Clin. lab. Haemat. 1982, 4, 239-246

A study of the cellular and humoral immune response in patients with myelofibrosis

E . V E L L E N G A , N . H . M U L D E R , T . H . T H E * & H.O. N I E W E G Divisions of' Haemutology and *Clinical Immunology, Department of Medicine, University of Groningen, The Nether1and.Y

Accepted for publication 23 November 198 1

Summary There was evidence of impaired cellular immunity in 10 patients with myelofibrosis. In-vitro lymphocyte transformation with phytohaemagglutin, concanavalin A, and dinitrochlorobenzene skin reaction were diminished. Signs of impaired humordl activity were also found. the primary response to a-Helix pomatia haemocyanin being impaired, particularly in the immunoglobulin-A class. Moreover three patients had a benign paraproteinaemia. Immune- complexes (IC) could be demonstrated with various test systems. The indirect granulocyte phagocytosis test was positive in SO";, the Clq-binding in 7 0 ; , and the polyethylene glycol precipitation test in 50";. In most patients complement levels were normal, although the patient with the most advanced disease had low C3A, C3 and C4 levels accompanied by high levels of IC. No correlation could be shown between impaired immune response or levels of IC when they were related to spleen diameter or degree of anaemia. Some relation however existed between disturbed immune response and IC when they were related to time elapsed since diagnosis. It is suggested that the impaired immune response is the result of primary bone marrow disease and that the presence of IC may reflect the extent of fibrosis.

Keywords : myelofibrosis, immune response

There is evidence for immunological dysfunction in patients with myelofibrosis. Immune complexes (IC) occur in the sera of these patients (Lewis & Pegrum 1978) and a high frequency of autoantibodies has been reported (Long et al. 1976).

The relevance of these immunological abnormalities is not clear. They could be of significance in the origin of the disease, as has been suggested in NZW-mice (Adler et ul. 1974). On the other hand they could be the result of lymphoid participation in the clonal abnormality that leads to MF (Jacobson, Salo & Fialkow 1978). To determine the extent and relevance of both cellular and humoral dysfunction, we studied them quantitatively and determined IC and complement levels in 10 patients with MF.

Correspondence: E. Vellenga, Division of Haematology, Department of Medicine, University of Groningen. Oostersingel 59, 97 13 EZ Groningen, The Netherlands.

014 I 9854/82/090&0239%02.00 (0 1982 Blackwell Scientific Publications 239

240 E. Vellenga et al.

Patients and methods

Pertinent data on the patients with myelofibrosis are given in Table 1. None of the patients had received cytostatic therapy.

The diagnosis of M.F. depended on histology of the bone marrow obtained by Jamshidi needle biopsy of the iliac crest and on the presence of extra-medullary erythropoiesis, either histologically proven or established by the Indium’ l 1 scan (Royer et al. 1978). Control patients were age and sex matched.

Table 1. Characteristics of 10 patients with myelofibrosis

Splenic size below costal margin Hb Lymphocytes Monocytes

Patients (cm) (g/dl) (109/1) (1

Disease duration since diagnosis

(Yr)

1 2 3 4 5 6 7 8 9 10

Normal

20 16 17 15 14 16 17 15 13 12

14.7 3 .O 11.8 2.5 10.0 2.9 10.9 2.0 12.7 3.5 8.0 1 .o 8.8 2.0

10.7 1.8 10.0 I .7 8.1 1.8

12-16 1.5-4.0

0.3 0.2 0.5 0.4 0.2 0.3 0.2 0.3 0.4 0.3

0.2- 0.9

3 3 4 3 3 4 1 1

0.5 1

I M M U N E S T I M U L A N T S

The mitogens used for testing immune capacity in vitro were Phytohaemagglutinin (PHA) from Burroughs Wellcome, in two concentrations 1 and 5pg/ml; Concanavalin A (Con A) from Calbiochem, 1 and 10 pg/ml; Poke Weed Mitogen (PWM) from Gibco-Grand Island, NY, 10 pg/ml.

In the mixed lymphocyte culture (MLC) 1 x lo5 patients cells were mixed with the same number of a pool of lymphocytes from normal donors. The pooled cells were treated with mitomycin C-M-0503, 32 pg/ml (Sigma Chemicals) and washed three times.

L Y M P H O C Y T E T R A N S F O R M A T I O N T E S T S

Mononuclear leucocytes were isolated from heparinized peripheral blood by Ficoll-Isopaque gradient centrifugation. Cultures were performed with a microculture technique in round bottom microwells. With mitogens as stimulants, 3 x lo4 lymphocytes in 0.1 ml of medium were cultured for 4 days at 37’C with 5”/, COz in humidified air; with antigens lo5 lymphocytes in 0.1 ml were cultured for 6 days

An immunological study in myeiojhrosis 241

All cultures were performed in RPMI 1640, buffered with Hepes and supplemented with 25% human serum. DNA synthesis was measured by radioactive thymidine incorporation (0.5 pCi of ’H thymidine, specific activity 400 pCi/mmol) over 26 h. Cultures were harvested by an automatic cell culture harvester (Skatron, Oslo) filtered through glass fibre filters and dried for 1 h at 60 C.

Radioactivity was measured in a Packard liquid scintillation counter and the results expressed as disintegrations per minute per culture (d.p.m.) after correction for background activity.

2,4 Dinitrochiorobenzene (DNCB) skin tests

The sensitization and challenge techniques have been described by Bleumink (Bleumink et al. 1974). Two thousand microgrammes of DNCB was applied on an area marked by a 2-cm polyethylene ring on the volar side of the arm. Challenges were performed by patch testing on the back 2 weeks after sensitization with 30, I0 and 3 pg of DNCB in acetone and the reactions after 48 h were graded 0-4. The results given are the sum of the reactions thus obtained.

cz Helix pomatia haemocyunin ( H P H ) urztibody response

HPH was provided by the Department of Biochemistry, University of Groningen in purified form, as previously described (de Gast, The & Snijder 1973). The immunization dose was 1.0 mg subcutaneously (s.c.) in the deltoid region. There were no adverse effects of immunization in any of the subjects.

Indirect Elisu Technique

Round bottom microtitre plates (Greiner, M 24-AR), containing 0.1 mg HPH in 0.1 ml M NaHCO, pH 9.6, were incubated with 0.02% NaN, overnight at 4 C. Before use, the plates were intensively washed with tap-water and were then incubated for 45 min at 37’C with 0.1 ml aliquots of doubling serum dilutions, starting with 1 :lo, in phosphate-buffered saline (PBS, 0.01 N, pH 7.2), containing 2”/, bovine serum albumen (BSA) and 0.05% Tween 20. After washing, 0.1 ml of rabbit anti human IgM or IgG (Central Blood Transfusion Service, Amsterdam, batch numbers KH 15-17 A01 and KH 16-1 13-A1) or IgA antiserum (prepared in our laboratory), diluted 1 :5000 in PBS with 2% BSA, were added and incubated for 45 min. at 37%C. After washing, 0.1 ml of peroxidase-labelled sheep antirabbit antiserum (Institut Pasteur, batch No. 09-7502), diluted 1 :2500 in PBS with 2% BSA was added, for 45 min, at 37 C. After washing, 0.2ml of a 0.08% 5- aminosalicylic acid and 0.05% hydrogen peroxide solution (9:1), pH 6.0 was allowed to react for 1 h, after which either the results were read immediately or the reaction was stopped by adding 0.01 ml 1.0 N NaOH.

The highest dilution with a clearly brown colour was taken as the titre. A positive and negative serum pool were always included in each series of

242 E. Vellenga et al.

measurements as a standard. The negative pool should have had a titre of less than 1 :40, if not, the series was discarded. The positive pool had a titre of 1 :640 and only small titre variations were allowed and corrected for. Findings are expressed as titre steps of two-fold serum dilutions, starting with 1 :40.

The C lq-Elisa and indirect granulocyte phagocytosis test (IGFT) were performed as previously described (Van der Giessen, Dokter-Fokkens & The 1980, Van Wingerden et al. 1979).

The polyethylene glycol precipitation (PEG) tests were performed with 50 pl serum, dissolved with 150 ~ 1 0 . 0 8 M CaCl, and 1 ml5% polyethylene glycol in 0.1 M borate buffer, pH 8.3.

The mixture is left at 4'C overnight and certrifugated for 3 min at 3500 g. The clear supernatant is removed and the pellet dissolved in 1 mlO.1 N NaOH. Optical density is measured in a 0.5cm cuvette at 280nm. Immunoglobulins were quantified with radial immuno-diffusion using Tri Partigen plates (Behring- Werke). Paraproteins were indentified by immuno-electrophoresis with monospe- cific antisera against IgG, IgA, IgM, kappa and lambda light chains. The complement components (C3, C3a, C4) were measured by radial immuno diffusion in agar gel, using L-partigen Plates (Behring , Marburg/Lahn). The results were expressed as a percentage of the standard. Statistical analysis was performed by Student's t-test. P values less than 0.05 were considered significant.

Results

M I T O G E N I C S T I M U L A T I O N I N V I T R O (Table 2)

The in-vitro response to PHA-1 (P < 0.001) and 5 (P < 0.01) and to Con A-1 (P < 0.01) showed a significantly diminished response. This difference was not present with the mitogens con A-10 and PWM, although some patients ( 5 , 6, 7) had a very low response. The reaction of patients lymphocytes to a pool of donor lymphocytes was not suppressed. When patients were divided into recently (< 1 year) diagnosed or not (> 1 year), the latter group only showed a significant difference with PHA 1, and 5 ( P < O . O O l , 0.01) PWM (P<O.OOl) and MLC (P < 0.001).

D N C B S K I N R E A C T I O N (Table 2)

The skin reaction with DNCB was tested in nine of the 10 patients. One patient (8) showed a normal response, the others being depressed.

I M M U N O G L O B U L I N (IS) L E V E L S

The quantitative Ig-levels were determined in nine of 10 patients. Three patients had a benign paraproteinemia on the basis of low plasma cell percentage in the bone marrow ( < 5 7 3 , normal skeletal X-rays and absence of Bence Jones protein in urine concentrated 50 times, and stability of paraprotein levels over a prolonged

A n immunological study in myelojihrosis 243

Table 2. In-vitro and in-viva study of the cellular immunity

PHA W m l ) con A (jcgiml)

Patients 5 1 1 0 1 PWM MLC DNCR

I 2 3 4 5 6 7 8 9 10

Normal SD

37.1 17.4 38.6 32.7 11.7 3.3

15.8 26.4 29.9 60.7

0.42 3.1 5.5 6.3 1.3 5.6 3.1

40.3 22.6 9.8

0.51 10.5 37.1 51.4 2.1 0.5 2.8

36.2 5.2 6 .0

57 23.2 17.7 k 12.8 -t 18.4 .+ 12.6

0.4 7.3 0.6

13.1 0.6

12.7 1.3 1.8 n.t. n.t.

12.7 k 10.8

18.1 15.0 5.3

11.7 0.02 3.2 0.2

20.9 33.2 34.0

22.5 i 1.7

~

9.1 I X 1 0.4 1 X 2 0.0 I 2 2.0 1

X 0 44.0 11 10.1 n.t. 45.0 3

21.2 4 10 10.7

Counts are given in desintegration per min x lo3 for stimulated cultures minus unstimulated cultures (background). X : in these experiments background activity extremely high. n.t. : not tested.

period. In six patients without paraproteinemia one each had a decreased level of IgG, IgM and IgA.

A N T I B O D Y R E S P O N S E T O a - H E L I X P O M A T I A H A E M O C Y A N I N ( H P H ) (Table 3) The mean values for IgG, IgM and IgA anti-HPH titres 2 weeks after primary immunization are 5.4;4.9;4.1. Mean titres in the patients were lower in all classes.

Table 3. Immunoglobulins and anti-HPH antibody response in 10 patients with myelofibrosis

IgG IgM IgA HPH- antibody response Patients rng/100ml mg/100ml mg/100ml IgG IgM IgA

1

3 4 5 6* 7 8t 9 10::

7 i

I122 n.t. 1298 1496 1562 4674

902 1210 2000 2604

24 1 n.t. 45

150 182 72

120 150 350 208

90 n.t. I65 323 257 112 394

1 I64 200

16

4 5

12 6 1 5 5 4 9 3

5 5 4 8 4 4 6 4 5 4

3 5 5 6 I 5 5 3 7 1

Normal 1 0 0 s 1 800 8 s 170 120 400 6.3 5.7 5.7 Range 2- 10 2 -9 3 8

*IgG- i.; tlgA- 1.; $IgA- ) paraprotehernia; n.t.: not tested.

244 E. Vellenga et al.

Table 4. Levels of immune complexes and complement components.

Patients IGFT Clq PEG c3 C3a c 4

I 2 3 4 5 6 7 8 9 10

Normal SD

60 10 9

44 28

203 21 2 4 6

< 20

I000 330 5 50

1000 300 400

20 580 40

~

< 40

70 X

1 I8 136 84 42 72 90 92 80

88* & 12.8

63 X 94

164 51 57 88

I66 149 102

98* & 26.0

90 X

124 128 98 56

134 146 100 I10

94.7* t 24.7

ICFT: indirect granulocyte phagocytosis test score; C 1 q : C Iq-Elisa titre; PEG : polyethylene glycol prcciptation test; SD: standard deviation; X : not tested. *complement levels are expressed as a percentage of the standard.

Significance was reached for IgA ( P < 0.05). IgG ( P < 0.5) and IgM ( P < 0.1) were not significantly lower.

I M M U N E C O M P L E X E S ( I C ) A N D C O M P L E M E N T L E V E L S (Table 4)

Increased levels of IC, tested with IGFT were found in five of the 10 patients. The Clq method was positive in seven of the 10 patients and the PEG in five of the 10. No positive correlations existed between the different methods used for detection of IC. In nine patients complement levels were determined. Decreased levels were found especially in patient 6, while in patients 1 and 5 only diminished levels of C3A were found. The others had normal or increased complement levels.

Discussion

An impaired in-vitro lymphocyte transformation test was found in patients with M F following stimulation with PHA and con A. This partially confirms the earlier report of Di Bella (Di Bella & Brown 1978). As the response to these mitogens involves T cells, this points to a disturbance of cellular immunity, which is supported by the results of the i n -v im DNCB skin reaction. The impaired cellular immune response could be due to a number of causes i.e. to a decreased number of T cells, a very low number of monocytes, which can influence these tests (Rosenstreich, Farrar & Dougherty 1976), or to an intrinsic defect of T-cells.

In this study the patients had a normal number of lymphocytes, and Lewis & Pegrum (1978) reported a normal percentage of T cells in patients with MF. Although the monocyte count was not severely depressed, the depressed DNCB

An immunological study in myelojhrosis 245

reactions suggest a possible role of monocyte malfunction. However, for in-vitro tests only a minimal number of monocytes are required and it is therefore probable that the variable disturbances in T cell function must be ascribed to an intrinsic abnormality in some subsets of T cells.

Ig-deficiency was only occasionally observed in the patients without benign paraproteinemia. The frequency of Ig-deficiency appears to vary in different MF patient groups. Di Bella & Brown (1978) reported Ig-deficiency in two of nine patients and Selroos (Selroos, Skrifvars Wasastjerna 1973) in five of 12 patients.

In our patients B-cell capacity, as tested with HPH, was depressed in the IgA class. In other classes the mean values were lower than normal, but the difference was not significant.

In plasma cell dyscrasia an impaired B cell response to the IgA class has been described (Weits et al. 1979) but this occurred mostly in patients with a high paraprotein level. In our patients an impaired humoral response occurred regardless of the presence of a paraproteinemia. It is therefore improbable that plasma cell dyscrasia in our patients is the only factor responsible for the impaired B cell response. It is possible that both the impaired response and the plasma cell dyscrasia are related to a disturbance in B cell function, inherent to the underlying disease. The consequence of this impaired humoral response will be an increased risk of bacterial infection, as has been decribed in M F (Ward Block 1971) and this may affect the clinical course of the patients.

IC were found with IGFT in 5 of the 10 patients. Lewis & Pegrum (1977) described a positive test in all patients. This discrepancy could be due to the phase of the disease in which the patients were studied. We found a low frequency of these IC in patients, especially in those, who were recently diagnosed, while the patient (6) with the most advanced disease had the highest level. These results are comparable to those of Gordon et al. (1980) who detected IC with the Raij-cell technique. The depressed complement levels in patient 6, including low C3a, C3 and C4, could point to activation of the classical pathway of the complement system, probably induced by IC. However these findings are not consistent with previous results (Gordon et al. 1980) where activation of the alternative pathway of the complement system has been found.

With the Clq method, there was also a low frequency of positive tests in recently diagnosed patients. When comparing the recently diagnosed patients and the others, only the latter showed a severe disturbance in cellular immunity. However the DNCB skin test was also disturbed in newly diagnosed patients, which could indicate other factors such as monocyte function which might influence this test.

Although we found some correlation between IC levels and time since diagnosis, this may have little relation to the duration of the disease, since patients may have few symptoms and signs in the early years (Ward & Block 1971). No correlation could be shown between the impaired immune response and levels of IC on the basis of spleen diameter.

The presence of IC could be relevant to the extent of fibrosis. IC can initiate the release reaction of platelets (Becker & Henson 1973) and during this reaction

246 E. Vellenga et al.

different factors are released, e.g. the platelet derived growth factor (Ross & Vogel 1978). This factor can stimulate the replication of mesenchymal cells and increase their collagen synthesis resulting in fibrosis (Burke & Ross 1977).

The clinical relevance of this hypothesis is that elimination of IC by chemotherapy or plasmaphoresis may be beneficial.

Acknowledgement

We are grateful to Dr J. Marrink and Dr W. V. Son for technical assistance.

References ADLER S.S., DAY S.B., HILL K., HUNSTEIN W.. SPECK B. & TABIN M.S. (1974) Experimental models

BECKER E.L. & HENSON P.M. (1973) Adv. Immunol. 17, 94 BLEUMINK E., NATER J.P., SCHRAFFORDT KWPS H. &THE TH. (1974) A standard method for DNCB

BURKE J.M. & ROSS R. (1977) Collagen synthesis by monkey arterial smooth muscle cells during

DI BELLA N.J. & BROWN G.L. (1978) Immunologic dysfunction in the myeloproliferative disorders.

GAST G.C. DE, THE T. & SNIJDERS J.A.M. (1973) The human immune response to a-haemocyanin of

GORDON B., COLEMAN M., ROMANO P. & DAY N.F. (1980) Circulating immune complexes associated

JACOBSEN R.J., SALOA. & FIAKOW P. (1978)Agnogenic Myeloid Metaplasia: A Clonal Proliferation of

LANG J.M., OBERLINC F., MAYER S. & HEID E. (1976) Auto-immunity in primary myelofibrosis.

LEWIS C.M. & PEGRUM G.D. (1977) Immune complexes in myeloproliferative disorders. Luncer ii,

LEWIS C.M. & PECRUM G.D. (1978) Immune complexes in myelofibrosis a possible guide to

ROJER R.A., MULDER N.H., NIEWEG H.O., BEEKHUIS H., PIERS D.A. & WOLDRINC M.G. (1978) In

ROSENSTREICH D.L. FARRAR J.J. & DOUCHERTY S. (1976) Absolute macrophage dependency of T-

ROSS R. & VOGEL A. (1978) The platelet-derived growth factor. Cell 14, 203 SELROOS 0. SKRIFVARS B. & WASASTJERNA C. (1973) Immunological deficiency in myelofibrosis.

Scand. J . Hemat. 11, 307 VAN DER GIESSEN M., DOKTER-FOKKENS J. & THE TH. (1980) A solid phase Clq-binding assay for

measuring circulating immune complexes (C lq-Elisa). Immuno-Enzymatic Assq . Techniques, ed. R. Malvano, p. 223. Martinus Nijhoff, Hague

VAN WINGERDEN I., THE TH., VAN DER GIESSEN M. & RHUMKE P.H. (1979) Demonstration of circulating immune complexes in melanoma patients by means of a granulocyte fagocytosis test. Proceedings of the 26th Colloquium Proptides of the Biological Fluids, ed. H. Peeters, p. 345. Pergamon Press, Oxford

WARD H.P. & BLOCK M.H. (1971) The natural history of agnogenic myeloid metaplasia and a critical evaluation of its relationship with the myeloproliferative syndrome. Medicine 50, 357

WEITS J., IX GAST G.G., THE TH., ESSELINK M.F.. VAN DER GIESSEN M. & MANDEMA E. (1979) Class and subclass specific antibody response to haemocyanin in non-malignant paraproteins. J . luh cliti. Med. 94, 458

of the myelofibrosis osteosclerosis syndrome. Advances in Biosciences 16, 295

sensitization testing in patients with neoplasma. Cancer 33, 91 1

proliferation and quiescence in culture. Exp. Cell Res. 107, 387

Cancer 42, 149

Helix pomatia. Acta med. scund. 194, 303

with hypocomplementia in myelofibrosis (Abstract). Clin. Res. 28, 492

Hematopoietic stem cells with secondary myelofibrosis. Blood 51, 189

Biomedicine 25, 39

1151

management. Br. J . Haemat. 39, 233

scintigraphy and splenic erythropoiesis. Acta med. scand. 203, 48 1

lymphocyte activation by mitogens. J . Immunol. 116, 131