct/min
calculated
for
each
triplicate
culture
were
compared
to
daily
normal
controls
from
the
Laboratory's
panel,
as
well
as to
normal
results
obtained
during
the
whole
period
of
the
study.
PBMC
from
the
first
sample
of
29
GHL,
27
AIDS,
44
HC
and
44
BBV
were
available
for
analysis
of
these
results.
Statistics.
Mean
relative
and
absolute
numbers
of
T
lymphocyte
populations,
lymphocyte
counts
and
OKT4+/OKT8+
ratios
for
each
clinical
entity
were
compared
by
using
the
standardized
normal
deviate
mean
test
or
variance
analysis
(Armitage,
1983).
Because
of
the
large
variations
observed
in
day
to
day
experiments,
proliferative
responses
were
converted
into
qualitative
variables
for
comparisons
between
each
group.
Thus,
the
cut-off
between
'normal'
and
'reduced'
responses
was
chosen
as
the
value
equal
to
the
mean
minus
one
standard
deviation
of
normal
controls.
Chi
square
analysis
was
then
used
to
detect
differences
in
proliferative
responses.
Since
several
correlated
variables
could
be
significantly
different
between
groups,
stepwise
discriminant
analysis
was
performed
in
order
to
determine
which
immunological
variables
were
the
most
discriminant
between
the
groups
studied.
Qualitative
variables
were
included
according
to
D.
B.
Suits'
rules
(Suits,
1957).
This
analysis
was
performed
by
using
the
'Stepwise
discriminant
analysis'
program
(BMD
07
M,
Health
Sciences
Computing
Facility,
UCLA)
as
revised
on
April
1972,
on
a
CIBER
550
(Control
Data
Corporation)
computer.
RESULTS
All
the
results
from
this
study
are
summarized
in
Figs
1
&
2.
Selection
of
controls
All,
but
one,
of
the
GHL
patients
were
male
homosexuals.
Since
an
AIDS
like
pattern
of
abnormalities
has
already
been
described
in
symptom
free
homosexuals
(Kornfeld
et
al.,
1982;
Stahl
et
al.,
1982;
Wallace
et
al.,
1982;
Ammann
et
al.,
1983),
results
obtained
in
GHL
should
be
compared
to
those
observed
in
a
homosexual
control
population.
It
was,
however,
difficult
to
determine
what
might
constitute
such
a
control
population.
We
therefore
chose
individuals
from
514
M.
Cavaille-Coll
et
al.
3
0
(a)
100
(b)
1.0
T%5
T
2
!
05
LW5
0
LW
LWJ
L±
LWLW
L
OKT3`
OKT4*OKT8
OKT3+OKT4+
OKT8+
Fig.
1.
Phenotypes
of
peripheral
blood
lymphocytes
(PBL)
in
patients
and
control
group.
(a)
Mean
and
s.d.
of
absolute
counts
of
lymphocytes
and
their
subsets.
(b)
Mean
and
s.d.
of
percentages
of
T
lymphocyte
subsets.
Hatched
areas
represent
intervals
of
1
s.d.
from
mean
of
normal
blood
bank
volunteers.
O=HC
(n=
34);
*=GHL
(n=
31);
0
=AIDS
(n=27).
%
individuals
with
normal
responses
0
25
50
75
100
r
--I
----I
I-
----
nl
PHA
Con
A
PWM
MLR
-4
(n:39)
(n=23)
(n
0)
(n=30)
(n=
19)
(n=23)
(
n-231
Fig.
2.
In
vitro
proliferative
responses
to
mitogens
and
allogeneic
lymphocytes:
percentage
of
normal
responding
cases
in
patient
and
control
groups
and
95%
confidence
limit.
U
=
BBV
(n
=44);
U
=
HC
(n
=
44);
U
=
GHL
(n
=
29);
0
=
AIDS
(n
=
27).
different
groups
of
comparable
age
and
life
style
with
the
patients.
Biological
results
of
sexual
and
non-sexual
friends
of
AIDS
patients
did
not
differ
significantly
with
each
other.
This
group
of
'contacts'
tended,
however,
to
have
higher
percentages
of
OKT3+
and
OKT4+
cells
than
the
controls
from
the
STD
clinic,
although
absolute
counts
did
not
differ
significantly.
In
spite
of
these
differences,
we
chose
not
to
divide
this
homosexual
control
group
(HG)
on
the
basis
of
immunological
parameters
alone,
all
the
more
since
it
was
uncertain
to
what
extent
patients
from
the
clinic
had
themselves
been
sexual
or
non-sexual
friends
of
AIDS
patients.
Up
to
now,
none
of
these
HC
has
developed
GHL
or
AIDS
(Table
2).
The
HC
differed
from
the
BBV
mainly
by
lower
percentages
and
absolute
counts
of
OKT4+
lymphocytes
(P
<
10-).
However,
due
to
the
variability
of
results
from
one
individual
to
another,
and
to
a
slight
but
not
significant
increase
in
the
mean
peripheral
blood
lymphocyte
counts
(PBL)
in
el
;!.
W/1-
el
E
PI
10;1;1
I
.1;1
I
.,
II";,
II,'
.,
I
I
V'11141*
T
cell
subset
analysis
in
AIDS
risk
patients
Table
2.
Follow-up
of
HC
and
GHL
Range
of
follow-up
period
between
initial
sampling
Group
and
January
1984
Status
HC
5-7
months
No
cases
of
AIDS
or
GHL
(n
=
44)
1
case
of
metastatic
carcinoma
at
6
months
from
initial
sampling
GHL
4-21
months
2
cases
of
AIDS
at
4th
and
(n
=42)
8th
month
from
initial
sampling
5I5
HC,
this
was
accompanied
only
by
a
significant
decrease
of
the
percentages
of
OKT3
+
cells
(P<
10-6)
but
not
of
their
absolute
counts.
Similarly,
a
deficit
in
T
cell
proliferative
responses
was
observed
more
often
in
this
group
than
in
BBV
for
PHA
(P<2
x
10-2),
Con
A
(P<
5
x
10-7),
but
not
for
the
MLR.
GHL
patients
Whatever
the
test
utilized,
an
important
dispersion
of
individual
results
was
observed
among
these
patients
indicating
that
this
may
be
a
heterogeneous
group.
Distribution
of
T
cell
subsets
was
abnormal
when
compared
to
BBV.
Lower
percentages
and
absolute
counts
of
OKT4+
cells
(P
<
l0-5)
and
increased
OKT8+
values
(P
<
10-2)
led
to
an
abnormal
OKT4+/OKT8+
ratio
(0-8
+
0-5
vs
1-6
+
0-5;
P
<
10-9).
However,
GHL
patients
differed
from
HC
only
by
higher
OKT3+
and
OKT8+
percentages
(P
<
l0-3).
Only
when
comparisons
were
done
controlling
the
levels
of
PBL
counts,
did
the
increase
in
absolute
counts
of
OKT8+
cells
reach
significance.
The
frequency
of
low
responders
to
proliferative
stimuli
was
no
more
important
in
GHL
patients
than
in
HC.
AIDS
patients
Like
those
described
in
other
studies
(Gottlieb
et
al.,
1981;
Masur
et
al.,
1981;
Mildvan
et
al.,
1982;
Friedman-Kien
et
al.,
1982;
Rogers
et
al.,
1983;
Ammann
et
al.,
1983)
our
patients
with
AIDS
had
a
decrease
in
the
number
of
PBL
(P
<
l0-3),
T
lymphocytes
(P
<
10-3),
and
especially
in
OKT4+
cells
(P
<
10-9)
with
a
marked
reduction
of
OKT4+/OKT8
+
ratio
(0-9
+
1-0;
P
<
10-3),
when
compared
to
BBV.
Differences
in
PBL
(P<
102),
OKT4+
cells
(P<
10-2)
and
OKT3+
cell
counts
(P<
5
x
10-2)
were
still
noted
as
compared
to
HC.
No
significant
differences
between
values
observed
in
patients
with
01
or
with
KS
were
observed
although
01
had
usually
lower
counts
of
PBL,
T
lymphocytes
and
OKT4+
subset
cells
(data
not
shown).
Although
approximately
two-thirds
of
these
patients
had
subnormal
or
null
proliferative
responses
(P
<
10-4-10-6
as
compared
to
BBV)
these
proportions
were
statistically
different
from
those
observed
in
HC
only
for
PHA
(P
<
5
x
10
3)
and
MLR
(P
<
2
x
10-2).
Comparison
of
GHL
vs
AIDS
patients
Similar
patterns
of
abnormalities
occured
in
both
syndromes,
but
AIDS
patients
could
be
characterized
by
the
more
profound
reduction
in
the
counts
of
OKT4+
cells
(P
<
25
x
10-2)
without
modification
of
the
OKT8
+
cell
subset,
which
was
reflected
by
low
T
lymphocyte
(P
<
10-2)
and
PBL
(P<
10-2)
counts.
Mean
OKT4+/OKT8+
ratios
were
decreased
in
both
patients
groups
as
compared
to
BBV,
but
the
individual
values
demonstrated
a
great
dispersion
and
did
not
discriminate
between
them.
516
M.
Cavaille-Coll
et
al.
The
frequency
of
low
responders
in
functional
tests
was
only
higher
for
PHA
(P
<
2
x
10-2)
and
MLR
(P<
2
x
10-2)
responses.
Stepwise
discriminant
analysis
An
initial
stepwise
discriminant
analysis
among
the
four
groups
correctly
classified
only
59%
of
the
subjects.
We
therefore
decided
to
compare
them
two
by
two.
Because
proliferative
response
data
were
obviously
not
discriminant
between
HC,
GHL
and
AIDS,
analysis
was
performed
without
those
variables.
Results
are
shown
in
Table
3.
When
HC
or
GHL
were
compared
to
BBV,
the
sole
most
discriminant
variable
was
the
percentage
of
OKT4+
cells,
leading
to
a
correct
classification
of
85-90%
of
the
subjects.
Table
3.
Stepwise
discriminant
classification
results
%
of
correctly
Analysis
Midpoint
value
classified
cases
Discriminant
for
Group
I
Group
2
variables
classification§
Group
1
Group
2
Total
HC
Vs
BBV
%T4+
43%
82
90
87
GHL
Vs
BBV
%T4+
42%
77
92
86
GHL
Vs
HC
%T8+
41%
55
74
65
GHL
Vs
HC
%T3+*
71
71
71
GHL
vs
AIDS
No.
T3+/p1
1,200
64
74
69
AIDS
Vs
BBV
%T4+
42%
74
92
85
AIDS
Vs
BBV
T4+/T8+
74
96
88
AIDS
Vs
BBV
%T8+
78
96
90
AIDS
Vs
BBV
No.
PBL/plt
81
98
92
AIDS
Vs
HC
No.
T4+/kl
500
67
59
62
AIDS
Vs
HC
%T3+t
67
65
66
Discriminant
formula:
*
0-07
(%T8+)+0004
(%T3+)-5
6=0
t
0-24
(%T4+)-3-78
(T4+/T8+)-0.12
(%T8+)
+0-001
(lymphocytes/Pl)
-2-41
=0
t
0
0035
(T4+/.u)-0-049
(%T3+)+
1
51
=0
§
As
obtained
by
resolution
of
discriminant
equation
at
first
step
(one
variable
entered).
Comparison
of
GHL
to
HC
showed
that
OKT8+
and
OKT3+
percentages
were
the
most
discriminant,
but
this
analysis
ended
up
by
correctly
classifying
only
two-thirds
of
the
subjects.
The
same
method
applied
to
AIDS
vs
BBV
showed
that
%
OKT4+
and
OKT8+
ratio
and
PBL
counts
were
the
most
discriminant
variables,
classifying
correctly
92%
of
the
cases.
On
the
other
hand
when
AIDS
and
HC
groups
were
compared,
absolute
OKT4+
and
relative
OKT3+
numbers
were
retained
as
discriminant
variables,
but
this
produced
correct
classification
in
only
66%.
Finally,
the
sole
discriminant
variable
between
GHL
and
AIDS
was
absolute
T
cell
count.
Worthy
of
note,
however,
is
the
fact
that
most
of
the
eleven
GHL
patients
who
were
uncorrectly
classified
as
AIDS
by
this
analysis,
suffered
at
that
time
from
the
most
pronounced
form
of
what
are
claimed
to
the
prodroms
of
AIDS:
fever,
diarrhoea,
weight
loss,
etc.
The
two
individuals
among
the
GHL
who
subsequently
developed
AIDS
came
from
this
group
(Table
2).
Moreover,
when
taken
alone,
none
of
the
major
discriminant
variables
selected
by
this
analysis
could
be
used
to
correctly
classify
most
of
the
HC
or
the
GHL
patients
(Table
4),
and
at
least
26%
of
AIDS
patients
did
not
present
with
one
of
these
abnormalities.
T
cell
subset
analysis
in
AIDS
risk
patients
Table
4.
Classification
of
patients
and
controls
according
to
the
major
discriminant
parameters
HC
GHL
AIDS
Groups
studied
No.
%
No.
%
No.
%
OKT3+
cells
<1,200
/1l
17
50
11
35
20
74
OKT4+
cells
<500
/yl
14
41
12
39
18
67
%
OKT4+
cells
<42%
26
76
23
74
20
74
%
OKT8+
cells
>41%
9
26
17
55
16
59
Total
number
of
subjects
in
this
group
34
31
27
517
DISCUSSION
Immunological
abnormalities
have
been
described
in
many
clinical
disorders
found
in
homosexual
men,
even
in
symptom
free
individuals
(Ammann
et
al.,
1983;
Fahey,
Detels
&
Gottlieb,
1983;
Pinching
et
al.,
1983).
The
purpose
of
this
study
was
to
delineate
a
cellular
immune
reponse
profile
characteristic
of
GHL
and
to
compare
this
profile
with
those
of
a
normal
control
population,
a
homosexual
control
population
and
a
group
of
AIDS
patients.
This
study
reflects
the
situation
in
essentially
male
homosexual
patients
with
GHL
or
AIDS
who
moreoever
had
a
history
of
multiple
STD
(J.
B.
Brunet,
Ministere
de
la
Sante,
personal
communication).
Therefore
it
was
appropriate
to
select
apparently
healthy
homosexual
volunteers
recruited
among
the
friends
of
patients,
as
well
as
homosexual
men
consulting
a
STD
clinic
which
has
referred
many
GHL
and
AIDS
patients
to
our
hospitals.
Short
of
a
strict
case-control
study,
the
group
assembled
here
gives
the
best
representation
of
the
life
style
of
the
majority
of
our
patients.
In
both
control
groups
frequent
significant
abnormalities
of
cellular
immunity
were
observed.
For
example,
T
lymphopenia
occured
in
50%
of
these
subjects,
decreased
OKT4+
cell
counts
in
41%,
while
the
%
of
OKT4+
cells
were
decreased
in
76%
and
%
of
OKT8+
augmented
in
26%
of
them.
Reduced
proliferative
responses
were
also
frequent
in
this
group.
These
results
are
comparable
with
those
of
similar
populations
in
high
(Kornfeld
et
al.,
1982;
Stahl
et
al.,
1982;
Wallace
et
al.,
1982;
Detels
et
al.,
1983)
and
low
incidence
areas
for
AIDS
and
GHL
in
Europe
and
the
United
States
(Rinaldo
et
al.,
1984;
Pinching
et
al.,
1983).
The
disturbances
reported
here
have
also
been
described
in
a
variety
of
infections
with
viruses
from
the
Herpes
group
(Reinherz
et
al.,
1980;
Rinaldo
et
al.,
1980;
Rubin,
Carney
&
Schooley,
1981)
which
are
known
to
occur
with
high
incidence
in
this
population
(Drew
et
al.,
1981;
Mintz
et
al.,
1983).
However,
these
transformations
are
usually
transitory
unless
there
is
reinfection
and
they
are
associated
with
an
increase
in
the
absolute
number
of
T
lymphocytes
and
OKT8
+
cells
in
the
peripheral
blood
(Reinherz
et
al.,
1980;
Rinaldo
et
al.,
1980;
Rubin
et
al.,
1981).
Similar
data
on
viral
infections
of
male
homosexuals
in
France
are
not
yet
available.
Among
the
HC
subgroup
of
16
AIDS
friends,
one
sexual
contact
had
serological
evidence
of
active
Epstein-Barr
(EB)
virus
infection
without
lymphadenopathy
or
other
symptoms.
The
remaining
15
had
a
normal
positive
serological
profile
indicating
past
exposure
to
the
virus
(A.
Messiah
&
W.
Rozenbaum,
unpublished
observation).
The
28
individuals
from
the
STD
clinic
were
not
serologically
screened
for
active
EB
virus
or
cytomegalovirus
(CMV)
infection.
However,
a
parallel
study
of
homosexual
males
from
the
same
clinic
found
94%
positive
for
anti-CMV,
IgG
(F.
Brun-Vesinet,
personal
communication,
1984).
Therefore
the
abnormal
base
values
for
the
symptom
free
homosexual
population
may
be
a
reflection
of
a
high
prevalence
of
these
infections
within
this
group.
It
is
possible
that
such a
state
of
relative
immune
suppression
is
necessary
for
the
development
of
an
AIDS
agent,
whether
these
anomalies
are
themselves
the
consequence
of
infections
or
of
a
particular
immunological
terrain.
Alternatively,
they
may
reflect
the
high
prevalence
within
the
5I8
M.
Cavaille-Coll
et
al.
homosexual
population
of
a
putative
aetiological
agent
for
AIDS
which
disrupts
in
vitro
immunological
parameters
of
otherwise
assymptomatic
individuals.
However,
the
absence
of
any
of
these
immunologic
abnormalities
in
another
high
risk
groups,
namely
Haitian
immigrants
living
in
New
York,
would
argue
against
these
latter
two
hypothesis
(Nicholas
et
al.,
1983).
Statistical
analysis
of
T
cell
parameters
of
the
two
patients
groups
compared
to
BBV
showed
a
decrease
in
OKT4+
cell
counts
and
percentages,
the
greatest
reduction
being
found
in
AIDS.
OKT8
cells
were
augmented
only
in
GHL.
Therefore
AIDS
tended
to
be
associated
with
lower
PBL
and
T
lymphocyte
count.
The
most
severe
functional
defects
were
moreover
found
in
AIDS
patients
where
proliferative
responses
to
all
stimuli
were
abnormal,
although
many
GHL
patients
were
also
found
to
display
low
responsiveness.
Compared
to
HC,
however,
the
distinct
T
cell
immunological
profile
of
GHL
consists
primarily
of
a
statistically
significant
increase
in
the
OKT8+
lymphocyte
subset
which
is
close
to
what
has
been
described
as
T
cell
augmentation
(Fahey
et
al.,
1983).
This
group
is
further
clearly
distinguished
from
AIDS
by
a
lesser
reduction
of
the
OKT4+
subset
with
conservation
of
PBL
and
T
lymphocyte
counts.
It
is
notable
that
in
GHL
patients,
the
two
who
subsequently
developed
AIDS
had
presented
with
the
most
severe
abnormalities
from
the
time
of
their
first
sample.
Since
GHL
most
likely
constitute
a
heterogeneous
group
of
which
only
a
fraction
is
at
risk
for
developing
AIDS,
these
observations
therefore
raise
the
problem
of
the
biological
definition
of
a
subgroup
of
patients
who
display
the
more
pronounced
immune
defects
in
vitro.
Only
epidemiological
and
long
term
follow-up
studies
will
determine
whether
other
patients
sharing
these
features
evolve
in
the
same
manner.
This
does
not
exclude
that
GHL
and
AIDS
may
be
considered
as
a
successively
more
severe
manifestation
of
the
same
disorder.
Indeed,
if
this
is
so,
one
would
expect
to
observe
such
a
continuity
of
the
laboratory
parameters
between
these
two
groups.
The
mere
fact
that,
within
each
group,
initial
investigations
occurred
at
points
which
differed
from
one
individual
to
the
other
with
respect
to
the
course
of
the
disease,
further
explains
the
dilution
of
the
boundaries
of
each
stage.
Recognition
of
different
immunologic
patterns
in
AIDS
and
other
related
disorders
would
have
important
implications
in
relation
to
diagnosis,
prognosis
and
treatment.
However,
even
stepwise
discriminant
analysis
could
not
frankly
delineate
discrete
groups
among
homosexual
men
either
well
or
ill.
Thus
when
a
homosexual
patient
first
consults
for
GHL
or
fear
of
AIDS,
in
the
absence
of
clear
clinical
signs
in
favour
of
the
latter,
a
single
investigation
of
T
lymphocyte
immunological
parameters
is
unlikely
by
itself
to
provide
substantial
diagnostic
or
prognostic
evidence.
We
are
indebted
to
Professor
A.
J.
Valleron
and
to
Dr
J.
Y.
Mary
who
graciously
provided
computer
facilities
and
thoughtful
advice.
This
work
was
supported
by
a
grant
from
the
Ministere
de
l'Industrie
et
de
la
Recherche
(No.
83-C.
1414)
and
funded
in
part
by
the
University
Paris
VI.
REFERENCES
AMMANN,
A.J.,
ABRAMS,
D.,
CONANT,
M.,
CHUDWIN,
D.
COWAN,
M.,
VOLBERDING,
P.,
LEWIS,
B.
&
CASAVANT,
C.
(1983)
Acquired
immune
dysfunc-
tion
in
homosexual
men:
immunologic
profiles.
Clin.
Immunol.
Immunopath.
27,
315.
ARMITAGE,
P.
(1983)
Statistical
methods
in
Medical
research.
Blackwell
Scientific
Publications,
Oxford.
CENTERS
FOR
DISEASE
CONTROL
TASK
FORCE
(1982a)
Opportunistic
infections
and
Kaposi's
sarcoma
among
Haitians
in
the
United
States.
MMWR,
31,
353,
360.
CENTERS
FOR
DISEASE
CONTROL
TASK
FORCE
(1982b)
Update
on
acquired
immune
deficiency
syndrome
(AIDS)
among
patients
with
hemophilia
A.
MMWR,
31,
644.
CENTERS
FOR
DISEASE
CONTROL
TASK
FORCE
(1982c)
Special
report:
epidemiologic
aspect
of
the
current
outbreak
of
Kaposi's
sarcoma
and
opportunistic
infections.
N.
Engl.
J.
Med.
306,
248.
CENTERS
FOR
DISEASE
CONTROL
TASK
FORCE
(1982d)
Persistent
generalized
lymphadenopathy
among
homosexual
males.
MMWR,
31,
249.
DETELS,
R.,
FAHEY,
J.L.,
SCHWARTZ,
R.S.,
GREENE,
B.R.,
VISHER,
B.R.
&
GOTILIEB,
M.S.
(1983)
Rela-
tion
between
sexual
practices
and
T-cell
subsets
in
homosexual
active
men.
Lancet,
i,
609.
DREW,
W.L.,
MINTz,
L.,
MINER,
R.C.,
SANDS,
M.
&
KETTERER,
B.
(1981)
Prevalence
of
cytomegalovirus
infection
in
homosexual
men.
J.
Infect.
Dis.
143,
188.
FAHEY,
J.L.,
DETELS,
R.
&
GOTTLIEB,
M.S.
(1983)
Immune
cell
augmentation
(with
altered
T-subset
ratio)
is
common
in
healthy
homosexual
men.
N.
Engl.
J.
Med.
308,
842.
T
cell
subset
analysis
in
AIDS
risk
patients
5I9
FRIEDMAN-KIEN,
A.E.,
LAUBENSTEIN,
L.J.,
RUBEN-
STEIN,
P.,
BUIMOvICI-KLEIN,
E.,
MARMOR,
M.,
STAHL,
R.,
SPIGLAND,
I.,
KIM,
K.S.
&
ZOLLA-
PAZNER,
S.
(1982)
Disseminated
Kaposi's
sarcoma
in
homosexual
men.
Ann.
Intern.
Med.
96,
693.
GOTTLIEB,
M.S.,
ScHRoFF,
R.,
SCHANKER,
H.M.,
WEISMAN,
J.D.,
FAN,
P.T.,
WOLF,
R.A.
&
SAXON,
A.
(1981)
Pnewnocystis
carinii
pneumonia
and
muco-
sal
candidiasis
in
previously
healthy
homosexual
men:
evidence
of
a
new
acquired
cellular
immuno-
deficiency.
N.
Engl.
J.
Med.
305,
1426.
KLATZMANN,
D.,
GLUCKMAN,
J.C.,
FOUCAULT,
C.,
BENSUSSAN,
A.,
ASSOGBA,
U.
&
DUBOUST,
A.
(1983)
Suppression
of
Lymphocyte
Reactivity
by
blood
transfusions
in
uremic
patients.
I.
Proliferative
responses.
Transplantation,
35,
332.
KORNFELD,
H.,
VAN
DE
STOwE,
R.A.,
LANGE,
M.,
REDDY,
M.M.
&
GRIEco,
M.H.
(1982).
T-lympho-
cyte
subpopulations
in
homosexual
men.
N.
Engl.
J.
Med.
307,
729.
MASUR,
H.,
MICHELIS,
M.A.,
GREENE,
J.B.,
ONOVATO,
I.,
VAN
DE
STOwE,
R.A.,
HOLZMAN,
R.S.,
WORMSER,
G.,
BRETTMAN,
L.,
LANGE,
M.,
MURRAY,
H.W.
&
CUNNINGHAM-RUNDLES,
S.
(1981)
An
outbreak
of
community-acquired
Pneumocystis
carinii
pneu-
monia:
initial
manifestation
of
cellular
immune
dysfunction.
N.
Engl.
J.
Med.
305,
1431.
METROKA,
C.E.,
CUNNINGHAM-RUNDLES,
S.,
POL-
LACK,
M.S.,
SONNABEND,
J.A.,
DAVIS,
J.M.,
GOR-
DON,
B.,
FERNANDEZ,
R.D.
&
MOURADIAN,
J.
(1983)
Generalized
lymphadenopathy
in
homosex-
ual
men.
Ann.
Int.
Med.
99,
585.
MILDVAN,
D.,
MATHUR,
U.,
ENLOW,
R.W.,
ROMAIN,
P.L.,
WINCHESTER,
R.J.,
COLP,
C.,
SINGMAN,
H.,
ADELSBERG,
B.R.
&
SPIGLAND,
I.
(1982)
Oppor-
tunistic
infections
and
immune
deficiency
in
homo-
sexual
men.
Ann.
Intern.
Med.
96,
700.
MINTZ,
L.,
DREw,
W.L.,
MINER,
R.C.
&
BRAFF,
E.H.
(1983)
Cytomegalovirus
infections
in
homosexual
men;
an
epidemiological
study.
Ann.
Int.
Med.
99,
326.
NICHOLAS,
P.,
MASCI,
J.,
DE
CATALOGNE,
J.,
SOLO-
MON,
S.,
BEKESI,
J.G.
&
SELIKOFF,
I.J.
(1983)
Immune
competence
in
Haitians
living
in
New
York.
N.
Engl.
J.
Med.
309,
1187.
OPPENHEIM,
J.J.
&
SCHECHTER,
B.
(1976)
Lymphocyte
transformation.
In
Manual
of
Clinical
Immunology
(ed.
by
N.
R.
Rose
&
H.
Friedman).
Am.
Soc.
Microbiol.
Washington,
DC.
PINCHING,
A.J.,
MCMANUS,
J.J.,
JEFFRIES,
D.J.,
MOSHTAEL,
O.,
DONAGHY,
M.,
PARKIN,
J.M.,
MUN-
DAY,
P.E.
&
HARRIS,
J.R.W.
(1983)
Studies
of
cellular
immunity
in
male
homosexuals
in
London.
Lancet,
Ai,
126.
REINHERZ,
E.L.,
O'BRIEN,
C.,
ROSENTHAL,
P.
&
SCHLOSSMAN,
S.F.
(1980)
The
cellular
basis
for
viral-induced
immunodeficiency:
analysis
by
monoclonal
antibodies.
J.
Immunol.
125,
1269.
REINHERZ,
E.L.
&
SCHLOSSMAN,
S.F.
(1980)
Regula-
tion
of
the
immune
response-inducer
and
suppres-
sor
T-lymphocyte
subsets
in
human
beings.
N.
Engl.
J.
Med.
303,
370.
RINALDO,
C.R.,
CARNEY,
W.P.,
RICHTER,
B.S.,
BLACK,
P.H.
&
HIRSCH,
M.S.
(1
980)
Mechanisms
of
immunosuppression
in
cytomegaloviral
mononuc-
leosis.
J.
Infect.
Dis.
141,
488.
RINALDO,
C.,
LYTER,
D.,
KINGSLEY,
L.,
DEBIASIO,
R.,
JIA-QI,
Z.,
TIEN,
L.,
JENSEN,
F.,
ATCHISON,
R.,
ZERBE,
A.
&
RABIN,
B.
(1
984)
Virologic,
immunolo-
gic
and
epidemiologic
associations
with
AIDS
among
gay
males
in
a
low
incidence
area.
Ann.
N.
Y.
Acad.
Sci.
(In
press.)
ROGERS,
M.F.,
MORENS,
D.M.,
STEWART,
J.A.,
KAMINSKI,
R.M.,
SPIRA,
T.M.,
FEORINO,
P.M.,
LARSEN,
S.,
FRANCIS,
D.,
WILSON,
M.,
KAUFMAN,
L.
AND
THE
TASK
FORCE
ON
ACQUIRED
IMMUNE
DEF-
CIENCY
SYNDROME
(1983)
National
Case.
Control
study
Kaposi's
sarcoma
and
Pneumocystis
carinii
pneumonia
in
homosexual
men.
II.
Laboratory
results.
Ann.
Intern.
Med.
99,
151.
RUBIN,
R.H.,
CARNEY,
W.P.
&
SCHOOLEY,
R.T.
(1981)
The
effect
of
infection
on
T-lymphocyte
subpopula-
tions:
a
preliminary
report.
Int.
J.
Immunopharma-
col.
3,
307.
SHEEHY,
M.J.,
MAWAS,
C.
&
CHARMOT,
D.
(1979)
Specific
inhibition
of
human
lymphocyte
response
by
primed
autologous
lymphocyte.
I.
Evaluation
of
MLR
inhibition
as
a
model
for
suppression.
J.
Immunol.
122,
2198.
STAHL,
R.E.,
FRIEMAN-KIEN,
A.,
DUBIN,
R.,
MAR-
MOR,
M.
&
ZOLLA-PASNER,
S.
(1982)
Immunologic
abnormalities
in
homosexual
men:
relationship
to
Kaposi's
sarcoma.
Am.
J.
Med.
73,
171.
SUITS,
D.B.
(1957)
Use
of
dummy
variables
in
regres-
sion
equations.
J.
Am.
Stat.
Ass.
52,
548.
WALLACE,
J.I.,
CORAL,
F.S.,
RIMM,
I.J.,
LANE,
H.,
LEVINE,
H.,
REINBERG,
E.L.,
SCHLOSSMAN,
S.F.
&
SONNABEND,
J.
(1982)
T-cell
ratios
in
homosexuals.
Lancet,
i,
908.
