FASELY DECREASED HBA_1c IN A TYPE 2 DIABETIC PATIENT TREATED WITH DAPSONE

Abstract

Objective

To discuss a case of a falsely low hemogloin A_1c (HbA_1c) in a transplant patient treated with dapsone and its implications. HbA_1c is widely used as a measure of glycemic control in diabetic patients. With the increasing transplant population, it is important to be mindful of medications used in this population that can affect HbA_1c and to use other measures of glycemic control to guide treatment decisions.

Methods

We present details of the case and review the relevant literature.

Results

A 61-year-old patient received a liver transplant in 2012 and subsequently was noted to have a falling HbA_1c despite evidence of hyperglycemia based on finger-stick glucose and fructosamine measurements. Review of the medical records revealed that the discordance between HbA_1c and fingerstick glucose levels developed after initiation of dapsone therapy. Dapsone may lead to a falsely low HbA_1c via several mechanisms. Upon cessation of dapsone therapy, the patient’s HbA_1c returned to pre-dapsone levels.

Conclusion

It is important to be aware of medications commonly used in transplant patients that may lead to a falsely low HbA_1c level so that incorrect treatment decisions are not made. Fructosamine correlates with HbA_1c and can be used as a measure of glycemic control in transplant patients when HbA_1c cannot be used.

CASE REPORT

A 61-year-old Chinese woman underwent a liver transplant in November 2012 for cirrhosis secondary to hepatitis C. Her past medical history was notable for portopulmonary hypertension, leukocytoclastic vasculitis, hyperlipidemia, and type 2 diabetes mellitus (T2DM). She was diagnosed with T2DM several years prior to transplant and was managed with glyburide alone. Her diabetic control varied due to her liver disease, and her hemoglobin A_1c (HbA_1c) level was generally between 6.5 and 8.0%. In patients with cirrhosis, glycemic control is frequently unsatisfactory for several reasons, including impaired glucose homeostasis and impaired glucagon catabolism, which is accompanied by an increased risk of hypoglycemia due to decreased hepatic insulin extraction and decreased hepatic glycogen stores (1,2). In addition, treatment of diabetes in patients with cirrhosis is complex for several reasons, including malnutrition, impaired ability to metabolize certain oral antidiabetic drugs, hepatotoxicity of certain oral antidiabetic drugs, and the risk of hypoglycemia (1,2).

Following her transplant, the patient required insulin for several months to achieve glycemic control. Her fingerstick glucose levels were closely monitored to assist with insulin titration, as her immunosuppressant regimen (including prednisone) was often changed. Once her prednisone was tapered to a maintenance dose of 5 mg, she was able to transition off insulin to metformin and glyburide. On this regimen, the patient reported fasting glucose levels of 120 to 160 mg/dL, yet her HbA_1c decreased from 7.1 to 4.9%. Her metformin and glyburide doses were being reduced based on HbA_1c results approximately every 6 weeks. Glyburide was stopped based on an HbA_1c of 4.3%; however, the patient was reporting fasting glucose levels of 160 to 184 mg/dL and pre-dinner glucose levels of 270 to 285 mg/dL. It became increasingly apparent that there was a discordance between her fingerstick glucoses and HbA_1c levels. Repeat testing showed a fructosamine level of 470 μmol/L (range, 190 to 270 μmol/L), fasting glucose level of 140 mg/dL, and HbA_1c of 4.2% (Table 1).

Table 1.

Summary of Patient Data: Medications and Laboratory Values

	Months 0–40	Months 42–44	Month 45	Month 47	Month 48	Month 49	Month 50	Month 52a	Month 54	Month 56	Month 62
Key medications	Glyburide	Insulin	Insulin	Glyburide and dapsone started March 2013	Glyburide and metformin, dapsone	Glyburide and metformin, dapsone	Glyburide and metformin, dapsone	Metformin, dapsone (glyburide stopped for <1 week)	Glyburide and metformin, NO dapsone	Glyburide and metformin, septra SS, NO dapsone	Glyburide and metformin, insulin, septra SS, NO dapsone
Fructosamine (range, 190–270 μmol/L)								407		392
Glucose (range, 70–99 mg/dL)			120–150 (fasting)	80–140 (fasting)	100–160 (fasting)	120–150 (fasting) and 200 (pre-dinner)	120–150 (fasting) and 200 (pre- dinner)	166–184 (fasting) and 271–285 (pre- dinner)		110–140 (fasting) and 160–200 (pre-dinner)	110–154 (fasting)
Hemoglobin A1c (range, 4.3–5.6%)	7.0–8.3		7.1	4.9	4.8	5.2	4.6	4.2	6.8	7.8	7.0

^aDapsone was stopped after month 52 and has not been restarted since.

An endocrinology eConsult was ordered. Review of the medical records revealed that the patient was started on dapsone 6 months earlier. Prior to dapsone, the patient’s HbA_1c on insulin alone was 7.1%. One month after initiation of dapsone, her HbA_1c was 4.9% on metformin and glyburide. It was felt that the falsely low HbA_1c was due to hemolysis induced by dapsone therapy. Dapsone was stopped.

Literature Search

A systematic review of the literature was performed. The terms “dapsone and hemoglobin A_1c” were entered into PubMed to search for relevant articles. Twelve articles were returned in the search results. All 12 articles and their references were reviewed.

DISCUSSION

Measurement of HbA_1c was incorporated into clinical practice in the 1980s after studies demonstrated that it correlates with blood glucose measurements (3,4). Today, it is used as an indicator of glycemic control and as a diagnostic test for diabetes mellitus (4). HbA_1c is formed by nonenzymatic condensation of glucose with the N-terminal valine residue of the β chains of hemoglobin (4,5). Therefore, HbA_1c reflects the glucose level an erythrocyte has been exposed to during its lifespan and is a measure of glycemic control over the last 3 months, with the immediately preceding 30 days contributing 50% to the HbA_1c (4). These measurements help guide treatment decisions. However, HbA_1c values depend both on erythrocyte lifespan and assay methods (5). Therefore, conditions and medications that affect red blood cell turnover and erythrocyte lifespan lead to false measurements of HbA_1c (6). Dapsone can lead to a falsely low HbA_1c (4).

In our patient, there was a dramatic decrease in HbA_1c after dapsone was started (despite evidence of hyperglycemia based on fingerstick glucose levels) (Table1). Of note, the patient was not experiencing any hypoglycemia during this time, but the doses of her diabetes medications were being reduced based on the HbA_1c. Six months later, a large discrepancy was noted between HbA_1c and fructosamine (Table 1). The accuracy of the HbA_1c was questioned, and it was thought that the dapsone was causing a falsely low HbA_1c. After cessation of the medication, her HbA_1c returned to pre-dapsone values, and furthermore, her HbA_1c values now correlated with fructosamine (Table 1). In addition, based on the score calculated using the Naranjo algorithm (7), this was a probable adverse drug reaction.

There are several case reports of falsely low HbA_1c levels in patients on dapsone, including reports of patients infected with the human immunodeficiency virus (8,9), patients with autoimmune and skin conditions (5,10–14), transplant patients (15), and patients on dapsone therapy for other reasons, such as necrobiosis lipoidica and leprosy (12,16–18).

Dapsone leads to a falsely low HbA_1c via 3 mechanisms. One of the mechanisms by which dapsone can cause a misleadingly low HbA_1c is by inducing hemolysis. While we do not have measures of hemolysis in our patient, her hemoglobin dropped about 1 g/dL from baseline after initiation of dapsone, and 1 month after cessation of the drug, her hemoglobin increased by about 2 g/dL and has remained at this level, suggesting that the dapsone was inducing hemolysis. In addition, prior case reports demonstrated that patients on dapsone have elevated reticulocyte counts and elevated lactate dehydrogenase levels, which are indicative of hemolysis (5,9,11,12,14,17,18). Hemolysis does not always result in anemia (9), although most patients on dapsone do experience a 1 to 2 g/dL reduction in hemoglobin (5,8,12). Therefore, hemolysis can lead to a misleadingly low HbA_1c by reducing erythrocyte lifespan. Second, dapsone promotes the oxidation of hemoglobin to methemoglobin, which interferes with the high-performance liquid chromatography assay used to measure HbA_1c, since it will not spike in the correct HbA_1c location and hence, will give a falsely low value (4,8,12–14,16). In addition, diminished activity of nicotinamide adenine dinucleotide-methemoglobin reductase leads to decreased breakdown of methemoglobin and further affects the HbA_1c measurement (13). Finally, dapsone is thought to reduce erythrocyte survival independent of its hemolytic effect (4,13).

Given the multiple mechanisms by which dapsone leads to a falsely low HbA_1c, another measure of glycemic control is necessary in patients taking this medication. Fructosamine measures glycated serum proteins and therefore can serve as an alternative measure of glycemic control over a 2- to 3-week period (9,13,14). However, because fructosamine measures glycosylation of serum albumin, it is less reliable in patients with hypoalbuminemia (9,13).

Dapsone is being increasingly used for prophylaxis against Pneumocystis jiroveci pneumonia in the growing transplant population, especially when patients are unable to tolerate trimethoprim-sulfamethoxazole (19). Our patient was on dapsone for this reason. A falsely low HbA_1c in a transplant patient on dapsone has been demonstrated (15). Therefore, providers should be attentive to possible falsely low HbA_1c values and should be careful when using it for glycemic monitoring in the setting of dapsone use. In this case, use of the falsely low HbA_1c to guide therapy led to a medical error and inappropriate adjustments in the patient’s therapy. However, Tharavanij et al (20) demonstrated that fructosamine correlates with HbA_1c in islet transplant patients and therefore can be used to evaluate glycemic control in transplant patients. When our patient was on dapsone, fructosamine should have been used to monitor her glycemic control.

Case Follow-up

As noted above, the patient had a dramatic decrease in her HbA_1c after initiation of dapsone therapy despite evidence of hyperglycemia by fingerstick glucoses and fructosamine. Use of the HbA_1c to guide therapy led to incorrect treatment decisions. Her HbA_1c returned to predapsone levels after cessation of therapy and became consistent with fructosamine and fingerstick glucoses levels. Currently, she is doing well on oral agents and very-low-dose insulin glargine.

CONCLUSION

This case illustrates the fact that the HbA_1c measurement may be falsely low when certain medications—such as dapsone—are being used. With the growing transplant population, it is critical to be mindful of instances when the HbA_1c may be falsely low in this population so that incorrect treatment decisions are not made. In settings where HbA_1c cannot be used, fructosamine serves as a good alternative. Therefore, in transplant patients receiving dapsone, we recommend following fructosamine levels.

Abbreviation

HbA_1c

hemoglobin A_1c