Decrease of the okt8 positive t cell subset in polymyalgia rheumatica.

1472
DECREASE OF THE OKT8 POSITIVE T C E L L
SUBSET IN POLYMYALGIA RHEUMATICA
Lack of Correlation with Disease Activity
CHARIF BENLAHRACHE, PATRICK SEGOND, LOUIS AUQUIER,
and JEAN-PIERRE BOUVET
Peripheral T cell populations were investigated in
35 patients suffering from polymyalgia rheumatica. The
total number of T cells was low compared with those of a
control group of similar age (P < 10 -?. This decrease
was demonstrated by using both classic E-rosette and
monoclonal antibody techniques (OKT3, and OKT4
OKTS) and was shown to be secondary to a selective TS
defect (P <
There was no correlation between the
decrease in T8 (a cytotoxic suppressor T cell subset) and
steroid therapy, disease activity, and temporal arteritis,
nor between this decrease and the T y percentage and
the presence of circulating immune complexes (CIC).
The T y cell percentage was low in the patient group (P
< lo-’) and correlated with the presence of detectable
CIC (P < 0.05). In contrast to the TS and T y defects,
concanavalip A-stimulated cells from 5 selected patients
were found capable of suppressing in vitro anti-trinitrophenyl response. This suppression was found in both
autologous and allogeneic experiments. From these data
one can assume that an immune anomaly (T8 defect)
could be the origin of CIC and the disease occurrence.
+
From the Department of Rheumatology. Hbpital Ambroise
Pare, Faculte de MCdecine Paris-Ouest. the Department of Haematology, Facultt de Medecine Paris-Sud. and the Laboratoire d’lmmunologie Microbienne, Institut Pasteur, Paris, France.
Charif Benlahrache, MD: recipient of a fellowship award
from the College de Medecine des Hbpitaux de Paris: Patrick
Segond, MD: Assistant Professor in Immunology, Facultk de Medecine Paris-Sud; Louis Auquier, MD: professor of Rheumatology.
Head of the Department of Rheumatology. Faculte de MCdecine
Paris-Ouest; Jean-Pierre Bouvet, MD. PhD: Charge de Recherches
at the Institut National de la Sante et de la Recherche Medicale
(INSERM).
Address reprint requests to Dr. Jean-Pierre Bouvet. Laboratoire d’Immunologie Microbienne, Institut Pasteur. rue du Docteur
Roux, 25, 75724 PARIS Cedex, France.
Submitted for publication February 8. 1983; accepted in
revised form July 19, 1983.
Arthritis and Rheumatism, Vol. 26, No. 12 (December 1983)
Polymyalgia rheumatica (PMR) is a relatively
common disease in aged patients (1). Although the
clinical symptoms and the response to steroid therapy
are well documented (2), the etiology and pathogenesis
of the disease remain unknown. However, a frequent
association with giant cell ajteritis (GCA) (1) points to
a vascular origin of the disease. Immunoglobulin deposits are found in the media of affected arteries,
suggesting an anti-vascular autoimmunity (3) or an
immune complex mediated vasculitis. Hazleman and
coworkers reported enhanced in vitro proliferative
responses of peripheral blood lymphocytes to homologous artery walls (4). Unfortunately these results were
not confirmed by Papaioannou and coworkers ( 5 ) and
recent studies have investigated mainly the immune
complex hypothesis. Circulating immune complexes
(CIC) are often found in patients; the percentages
depend partly on the methods of measurement (6-8).
The highest percentage is 94, which is found when the
Raji cell radioimmunoassay is used (8).
However, these data provide only indirect evidence for an immune vasculitis since similar deposits
were found in other diseases lacking signs of vasculitis
(3), and CIC have also been detected in a wide variety
of nonimmune diseases as well as in normal subjects
(9,lO). Additional data are therefore required to confirm the immunologic origin of the disease.
The recent development of monoclonal antibodies which react with specific antigenic determinants on human T lymphocytes provides a means to
readily assess the immune status of individuals (1 1).
These antibodies have been widely used in the study of
T cell imbalances in autoimmune diseases (12). A
decrease in OKT8 positive cells, a cytotoxic/suppressor T cell subset, was found in peripheral blood in
DECREASE OF OKT8 IN PMR
1473
Table 1. Clinical findings and T lymphocyte subsets in patients*
Patient
I
2
3
4
5
6
7
8
9
10
I1
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Duration from
disease onset
(years)
Steroid
therapy
(mdday)
3
3
3
5
6
2
0.3
9
4
0.16
2
2
I
2
2
5
9
3
0.16
0.16
0.16
I .4
5
6
0.5
0.3
0.16
0.3
3
3
1.5
0.6
0.1
0.5
0.2
0
0
25
-
19
t
10
17
II
17
15
45
19
16
32
9
33
16
40
in
7
6
I5
130
54
52
40
16
2n
10
20
I19
21
20
94
37
40
60
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
112
+
0
0
6
20
0
0
n
15
15
6
7
6
0
0
15
10
10
15
n
0
0
10
10
15
15
8
6
17
15
30
20
0
ESRt
no
CICf
Temporal
vasculitis
-
+
+
-
+
+
+
-
+
+
96
%
E-RFCt
T3
65
71
72
61
36
59
72
54
68
55
66
45
75
4n
75
65
40
63
48
72
58
47
82
7n
59
60
53
59
65
60
47
65
78
50
59
71
64
74
65
40
56
7n
51
75
52
68
42
76
46
71
68
4n
60
58
69
56
50
79
75
54
63
55
5n
76
63
49
* - = absence of; + = presence of; ND = not determined.
t ESR = erythrocyte sedimentation rate; ClC = circulating immune complexes; E-RFC
rheumatoid arthritis (13,14), multiple sclerosis (15,16),
Sjogren’s syndrome (17), and other diseases (12). In
the majority of cases the decreases seemed t o be
correlated with the disease activity. I n the present
study we have investigated lymphocyte subsets in
polymyalgia rheumatica patients for t h e presence of
similar abnormalities.
PATIENTS AND METHODS
Patients. Thirty-five patients (7 men and 28 women)
with polymyalgia rheumatica were selected for the study.
The diagnosis was based on the ptesence of 3 or more
criteria proposed by Bird and coworkers (2). The mean age
was 72.6 years 2 7.8 SD. The presence of a temporal
arteritis was found in 7 patients. The patients were not
investigated at the same stage of the disease, thus allowing
for the study of the influence of steroid therapy and disease
61
71
53
54
%
T4
+8
75
65
68
sn
42
52
67
56
72
52
65
3n
73
41
69
63
46
59
45
66
52
46
80
73
51
58
52
50
71
55
54
63
69
5n
52
%
%
%
T4
TS
TY
50
48
53
49
36
49
55
46
61
48
51
34
50
32
56
53
42
22
23
28
26
I0
13
10
7
53
50
55
46
44
53
56
49
61
48
50
61
49
45
54
52
46
43
23
12
n
10
n
6
6
5
4
2
I1
3
25
2
2
3
2
11
35
n
25
16
9
16
11
24
12
11
24
31
16
n
14
13
14
16
7
16
31
in
15
10
6
9
7
12
11
13
13
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
= E-rosette forming cells
activity (Table 1). The median duration from the onset of
symptoms was 2 years (mean: 2.41 2 2.4). Twenty-four
patients were undergoing steroid therapy (mean dosage:
8.48 ? 7.42 mg/day). Disease activity was estimated by the
erythrocyte sedimentation rate (ESR) and by serum levels of
haptoglobin and orosomucoid: mean ESR was 36.9 2 32.3
mm. haptoglobin was 2.41 % 1.12 mg/ml (normal value:
0.971, and orosomucoid was 1.08 2 0.37 mg/ml (normal
value: 0.52). The results of these 3 tests for biologic
inflammation were found to vary according to a linear
positive correlation: ESR-haptoglobin (P < 2 x 10-7,
ESR-orosomucoid (P <
orosomucoid-haptoglobin
(P< 10-61.
Control subjects. Thirty-five subjects (7 men and 28
women) of similar age (77.2 % 8.9 years) without any
infectious, inflammatory, or neoplastic disease, were selected as controls. This control group included patients hospitalized for hemiplegia or osteoporosis. All controls had normal
values for ESR, haptoglobin (0.59 to 1.44 mg/ml), and
orosomucoid (0.40 to 0.68 mg/ml).
BENLAHRACHE ET AL
1474
Assay for immune complexes. The procedure for the
detection o f circulating immune complexes was carried out
on fresh sera, hsing a simple qualitative technique: the PEGC l q immunodiffusion test (18). CIC were precipitated with
5% polyethylene glycol (PEG), washed with 2.5% PEG, and
after dissolution of the pellet with saline, investigated for the
presence of IgG and complement by immunodiffusion ass.ay.
Positive sera showed precipitating lines in the presence of
anti-Fcy and anti-CIq sera .(Betiring).
Mononuclear cell isolation. Purified mononuclear cell
(MNC) suspensions were obtained from hepatinized blood
by standard Ficoll-Hypaque density cedtrifugation and kept
frozen in liquid nitrogen after being treated with dimethylsulfoxide..
E-rosettes. Rosetting of mononuclear cells was performed using sheep red blood cells treated with neuraminidase. and the percentage of rosette forming cells (E-RFC)
was established from 200 cells. E-RFC were purified by
centrifugation on Ficoll-Hypaque followed by lysis of erythrocytes with Tris-ammonium chloride (19). Purified E-RFC
were used for T y cell percentage estimation as described in
the following section.
Ty cells. T cells that bear Fc receptor specific for IgG
were identified with the erythrocyte-antibody rosette technique performed on E-RFC preparations. These cells (Ty)
formed rosettes with ox erythrocytes sensitized with IgG
from rabbits hyperimmunized with ox erythrocytes (19). The
CYR
C
FUR
C
hyperimmune serum was a gift from Dr. N. Chiorazzi of the
Rockefeller University (NY).
B cells. Membrane immunoglobulin positive cells (B
cells) were identified on MNC preparations after 1 hour
incubation at 37°C. Membrane immunoglobulins were revealed by immunofluorescence examination using fluoresceinconjugated F(ab’Iz goat anti-human Ig (Kallestad).
MOUOCIOMI
antibodies. Membrane markers specific
for total peripheral T cells (T3) or for T cell subsets (T4 and
T8) were identified by indirect immunofluorescence studies
with commercial monoclonal antibodies (Ortho Pharmaceutical Corporation). In the peripheral blood OKT3 identifies
all T cells (T3), OKT4 is specific for T helper cells (T4). and
OKT8 for suppressorkytotoxic cells (T8) (1 1). Stock solutions of antibodies were made in phosphate buffered saline
(0.5 mg/ml) and diluted in appropriate volumes of RPMI
medium containing 10% fetal calf serum (Gibco, Grand
Island, NY) and 0.02% sodium azide: these dilutions were
1:30 (OKT3). 1 : 5 0 (OKT4), and 1:100 (OKT8), aliquoted
and stored at -80°C.
Fluorescence studies. Suspensions of MNC (5 x lo4
cells per well) were washed 3 times in Linbro plates with an
RPMI-fetal calf serum solution. They were incubated for 30
minutes at 4°C with 2 p1 of diluted monoclonal antibody, and
after 3 additional washes, positive cells were revealed by
indirect immunofluorescence using 1 :20 diluted fluoresceinated goat F(ab’)z anti-mouse IgG antibody (Cappel Labora-
C
CYR
ma
F
-p.f+ :f i- 4
..
..
:
8
...
I
1
t
-&
E- RFC
..
L
144
14
1s
Figure 1. Petcentages of T lymphocytes in polymyalgia rheumatica patients (PMR) and controls (C):
significant decreases are apparent between the patients and the controls when measured by 4 different
methods. Dots represent individual values, horizontal lines correspond to means (long tines) and
standard deviations (short lines). E-RFC = cells forming rosettes with sheep erythrocytes. T3 = cells
positive with OKT3 antibody. T4+8 = cells positive with OKT4 andor OKTE (simultaneous labeling
experiments).T4+TH = sum of cells positive with OKT4 and OKTH (distinct labeling experiments).
1475
DECREASE OF OKT8 IN PMR
tory, Cochranville, PA). Washed fluorescent cells were
suspended in a buffered solution, set on slides, and examined
with a Leitz Orthoplan microscope equipped for epifluorescence. Percentages were established on 200 cells.
Assay of suppressor cell function. The capacity of
freshly isolated lymphocytes to inhibit the in vitro antibody
response was investigated in 5 patients and compared with
that of 4 aged and 15 young controls. The patients were
selected for their high T4:T8 ratios: 7.62 (patient 9), 3.43
(patient 27), 4.36 (patient 29), and 6.43 (patient 31), or for
their high disease activity (patient 34). The suppressive
function of M N C was measured by their ability to inhibit the
in vitro response to trinitrophenyl (TNP) antigen after having
been stimulated by concanavalin A (Con A) (20). Suppressor
cell macrophages were eliminated by nylon adherence as
previously described (21). Suppression was induced either
by direct addition of Con A to the autologous TNP-polyacrylamide (TNP-PAA) stimulated culture or by transfer of
patient cells, incubated with Con A. to a TNP-PAA stimulated 2-day culture from a normal young subject. Assays for
anti-TNP response were performed at day 7 , using a conventional agarose plaque technique (20) and the percentage of
suppression was calculated by comparing cells incubated
with and without Con A.
Statistical analysis. Statistical methods used were
Student's t-test, linear correlation, and partial correlation
tests. When data did not show a Gaussian distribution,
nonparametric tests (chi-square and Mann-Whitney) were
used.
RESULTS
Presence of circulating immune complexes. CIC
were detected in 24 of 35 patients (68.5%) and in 4 of
32 controls (12.5%). The difference between the 2
groups was clearly significant ( P < lop4). No correlation was found between CIC and the age of patients,
disease duration, presence of steroid therapy, presence of temporal arteritis, nor with results of a n y of
the 3 tests for biologic inflammation (Table 1).
E-RFC and Ty. The percentage of E-RFC was
found to be clearly lower in patients (60.86 2 11.17%)
than in the controls (68.83 5 6.39%) (Figure 1 ) . The
difference was significant ( P < lop3). The percentages
of Ty were established from purified E-RFC from 22
patients and 20 controls. Patients were found to have a
lower percentage of Ty cells (median: 6%) than the
controls (median: 15%); the difference was highly
significant ( P < lo-')).
Similar results were found when the percentages of T y were calculated from total MNC by multiplying individual Ty percentages with corresponding
E-RFC percentages. E-RFC and Ty did not show any
statistical correlation with clinical data or with biologic
tests for inflammation. E-RFC were present in the
%
aa
8
PMR
C
PMR
C
..
PMR
C
7
7c
6c
...
6
w
5
5c
U
3
(
n
...
4
3
8 *
..
W
Figure 2. T cell subsets in polymyalgia rheumatica patients (PMR) and controls (C): decreases of T8 and Ty
percentages and high T4R8 ratio in patients. The percentages of T4 (cells positive with OKT4) and T8 (cells
positive with OKT8) were established from total mononuclear cell preparations. The percentages of Ty(T cell
bearing Fcy receptor) were established from purified T cell preparations. Dots enclqsed in boxes correspond
to 25% and 75% of individual values. The horizontal lines in the boxes correspond to the medians.
BENLAHRACHE ET A L
1476
same percentages in patients with and without CIC. In
contrast, among the patients investigated for Ty percentage, the 5 subjects without CIC had higher levels
of Ty (median: 10.00%) than those with detected CIC
(median: 6.17%) (P < 0.05) and this was also noticed
when percentages were calculated on total MNC ( P <
0.05).
B lymphocytes. Similar percentages of B cells
were found in patients (5.77 2 1.69%) and controls
(6.03 ? 2.31%). The difference was not significant.
T cell percentage as defined by monoclonal antibodies. Using the OKT3 monoclonal antibody, we
found a lower percentage of T cells (T3) in patients
(61.57 4 10.60%) than in controls (66.89 +- 6.45%).
The difference was significant ( P < 0.01) (Figure 1).
Comparable results were found when the total T cell
percentage was established by simultaneous labeling
with OKT4 and OKT8 antibodies. In this case the
positive cells (T4t-8) corresponded to the sum of
T4-8+, and T4+8+ subsets. They were lower in
patients (58.74
10.60%) than in controls (66.17 ?
6.03%) (P < 0.001) (Figure 1). The sums of individual
*
% T3
percentages of OKT4 (T4) and OKT8 (T8) positive
cells (T4+T8) also showed lower levels in patients
(66.63 2 11.8%) than in controls (74.03 t 6.67%) (P<
0.001). The increased levels of T4+T8, when compared with T3 and T4+8, were because of a double
count of T4+8+ subset in the sum T4+T8.
A close linear positive correlation was found
among the 4 different methods for measuring T cell
percentages (P < lop9). As found with E-RFC, T3,
T4+8, and T4+T8 did not correlate with clinical data
nor with CIC, biologic inflammation, or Ty.
T cell subsets. Percentages of T4 were not found
to be statistically different in patients (49.54 2 6.73%)
and controls (47.17 ? 4.55%) (Figure 2).
In contrast, percentages of T8 were found to be
dramatically lower in patients (17.26 -+ 7.70%) when
compared with controls (26.60 3.67%) ( P < lop9).
The T4:T8 ratio was similarly modified; it was clearly
higher in patients (3.415 -+ 1.506) than in controls
(1.801 2 0.264) (P <
Both the T8 and T4:T8
ratio were not found to correlate with clinical data,
biologic inflammation, CIC, or Ty. On the contrary,
*
% E RFC
00-
O0l
80 *
0
i”
me
10
me*
0.
60.
* *
e.
A
50.
*e
40.
m
10
.
M
30
%T*
%-
0
10
20
30
lG.
Figure 3. A, Positive correlation between percentages of T3 and T8. 9, Lack of correlation between
percentages of E-rosette forming cells (E-RFC) and TG (Ty). = individual values of circulating immune
complex (CIC) positive polymyalgia rheurnatica patients. = values of CIC negative patients. The solid bar
is the regression line.
+
DECREASE OF OKT8 IN PMR
1477
Table 2. Con A-induced suppression of the in vitro anti-TNP
response*
Number of anti-TNP-PFC per lob
cells
Patient
Culture with
Con A
Culture without
Con A
9
27
31
34
2
0
9
0
15 young
21.7
-t
16.2
allogeneic cells from normal young patients. Using this
method, the suppressive capacity of patient cells,
except for 1 case (patient 27), was found to be normal
and similar to that of aged controls (Table 3).
%
suppression
DISCUSSION
8
62
99
14
75
100
89
299 t 161
90.8 t 9.8
This study was undertaken to determine whether there could be lymphocyte abnormalities which
would support the hypothesis of an immune mechanism as the origin of polymyalgia rheumatica. The results indicate a decrease in detected T y cells (P <
lop6)(Figure 2) which was found to correlate with the
presence of CIC, and mainly, a decrease in peripheral
(Figure 1) which correlated with a
T cells (P <
major decrease of T8 subset ( P < lop9)(Figure 2). No
statistical relationship was found between these two
groups of abnormalities, nor with clinical data, biologic inflammation, or steroid therapy. The results obtained in the control group did not show the T cell
imbalance usually found in aged persons (22,23). This
is because of the low ma1e:female ratio (1:5) in our
control group; the T cell abnormality occurs mainly in
aged males. The comparison with such a sex-matched
control group avoided bias differences due to the age
of the patients, cells freezing, or any unexpected
technical artifact.
The presence of elevated levels of CIC in 8 of 10
patients with polymyalgia rheumatica alone was found
by Papaioannou and coworkers (6). In these patients
the CIC levels did not seem to correlate with the
disease activity, contrasting with the results obtained
in patients with giant cell arteritis in whom CIC and
disease activity seemed clearly correlated (6). Using
another method of CIC detection, Park and coworkers
100
controls
* Con A = concanavalin A; anti-TNP-PFC
plaque forming cells.
=
anti-trinitrophenyl
they correlated in the patient group with E-RFC ( P <
lop9)and T3 ( P <
(Figure 3).
Indirect estimation of the T cell subset positive
for both OKT4 and OKT8 antigens (T4+8+) was made
by subtracting the individual values of T4+8 from the
corresponding sum T4+T8. No significant difference
was found in patients (8.4 +- 4.4%) compared with
controls (7.5 -+ 3.9%).
Suppressor cell function. To correlate T8 defect
with a functional T cell status, we investigated the
capacity of T cells from 5 selected patients to inhibit
autologous or allogeneic in vitro anti-TNP response.
In autologous experiments, nylon-passed mononuclear cells from 4 patients exerted, after Con A stimulation, a significant inhibition of the specific plaque
forming cell response similar to that of 15 different
young controls (Table 2). Since some control responses were poor (patients 9 and 34), the suppressive
capacity of patient Con A incubated cells was also
investigated by inhibiting TNP-stimulated cultures of
Table 3. Suppression of the in vitro anti-TNP response of cells cultured from young controls by Con
A incubated cells from patients and aged controls*
Number of TNP-PFC per 10' cells
Subjects
Con A
incubated cells
Cells without
Con A
suppression
Patient 9
Experiment I (
Aged control
Patient 27
Experiment 2
Aged control
Patient 31
Experiment 3
Aged control
(Patient 29
Experiment 4 Patient 34
[Aged control
I1
3
27
27
10.8
15
160
129
I18
73
34
31
170
99.6
66
777
39 1
359
85
91
13
84
89
77
79
67
67
i
%
* Except for patient 27, all patients demonstrated normal suppression when Con A incubated cells
were added to TNP-stimulated cells from normal young controls. TNP-PFC = anti-trinitrophenyl
plaque forming cells; Con A = concanavalin A.
1478
(7) investigated 2 groups of polymyalgia rheumatica
and/or giant cell arteritis patients and found 44%
positive cases in patients with active disease and only
23% in the inactive group. However, in this study the
polymyalgia rheumatica patients were not investigated
separately and the relatively low percentage of positivity could be due to a lack of sensitivity of the technique
(no positivity was found in controls). In contrast,
Espinoza and coworkers (8) found 94% positive cases
by the use of the same method as Papaioannou (6) and
showed an obvious correlation with the presence of
CIC and disease activity.
This discrepancy in results of CIC investigations, which has also been noted in other diseases
(24,25). casts doubt on the reliability of different
techniques. It is indeed possible that a particular
method could, by chance, detect mainly the pathogenic molecules among the CIC molecule population.
However the well-known interferences between aggregated IgG and CIC detection could also explain conflicting results, particularly in these patients who usually have increased levels of IgG during periods of
disease activity.
In the present study, CIC were detected in
67.7% of patients; this was statistically different from
The presence in vivo of
the control group ( P <
such complexes was suggested by their correlation
with the decrease in Ty cells (P < 0.05). These cells
are defined and characterized by the presence of a
membrane receptor specific for the Fc part of IgG.
This Fc receptor reacts with CIC, and this could
therefore inhibit the detection of Ty rosettes. In addition, the interaction with CIC could also trigger cellular modifications of Ty, leading to the loss of their Fcy
receptor (26). Thus, the apparent decrease of Ty cells
could be due to the presence in vivo of CIC acting by
two mechanisms: masking effect on the Fc receptor
and triggering a transformation in another T cell subset.
Thus, the total number of E-RFC might not be
influenced by such a decrease of Ty, as was shown by
the lack of correlation between the Ty percentage and
the E-RFC. The decrease of Ty cells, even in patients
with no CIC, might be due to another mechanism or to
a low level of CIC, undetectable by our method.
Although this method of detection is not quantitative,
the lack of correlation with disease activity, also
observed for the T y percentage, did not favor the
hypothesis of an immune complex mechanism as the
origin of polymyalgia rheumatica. Nevertheless this
hypothesis is not ruled out since pathogenic CIC
BENLAHRACHE ET AL
molecules are probably not correlated to the total level
of CIC and might be present even in CIC negative
patients. The decrease in the peripheral total T cells
was shown by 4 different methods: E-RFC, T3, T4+8,
and the sum T4+T8. The 4 measurements provided
similar results in each patient and were found to be
highly correlated ( P <
This decrease was relatively low (5 to 8% of MNC) and highly correlated to a
selective defect in T8 subset (P< lop9). The normal
percentages of T4, T4+T8, and B cell populations and
the high negative correlation with T4:T8 ratio also
suggest that T cell decrease was not due to variations
in the non-T population but was secondary to the
major T8 defect.
The T8 subpopulation was found clearly lower
in patients when compared with controls (P <
This decrease was correlated to the total T cell decrease and was of similar size. No correlation was
found with the other clinical or biologic data, nor with
the decrease of Ty. Though both T y and T8 subsets
include cells with suppressor activity, this lack of
correlation was not inconsistent since T y detection
was probably partially inhibited by CIC, and since the
two populations have been shown to be clearly different (27). T8 marker defines a T cell population which
mainly includes T3 positive cells with suppressor or
cytotoxic activities (1 1).
In contrast, the precise nature of T y cells is
controversial. Reinherz and coworkers first established the presence of a myeloid marker (OKMI)
together with the absence of the T3 marker at the
surface of these cells (27) and concluded that T y cells
may be non-T in origin. Further studies have modified
these conclusions by showing the existence of T3
positive cells in a monoclonal T y clone (28), and also
the presence of other markers specific for T cells in the
T3 negative-OKM1 positive Ty population (29). Recent reports suggest the existence of different Ty
subsets belonging to the T population and having
OKMl plus T4 or T8 markers; these cells should be
related to natural killer cells (30).
The decrease of T8 cells in polymyalgia rheumatica patients was comparable to those observed
in other diseases, though their mechanisms might be
different. In rheumatoid arthritis patients, the T8 defect seems to be moderate in the peripheral blood and
to contrast with a relatively high percentage of T8 in
the synovial fluid and synovial tissue cell suspensions
(13,14,31). A possible explanation of these data is that
cytotoxic T8 would migrate to the synovial lesions and
therefore be decreased in peripheral blood during
1479
DECREASE OF OKT8 IN PMR
periods of disease activity (14). However, this hypothesis needs additional data since opposite results were
reported showing a low T4:T8 ratio in lymphocytes
from the rheumatoid synovial membrane when examined in situ (32).
In multiple sclerosis, a similar T8 decrease was
found and it also clearly correlated with disease activity ( 1 5 ~ 6 )In
. leprosy, a decrease in T8 subset has been
found only in patients with associated nodosum leprosum eruption (33), a vascular skin lesion possibly due
to an Arthus phenomenon (34) with local deposits of
immune complexes (35). Such a relationship between
T8 percentage and this particular form of the disease
has suggested a primitive T8 defect as the origin of an
abnormal response t o the pathogen (33).
In polymyalgia rheumatica, the drop in T8
proportion did not correlate with disease activity nor
with the presence of CIC. This suggests that T8 are not
responsible for the local lesions nor is this subset
decrease a simple consequence of the disease. The T8
defect might preexist and favor the occurrence of the
disease. Our data showing a normal suppressor cell
function in 5 patients with abnormal T4:T8 ratios
indicate that such a defect in T8 does not affect the
suppressor cell subpopulation.
Although these results need additional functional data, they point to the problem of correlations
between functional experiments and numerations of
subsets. In normal subjects, Chudwin and coworkers
(36) found a good correlation between T lymphocyte
subset enumerations and phytohemagglutinin o r mixed
lymphocyte culture responses, though poor correlations were found in immunodeficient patients. In patients with acute polyradiculoneuritis (37) or lupus
(38), normal T suppressor cell functions were found,
contrasting with abnormal proportions of T cell subsets. T h e possibility of a cytotoxic precursor cell
defect was suggested t o explain these unexpected
results; however, a comparison with results obtained
on human T cell lines shows that T8 includes cells with
different immune functions whose selective defects are
actually explored by using the same surface marker
(39).
In the absence of a better delineation of T cell
subsets, the precise nature of the T8 defect remains
obscure in polymyalgia rheumatica patients. It is likely
that this defect corresponds t o a n abnormal immune
status that might favor the occurrence of the disease.
Participation of pathogenic immune complexes secondary t o this abnormal status might be the cause of
local lesions.
AKNOWLEDGMENTS
We wish thank Dr. M. A. Bach and Df. S. Iscaki for
their advice and critical reading of the manuscript, A. Portier
for his excellent technical assistance, and Mole. C. Gauvin
for secretarial support.
REFERENCES
1. Bengtsson BA, Malmvall BE: The epidemiology of giant
cell arteritis including temporal arteritis and polymyalgia
rheumatica: incidences of different clinical presentations
and eye complications. Arthritis Rheum 24:899-904.
1981
2. Bird HA, Esselinckx W, Dixon ASJ, Mowat AG, Wood
PHN: An evaluation of criteria for polymyalgia rheumatics. Ann Rheum Dis 38:434-439. 1979
3. Liang GC, Simkin PA, Mannik M: Immunoglobulins in
temporal arteries: an immunofluorescent study. Ann
Intern Med 81:19-24, 1974
4. Hazleman BL, MacLennan ICM, Esiri MM: Lymphocyte proliferation to artery antigen as a positive diagnostic test in polymyalgia rheumatica. Ann Rheum Dis
34: 122-127. 1975
5. Papaioannou CC, Hunder GG, McDuffie FC: Cellular
immunity in polymyalgia rheumatica and giant cell arteritis: lack of response to muscle or artery homogenates.
Arthritis Rheum 22:740-745, 1979
6. Papaioannou CC, Gupta RC. Hunder GG, McDuffie FC:
Circulating immune complexes in giant cell arteritis and
polymyalgia rheumatica. Arthritis Rheum 23: 1021-1025,
1980
7. Park JR, Jones JG, Harkiss GD. Hazleman BL: Circulating immune complexes in polymyalgia rheumatica and
giant cell arteritis. Ann Rheum Dis 40:360-365, 1981
8. Espinoza LR, Bridgeford P, Lowenstein M, Bocanegra
TS, Vasey FB, Germain BF: Polymyalgia rheumatica
and giant arteritis: circulating immune complexes. J
Rheumatol 9556-560, 1982
9. Levinsky RJ: The measurement of immune complexes:
is it clinically useful? Immunology Today 2:94-97, 1981
10. Plotz PH: Studies of immune complexes. Arthritis
Rheum 25:1151-1155, 1982
11. Reinherz EL, Schlossman SF: The differentiation and
function of human T lymphocytes. Cell 19:821-827, 1980
12. Bach MA, Bach JF: The use of monoclonal anti-T cell
antibodies to study T cell imbalances in human diseases.
Clin Exp Immunol 45:449-456, 1981
13. Veys EM, Hermanns P. Schindler J, Kung PC, Goldstein G, Symoens J , Van Wauwe J: Evaluation of T cell
subsets with monoclonal antibodies in patients with
rheumatoid arthritis. J Rheumatol 9:25-29, 1982
14. Fox RI, Fong S, Sabharwal N, Carstens SA, Kung PC,
Vaughan JH: Synovial fluid lymphocytes differ from
peripheral blood lymphocytes in patients with rheumatoid arthritis. J Immunol 128:351-354, 1982
IS. Reinherz E, Weiner HL, Hauser SL, Cohen JA, Distaso
1480
JA, Schlossman SF: Loss of suppressor T cells in active
multiple sclerosis: analysis with monoclonal antibodies.
N Engl J Med 303:125-129, 1980
16. Bach MA, Phan-Dinh-Tuy F, Tournier E, Chatenoud L,
Bach JF, Martin C, Degos JD: Deficit of suppressor T
cells in active multiple sclerosis. Lancet ii: 1221-1223,
1980
17. Fox RI, Carstens SA, Fong S, Robinson CA, Howell F,
Vaughan JH: Use of monoclonal antibodies to analyze
peripheral blood and salivary gland lymphocyte subsets
in Sjogren’s syndrome. Arthritis Rheum 25:419-426,
1982
18. Grangeot-Keros L, Segond P, Capel F, Iscaki S, Pillot J:
Detection of immune complexes: a simple assay based
on characterization of the in vivo bound Clq (PEG-Clq
immunodiffusion test). J Immunol Methods 23:349-362.
1978
19. Hallberg T , Gurner BW, Coombs RA: Opsonic adherence of sensitized ox red blood cells to human lymphocytes as measured by rosette formation. Int Arch Allergy Appl Immunol 44:500-513, 1973
20. Galanaud P, Crevon MC, Delfraissy JF. Segond P,
Wallon C, Dormont J: Regulation of the primary in vitro
antibody response in human peripheral blood lymphocytes: different effects of mitogen-induced and spontaneous T suppressor cells. Clin Exp Immunol 38: 106-1 IS,
1979
21. Segond P, Salliere D. Galanaud P, Desmottes RM.
Massias P. Fiessenger JN: Impaired primary in vitro
antibody response in progressive systemic sclerosis patients: role of suppressor monocytes. Clin Exp Immunol
47: 147-154, 1982
22. Nagel JE. Chrest FJ, Adler WH: Enumeration of T
lymphocyte subsets by monoclonal antibodies in young
and aged humans. J Immunol 127:2086-2088, 1981
23. Mascart-Lemone F, Delespesse G, Servais G, Kunstler
M: Characterization of immunoregulatory T lymphocytes during ageing by monoclonal antibodies. Clin Exp
Immunol 48:148-154, 1982
24. Fiist G, Fekete B, Angyal I. Jakab A, Pal A, Meretey K ,
Falus A, Torok K, Szegedi G, Kavai M, Puskas E,
Csecsi-Nagy M, Szabo I , Lenkey A, Misz M: Evaluation of different methods for detecting circulating immune complexes: studies in patients with lung cancer. J
Immunol Methods 46:259-276, 1981
25. McDougal JS, Hubbard M, McDuffie FC, Strobel PL,
Smith SJ. Bass N , Goldman JA, Hartman S, Myerson G ,
Miller S, Morales R, Wilson CH Jr: Comparison of five
assays for immune complexes in the rheumatic diseases:
an assessment of their validity for rheumatoid arthritis.
Arthritis Rheum 2S:llS6-1166, 1982
26. Moretta L, Mingari MC, Romanzi CA: Loss of FC
receptors for IgG from human T lymphocytes exposed
to IgG immune complexes. Nature 272:618-620, 1978
BENLAHRACHE ET A L
27. Reinherz EL, Moretta L, Roper M, Breard JM, Mingari
MC, Cooper MD, Schlossman SF: Human T lymphocyte subpopulations defined by Fc receptors and monoclonal antibodies: a comparison. J Exp Med 151:969974, 1980
28. Strong DM, Pandolfi F, Slease RB, Budd JE, Woody
JN: Antigenic characterization of a T-CLL with heteroantisera and monoclonal antibodies: evidence for the T
cell lineage of an Ia-positive, Fc-IgG-positive, suppressor-cell subpopulation. J Immunol 126:22OS-2208, 1981
29. Fox RI, Thompson LF, Huddlestone JR: T y cells express T lymphocyte-associated antigens. J Immunol
126:2062-2063, 1981
30. Van de Griend RJ, Ten Berge I, Tanke HJ, Roos D,
Schellekens PTA, Melief CJM, Zeijlemaker WP, Astaldi
A: Characterization of two subsets of human T y cells. J
Immunol 128:1979-1985, 1982
31. Burmester GD, Yu DTY, Irani AM, Kunkel HG, Winchester RJ: Ia+ T cells in synovial fluid and tissues of
patients with rheumatoid arthritis. Arthritis Rheum
24:1370-1376, 1981
32. Meijer CJLM, de Graaf-Reitsma CB, Labefer GJM.
Cats A: In situ localization of lymphocyte subsets in
synovial membranes of patients with rheumatoid arthritis with monoclonal antibodies. J Rheumatol 9:359-365,
1982
33. Bach MA, Chatenoud L, Wallach D, PhanDinh Tuy F,
Cottenot F: Studies on T cell subsets and functions in
leprosy. Clin Exp Immunol 44:491-500, 1981
34. Wemembu SNC, Turk JL, Waters MFR, Rees RJW:
Erythema nodosum: a clinical manifestation of the
Arthus phenomenon. Lancet ii:933-935, 1969
35. Bjorvatn B, Barnetson RS, Kronvall G, Zubler RH,
Lambert PH: Immune complexes and complement hypercatabolism in patients with leprosy. Clin Exp Immunol 26:388-396, 1976
36. Chudwin DS, Cowan MJ, Wara DW, Ammann AJ:
Patients with abnormal proportions of T-lymphocyte
subsets have reduced in vitro cellular immunity. Clin
Immunol Immunopathol 26: 126-136, 1983
37. Hughes RAC, Aslan S, Gray IA: Lymphocyte subpopulations and suppressor cell activity in acute polyradiculoneuritis (Guillain-Barre syndrome). Clin Exp Immunol
5 1:448-454, 1983
38. Tsokos GC, Balow JE: Phenotypes of T lymphocytes in
systemic lupus erythematosus: decreased cytotoxicl
suppressor subpopulation is associated with deficient
allogeneic cytotoxic responses rather than with concanavalin A-induced suppressor cells. Clin Immunol Immunopathol 26:267-276, 1983
39. Mingari MC, Moretta A, Pantaleo G, Moretta L: Surface
markers of resting and activated human T cells: functional implications and experimental limits. Semin Immunopathol S:477-488, 1982