Immunologic Tolerance to Intravenously Injected Insulin
http://www.100md.com
《新英格兰医药杂志》
To the Editor: Rosalyn Yalow, the inventor of radioimmunoassay, found that insulin treatments led to the production of antibodies against insulin.1 Initially, this phenomenon was thought to be due to slight immunogenicity induced by the refining of preparations or the difference in amino acid sequences between species. However, even today, when genetically engineered preparations of human insulin are used, anti–human insulin IgG subclasses still are frequently detected in patients treated with insulin. Why therapeutically used insulin molecules, despite having exactly the same primary structure as endogenous insulin, are immunogenic has not been fully clarified.
We describe the induction of immunologic tolerance with the use of intravenously injected insulin in a severely insulin-allergic patient with diabetes. In April 2003, a 24-year-old man with a three-year history of poorly controlled type 1 diabetes and with insulin allergy was referred to our institution. Skin tests performed with all forms of commercially available insulin and insulin analogues were positive, but skin tests were negative for solvent and additives. Attempted desensitization with the subcutaneous administration of insulin, with continuous subcutaneous injection of insulin lispro, or with oral antiallergic agents did not prevent frequent life-threatening allergic symptoms (e.g., systemic urticaria, shock, and dyspnea), especially after bolus injections with meals. However, when no allergic reactions were observed after intravenous injection of 0.05 U of regular insulin per patient, we gradually increased the dose.
We ultimately succeeded in administering all the required insulin intravenously with a portable parenteral-nutrition device (a central venous catheter, a subcutaneously embedded reservoir, and a portable infusion pump). All symptoms of an allergic reaction disappeared as soon as intravenous injection was begun. Oral antiallergic agents were no longer required. Within a year after the introduction of intravenously injected insulin, levels of anti–human insulin IgE returned to normal (decreasing from 2.51 to <0.34 U per milliliter), as did levels of anti–human insulin IgG bound/total (B/T) (from 42.6 to 9.2 percent), without any adverse effect on glucose control (Figure 1).
Figure 1. Blood Levels of Glycosylated Hemoglobin, Anti–Human Insulin Antibodies, and Antiallergic Agents over Time.
Conventional intermittent subcutaneous injections of insulin were switched to continuous subcutaneous injections of insulin lispro in March 2003. In September 2004, continuous intravenous injections of regular insulin (Humulin R, Eli Lilly) were introduced (at a dose of 100 U per 500 ml, given by means of a central line), owing to exacerbation of allergic symptoms. A portable pump delivered 5 to 10 ml per hour, adjusted according to self-monitored blood glucose levels. In addition, undiluted regular insulin (approximately 25 U) was injected intravenously before meals. The figure shows blood levels of glycosylated hemoglobin (Panel A), anti–human insulin IgG bound/total (B/T) (Panel B), and anti–human insulin IgE (Panel C). The levels of anti–human insulin IgG and IgE declined sharply once treatment with intravenous insulin began, and normal levels were reached 10 months after the start of intravenous treatment. The doses of the oral antiallergic agents olopatadine hydrochloride and prednisolone are also shown.
In July 2005, despite these changes in anti-insulin antibodies, subcutaneous injection of regular insulin still caused immediate allergic reactions in our patient. Thus, identical insulin molecules can behave in markedly different ways depending on the route of injection. Also, it is possible that the formation of anti–human insulin IgG is caused only by insulin molecules that are in contact with subcutaneous tissue. We assume that some modification of insulin, such as aggregation, leads to the immunologic reactions.2,3 We believe that the example of our patient not only indicates an option for the treatment of severe insulin allergy but also may help elucidate the pathogenesis of the immunologic response to therapeutic insulin.
Masato Asai, M.D., Ph.D.
Nagoya University Graduate School of Medicine and Hospital
Nagoya 466-8550, Japan
masato-a@mbc.ocn.ne.jp
Masanori Yoshida, M.D.
Toyota Memorial Hospital
Toyota 471-8513, Japan
Yoshitaka Miura, M.D., Ph.D.
Nagoya University Graduate School of Medicine and Hospital
Nagoya 466-8550, Japan
References
Berson SA, Yalow RS, Bauman A, Rothschild MA, Newerly K. Insulin-I131 metabolism in human subjects: demonstration of insulin binding globulin in the circulation of insulin treated subjects. J Clin Invest 1956;35:170-190.
Brange J, Andersen L, Laursen ED, Meyn G, Rasmussen E. Toward understanding insulin fibrillation. J Pharm Sci 1997;86:517-525.
Maislos M, Mead PM, Gaynor DH, Robbins DC. The source of the circulating aggregate of insulin in type I diabetic patients is therapeutic insulin. J Clin Invest 1986;77:717-723.
We describe the induction of immunologic tolerance with the use of intravenously injected insulin in a severely insulin-allergic patient with diabetes. In April 2003, a 24-year-old man with a three-year history of poorly controlled type 1 diabetes and with insulin allergy was referred to our institution. Skin tests performed with all forms of commercially available insulin and insulin analogues were positive, but skin tests were negative for solvent and additives. Attempted desensitization with the subcutaneous administration of insulin, with continuous subcutaneous injection of insulin lispro, or with oral antiallergic agents did not prevent frequent life-threatening allergic symptoms (e.g., systemic urticaria, shock, and dyspnea), especially after bolus injections with meals. However, when no allergic reactions were observed after intravenous injection of 0.05 U of regular insulin per patient, we gradually increased the dose.
We ultimately succeeded in administering all the required insulin intravenously with a portable parenteral-nutrition device (a central venous catheter, a subcutaneously embedded reservoir, and a portable infusion pump). All symptoms of an allergic reaction disappeared as soon as intravenous injection was begun. Oral antiallergic agents were no longer required. Within a year after the introduction of intravenously injected insulin, levels of anti–human insulin IgE returned to normal (decreasing from 2.51 to <0.34 U per milliliter), as did levels of anti–human insulin IgG bound/total (B/T) (from 42.6 to 9.2 percent), without any adverse effect on glucose control (Figure 1).
Figure 1. Blood Levels of Glycosylated Hemoglobin, Anti–Human Insulin Antibodies, and Antiallergic Agents over Time.
Conventional intermittent subcutaneous injections of insulin were switched to continuous subcutaneous injections of insulin lispro in March 2003. In September 2004, continuous intravenous injections of regular insulin (Humulin R, Eli Lilly) were introduced (at a dose of 100 U per 500 ml, given by means of a central line), owing to exacerbation of allergic symptoms. A portable pump delivered 5 to 10 ml per hour, adjusted according to self-monitored blood glucose levels. In addition, undiluted regular insulin (approximately 25 U) was injected intravenously before meals. The figure shows blood levels of glycosylated hemoglobin (Panel A), anti–human insulin IgG bound/total (B/T) (Panel B), and anti–human insulin IgE (Panel C). The levels of anti–human insulin IgG and IgE declined sharply once treatment with intravenous insulin began, and normal levels were reached 10 months after the start of intravenous treatment. The doses of the oral antiallergic agents olopatadine hydrochloride and prednisolone are also shown.
In July 2005, despite these changes in anti-insulin antibodies, subcutaneous injection of regular insulin still caused immediate allergic reactions in our patient. Thus, identical insulin molecules can behave in markedly different ways depending on the route of injection. Also, it is possible that the formation of anti–human insulin IgG is caused only by insulin molecules that are in contact with subcutaneous tissue. We assume that some modification of insulin, such as aggregation, leads to the immunologic reactions.2,3 We believe that the example of our patient not only indicates an option for the treatment of severe insulin allergy but also may help elucidate the pathogenesis of the immunologic response to therapeutic insulin.
Masato Asai, M.D., Ph.D.
Nagoya University Graduate School of Medicine and Hospital
Nagoya 466-8550, Japan
masato-a@mbc.ocn.ne.jp
Masanori Yoshida, M.D.
Toyota Memorial Hospital
Toyota 471-8513, Japan
Yoshitaka Miura, M.D., Ph.D.
Nagoya University Graduate School of Medicine and Hospital
Nagoya 466-8550, Japan
References
Berson SA, Yalow RS, Bauman A, Rothschild MA, Newerly K. Insulin-I131 metabolism in human subjects: demonstration of insulin binding globulin in the circulation of insulin treated subjects. J Clin Invest 1956;35:170-190.
Brange J, Andersen L, Laursen ED, Meyn G, Rasmussen E. Toward understanding insulin fibrillation. J Pharm Sci 1997;86:517-525.
Maislos M, Mead PM, Gaynor DH, Robbins DC. The source of the circulating aggregate of insulin in type I diabetic patients is therapeutic insulin. J Clin Invest 1986;77:717-723.