Recently, insulin analogues have come to be widely used in Japan. We therefore examined the prevalence of serum anti-insulin antibodies in type 2 diabetic patients treated with human insulin or other insulin analogues and compared the clinical features of those with and without antibodies. Fasting serum from type 2 diabetic patients treated with identical insulin preparations for the first time in more than 1 year (Human [N=22], Aspart [N=31], Rispro [N=23], and Glargine [N=23]) were used. Anti-insulin antibodies were measured by a radioimmunoassay using I¹²⁵-labeled human insulin. Rates of positive test results for antibodies were 0 % in Human, 58.1 % in Aspart, 39.1 % in Lispro, and 34.8 % in Glargine. In patients treated with analogues, the daily treating insulin dose was significantly higher (p=0.009) in those with antibodies than in those without antibodies, as demonstrated by an analysis of covariance adjusted for the body mass index, HbA1c, period of insulin treatment and presence of combined anti-diabetic drugs. The incremental dose amounted to 34.3 % of that in those without antibodies. Anti-human insulin antibodies were developed in patients treated with insulin analogues at a considerably high rate. In these patients, a higher dose of daily insulin treatment was needed than in those without antibodies.

After starting the immune checkpoint inhibitor pembrolizumab in October X-1 for postoperative recurrence of transverse colon cancer, a 77-year-old man with no history of diabetes mellitus visited our hospital in January X with complaints of general fatigue, decreased appetite, thirst, and heavy drinking. He was urgently admitted to the hospital with diabetic ketosis: random blood glucose 341 mg/dL, HbA1c 7.6 %, urinary ketone (2+), and pH 7.341. His fasting serum C-peptide level was 0.22 ng/mL, and his anti-GAD antibody was negative, so a diagnosis of fulminant type 1 diabetes mellitus was made. The patient was also found to have low cortisol levels with normal ACTH levels and was diagnosed with secondary hypoadrenocorticism after load tests. In addition, he had FT4 0.16 ng/dL, TSH 111.67 μIU/mL, and hypothyroidism. Intensive insulin therapy, hydrocortisone, and levothyroxine were administered sequentially. This is an extremely rare case in which pembrolizumab caused three endocrine disorders almost simultaneously. We consider this case to be informative regarding the management of endocrinologic immune-related adverse events, and we herein described the findings of this case with a review of the literature.

The patient was a 69-year-old man with type 2 diabetes who was being treated at a nearby hospital. The patient developed diabetic ketoacidosis with impaired consciousness and was brought transported to our hospital for emergency treatment. His blood C-peptide level was below detectable and tests for islet-related autoantibodies were negative. His blood glucose level was 1153 mg/dL. Initially, we suspected the onset of fulminant type 1 diabetes during treatment for type 2 diabetes, but the patient was diagnosed with ketosis-prone type 2 diabetes due to the recovery of his insulin secretory capacity after the improvement of DKA and the history of temporary insulin treatment for ketosis. It is difficult to distinguish KPD from soft drink ketosis and type 1 diabetes at the onset, and it is important to conduct detailed medical examinations and re-evaluate the insulin secretory capacity after the patient's condition stabilizes. In addition, cases have been reported in which the insulin secretory capacity declines each time that DKA recurs repeatedly or over time, and it is important to understand the clinical characteristics and to educate patients about the importance of weight management and outpatient treatment. Recently, KPD has been suggested as a possible subtype of type 1 diabetes. The elucidation of pathological conditions, such as disease enlightenment, the accumulation of further cases, and the results of genetic analyses are desired.