Ann. rheum. Dis. (1975), 34, 422

Hyperbasophilic immunoblasts in circulating blood in chronic inflammatory rheumatic and collagen diseases F. DELBARRE, A. LE GO AND A. KAHAN From Institut de Rhumatologie, Facdilte Medecine Pa)ris-Cochin

Delbarre, F., Le Go, A., and Kahan, A. (1975). Annals of the Rheumatic Diseases, 34, 422430. Hyperbasophilic immunoblasts in circulating blood in chronic inflammatory rheumatic and collagen diseases. The number of large circulating hyperbasophilic mononuclear cells-referred to as hyperbasophilic immunoblasts (HBI)-is often increased in collagen diseases and rheumatoid arthritis (RA) and grossly reflects the degree of disease activity. In contrast, in psoriatic arthropathy the percentage of HBI is within the normal range. HBI are mainly involved in immune reactions and may provide a valuable routine test for the assessment of the latter in disease states and for the prediction of relapse in chronic collagen diseases. Immunofluorescent techniques applied to samples from active autoimmune diseases have shown that a number of HBI are Ig-producing B-blasts. Moreover, in a few cases these intracytoplasmic immunoglobulins exhibited a rheumatoid factor-like activity, a finding which promises to yield additional information on the immunopathogenesis of RA.

In 1965 a patient was observed (by F.D.; Richard, 1975) with severe Fiessinger-Leroy-Reiter's syndrome in whom up to 5 % of the circulating leucocytes were hyperbasophilic mononuclear cells of the lymphoplasmocytoid type with mitotic figures. Since then we have studied these cells in over 230 patients with various forms of rheumatic disease. Hyperbasophilic immunoblasts (HBI) can be detected readily in smears of leucocyte concentrates, and stain well (May-Griinwald-Giemsa), with a spongy and strongly basophilic (or pyroninophilic) cytoplasm; the nucleus is usually eccentric, the nuclear-cytoplasmic ratio high, and the diameter 20-30 pm (Fig. 1). These atypical mononuclear cells, first described by Turk (1898), have been frequently seen in human blood during the course of bacterial (Gump and Fekety, 1967) or viral infections, such as infectious mononucleosis (Downey and McKinlay, 1923; Litwins and Leibowitz, 1951; Frank and Dougherty, 1953; Wood and Frenkel, 1967), although their significance was not at first realized. In the present study circulating hyperbasophilic mononuclear cells have been estimated, using morphological criteria, in patients with a variety of

rheumatic diseases. We have found an increased number of these cells in autoimmune diseases, such as rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE), but not in psoriatic arthritis with acute inflammatory manifestations. These findings suggest that the occurrence of HBI cells is related more to immune reactions than to inflammation. The results of immunofluorescent techniques applied to samples from patients with active RA and SLE confirm this, leading us to the conclusion that HBI estimation may provide a simple routine test for the assessment of immune reactions in disease states and provide a guide for therapy. Material and methods PATIENTS

207 peripheral blood samples from 174 adults with inflammatory rheumatic diseases (Table 1) were examined. Only patients with definite disease, whether acute or in remission, according to clinical and biological criteria, were included in this study. All SLE patients and some of the RA patients were receiving corticosteroid therapy. Ten patients who were treated with chlorambucil (8-10 mg daily) were studied separately.

Accepted for publication March 27, 1975. Requests for reprints to: Professeur F. Delbarre, Institut de Rhumatologie, H6pital Cochin Paris, 27 rue du Faubourg St.-Jacques, 75674 Paris Cedex 14.

Hyperbasophilic immunoblasts in chronic inflammatory rheumatic and collagen diseases 423

Table I Results of HBI test in different groups ofpatients Group studied Controls Normal

No. of cases 43

Osteoarthrosis Sciatica Rheumatoid arthritis Seropositive Seronegative SCAT(-) Latex FII+ SLE Scleroderma Ankylosing spondylitis Psoriatic arthritis Gout Acute viral infection

Total * >5/103

25

No. of samples 43

HBI/103 lymphocytes (range) (mean ± SD) 0-5 (2-2 + 1-5) 0-4 (2-1 + 1-1) 0-5 (2*5 ± 1*5)

25

10

10

8

8

86

76

51 17 8

1-28 2-25 1-16 2-60 2-10 0-4 0-5 0-11 4-68

58 18 10

30 13 17 18 17 17

40 21 17 24 19 17

234

267

% of cases with raised value*

0*0 0-0 57-0 56-9 66-6 40*0 82-5 38*1 0.0 0-0 31*5

lymphocytes.

The normal controls were 25 healthy persons from the hospital staff and eighteen patients in hospital with degenerative joint diseases. In addition, a group of seventeen children with acute viral infection (mumps 7, measles 7, rubella 3) was included in this study. Blood samples were obtained at varying intervals from the onset of the clinical symptoms.

were developed 12 days later in Microdol-X (for 5 min at 20°C) and stained through the emulsion with Giemsa stain. Only qualitative estimations were done, attention being focused on HBI which were found to be heavily labelled.

a

ASSESSMENT OF CELLS IN PERIPHERAL BLOOD

Leucocyte concentrates were prepared from blood collected with EDTA-Na2, centrifuged for 15 min at 1 500 r.p.m. in capillary tubes (3 mm internal diameter) and the buffy-coat was smeared on slides. Differential cell counts were made on May-Grunwald-Giemsa stained smears. At least 1 000 and usually 2000 lymphoid cells were counted; since the identification of HBI presents no difficulties, differential counts done by two independent observers on different smears from the same sample always gave similar results; therefore, HBI was estimated by any one observer and the number expressed as a percentage of the total lymphocyte count. Other techniques of leucocyte or

lymphocyte separation (plasma sedimentation, filtration through nylon fibres) have been tried and do not influence the results. The criteria used to identify HBI were the intense basophilia of the cytoplasm and the leptochromatic pattern of the nucleus which showed one or two nucleoli (more clearly seen in phase microscope). They were 1-5 to 3 times the diameter of a small lymphocyte and had the nuclear-cytoplasmic ratio of large lymphocytes (Fig 1). Staining with methyl-green-pyronine indicated a high content of RNA in their cytoplasm and nucleus. In HBI-rich samples, mitotic figures were often seen and the proliferative potential of HBI was occasionally assessed using a flash-labelling technique with 3H-thymidine (CEN Saclay, France) by incubating 106 lymphocytes (for 1 h at 37°C) in 1 ml autologous plasma and 1 ,Ci isotope. Radioautography of methanol fixed smears was done by dipping the slides in Ilford K5 nuclear emulsion. These

F IG. I

Morphology of HBJ in phase contrast microscopy

(xl 000). (a) Large blast cell with prominent nucleoli. (b)

Lymphocyte-like cell with dark grey cytoplasm and elongated mitochondria close to the nucleus. (c) Typical plasma cell with abundant mitochondria and well-definedarchoplasm

424 Annals of the Rheumatic Diseases

ELECTRON MICROSCOPY

HBI were easily identified with the electron microscope. In the least differentiated cells the cytoplasm showed many ribosomal clusters and a small amount of erga-

stoplasm (Fig. 2a); the nuclear chromatin was scanty or clumped. In those cells judged to be more differentiated, the ergastoplasm was well developed and showed some orientation around the nucleus (Fig. 2b).

FIG. 2 (a) Electron micrograph of HBI. Fixed n osmiumglutaraldehyde, stained with uranyl acetate and lead citrate. Large 'undifferentiated' cell. The chromatin is fine

and loose; one nucleolus is present; the cytoplasm contains

polyribosomal clusters and short pieces of reticulum localized in one part of the cell. x7000. (b) HBI with a

more differentiated appearance: the ergastoplasm shows the concentric orientations seen in a classical plasma cell but the nucleus is similar to that of the undifferentiated cell.

x10000

(a)

.im

:... 't ... R: .. 4

(b)

Hyperbasophilic immunoblasts in chronic inflammatory rheumatic and collagen diseases 425

IMMUNOFLUORESCENT STAINING

Peripheral blood lymphocytes were isolated by FicollContrix density flotation, washed in Hanks's medium and cell films were prepared on a cytocentrifuge (Shandon). Acetone-fixed smears were stained using an indirect technique. They were first incubated with monospecific goat antisera (Meloy Laboratory), were thoroughly washed, and then incubated with a fluorescein-isothiocyanate conjugated rabbit antigoat-lgG antiserum (FITCRaGG, Hyland Laboratory), washed again, and mounted in buffered glycerol. The antisera or conjugates were tested by immunoelectrophoresis and the specificity of the immunofluorescent staining was ascertained on human bone marrow samples from patients with paraprotein0 aemia. All the reactions were performed with the same batches of antisera. 0 5 10 20 15 25 Days Control reactions included blocking procedures by unconjugated RaGG and incubation with FITC-RaGG alone. When the conjugated sera were used in a dilution of F I G. 3 Levels of HBI in peripheral blood during the course 1: 50, as in the indirect technique, no specific staining was of viral infections. Patients with mumps, rubella, or measles observed; however, by lowering the dilution to 1 :4 a few were assumed to form a homogeneous group. Individual cells exhibiting cytoplasmic fluorescence were found. percentages of HBI were plotted and a schematic curve was Differential counts were made by alternate phase contrast fitted by eye and fluorescent microscopy for the percentage of positive cells, and at least 1000 lymphoid cells were counted. o RA SCAT latex Fll+ HBI[/10' o RA SCAT latex FilThirty patients (22 RA, 8 SLE) and ten normal donors lymiiphocytes A RA SCAT latex Fll. were evaluated for percentage of fluorescent cells, using 30 * SLE * Psoriatic arthritis antihuman antisera (specific for a, y, ,u, K, A), and 17 of these were further investigated using monospecific antiIgA, anti-IgG, anti-IgM, and FITC-RaGG in a dilution of 1:4. . .-

-,

20

Results * * P^ CAN The range of HBI concentration values in each group PAN * 021 is given in Table I. In normal subjects the number of D~ 'p HBI never exceeded 5/103 lymphocytes, considered to 10 AC be the upper limit of normal. All values from patients with degenerative arthritis were normal. In the group _ NPAN with acute viral infections, the percentage of HBI was ° ° 8 °variable and depended on the day of blood sampling *O°0 ioo PAN 0' (Fig. 3); the highest values were observed within 3-4 0 20 40 80 60 100 days of the onset of disease, while by the 15th day all ESR 1st hour the values were nearly normal. Raised values (up to 60 HBI/103 lymphocytes) were FIG. 4 Relationship between disease activity and HBI often observed in patients with RA and SLE, mainly shown by correlation between HBI level and ESR in RA in those with recent exacerbation of their disease; and SLEpatients clinical improvement was usually associated with a In SLE there was a correlation (r = 0-38; P < 0 02) fall in the percentage of HBI. The relationship between HBI and disease activity is illustrated by the between the percentage of HBI and gammaglobucorrelation between the HBI level and the erythrocyte linaemia. No other correlations were found either sedimentation rate (ESR) in patients with RA (r = with total leucocyte and lymphocyte counts, with 0 2666; P < 0 02) and SLE (r = 0-378; P = 0-02) (Fig. rheumatoid factor titre in RA, or with antinuclear 4). However, serial determinations of HBI and ESR factor titre in SLE. In 8 of 21 samples from patients in 8 patients did not show such a correlation, since the with nonprogressive scleroderma there was a changes in the percentage of HBI occurred earlier moderate increase of HBI. than those in the ESR value. Fig. 5 gives two typical Ten of fifteen patients with ankylosing spondylitis examples of dissociation between HBI and ESR. had clinical and laboratory evidence of active disease; Patients with clinically inactive SLE and normal ESR all had normal HBI values. Patients with gout and often had raised values of HBI (> 10/103 lymphocytes) moderately increased HBI all had inflammatory compared with RA patients (Table II), the difference synovitis, while those with normal HBI values had no between the two groups of patients being significant. joint involvement. In patients with psoriatic arthritis /

*

Ai a

_

*

0

/

0

u

OS

0

AA

A

_ M

A

0

A

426 Annals ofthe Rheumatic Diseases

60-

the percentage of HBI was always normal whereas ESR was raised in all but four; nine patients had a markedlyraised ESR (range 60-100mm) and multiple joint involvement.

40-

IMMUNOFLUORESCENT STAINING

, 204-V

0~

S020

0

Days

10

i,l

40 b i

60

v _

i0m

80

= [40 ° 9

-0-

(N

-O

A.-z~~~~-

J I 10\-VI

4 0 Months

8

12

16

20

FIG. 5 HBI level and ESR in two patients with SLE. (a) SLE in a girl hospitalized with fever, joint and pericardial symptoms, positive ANFandLE tests. In response to treatment (prednisone 40 mgfd) the percentage of HBI fell rapidly but remained abnormal for several weeks, while the ESR began to decrease only one month later. (b) SLE without clinical symptoms; over a period of 20 months ESR remained normal while ANFand LE cells were always present; the percentage of HBI was most often high. The arrow indicates a transient clinical relapse. The number of LE cells and HBI increased at the same time, while there was no change in the ESR value

Cells which were found to fluoresce never showed the morphological features of typical plasma cells, and by alternating phase contrast and fluorescent microscopy they were identified as medium and large HBI. Few, if any, fluorescent cells were observed in the peripheral blood of normal subjects and the percentage correlated well with that of HBI (Table III). In most RA patients the percentage of fluorescent cells was increased to the same extent as that of HBI, while in SLE patients and in two RA patients fluorescent cells tended to be less numerous than HBI. There was a wide disparity in the relative number of cells positive for IgA, IgG, and IgM (Table IV), but the number of samples was too small for interpretation. In RA patients (either seropositive or seronegative: Waaler-Rose test) and in two SLE patients, cells reacting with FITC-RaGG were found and their number did not always correlate with the total number of Ig positive cells. In a few cases, however, they probably represented all the positive cells (Table IV). EFFECT OF TREATMENT

Corticosteroids have been found to lower the percentage of HBI in the same way that they influence clinical improvement (Fig. 5a). This was not true for chlorambucil. In ten patients the percentage of HBI was estimated before treatment and again between 10

Table II Compar-ison of relationship between HBI and ESR in RA and SLE ESR Ist h

(mm)

Group

Range of HBI values

S-10/103 lymphocytes

% of cases

No. of cases

% of cases

7 6 2

41-2 30-0

1 10 0 0

50-0*

RA (40) SLE (15)

12 3 0 0

30-0 20-0

12 9 0 0

30-Ot 60-Ot

RA (21) SLE (5)

7 1

33-3 20-0

10 4 0 0

47*6

No. of cases

0-20

RA (17) SLE (20)

Ankylosing spondylitis (75) Psoriatic arthritis (4) 21-50

Ankylosing spondylitis (9) Psoriatic arthritis (6) >50

>10/10' lymphocytes

Ankylosing spondylitis (1) Psoriatic arthritis (8)

0

0 1

* P < 0 01 (exact probability calculated according to the principles for small series). t P < 0 05 (X2 test). Number of cases in each group in parentheses.

5.9*

80-0

Hyperbasophilic immunoblasts in chronic inflammatory rheumatic and collagen diseases 427

Table III Comparison between morphology and immunofluorescent (IF) staining No. of cases

Group

HBI/103 lymphocytes Giemsa staining

Controls RA Seropositive Seronegative SLE *

Correlation

P

IF staining

10

2*4 + 1.9

2-0 ±16

0-81