Defective T cell function in atopic dermatitis

Defective T cell function in atopic dermatitis Gary S. Racheiefsky, M.D., Gerhard Opelz, M.D., M. Ray Mickey, Ph.D., Masahiro Kiuchi, M.D., Paul I. Terasaki, Ph.D., Sheldon C. Siegel, M.D., and E. Richard Stiehm, M.D. Los Angeles, Cnlif.

The cellular immwne system of 37 patients with atopic dermatitis (AD) was assessed by menslrring peripheral blood T and B cells and the in vitro lymphocyte response to gmdcd doses of phytohemagglwknin (PHA) (baekgronnd and 6 conoentrations of PHA from 100 to 1.6 pg). These were then correlated with clinical severity, eosinophil counts, and serwm IgE levels. The IgE levels (1,481 ID + 852 SEM), eosinophil covnts (977 + .X43), rind absolute number of B cells (958 ? 183) were significantly (p < 0.05) higher than in age-matched controls (YO IU t 28, I&? + 79, and 480 ? 60, respectively), and each significantly (p < 0.05) correlated with the clinical severity. By contrast, percent B lymphocytes ($0 + l), percent (51 2 .S) and total (8,357 k 217J T cells did not difSer from controls. Eleven patients had low percent T cells (< 40%); clinical and laboratory evaluation in these patients did not difler from the remaining 86. Lymphocytes from AD patients had higher background deoxyribonucleic acid (DNA) synthesis than controls (suggestive of increased nwmber of 13 cells) and significantly depressed responses at the 10% PHA concentrations (6.3, 3.1, and 1.6 pg), which significantly correlated (p < 0.05) inversely with IgE levels. These studies suggest a .dtle defect in T lymphocyte function leading to increased B cells una increased IgE production.

There is considerable indirect evidence for an immune defect in atopic dermatitis (AD). Such patients have frequent bacterial skin infections,l severe viral skin infections (notably herpes simplex and vaccinia) ,2 and depressed delayed cutaneous hypersensitivity reactions that clinically resemble certain immunodeficiency syndromes. Decreased delayed-type hypersensitivity in AD has been noted to a variety of antigens, including old tuberculin (following BCG immunization) ,3 tetanus toxoid and staphylococcal and streptococcal antigens4 Lawlor and associates” in evaluating the phytohemagglutinin (PHA) skin test response in 50 patients with no known immunodeficiency, noted absent responses in only 3, 2 of whom had AD. Buckley and McGeadyO noted anergy to candida and streptococcal antigens From the Departments of Pediatrics, Surgery, and Biomathematics, School of Medicine, University of California. Supported by Grants Nos. AM 02375 and HD 06463 and Contract No. HIH-NC1 72-2008 kthin the Virus Cancer Program of the National Cancer Institute from the National Institutes of Health, Bethesda, Md. Computing assistance lras obtained from the Health Sciences Computing Facility, UCLA, supported by National Institutes of Health special research grant RR-3. Received for publication May 14, 1975. Accepted for publication Aug. 20, 1975. Reprint requests to: Dr. E. Richard Stiehm, Department of Pediatrics, UCLA Medical Center, Los Angeles, Calif. 90924. Vol. 57, No. 6, pp. 569-576

570

Rachelefsky

I. Laboratory individuals

TABLE

Laboratorv

J. ALLERGY CLIN.

et al.

characteristics

of 37 atopic

dermatitis

37 atopic dermatitis patients (meen f SEMI’

test

compared

Normal (mean f SEMI’

10,480 f 551

White blood count (cells/mm3) Eosinophil count (cells/mmJ) IgE level

patients

9,550 f 780 (N = 50) 182 f 79 (N = 50) 70 h 28 (N = 200)

977 f 143 1,482 f 252

B %?” Per cent Total (cells/mm*)

“SEM: one standard tNS : not significant.

to normal

Dt

NS
NS

20.3 f 1.3 958 It 123

16 f 2 480 3~ 60 (N = SO)

<0.05

51 f 2.2 2,351 f 217

56 zk 2 2,180 3~ 210 (N = 50)

NS NS

T cells: Per cent Total (cells/mma)

IMMUNOL. JUNE 1976

error of the mean.

in 6 of 10 children with AD. Lymph node morphology in 2 AD patients disclosed normal germinal centers but deficient thymic-dependent cortical areas.? Eczema is a consistent component of the Wiskott-Aldrich syndrome,8 is occasionally seen in ataxia telangiectasiag and X-linked agammaglobulinemia,lo and is associated with one recently described defect in chemotaxis.ll These observations, combined with the persistently high IgE levels, suggest abnormal cellular immunity in AD. Thus we evaluated the cellular immune response of 37 patients with AD by the use of B and T cell enumeration and the titrated in vitro PHA response and correlated these studies with clinical severity, eosinophil count, and IgE levels. MATERIALS AND METHODS Thirty-seven consecutive patients with AD were studied at the UCLA pediatric allergy clinic or the private practice of the authors (G. R., S. S.) in an 8-mo period. All had typical AD for at least 2 yr. Fourteen of the 37 patients also had asthma and allergic rhinitis, 12 of 37 had allergic rhinitis or asthma, and 11 had AD alone. The mean age of the patients was 9.6 yr (? 0.9 SEM; range, 3-24 yr) ; there were 18 females and 19 males. Twenty-three were 1 was receiving oral prednisone, and 13 were not being receiving topical corticosteroids, treated. A clinical score, modified from Clendenning and associates,12 was used to evaluate the intensity and severity of the illness. Ten features were analyzed, including: (1) erythema, (2) vesiculation, (3) crusting, (4) excoriation, (5) lichenification, (6) pigmentation, (7) pruritus, (8) loss of sleep, (9) disease involving 4 separate anatomic areas or less, and (10) disease involving more than 4 separate areas. Each feature was graded on a 1 to 4 scale (absent = 4, severe = l), and these were added; a score of 40 indicates no disease and 10 indicates maximally severe disease. Serum IgE levels were determined by a radioimmunoabsorbent technique with Phadebas” IgE test kit. Mean normal value was 70 international units (IU)/ml f 28 (SEM). A value greater than 300 IU/ml is considered elevated.* White blood counts and differentials were determined and total lymphocyte counts were calculated by multiplying the percent of *Personal communication February, 1975.

: Douglas

Heiner,

M.D.,

Harbor

General

Hospital,

Torrance,

Calif.,

VOLUME 57 NUMBER 6

T cell function

.

I

.

1000

dermatitis

571

. .

100

SERUM

FIG. 1. Comparison dermatitis patients. disease. Correlation

.

.

in atopic

.

10,000

Ig E &U./ml)

of serum IgE levels (IU/ml) with clinical A score of 40 indicates no disease and coefficient (R = -0.35) was statistically

severity scores in 37 atopic 10 indicates maximal severe significant (p < 0.05).

lymphocytes by total white count. Total peripheral blood eosinophils were determined by Pilot’s teclmiquera; a count greater than 500 cells/mma is considered elevated. Peripheral blood lymphocytes were isolated from heparinized venous blood by FicollHypaque gradient centrifugation. No more than 5v0 of the isolated cells were granulocytes. PHA stimulation of isolated lymphocytes was performed hy modification of the semimicro method of Sengar and Terasaki.14 PHA (PHA-M, Difco Laboratories, Detroit, Michigan), at 7 different concentrations (background and 100, 25, 12.5, 6.3, 3.1, and 1.6 $8) is added to triplicate aliquots of 105 lymphocytes suspended in 0.1 ml culture medium and cultured at 37” C in a 5?& CO, atmosphere. Ha thymidine (0.8 pg/culture) is added after 72 hr and the cultures continued for 16 more hr, at which time the DNA was precipitated and Ha thymidine uptake counted by liquid scintillation spectrophotometry. The results are expressed as counts per minute (CPM) + SEM. The background count represents spontaneous DNA synthesis, i.e., without the addition of PHA. At least one normal healthy nonallergic age-matched control was done simultaneously with each patient. T cell enumeration utilized the number of lymphocytes forming rosettes with sheep erythrocytes (E rosettes).15 Lymphocytes surrounded by at least 3 red blood cells were considered T cells. Total T cells were determined by multiplying their percentage by the total peripheral blood lymphocyte count. Normal values for T cells are 56% ? 2% (SEM) and 2,180 cells/mm3 + 210 (SEM). B lymphocytes were determined by the presence of surface membrane immunoglobulins, detected by immunofluorescent staining with fluorescent-conjugated polyvalent goat antihuman immune globulin.16 Normal values for B cells were 16% ? 2% (SEM) and normal total for B cells was 480 cells/mm3 t 60 (SEM). Correlation coefficients were computed using the logarithmic scale for cell counts, PHA response, and IgE levels. Multiple and partial correlation coefficients were calculated using Dixon’s program BMDP2R.17

572

Rachelefsky

J. ALLERGY CLIN.

et al.

IMMUNOL. JUNE 1976

40

100

1000 NO. OF B CELLS

10,000

(ddmm3)

FIG. 2. Comparison of total number of peripheral blood B lymphocytes with clinical severity scores. A score of 40 indicates no disease and 10 indicates maximum severity. Correiation coefficient (R = -0.40) was significant (p < 0.05).

RESULTS Analysis of cases by mode of therapy, age, sex, or duration of AD revealed no significant association with TgE levels, percent, or numbers of T or B cells or PHA response. The mean clinical severity score was 24.8 f 1.4 (XEM) . The mean eosinophil count (Table I) was 977 ? 143 (SEM) cells/mm3; 26 had counts greater than 500. There was a significant inverse correlation between the eosinophil count and the clinical severity score (R = -0.38, p < 0.05). The mean serum IgE level was 1,482 t 252 IIJ/ml; this also had a significant inverse correlation with the clinical severity score (R = -0.35, p < 0.05) (Fig. 1). The mean percent peripheral B cells was 20.3 + 1.3 (SEM) . Seven (29% ) patients had an elevated percent B cells (more than 28% ) ; none were decreased. The mean absolute B cells was 958 _+123 (SEM) cells/mm” and was increased in 21 subjects and inversely correlated with the clinical score (R = -0.40, p < 0.05) (Fig. 2). Although the eosinophil count, serum IgE levels, and total B cells varied inversely with the clinical score, further analysis indicated that only the eosinophil count and the IgE level taken together correlated more closely to clinical severity than either separately (coefficient of partial correlation, -0.51, p = 0.04). The inclusion of the total B cell count did not increase this level of correlation (coefficient of partial correlation = 0.24, p = 0.28).

VOLUME 57 NUMBER 6

T cell function

in atopic

TABLE II. Comparison (< 40%)

of 11 atopic dermatitis patients with low with 26 patients with normal percentage T lymphocytes Eosinophll count (cells/mm3)

Serum W W/ml)

Clinical score

percentage

dermatitis

T lymphocytes

s cdl

8 %

1

Total

573

.

t cell total

I I patients with decreased 96 T cells MekIll 22.8 1,958 SEM* 2.8 607

1,223 331

20.7 2.4

1,061 317

1,;;‘:

Zdpatients with normaI A Tcells Mean 25.3 1,368 SEM* 1.6 278

887 153

19.8 1.8

924 82

2,420 410

NS NS Pt *SEM: one standard error of the mean. tNS : not significant.

TABLE III. In vitro atopic

dermatitis

NS

NS

NS

lymphocyte response to PHA at ail concentrations patients with 50 normal controls

tested,

-co.05

comparing

37

Response (CPM X 10’) PHA (ad O$ 100 25

12.5 6.3 3.1 1.6

Number tasted

Atopic dermatitis (maan f SEW’

37 i: 35 3j

1.7 4;; 502 392 207

:i 36 ii 11

::

:66

4

I

Normal (mean f SEMI’

6.6 507 568 472 345 164 74

1.8 :i 29 26 15 8

Pf

<0.05 E NS <0.05 <0.05 <0.05

“SEM: one standard error of the mean. tNS: not significant. $Background or spontaneous DNA synthesis.

The mean percent T cells was 51% + 2 (SEM) . Eleven patients had depressed percent T cells (less than 40%) ; none were elevated. Separation of these 11 pascore, IgE tients from the other 26 revealed no difference (Table II) in severity levels, eosinophil count, percent and total B cells. Fifteen (40%) had increased number of null cells (more than 30% of lymphocytes without T or B cell markers). The number of these cells bore no relationship to clinical severity or PHA response. Examination of lymphocyte responsiveness (Table III) revealed a significantly (p < 0.05) higher spontaneous (background) DNA synthesis in the AD patients. Twelve patients had more than 2,000 CPM (normal 658 CPM + 188 XEM), all of whom had elevated total B cells. None of these patients had clinical or laboratory evidence of infection and there was no correlation of high background counts with decreased PHA response. 111 vitro lymphocyte stimulation by PHA concentrations of 100, 25, and 12.5 pg was normal in all of the patients. However, all of the AD patients demonstrated significantly (p < 0.05) depressed response at one or more of the lower PHA concentrations (6.3, 3.1, and 1.6 pg). The depressed response was correlated inversely at all 3 lower PHA concentrations with the serum IgE levels (correlation, -0.37, -0.42, and -0.49, p < 0.05) but not with clinical severity score (just barely insignificant), percent or total T and B cells, and eosinophil counts.

574

Rachelefsky

et al.

J. ALLERGY CLIN.

IMMUNOL. JUNE 1976

DISCUSSION Our findings demonstrate correlation between the severity of atopic dermatitis and the eosinophil count, serum IgE levels, and tota, number of peripheral blood B cells. In addition, there appears to be a subtle defect of T cell function-the response of in vitro peripheral blood lymphocytes to low doses of PHA. This diminished responsiveness is inversely correlated with serum IgE levels. Our finding of increased total B cells in AD is in agreement with the work of Cormane, Hugz, and Hamerlinckifi who observed an increase of periphreal blood B cells in 10 patients (normal 15.5% vs. 37.796) ; the increase was predominantly due to lymphocytes with complement receptors and IgE-bearing cells. Buckley,” however (in 10 patients with AD), observed normal amount of B cells as determined by complement receptors and membrane immunofluorescence. The elevated spontaneous DNA synthesis (background CPM) of the patients’ lymphocytes is confirmatory evidence for the increased amount of B cells. Of interest are data (manuscript in preparation) collected by 2 of the authors (G. 0. and P. I. T.) suggesting that spontaneous lymphocyte H3 thymidine uptake is related to the number of B cells present, i.e., the more B cells, the higher the background DNA synthesis. Their studies suggest that the high background is a result of stimulation of T cells by autologous B cells, and that usually this activity is regulated by T cells. The high background stimulation in our patients appeared not to be related to low-grade or chronic infection. Thus, the high background in AD patients could be taken as an indication of T cell deficiency. The decreased number of T cells in 11 of 37 of the present study is in agreement with Buckley and McGeady”; the 10 patients studied had a mean percent peripheral blood T cells significantly lower than normal. (Two of Buckley’s 10 patients had absent PHA responses while the responses to pokemeed mitogen and concanavalin A were normal.) Fjelde and Kopecka,l” using mitot,ic changes of whole blood leukocytes to PHA, observed a decreased response in patients with AD. The lymphocytes of 2 AD patients evaluated by Lobitz’ had a decreased in vitro response to PHA. Though IgE antibodies appear somehow to be related to the pathogenesis of AD, their exact relationship remains unclear. Studies in animals suggest a relationship between T cell function and IgE production. Synthesis of homocytotropic antibody in rodents, analogous to human IgE, is enhanced following thymic irradiation, thymectomg, or administration of antilymphocyte serum, all of which result in depletion of thymic or T cells. Transfer of thymocytes will terminate this enhanced antibody production.2+Z” The possibility that the elevated serum IgE levels result from reduction of T cell influence is supported by the observations in human neoplastic and immunodeficiency diseases. Waldmann and associates’” demonstrated IgE deficiency in chronic lymphocytic leukemia and multiple myeloma, both diseases of B cell proliferation, while elevated IgE levels were observed in patients with Hodgkin’s disease, a disease associated with abnormalities of cellular (T cell) immunity. In other cellular immunodeficiency syndromes, notably the Wiskott-Aldrich syndrome,25 DiGeorge syndrome,*F Nezelof’s syndrome,27 and Buckley’s syndrome,“’

VOLUME 57 NUMBER 6

T cell function

in atopic

dermatitis

575

(recurrent infections with impaired antibody and cellular immunity), elevated serum IgE levels have been demonstrated. Studies in different species have demonstrated that T cells play an active suppressive role that controls the rate of antibody production.20, w 3o Removal of T cells augments the antibody response to protein”l and polysaccharide antigens. Reconstitution with thymoc.ytes depresses this enhanced response.“2 An attractive hypothesis that associates elevated serum IgE level with a disorder of cell-mediated immunity and T cell function is that of T cells as important regulators of IgE antibody production. Depletion or limitation of thymicdependent regulator rclls that. function to terminate ongoing IgE antibody responses would have the effect of reducing the ability to terminate IgE antibody responses to most antigens. Though our study did not prove a, relationship of cell-mediated immunity to IgE levels and to clinical severity, it does indicate that T cell function is abnormal in patients with atopic dermatitis. Thus, the primary defect leading to the lesions of atopic dermatitis may reside in a subtle defect in T cell regulatory function. REFERENCES 1 Rostenberg, A. : Psychosomatic concepts in atopic dermatitis: A critique, Arch. Dermatol. 79: 692, 1959. 2 Baer, R. L., Atopic dermatitis, Med. Clin. North Am. 43: 765, 1959. prurigo (atopic dermatitis) in 3 Gudjonsson, H., Lodin, A., and Modee, J.: Besnier’s children, Acta Derm. Venereal. (Stockh.) 46: 159, 1966. 4 Rajka, G.: Delayed dermal and epidermal reactivity in atopic dermatitis, Acta Derm. Venereol. 47: 159, 1967. 5 Lawlor, G. J., Stiehm, E. R., Kaplan, M. S., Sengar, D. P. S. and Terasaki, P. I.: Phytohemagglutinin (PHA) skin test in the diagnosis of cellular immunodeficiency, J. ALLERGY

CLIN. [MMUNOL. 52: 31, 1973. 6 Buckley, R. H., and McGeady, 8. J.: Studies of cell mediated immune function in atopic eczema, J. ALLERGYCLIN. IMMUNOL. 53: 72,1974. (A&t.) 7 Lobitz, W. C.: Suppressed cell mediated immunity in two adults with atopie dermatitis, Br. J. Dermatol 86: 317, 1972. 8 Stiehm, E. R., and McIntosh, R. M.: Wiskott-Aldrich syndrome: Review and report of a large family, Clin. Exp. Immunol. 2: 179, 1967. 9 Reed, W. B., Epstein, W. L., Boder, E., and Sedgwick, R.: Cutaneous manifestations of ataxia-telangiectasia, J. A. M. A. 195: 746, 1966. 10 Petrrson, R. D. A.: Immunologic responses in infantile eczema, J. Pediatr. 66: 22-1, 1965. 11 Hill, H. R., Quie, P. G., Pabst, H. F., Ochs, H. D., et al.: Defect in neutrophil granulocyte chemotaxis in Job’s syndrome of recurrent “cold” staphyloc~occal abscesses, Lancrt, 2: 617, 1974. 12 Clendenning, W. E., Clack, W. E., Ogawa, M., and Ishizaka, K.: Serum IgE studies in atopic dermatitis, J. Invest. Dermatol. 61: 233, 1973. 13 Pilot, M. L.: Use of base in fluids for counting eosinophils, Am. J. Clin. Pathol. 20: 870, 1950. 14 Sengar, D. P. S., and Terasaki, P. 1.: A semi-micro mixed leukocyte culture tt,st, Transplantation 11: 260, 1971. of normal T cells in chronic 15 Wybran, J., Chandler, S., and Fudenberg, H. H.: Isolation lymplmtic leukemia, Lancet 1: 126, 1973. 16 Pernis, B., Form, L., and Amente, L.: Immunoglobulin spots on the surface of rabbit lymphocytes, J. Exp. Med. 132: 1001, 1970. 17 Dixon, W. J., editor: BMDP Biomedical computer programs, Berkeley, 1975, University of California Press. 18 Cormane, R. H., Hugz, S., and Hamerlinck, F. F.: lmmunoglohulin-and-complement bearing lymphocytes in allergic contact dermatitis and atopic dermatitis (eczema), Br. J. Dermatol. 90: 597, 1974.

576

Rachelefsky

et al.

J. ALLERGY CLIN.

IMMUNOL. JUNE 1976

19 Fjelde, A., and Kopecka, B.: Cell Transformation ant1 mitogenic effects in blootl leucoc:ytcJ cultures of atopic dermatitis patients, Acta Derm. Venereal. 47: 168, 1967. 20 Tada, T.: Regulatory disturbances in IgE antibody formation, Acta Pathol. Jap. 23: 907, 1973. 21 Tada, T., Taniguchi, M., and Okumura, K.: Regulation of homocytotropic antilmtly formation in rat. II. Effect of radiation, J. Immunol. 106: 1012, 1971. 22 Okumura, K., and Tada, T.: Regulation of homocytotropic antibody formation in thca rat. III. Effect of thymectomy and splenectomy, 6. Tmmunol. 106: 1019, 1971. 23 Okumura, K., and Tada, T.: Regulation of homocytotropic antibody formation in the rat. IV. Inhibitory effect of thymocytes on the homocytotropic antibody, J. lmmunol. 107: 1682, 1971. 24 Waldmann, T. A., Bull, J. M., Bruce, R. M., Broder, S., et al.: Serum immunoglolmlin E levels in patients with neoplastic disease, J. Immunol. 113: 379, 1974. 25 Raldmann, T. A., Polmxr, S. H., Balestra, 8. T., Jest,, M. C., et al.: lmn~unoglotmlin E in immunologic deficiency diseases, J. Immunol. 109: 304, 1972. 26 l’olmar, 8. II., Waldmann, T. A., and Terry, W. D.: 1gE in immunodcficic,ncy, Anl. J. Pathol. 69: 499, 1972. 27 Buckley, R. H., and Fiscus, S. A.: Serum IgD and IgE concentrations in immunodefieiency diseases, J. Clin. Invest. 55: 157, 1975. b: 28 Buckley, K. H., Wray, B. B., and Belmaker, E. 2.: Extreme hyperimmunoglobulinemia and undue susceptibility to infection, Pediatrics 49: 59, 1972. 29 Michael, J. G., and Bernstein, 1. L.: Thymus dependence of reaginic antil)ody formation in mice, J. Immunol. 111: 1600, 1973. 30 .Jarrett, E., and Ferguson, A.: Effect of T-cell depletion on the potrntiated reagin response, Nature 250: 420, 1974. 31 Kerbel, R. S., and Eidinger, I).: Enhanced immune responsivrness to a thymus independent, antigen early after adult thymectomy; Evidence for short-livrd inhihitory thymus-derived wlls, Eur. J. Immunol. 2: 114, 1972. 32 Baker, P. J., Stashak, P. W., Amsbaugh, D. F., and Prescott, B.: Evidenw for the existence of 2 functionally distinct t,ypes of cells which regulate thca :lntibotly response to type ITT pneumococcal polysaccharides, J. lmmunol. 105: 1581, 1970.