Insulin-Induced Lipoatrophy in Type I Diabetes - Diabetes Care

2 downloads 0 Views 380KB Size Report
THIERRY BRUE, MD, PHD. PHILIPPE VAGUE, MD. JEAN-JACQUES GROB, MD. BERNARD VIALETTES, MD. OBJECTIVE — To test the hypothesis that tumor ...
S H O R T

R E P O R T

Insulin-Induced Lipoatrophy in Type I Diabetes A possible tumor necrosis factor-a-mediated dedifferentiation of adipocytes CATHERINE ATLAN-GEPNER, MD PIERRE BONGRAND, MD, PHD CATHERINE FARNARIER, MD Luc XERRI, MD, PHD REGINE CHOUX, MD

JEAN-FRANCOIS GAUTHIER, MD THIERRY BRUE, MD, PHD PHILIPPE VAGUE, MD JEAN-JACQUES GROB, MD BERNARD VIALETTES, MD

OBJECTIVE — To test the hypothesis that tumor necrosis factor (TNF)-a may mediate the loss and the dedifferentiation of subcutaneous fat tissue in the insulin-induced lipoatrophies of a diabetic patient who presented extensive lesions. RESEARCH DESIGN A N D METHODS — An in vitro exploration of cytokine production by peripheral blood mononuclear cells (PBMC) from the reported case was performed and compared with the same explorations of PBMC from three nondiabetic subjects and three diabetic patients without lipoatrophic lesions. A proliferation test and an evaluation of TNFa and interleukin (IL)-6 production from PBMC in presence of insulin were studied. RESULTS — The production of TNF-a and IL-6 by the macrophages of the patient in presence of insulin were dramatically increased in comparison with control subjects. This process needed cooperation with other lymphoid cells and was abrogated by dexamethasone. CONCLUSIONS— In our reported case, a local hyperproduction of TNF-a from macrophages that was induced by the injected insulin could explain the dedifferentiation of the adipocytes of the subcutaneous tissue and the reversion that was induced by the local injection of dexamethasone.

I

nsulin-induced lipoatrophies are rare, mainly as a consequence of the generalization of human insulin usage (1-5). The pathophysiological mechanism is still unknown, even where an immunological cause is suspected, based on histological observations (the deposition of IgA and C3 [6]), the presence of high titers of antiinsulin autoantibodies, and the efficiency of local treatment with corticoids (7). Because of its function in the differentiation of adipocyte (8), tumor necrosis factor (TNF)-a may mediate the atrophy of the subcutaneous adipose tissue that takes

place in these lesions. We attempted to test this hypothesis in a patient showing severe lipoatrophic areas by studying the production of cytokines and the effect of different monoclonal antibodies on this production by peripheral blood mononuclear cells (PBMC) in vitro in presence of insulin.

RESEARCH DESIGN AND METHODS Case r e p o r t The patient was a 25-year-old woman with a 17-year history of type I diabetes.

From Service de Nutrition, Maladies Metaboliques et Endocrinologie (C.A.-G., B.V), Laboratoire d'lmmunologie (RB, C.F.), Laboratoire d'Anatomie Pathologique (L.X., R.C.), and Service de Dermatologie (J.J.G.), Hopital Sainte Marguerite, University of Marseille; Service d'Endocrinologie (T.B.) and Service de Nutrition et Maladies Metaboliques (RV), Hopital de la Timone, University of Marseille; and Service d'Endocrinologie (J.-F.G.), Hopital Saint-Louis, Paris, France. Address correspondence and reprint requests to C. Atlan-Gepner, MD, Service de Nutrition, Maladies Metaboliques et Endocrinologie, Hopital Sainte Marguerite, 270 Boulevard de Sainte Marguerite, 13009 Marseille, France. Received for publication 12 February 1996 and accepted in revised form 13 June 1996. CCM, complete culture medium; 1L, interleukin; PBMC, peripheral blood mononuclear cells; TNF, tumor necrosis factor.

DIABETES CARE, VOLUME 19, NUMBER 11, NOVEMBER

1996

She had proliferative diabetic retinopathy, and at the time of observation, the diabetes was well controlled (HbA lo 6.6%) with two daily injections of human insulin (I lumulin NPH, Lilly) (40 Ul/day). Some very large and profound lipoatrophic areas became apparent after 13 years on her amis, abdomen, and hips, which were injections sites exclusively, while she was being treated with porcine insulin. After she began taking human insulin, the lipoatrophies did not improve and new areas appeared. A histological examination was performed on tissue obtained under local anesthetic from a lesion of the left ami. Standard evaluation by light microscopy and immunohistochemical analysis showed no modification of epidermic and superficial dermic structures. There was no major inflammatory reaction, but some CD3+ and CD68 + cells were present. Adipose cells were small, and there was an increase in the nucleus-lipids droplet ratio, suggesting a state of dedifferentiation. laboratory findings showed anti-insulin antibodies (60.7% in liquid phase immunoprecipitation) and normal blood lymphocyte numeration and that the level of TNF-a in the sera was not increased. A functional in vitro analysis of PBMC from the patient was performed. Eight weeks after treatment by the injection of a mixture of insulin and dexamethasone (4 pg/U insulin) directly into one of the lipoatrophic areas, as described by Whitley (7), there was no modification of metabolic control (1 ibA l o 6.6%) and daily insulin requirement did not change (47 Ul/day), but there was a marked improvement of the locally treated lesions. All of the lipoatrophic areas completely disappeared after 12 months of sequential treatment. In vitro p r o d u c t i o n of cytokines IL-6 and TNF-a synthesis. PBMC were suspended in complete culture medium (CCM) with 10% of either the control subjects or the patients serum. The cells were distributed in 24-well culture plates (1.25 X 10b per well), and a stimulating agent was added either directly (indicated as PBMC plus insulin in Tables 1 and 2) or

1283

TNF-a and insulin-induced lipoatrophy

Table 1—TNF-a production Adherent cells plus insulin (pg • mh 1 • pg"1 • ml)

PBMC plus insulin (pg • mb 1 • pg-1 • ml)

is shown for IL-6 production. The production of IL-6 in presence of insulin by total PBMC of the patient was also dramatically increased (195,900 pg/ml compared with 14,407 pg/ml without insulin).

CONCLUSIONS— Insulin and related peptides are known to be immunogenic in the natural history of type I diabetes (9) and are able to generate allergy or immunological resistance to insulin (10). Insulin-induced lipoatrophies are considered an immunological adverse effect of insulin therapy, because of the prevalence In the first column, data are the ratios of the production of TNF-a by adherent cells in the presence of of high titers of anti-insulin autoantibodies insulin to the production of TNF-a by adherent cells without insulin. In the second column, data are the ratios of the production of TNF-a by adherent and nonadherent cells in the presence of insulin to the proand some histological observations (6). duction of TNF-a by adherent and nonadherent cells without insulin. For the patient, for instance, the Our hypothesis that lipoatrophy is the conproduction of TNF-a by PBMC without insulin was 88 pg/ml and with insulin was elevated to 608 pg/ml sequence of an immune reaction mediated (608/88 = 6.90). by TNF-a is substantiated by the following landmarks. TNF-a, synthesized mainly by after removing nonadherent cells after a duction in the presence of insulin to the activated macrophages (11) and sometimes 60-min incubation at 37°C: lipopolysac- production without insulin by either by T-cells, has pleiotropic effects. It is charide (Sigma, Coger, France) (20 adherent cells alone or total PBMC from known to induce a dedifferentiation of pg/ml), insulin (10 U/ml), dexamethasone the patient were compared with the pro- adipocytes (12). IL-6 is also a multifunc(200 pg/ml), and anti-7-interferon (mono- duction of the same cells of diabetic and tional cytokine, produced mainly by clonal anti-human 7-interferon, Gen- nondiabetic subjects. In our patient, TNF- monocytes but also by T-cells, that is zyme) or anti-HLA-DR monoclonal anti- a production by total PBMC was dramati- known for its implications in inflammabody (mouse monoclonal antibody, cally increased in presence of insulin (601 tion, viral infection, and autoimmune disImmunotech, Marseille, France) (4 and 88 pg/ml without insulin). This hyper- eases (13). In consideration of these data, pg/ml). TNF-a production without stimu- production was T-cell dependent, since the histological examinations, and the results lation was performed in one well with depletion of nonadherent cells (the results of an in vitro exploration in one of our CCM substituted for the stimulating agent. indicated in the first column of Tables 1 observations, we could suggest an hypothAfter 24 h incubation at 37°C and and 2) abrogated the hypersecretion (with- esis. The monocytes of this lipoatrophic centrifugation, supernatants were col- out insulin, 48 pg/ml; with insulin, 48 patient released large amounts of TNF-a in lected. pg/ml) and since the addition of monoclo- the presence of insulin. This high producnal antibodies anti-7-interferon induced a tion needed cellular cooperation with 65% decrease (from 221 to 77 pg/ml) of other lymphoid cells, "y-interferon and Determination of cytokine levels TNF-a levels. Furthermore, dexametha- major histocompatibility complex class II TNF-a and IL-6 were assayed with a solid phase immunoenzymometric assay sone reduced the in vitro production of molecules seemed to be involved in these TNF-a (from 221 to 78 pg/ml). The effect processes as proved by the inhibiting effect (Immunotech, Marseille, France). of the anti-major histocompatibility com- of specific monoclonal antibodies. The plex class II molecules is difficult to inter- high production of IL-6 could be a witness Calculations pret because it was variable, either a 71% to the T-cell and monocyte activation by Data presented in tables are ratios of the production of cytokines by cells with and decrease (from 221 to 64 pg/ml) or an insulin. The abrogation of in vitro TNF-a increase (from 126 to 290 pg/ml) in TNF- secretion by dexamethasone is the counterwithout insulin. a production. In Table 2, the same analysis part of the spectacular restitutio ad inteControl subjects Three nondiabetic subjects were tested as Table 2—IL-6 production normal control subjects, and three type I diabetic patients who did not present any Adherent cells lipoatrophies were also tested. Two of the plus insulin PBMC plus insulin latter presented high titers of anti-insulin 1 1 (pg • mh • pg" • ml) (pg • ml"1 • pg-1 • ml) antibodies (52.3% for diabetic subject 2 and 33.2% for diabetic subject 3 in Table Patient 0.40 13.59 1) to have positive antibody matched con- Nondiabetic subject 0.02 0.16 trol subjects. 0.08 0.51 Diabetic subject Patient Nondiabetic subject 1 Nondiabetic subject 2 Nondiabetic subject 3 Diabetic subject 1 Diabetic subject 2 Diabetic subject 3

1 0.62 0.07 0.21 0.65 0.33 0.60

RESULTS — The results of the determination of the production of TNF-a are given in Table 1. The ratio of TNF-a pro1284

6.9 0.52 1 1.32 0.93 0.67 0.89

In the first column, data are the ratios of the production of IL-6 by adherent cells in the presence of insulin to the production of IL-6 by adherent cells without insulin. In the second column, data are the ratios of the production of IL-6 by adherent and nonadherent cells in the presence of insulin to the production of IL-6 by adherent and nonadherent cells without insulin.

DIABETES CARE, VOLUME 19, NUMBER 11, NOVEMBER

1996

Atlan-Gepner and Associates

grum of adipose tissue after local treatment by this corticosteroid. The role of an autocrine production of TNF-a by adipocytes themselves cannot be excluded but has to be substantiated. Finally, the complete recovery of adipose tissue after local dexamethasone treatment suggests the persistence of preadipocytes in atrophic areas, and in view of this, dexamethasone-induced amelioration could be explained by both its immunological effect and its effect on differentiation. Besides its immunomodulating properties, dexamethasone is known to be a potent inducer of adipocyte differentiation (14,15). In conclusion, insulin-induced lipoatrophies are conceivably an autoimmune phenomenon that is mediated in some cases by an in situ local high production of TNF-a leading to a dedifferentiaton of the adipocytes of subcutaneous tissue. The dexamethasone-inhibition of the local release of this cytokine in the presence of insulin remains an effective treatment for this uncommon complication of type I diabetes.

2.

3. 4.

5.

6.

7.

8.

9. References 1. Blickle JF, Du Cailar G, Reville P: Lipodys-

DIABETES CARE, VOLUME 19, NUMBER 11, NOVEMBER

trophies atrophiques apparues au cours d'un traitement par insuline humaine. PresseMed 12:2534, 1983 Chantelau E, Reuter M, Schotes S, Starke AA: Severe lipoatrophy with human insulin: successfully treated by CSII (Letter). DiabetMed 10:580-581, 1993 Rosman MS: Fat atrophy in human insulin therapy. Diabetes Care 9:436-438, 1986 Goldmann JM, Wheeler MF: Lipodystrophy from recombinant DNA human insulin. AmJMed 83:195-196, 1987 Page MD, Bodansky HJ: Human insulin and lipoatrophy (Letter). Diabet Med 9:779, 1992 Reeves WG, Allen BR, Tattersall RB: Insulin-induced lipoatrophy: evidence for an immune pathogenesis. Br Med J 280:1500-1503, 1980 Whitley TH, Lawrence PA, Smith CA: Amelioration of insulin lipoatrophy by dexamethasone injection. JAMA 235:839840, 1976 Torti FM, Dieckman B, Beutler B, Cerami A, Ringold GM: A macrophage factor inhibits adipocyte gene expression: an in vitro model of cachexia. Science 229:867869, 1995 Delovitch TL, Semple W, Naquet P, Bernard NF Ellis J, Champagne P, Philipps ML: Pathways of processing of insulin by antigen-presenting cells. Immunol Rev

1996

106:195-222,1988 10. Naquet P, Ellis J, Tibenski D, Kenshole A, Singh B, Hodges R, Delovitch TL: T cell autoreactivity to insulin in diabetic and related non-diabetic individuals. J Immunol 140:2569-2578, 1988 11. Vassali P: The pathophysiology of tumor necrosis factors. Ann Rev Immunol 10:411-452,1992 12. Petrushke T, Hauner H: Tumor necrosis factor-a prevents the differentiation of human adipocyte precursor cells and causes delipidation of newly developed fat cellsj Clin Endocrinol Metab 76:742-747, 1993 13. Kishimoto T, Akira S, Taga T: Interleukin6 and its receptor: a paradigm for cytokines. Science 258:593-597,1992 14. Hauner H, Entenmann G, Wabitsch M, Gaillard D, Ailhaud G, Negrel R, Pfeiffer EF: Promoting effect of glucocorticoids on the differentiation of human adipocyte precursor cells cultured in a chemically defined medium. J Clin Invest 84:1663-1670, 1989 15. MacDougald OA, Cornelius P Lin FT, Chen SS, Lane MD: Glucocorticoids reciprocally regulate expression of the CCAAT/enhancer-binding protein alpha and delta genes in 3T3-L1 adipocytes and white adipose tissue. J Biol Chem 269:19041-19047, 1994

1285