Endocrine Abstracts

Hemoglobin A1c (HbA1c) is used for the long-term management of patients with diabetes mellitus (DM). Hemoglobin variants other than HbA1c and e-N-lysine-glycated HbA0 may cause analytical interference in determinations of HbA1c. Hemoglobin J is an abnormal hemoglobin, an alpha globin gene variant and present in various geographic locations. Hemoglobin J (depending on its type) has different characteristics and functions. For example hemoglobin J Capetown (α2 92Gln β2), the most commonly seen Hb J variant (CGG->CAG), is associated in the heterozygous state with increased oxygen affinity and polycythemia. Other variants like Hb J Sardegna will show a completely unremarkable clinical picture in the heterozygote. Hemoglobin J Bancock (beta 56 Gly->Asp) and J Baltimore (beta 16 Gly->Asp) have been described in combination with sickle hemoglobin. Recently, Valencia Clinical Hospital discovered a new variant named Hemoglobin J Valencia; it was discovered after routine glycemic testing was carried out on a person with Diabetes, with the results of the test coming back abnormally low within the parameters. We describe the case of a 39-year-old caucasian male with history of HIV who presented to our institution for elevated HbA1c, with normal fasting glycemia. He was first diagnosed with HIV two years ago, when infectious disease specialist determined HbA1c for the first time (as a protocol), obtaining 12%. Our patient was also receiving treatment with Efavirenz, Tenofovir and Emtricitabine (HIV therapy). Fasting glycemias were always normal, so the doctor recommended metformin 850 mg twice daily. No hemoglobinopathy was known or suspected, as the blood count was normal; {red blood cells =5.89 million/mm³ [reference interval (RI) =4.5-6 million/mm³}; hemoglobin =17.2 g/dl (RI =13-18 g/dl); hematocrit =51.4% (RI =40%-55%). Previous values of HbA1c were: 12.0% (29/04/2015), 11.5% (27/01/2016) and 11.8% (28/06/2016); and fasting plasma glucose concentrations were 89 mg/dL, 96 mg/dL and 93 mg/dL, respectively. As these results did not correlate between them, we gave him a continuous glucose monitoring system; he worn it for one week. Time in target was 95%; glucose trend, 93 mg/dL, and no low glucose event was detected. Taking all these data into account, we performed a test for identification of hemoglobin variants using HPLC, which presented: HbA0=56.2%; HbA2=3.3%; and the presence of probable HbJ =39.2%. As a result, real HbA1c was lower than we could first determine.