AJCP / Case Report
Misleading Hemoglobin A1c Levels in a Patient With Paroxysmal Nocturnal Hemoglobinuria Daniel Xia, MD,1 Randall McShine, MD,2 and Anders H. Berg, MD, PhD1 From the 1Department of Pathology, Beth Israel Deaconess Medical Center, Boston, MA, and 2Department of Adult Medicine, South Cove Community Health Center, Boston, MA. Key Words: Diabetes; Hemoglobin A1c; Paroxysmal nocturnal hemoglobinuria; Anemia; RBC half-life; RBC survival Am J Clin Pathol August 2014;142:261-265
CME/SAM
DOI: 10.1309/AJCPTK1HPR2KULJU
ABSTRACT Objectives: We report a case of a patient with diabetes mellitus and unexpectedly low hemoglobin A1c results associated with paroxysmal nocturnal hemoglobinuria (PNH). We review the impact of shortened RBC half-life on the interpretation of hemoglobin A1c levels. Methods: Patient history and laboratory test results were obtained from electronic medical records and analyzed. Results: The patient’s hemoglobin A1c declined in parallel to worsening anemia after the diagnosis of PNH. However, elevated serum glucose (random), fructosamine, and glycated albumin suggest ongoing hyperglycemia. Together, these results argue that the decline in hemoglobin A1c was due to decreased RBC survival secondary to PNH. Conclusions: Hemoglobin A1c levels must be interpreted with caution in patients with hematologic diseases that change RBC survival. Serum fructosamine and glycated albumin measurements are alternative measures of timeaveraged blood glucose control and may be useful in this subset of patients.
© American Society for Clinical Pathology
Upon completion of this activity you will be able to: • describe the pathophysiology of paroxysmal nocturnal hemoglobinuria. • list the hematologic factors that can influence RBC turnover. • predict the impact of hematologic factors on hemoglobin A1c levels. • describe some of the limitations of hemoglobin A1c, glycated albumin, and serum fructosamine testing. The ASCP is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The ASCP designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 Credit ™ per article. Physicians should claim only the credit commensurate with the extent of their participation in the activity. This activity qualifies as an American Board of Pathology Maintenance of Certification Part II Self-Assessment Module. The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose. Questions appear on p 277. Exam is located at www.ascp.org/ajcpcme.
Diabetes mellitus is a group of chronic metabolic diseases defined by persistently high blood glucose. It is caused by defects in the synthesis of insulin by the pancreas and/ or end-organ insulin resistance. When poorly managed, the macrovascular (eg, myocardial infarction) and microvascular (eg, diabetic retinopathy, nephropathy, and neuropathy) complications of the disease result in significant morbidity and mortality for the patient,1 and the economic burden associated with diabetes mellitus contributes significantly to total health care costs.2 While random one-time blood glucose levels can be helpful in the diagnosis of diabetes, it is the average blood glucose levels that correlate best with the risk of complications.3 The most common test used to evaluate average glucose in the laboratory is hemoglobin A1c (Hgb A1c). Hgb A1c is formed by the nonenzymatic spontaneous posttranslational
261
Am J Clin Pathol 2014;142:261-265 DOI: 10.1309/AJCPTK1HPR2KULJU
261 261
Xia et al / Hemoglobin A1c in a Patient With Anemia
modification of the N-terminal amino acid of the b subunit of hemoglobin by blood glucose.4 Because hemoglobin is synthesized when an erythrocyte is first made and is not degraded until the RBC is removed from the circulation, the amount of Hgb A1c produced (measured as a percentage of total hemoglobin) depends on (1) blood glucose levels and (2) the amount of time an RBC has been in circulation (mean of 24 days, half-life of 29 days, and maximum life span of 120 days in healthy patients).5,6 It is important to note that all proteins are targets for modification by glucose; nonenzymatic glycation takes place not only on the N-terminus of the b subunit of hemoglobin but also on other solvent-accessible lysine side chains of hemoglobin and other proteins. In addition to clinical assays for glycated hemoglobin, several other clinical tests for serum glycated protein may be used to assess patients’ blood glucose control, including assays for serum fructosamine and glycated albumin.7 The significance of Hgb A1c as a laboratory test derives from its association with the risks of diabetic complications. In both the Diabetes Control and Complication Trial (DCCT) and the United Kingdom Prospective Diabetes Study, strict glycemic control and lower percent Hgb A1c were associated with lower rates of microvascular complications.8,9 These findings support the routine use of Hgb A1c in medical decision making for the outpatient management of diabetic patients. Not surprisingly, Hgb A1c can also be used to estimate average glucose. In 2002, Rohlfing et al10 retrospectively analyzed patient data from the DCCT trial and demonstrated a significant linear relationship between Hgb A1c values and time-averaged blood glucose concentrations (r2 = 0.67). In the A1c-Derived Average Glucose (ADAG) prospective study from 2008, Nathan et al11 carefully measured Hgb A1c and time-averaged glucose levels in 507 selected diabetic patients. This study also found a linear relationship between Hgb A1c and average glucose, as well as a stronger correlation (r2 = 0.84). The stronger correlation observed in the ADAG study may be explained by its study design, which excluded patients with anemia. Because of its dependence on RBC life span, testing for Hgb A1c in patients with anemias and RBC abnormalities is unreliable and challenging to interpret.12-17 As such, it does not necessarily follow that Hgb A1c should be used to estimate average glucose in all diabetic patients. In this report, we describe a diabetic patient with paroxysmal nocturnal hemoglobinuria (PNH) and declining Hgb Hgb A1c levels. We discuss the limitations of Hgb A1c testing, with emphasis on the complex and fascinating interplay between hematologic diseases (especially different forms of anemia) and percent Hgb A1c. A link is drawn between this patient’s PNH and his declining Hgb A1c levels, which falsely underestimate his recent average glucose. 262 262
Am J Clin Pathol 2014;142:261-265 DOI: 10.1309/AJCPTK1HPR2KULJU
Case Report The patient is a 67-year-old man with a medical history of type 2 diabetes mellitus (no significant microvascular and macrovascular complications), hypertension, hyperlipidemia, and hemolytic anemia secondary to PNH; he sought care at the office of his primary care physician (PCP) for routine follow-up. Laboratory results from the Beth Israel Deaconess Medical Center (BIDMC) since late 2010 show declining hematocrit/hemoglobin and increasing reticulocytosis in the setting of elevated lactate dehydrogenase and low haptoglobin levels—consistent with a compensated hemolytic anemia ❚Figure 1❚. Notably, the patient’s Hgb A1c levels as determined by an immunoassay method (Tina-quant Hemoglobin A1C Gen.2 on the Cobas Integra 800; Roche Diagnostics, Indianapolis, IN) have also declined steadily over this period, from more than 90 mmol/mol to a most recent level of 51 mmol/mol ❚Figure 2❚. The improving trend in Hgb A1c does not agree with corresponding fingerstick, random, and fasting blood glucose levels obtained by the PCP’s office and at the BIDMC ❚Table 1❚. In contrast to Hgb A1c, the latter set of laboratory results suggests suboptimal control of blood glucose over this same period. The patient’s anemia was originally noted on routine laboratory tests performed by his PCP in late 2010. He was asymptomatic at that time, received 1 month of oral iron supplementation with no improvement in hemoglobin, and was referred to the hematology service at an outside hospital in January 2011 for consultation. Flow cytometry demonstrated loss of CD55 and CD59 on the surface of RBCs and neutrophils—findings diagnostic of PNH. A bone marrow biopsy performed around this time showed hypercellular trilineage maturing hematopoiesis. Since PNH involves uncontrolled complement-mediated hemolysis of erythrocytes, the anticomplement immunotherapy eculizumab (Alexion Pharmaceuticals, Cheshire, CT) was offered. Following a discussion among the patient, his family, and the hematology team, the patient declined eculizumab. To further investigate the apparent discrepancy between the downward trend in Hgb A1c and the other measures of blood glucose control from late 2010 to the present, serum fructosamine and glycated albumin levels were measured on this patient’s most recent blood sample (November 20, 2012). Serum fructosamine and glycated albumin are measures of nonenzymatic glycation of serum proteins, and the half-lives of these proteins are not affected by RBC hemolysis. Serum fructosamine was measured using a nitroblue tetrazolium colorimetric assay method on a Roche Modular P800 automated analyzer (Roche Diagnostics); glycated albumin was measured by boronate affinity column chromatography (Quest Diagnostics, San Juan Capistrano, CA). The patient’s serum fructosamine, glycated albumin, and © American Society for Clinical Pathology
100 80 60 40 20
120 100 80 60 40 20 0
7/
0 3/26/08 3/26/09 3/26/10 3/26/11 3/26/12
0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00
29 9 / /1 0 2 1 1 9 /1 /2 0 9 1 / /1 0 29 3 / /1 1 29 5 / /1 1 29 7 / /1 1 29 9 / /1 1 2 1 1 9 /1 /2 1 9 1 / /1 1 29 3 / /1 2 29 5 / /1 2 29 7 / /1 2 29 9 / /1 2 29 /1 2
120
Hemoglobin A1C (mmol/mol)
Hemoglobin (g/dL)
140
Reticulocytes (Proportion)
AJCP / Case Report
❚Figure 1❚ Patient hemoglobin (solid line) and reticulocytes (dotted line) over time.
❚Figure 2❚ Patient hemoglobin A1c levels over time.
multiple measurements of fasting and random blood glucose on different occasions were all higher than their reference ranges/recommended clinical thresholds (Table 1 shows the most recent set of laboratory values). In contrast, the patients’ current Hgb A1c value was 51 mmol/mol, a value near the lower limit of the assay reference range (48-59 mmol/mol). The patient is currently being followed for PNH by his PCP and is receiving folate supplementation. He received 2 units of packed RBCs during an admission in November 2012 for a low hematocrit but has not experienced symptoms related to thrombophilia and anemia.
❚Table 1❚ Laboratory Valuesa
Discussion Measurements of Hgb A1c may be affected by either analytic inaccuracies or physiologic factors. Hgb A1c levels can be determined by a variety of techniques, including ion exchange chromatography, boronate affinity chromatography, immunoassay, electrophoresis, and mass spectrometry.3,18-20 Years ago, there were widespread problems with assay standardization and poor comparability between different commercial assay methods.4 Now, most available commercial assay methods produce closely comparable results because of cooperation by assay manufacturers with the National Glycohemoglobin Standardization Program. Even when Hgb A1c values are measured accurately, however, the values may be proportionately influenced in patients with abnormally aged circulating erythrocytes. Nonenzymatic glycation of hemoglobin takes place and accumulates over the course of the life span of an RBC. As such, lower mean RBC/hemoglobin age leads to lower Hgb A1c levels and falsely low estimated average glucose levels (in comparison to the actual average blood glucose levels).11,21,22 Conversely, higher mean RBC/hemoglobin age leads to higher Hgb A1c levels and falsely high estimated average glucose levels. The factors that decrease mean RBC © American Society for Clinical Pathology
Characteristic
Value (Reference Range)
Hemoglobin A1c, mmol/mol Estimated average glucose, mmol/L Random glucose, mmol/L Fructosamine, mmol/L Glycated albumin, % Hemoglobin, g/dL Hematocrit, proportion Mean corpuscular volume, fL WBC, 109/L Platelets, 109/L Reticulocytes, proportion LDH, U/L Haptoglobin, mmol/L Total bilirubin, mmol/L Direct bilirubin, mmol/L Serum urea nitrogen, mmol/L Creatinine, mmol/L Estimated GFR, mL/min/1.73 m2
51 (48-59) 5.6 (5.1-6.8) 11.0 (3.9-5.6) 0.34 (0.19-0.27) 1.5 (0.8-1.4) 54 (140-180) 0.17 (0.40-0.52) 116 (82-98) 3.8 (4.0-11.0) 204 (150-440) 0.16 (0.005-0.015) 3,850 (94-250)
Misleading Hemoglobin A1c Levels in a Patient With Paroxysmal Nocturnal Hemoglobinuria Daniel Xia, MD,1 Randall McShine, MD,2 and Anders H. Berg, MD, PhD1 From the 1Department of Pathology, Beth Israel Deaconess Medical Center, Boston, MA, and 2Department of Adult Medicine, South Cove Community Health Center, Boston, MA. Key Words: Diabetes; Hemoglobin A1c; Paroxysmal nocturnal hemoglobinuria; Anemia; RBC half-life; RBC survival Am J Clin Pathol August 2014;142:261-265
CME/SAM
DOI: 10.1309/AJCPTK1HPR2KULJU
ABSTRACT Objectives: We report a case of a patient with diabetes mellitus and unexpectedly low hemoglobin A1c results associated with paroxysmal nocturnal hemoglobinuria (PNH). We review the impact of shortened RBC half-life on the interpretation of hemoglobin A1c levels. Methods: Patient history and laboratory test results were obtained from electronic medical records and analyzed. Results: The patient’s hemoglobin A1c declined in parallel to worsening anemia after the diagnosis of PNH. However, elevated serum glucose (random), fructosamine, and glycated albumin suggest ongoing hyperglycemia. Together, these results argue that the decline in hemoglobin A1c was due to decreased RBC survival secondary to PNH. Conclusions: Hemoglobin A1c levels must be interpreted with caution in patients with hematologic diseases that change RBC survival. Serum fructosamine and glycated albumin measurements are alternative measures of timeaveraged blood glucose control and may be useful in this subset of patients.
© American Society for Clinical Pathology
Upon completion of this activity you will be able to: • describe the pathophysiology of paroxysmal nocturnal hemoglobinuria. • list the hematologic factors that can influence RBC turnover. • predict the impact of hematologic factors on hemoglobin A1c levels. • describe some of the limitations of hemoglobin A1c, glycated albumin, and serum fructosamine testing. The ASCP is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The ASCP designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 Credit ™ per article. Physicians should claim only the credit commensurate with the extent of their participation in the activity. This activity qualifies as an American Board of Pathology Maintenance of Certification Part II Self-Assessment Module. The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose. Questions appear on p 277. Exam is located at www.ascp.org/ajcpcme.
Diabetes mellitus is a group of chronic metabolic diseases defined by persistently high blood glucose. It is caused by defects in the synthesis of insulin by the pancreas and/ or end-organ insulin resistance. When poorly managed, the macrovascular (eg, myocardial infarction) and microvascular (eg, diabetic retinopathy, nephropathy, and neuropathy) complications of the disease result in significant morbidity and mortality for the patient,1 and the economic burden associated with diabetes mellitus contributes significantly to total health care costs.2 While random one-time blood glucose levels can be helpful in the diagnosis of diabetes, it is the average blood glucose levels that correlate best with the risk of complications.3 The most common test used to evaluate average glucose in the laboratory is hemoglobin A1c (Hgb A1c). Hgb A1c is formed by the nonenzymatic spontaneous posttranslational
261
Am J Clin Pathol 2014;142:261-265 DOI: 10.1309/AJCPTK1HPR2KULJU
261 261
Xia et al / Hemoglobin A1c in a Patient With Anemia
modification of the N-terminal amino acid of the b subunit of hemoglobin by blood glucose.4 Because hemoglobin is synthesized when an erythrocyte is first made and is not degraded until the RBC is removed from the circulation, the amount of Hgb A1c produced (measured as a percentage of total hemoglobin) depends on (1) blood glucose levels and (2) the amount of time an RBC has been in circulation (mean of 24 days, half-life of 29 days, and maximum life span of 120 days in healthy patients).5,6 It is important to note that all proteins are targets for modification by glucose; nonenzymatic glycation takes place not only on the N-terminus of the b subunit of hemoglobin but also on other solvent-accessible lysine side chains of hemoglobin and other proteins. In addition to clinical assays for glycated hemoglobin, several other clinical tests for serum glycated protein may be used to assess patients’ blood glucose control, including assays for serum fructosamine and glycated albumin.7 The significance of Hgb A1c as a laboratory test derives from its association with the risks of diabetic complications. In both the Diabetes Control and Complication Trial (DCCT) and the United Kingdom Prospective Diabetes Study, strict glycemic control and lower percent Hgb A1c were associated with lower rates of microvascular complications.8,9 These findings support the routine use of Hgb A1c in medical decision making for the outpatient management of diabetic patients. Not surprisingly, Hgb A1c can also be used to estimate average glucose. In 2002, Rohlfing et al10 retrospectively analyzed patient data from the DCCT trial and demonstrated a significant linear relationship between Hgb A1c values and time-averaged blood glucose concentrations (r2 = 0.67). In the A1c-Derived Average Glucose (ADAG) prospective study from 2008, Nathan et al11 carefully measured Hgb A1c and time-averaged glucose levels in 507 selected diabetic patients. This study also found a linear relationship between Hgb A1c and average glucose, as well as a stronger correlation (r2 = 0.84). The stronger correlation observed in the ADAG study may be explained by its study design, which excluded patients with anemia. Because of its dependence on RBC life span, testing for Hgb A1c in patients with anemias and RBC abnormalities is unreliable and challenging to interpret.12-17 As such, it does not necessarily follow that Hgb A1c should be used to estimate average glucose in all diabetic patients. In this report, we describe a diabetic patient with paroxysmal nocturnal hemoglobinuria (PNH) and declining Hgb Hgb A1c levels. We discuss the limitations of Hgb A1c testing, with emphasis on the complex and fascinating interplay between hematologic diseases (especially different forms of anemia) and percent Hgb A1c. A link is drawn between this patient’s PNH and his declining Hgb A1c levels, which falsely underestimate his recent average glucose. 262 262
Am J Clin Pathol 2014;142:261-265 DOI: 10.1309/AJCPTK1HPR2KULJU
Case Report The patient is a 67-year-old man with a medical history of type 2 diabetes mellitus (no significant microvascular and macrovascular complications), hypertension, hyperlipidemia, and hemolytic anemia secondary to PNH; he sought care at the office of his primary care physician (PCP) for routine follow-up. Laboratory results from the Beth Israel Deaconess Medical Center (BIDMC) since late 2010 show declining hematocrit/hemoglobin and increasing reticulocytosis in the setting of elevated lactate dehydrogenase and low haptoglobin levels—consistent with a compensated hemolytic anemia ❚Figure 1❚. Notably, the patient’s Hgb A1c levels as determined by an immunoassay method (Tina-quant Hemoglobin A1C Gen.2 on the Cobas Integra 800; Roche Diagnostics, Indianapolis, IN) have also declined steadily over this period, from more than 90 mmol/mol to a most recent level of 51 mmol/mol ❚Figure 2❚. The improving trend in Hgb A1c does not agree with corresponding fingerstick, random, and fasting blood glucose levels obtained by the PCP’s office and at the BIDMC ❚Table 1❚. In contrast to Hgb A1c, the latter set of laboratory results suggests suboptimal control of blood glucose over this same period. The patient’s anemia was originally noted on routine laboratory tests performed by his PCP in late 2010. He was asymptomatic at that time, received 1 month of oral iron supplementation with no improvement in hemoglobin, and was referred to the hematology service at an outside hospital in January 2011 for consultation. Flow cytometry demonstrated loss of CD55 and CD59 on the surface of RBCs and neutrophils—findings diagnostic of PNH. A bone marrow biopsy performed around this time showed hypercellular trilineage maturing hematopoiesis. Since PNH involves uncontrolled complement-mediated hemolysis of erythrocytes, the anticomplement immunotherapy eculizumab (Alexion Pharmaceuticals, Cheshire, CT) was offered. Following a discussion among the patient, his family, and the hematology team, the patient declined eculizumab. To further investigate the apparent discrepancy between the downward trend in Hgb A1c and the other measures of blood glucose control from late 2010 to the present, serum fructosamine and glycated albumin levels were measured on this patient’s most recent blood sample (November 20, 2012). Serum fructosamine and glycated albumin are measures of nonenzymatic glycation of serum proteins, and the half-lives of these proteins are not affected by RBC hemolysis. Serum fructosamine was measured using a nitroblue tetrazolium colorimetric assay method on a Roche Modular P800 automated analyzer (Roche Diagnostics); glycated albumin was measured by boronate affinity column chromatography (Quest Diagnostics, San Juan Capistrano, CA). The patient’s serum fructosamine, glycated albumin, and © American Society for Clinical Pathology
100 80 60 40 20
120 100 80 60 40 20 0
7/
0 3/26/08 3/26/09 3/26/10 3/26/11 3/26/12
0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00
29 9 / /1 0 2 1 1 9 /1 /2 0 9 1 / /1 0 29 3 / /1 1 29 5 / /1 1 29 7 / /1 1 29 9 / /1 1 2 1 1 9 /1 /2 1 9 1 / /1 1 29 3 / /1 2 29 5 / /1 2 29 7 / /1 2 29 9 / /1 2 29 /1 2
120
Hemoglobin A1C (mmol/mol)
Hemoglobin (g/dL)
140
Reticulocytes (Proportion)
AJCP / Case Report
❚Figure 1❚ Patient hemoglobin (solid line) and reticulocytes (dotted line) over time.
❚Figure 2❚ Patient hemoglobin A1c levels over time.
multiple measurements of fasting and random blood glucose on different occasions were all higher than their reference ranges/recommended clinical thresholds (Table 1 shows the most recent set of laboratory values). In contrast, the patients’ current Hgb A1c value was 51 mmol/mol, a value near the lower limit of the assay reference range (48-59 mmol/mol). The patient is currently being followed for PNH by his PCP and is receiving folate supplementation. He received 2 units of packed RBCs during an admission in November 2012 for a low hematocrit but has not experienced symptoms related to thrombophilia and anemia.
❚Table 1❚ Laboratory Valuesa
Discussion Measurements of Hgb A1c may be affected by either analytic inaccuracies or physiologic factors. Hgb A1c levels can be determined by a variety of techniques, including ion exchange chromatography, boronate affinity chromatography, immunoassay, electrophoresis, and mass spectrometry.3,18-20 Years ago, there were widespread problems with assay standardization and poor comparability between different commercial assay methods.4 Now, most available commercial assay methods produce closely comparable results because of cooperation by assay manufacturers with the National Glycohemoglobin Standardization Program. Even when Hgb A1c values are measured accurately, however, the values may be proportionately influenced in patients with abnormally aged circulating erythrocytes. Nonenzymatic glycation of hemoglobin takes place and accumulates over the course of the life span of an RBC. As such, lower mean RBC/hemoglobin age leads to lower Hgb A1c levels and falsely low estimated average glucose levels (in comparison to the actual average blood glucose levels).11,21,22 Conversely, higher mean RBC/hemoglobin age leads to higher Hgb A1c levels and falsely high estimated average glucose levels. The factors that decrease mean RBC © American Society for Clinical Pathology
Characteristic
Value (Reference Range)
Hemoglobin A1c, mmol/mol Estimated average glucose, mmol/L Random glucose, mmol/L Fructosamine, mmol/L Glycated albumin, % Hemoglobin, g/dL Hematocrit, proportion Mean corpuscular volume, fL WBC, 109/L Platelets, 109/L Reticulocytes, proportion LDH, U/L Haptoglobin, mmol/L Total bilirubin, mmol/L Direct bilirubin, mmol/L Serum urea nitrogen, mmol/L Creatinine, mmol/L Estimated GFR, mL/min/1.73 m2
51 (48-59) 5.6 (5.1-6.8) 11.0 (3.9-5.6) 0.34 (0.19-0.27) 1.5 (0.8-1.4) 54 (140-180) 0.17 (0.40-0.52) 116 (82-98) 3.8 (4.0-11.0) 204 (150-440) 0.16 (0.005-0.015) 3,850 (94-250)
