Circulating 1,5-Anhydroglucitol Levels in Adult Patients with Diabetes Reﬂect Longitudinal Changes of Glycemia a U.S

Clinical Care/Education/Nutrition ORIGINAL ARTICLE

Circulating 1,5-Anhydroglucitol Levels in Adult Patients With Diabetes Reﬂect Longitudinal Changes of Glycemia A U.S. trial of the GlycoMark assay

1 2 JANET B. MCGILL, MD ERIKA B. AMMIRATI, BS glycemia is of paramount importance be- 1 2 THOMAS G. COLE, PHD THERESA GAUTILLE, RN cause good metabolic control can reduce 1 2 WILLIAM NOWATZKE, PHD MARK J. SARNO, BA 1 the incidence of micro- and macrovascu- SHANNON HOUGHTON, MD lar complications (2–5). In recent years, various biochemical markers have been cleared by the U.S. OBJECTIVE — 1,5-Anhydroglucitol (1,5AG) is a major circulating polyol arising primarily Food and Drug Administration (FDA) for from ingestion and excreted competitively with glucose. Japanese studies have demonstrated assessment of glycemia. These include reduced concentrations of 1,5AG in serum in hyperglycemic patients in comparison with eu- self-monitored blood glucose methods glycemic subjects and a gradual normalization of 1,5AG values for patients responding to and assays for HbA1c and fructosamine. antihyperglycemic therapies. In this first U.S. study, we assessed the ability of 1,5AG measure- These tests differ in the time period in ments to monitor glycemic control in a cohort of 77 patients with diabetes (22 with type 1 which measured values reflect glycemia. diabetes, 55 with type 2 diabetes) who presented with suboptimal glycemic control at baseline Random glucose measurements convey a (defined as HbA Ն7%). 1c “snapshot” of ambient circulating glucose RESEARCH DESIGN AND METHODS — Each patient received therapies consisting of but do not report the consistency of gly- combinations of diabetes education, nutritional counseling, and addition or dose adjustment of cemic control or excursion thereof. In various insulins or oral antihyperglycemic medications. Therapy was targeted to reduce mean contrast, HbA1c and fructosamine mea- Ն HbA1c by 1.0% over the monitoring period. 1,5AG, HbA1c, fructosamine, and random glucose surements reflect time-averaged glycemia measurements were performed at baseline and at 2, 4, and 8 weeks after the initiation of therapy. in the past 2–3 months and in the past 2–3 weeks, respectively. Of these two RESULTS — 1,5AG, fructosamine, and glucose values progressed significantly toward eug- measures, HbA has been identified in lycemia by week 2 of monitoring (Wilcoxon’s signed-rank test, P Ͻ 0.05), with median changes 1c Ϫ Ϫ multiple studies as the most valid predic- of 93, 7, and 13% for 1,5AG, fructosamine, and glucose, respectively. In contrast, HbA1c values did not respond significantly to therapy until week 4. On an individual patient basis, tor of risk of complications (2–6). How- 89.6% of patients displayed longitudinal changes of 1,5AG from baseline to week 8 in concor- ever, despite the demonstrated use of HbA measurements, the slow rate of dance with HbA1c. 1,5AG was also highly correlated with HbA1c and fructosamine (Spearman 1c ␳ϭϪ0.6459 and Ϫ0.6751, respectively; both P Ͻ 0.0001). change may contribute to delays in mod- ification of therapy. Thus, a marker that CONCLUSIONS — We conclude that 1,5AG responds sensitively and rapidly to changes in responds rapidly and significantly to glycemia and monitors glycemic control in accordance with established markers. changes in glycemia, that is metabolically stable, that demonstrates low biological Diabetes Care 27:1859–1865, 2004 variability, and that can be easily measured would be useful in management of patients with diabetes. he incidence of diabetes is clearly a the total population), and this was corre- 1,5-Anhydroglucitol (1,5AG) has serious health concern. During the lated with a 61% increase in obesity dur- been recently proposed as a marker con- T period 1991–2000, the incidence in ing the same period (1,2). For those forming to these criteria (7). 1,5AG was the U.S. increased by 49% (4.9–7.3% of afflicted with diabetes, control of their first discovered in the plant family Po- ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● lygala senega in 1888. The structure was From the 1Washington University School of Medicine, St. Louis, Missouri; and 2Vision Biotechnology identified in 1943, and the presence of the Consulting, Encinitas, California. compound in human blood (8) and cere- Address correspondence and reprint requests to Mark Sarno, Vision Biotechnology Consulting, 315 S. brospinal fluid (9) was established in Coast Hwy. 101, Suite U, PMB 144, Encinitas, CA 92024. E-mail: [email protected]. 1972 and 1973, respectively. Research Received for publication 9 February 2004 and accepted in revised form 5 May 2004. J.B.M. and T.G.C. have received grant support from Tomen America. E.B.A. and M.J.S. are paid consult- studies have shown that 1,5AG originates ants for Tomen America. mostly from foods with a mean intake of T.G.C. and W.N. are currently affiliated with Linco Research, St. Louis, Missouri. ϳ4.4 mg/day, that the closed pyran ring Abbreviations: 1,5AG, 1,5-anhydroglucitol; FDA, Food and Drug Administration; PROD, pyranose structure confers metabolic stability, that oxidase. A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion the rate of intake is matched by the daily factors for many substances. excretion rate, and that a bodily pool of © 2004 by the American Diabetes Association. ϳ500–1,000 mg of 1,5AG is constantly

DIABETES CARE, VOLUME 27, NUMBER 8, AUGUST 2004 1859 1,5-Anhydroglucitol in diabetes

maintained (10). The body pool may orig- RESEARCH DESIGN AND Biochemical measurements inate from an accumulation of small METHODS — The experimental pro- Serum 1,5AG was measured with the amounts of retained dietary 1,5AG or tocol for this study was approved by the GlycoMark assay (Tomen America, New from biosynthesis. There is some evi- Washington University Medical Center York, NY) as automated on a Hitachi 917 dence to support de novo biosynthesis of Human Studies Committee. Written in- analyzer (Roche Diagnostics, Indianapo- 1,5AG in the amount of ϳ0.5 mg/day formed consent was obtained from all pa- lis, IN). The method uses pyranose oxi- (10), but this has neither been extensively tients, and all participants were either dase (PROD) as an oxidase of 1,5AG to investigated nor confirmed. previously or newly diagnosed with either oxidize the second position hydroxyl 1,5AG is well absorbed in the intes- type 1 or type 2 diabetes as defined by the group of 1,5AG and to detect the gener- tine and distributes to all organs and tis- American Diabetes Association criteria ated hydrogen peroxide by colorimetry. sues (11). Renal reabsorption of 1,5AG is (16). Patients who were pregnant or lac- Because PROD also reacts with glucose, 99.9% and is competitively inhibited by tating or had a history of severe hypogly- the sample is pretreated by enzyme reac- excessive excretion of urinary glucose cemia or liver dysfunction, various tion using glucokinase, and glucose is (glucosuria). With this finding, Japanese hematological abnormalities (including changed into a nonreactive substance by research groups demonstrated reduced significant anemia), unstable or advanced phosphorylation. This makes the reaction concentrations of 1,5AG in serum of hy- renal disease or significant proteinuria, specific for 1,5AG. The reactions are per- ␮ perglycemic patients in comparison with unstable retinopathy, or recent retinal formed at 37°C. At time zero, 4 l of stan- ␮ euglycemic subjects (12,13). Addition- procedure were excluded. Hypoproteine- dard (50 g/ml 1,5AG), control, reagent ␮ ally, a gradual normalization of 1,5AG mia was not an exclusion criterion. Pa- blank, or sample is added to 120 lofthe values for patients responding to antidia- glucokinase-containing pretreatment re- tients were recruited from the outpatient ␮ betic therapies (14) has been demon- clinic population of the investigator, clin- agent. After a 5-min incubation, 50 lof strated, and studies have shown that ics associated with the principal investi- coloring reagent (containing PROD) is 1,5AG measurements reflect glycemic gator’s hospital or institution, or from a added, and the absorbance is measured at status over the previous 48 h to 2 weeks. volunteer database. Each patient pre- 546 and 700 nm. Absorbance at both Dietary variation does not appreciably af- sented at baseline with suboptimal glyce- wavelengths is again measured after a 10- fect the efficacy of such measurements be- mic control, defined as HbA Ն7.0%. min incubation. The absorbance at 700 cause the content of 1,5AG is similar in 1c nm is subtracted from that at 546 nm to The study cohort was comprised of 38 various starches (mean, 2.5 Ϯ 1.1 ␮g/g), correct for background. The kinetic dif- men and 39 women. The cohort ethnic meats and seafood (0.9 Ϯ 0.6 ␮g/g), veg- ference in corrected absorbance at 546 distribution included 58 Caucasian pa- etables (0.4 Ϯ 0.2 ␮g/g), fruits (0.7 Ϯ 0.6 nm is then calculated. Concentrations of tients, 18 African-American patients, and ␮g/g), and beverages (0.8 Ϯ 0.7 ␮g/g) 1,5AG in the reagent blank, control, or 1 patient of Hispanic origin. Mean subject (10). Only raw soybeans have been dem- sample are determined by comparison age was 50 Ϯ 11 years (men 53 Ϯ 10, onstrated to have significantly enriched Ϯ with a two-point calibration based on the levels of 1,5AG, although processed soy- women 47 12). Sex-specific mean body corrected kinetic absorbance of the re- weights were 100 Ϯ 21 kg (men) and beans (e.g., tofu, soy sauce) have 1,5AG Ϯ agent blank and standard. content essentially equivalent to all other 100 7 kg (women). Overall, 22 patients The assay for 1,5AG displayed an an- starches (10). Based on these early analyt- with type 1 diabetes (8 men and 14 alytical sensitivity of 0.2 ␮g/ml, was linear ical and clinical findings, an automated women) and 55 patients with type 2 dia- to 113 ␮g/ml, and demonstrated intra- assay using an enzymatic methodology betes (30 men and 25 women) were re- and interassay coefficients of variation of was developed and has been commer- cruited for the study. Ͻ4%. Hemoglobin, triglycerides, and bil- cially available in Japan since 1991 (15). A This was a longitudinal trial of 8 irubin did not interfere with the assay at domestic version of this assay (Glyco- weeks’ duration comparing changes of se- concentrations of up to 125 mg/dl, 1,153 Mark) has been under evaluation in clin- rum 1,5AG to whole-blood HbA1c, serum mg/dl, and 53 mg/dl, respectively. Assay ical trials in the U.S. and has recently been fructosamine, and random serum glucose results were also unaffected by glucose cleared for marketing by the FDA as a tool measurements. After screening, each pa- concentrations up to 1,000 mg/dl. The for intermediate-term monitoring of gly- tient received a combination of antihyper- sex-specific 5th to 95th percentile refer- cemia. glycemic treatments, including diabetes ence ranges established for the assay, It is the objective of this study to and nutritional education, various insulin based on analysis of samples from 224 present the results of the first U.S. clinical therapies, thiazolidenediones, sulfonyl- healthy individuals, were 10.7–32.0 study evaluating the ability of the Glyco- ureas, glitinides, ␣-glucosidase inhibi- ␮g/ml (men) and 6.8–29.3 ␮g/ml Mark assay to respond to and reflect tors, and metformin. Therapy was (women). The reference ranges and asso- changes in glycemia in a cohort of type 1 targeted toward reducing HbA1c by at ciated interindividual biological variation and type 2 diabetic patients with subop- least 1.0% on average over the monitoring parameters are essentially equivalent to timal glycemic control who are being period. Blood was drawn for biochemical those established for the assay in Japan managed aggressively with antihypergly- measurements at baseline and at 2, 4, and and reflect an insensitivity to cultural di- cemic treatments. Specific attention is 8 weeks. At these time points, individual etary differences. given to addressing the question of treatment regimens were adjusted as nec- HbA1c was determined from whole- whether the assay can reflect changes in essary to achieve progression toward eu- blood samples via turbidimetric inhibi- accordance with the established method glycemia. Serum 1,5AG values were not tion immunoassay (Tina-Quant; Roche of choice, HbA1c. used in patient management. Diagnostics) as automated on the Hitachi

1860 DIABETES CARE, VOLUME 27, NUMBER 8, AUGUST 2004 McGill and Associates

917. The Roche Frucostamine and Gluco- concomitant medications was monitored tween variables were assessed. 1,5AG was Quant tests were used to measure fruc- throughout the study. No patients in the most closely associated with HbA1c and tosamine and glucose, respectively, in cohort were found to have used any med- fructosamine, with Spearman ␳ values of serum samples. Both assays were also au- ication that would interact with antihy- Ϫ0.6459 and Ϫ0.6751, respectively Ͻ tomated on the Hitachi 917. Intra- and perglycemic treatments or interfere with (both P 0.0001). HbA1c and fruc- interassay coefficients of variation were the outcome measures. Three serious ad- tosamine were also highly associated (␳ϭ Ͻ1.0 and 3.0%, respectively, for the verse events occurred during the study. 0.6955). Correlations of glucose with ␳ϭϪ ␳ϭ HbA1c, fructosamine, and glucose assays. One female patient underwent a discec- 1,5AG ( 0.3358), HbA1c ( All biochemical assays were per- tomy and spinal fusion for a preexisting 0.3334), and fructosamine (␳ϭ0.3529) formed at the Core Laboratory for Clinical disc herniation, one male patient under- were lower due to the diurnal and situa- Studies at Washington University School went catheter ablation for atrial tachycar- tional variability of random glucose of Medicine. The facility is a secondary dia, and one female patient was measurements. reference laboratory for the National Gly- hospitalized for insulin dose adjustment Longitudinal changes in 1,5AG, cohemoglobin Standardization Program. due to chronic hyperglycemia. None of HbA1c, fructosamine, and glucose values these serious adverse events or any of the were assessed (Table 1). Whole-blood Data analysis minor adverse events impacted the accu- HbA1c was reduced by 1.3% from base- Demographics, baseline characteristics, racy of any of the biochemical measure- line to week 8, indicating that the thera- baseline medical histories, concomitant ments. Forty-three of the cohort of 77 peutic regimens used had achieved the medications, antihyperglycemic treat- patients reported at least one hypoglyce- study design target. However, mean and ments, adverse events, and hypoglycemic mic event during the study. Of these, only median absolute percent changes versus events were analyzed descriptively. The two patients reported hypoglycemic baseline are markedly higher for 1,5AG Wilk-Shapiro test was used to determine events requiring assistance. Again, none than for HbA1c, fructosamine, and glu- the distribution of values of all biochem- of the events reported impacted the bio- cose (Fig. 1). By week 2 of treatment, ical measurements. The Spearman’s non- chemical outcome variables. 1,5AG values had already increased by parametric analysis was used to 57.9%, reflecting the sensitive and rapid determine association between variables, Therapeutic regimens response of 1,5AG to changes in blood and Wilcoxon’s signed-rank test was used All patients received diabetes education glucose. For HbA1c, significant changes to determine the significance of changes and nutritional counseling at baseline and were not observed until 4 weeks. By week versus baseline values at individual mon- throughout the study. For those patients 8, 1,5AG mean values had increased by itoring time points (P Ͻ 0.05 was consid- with type 1 and type 2 disease already ϳ160% versus diminution of values of ered significant). To determine the degree receiving insulin therapy at baseline, the 12.1–35.3% for the established markers of concordance between longitudinal insulin regimen was modified by chang- of glycemia. changes in 1,5AG and HbA1c, individual ing the dosage of insulin, changing the patients were scored as to the direction of type of insulin, changing the schedule of Concordance of longitudinal changes change in measured values between base- insulin delivery, or initiating or changing Longitudinal changes in glycemic control line and week 8. Concordance was de- the combination of insulins. Some pa- were evaluated on an individual patient fined as either increases in 1,5AG with tients with type 2 diabetes who were not basis (Fig. 2). HbA1c and 1,5AG values at corresponding decreases of HbA1c or, previously receiving insulin initiated in- baseline and week 8 were assessed indi- conversely, decreases in 1,5AG values sulin therapy under the study protocol. vidually for indications of progression to- with corresponding increases in HbA1c. Oral antihyperglycemic regimens were ward, or away from, euglycemia. The Concordance of 1,5AG and fructosamine added or modified for patients with type 2 concordance of these indications was and fructosamine and HbA1c were simi- diabetes. Throughout the monitoring pe- then tabulated. In the cohort of 77 pa- larly determined. riod, these regimens were continually op- tients, the longitudinal changes of 1,5AG timized to achieve the average HbA1c were in concordance with changes of Ն Software target decrease of 1%. Of the oral anti- HbA1c in 69 (89.6%) of 77 patients. Sev- The JMP version 5.0 software (SAS Insti- hyperglycemic medications, metformin enty-five of the 77 patients were classified tute, Cary, NC) was used for all statistical and various glitinides and sulfonylureas, as responders to therapy based on dimi- investigations. in combination with insulin therapies, nution of HbA1c values with time. The were the most frequently used and were two patients classified as nonresponders RESULTS observed to best induce progression to- were both men with type 2 diabetes. In ward euglycemia. these patients, 1,5AG values decreased Baseline measures, concomitant with time and HbA1c values increased, medications, and adverse and Distribution of biochemical marker thus indicating 100% concordance in hypoglycemic events values, correlation, and longitudinal these patients. Baseline physical examinations were per- changes Concordance between longitudinal formed and medical histories taken. The The distribution of values for 1,5AG, changes of 1,5AG and HbA1c in the inter- cohort of 77 patients meeting the entry HbA1c, fructosamine, and random glu- mediate time points of the study was also criteria had no preexisting conditions that cose measurements was found to be non- assessed. Concordance ranged from 58.4 would compromise study procedures gaussian, thus supporting the need for to 76.6%, with earlier time points display- and/or biochemical measurements. Use of nonparametric analyses. Correlations being lower concordance. Concordance

DIABETES CARE, VOLUME 27, NUMBER 8, AUGUST 2004 1861 1,5-Anhydroglucitol in diabetes

Table 1—Longitudinal changes of biochemical variables

Fructosamine ␮ ␮ Time point/statistic 1,5AG ( g/ml) HbA1c (%) ( mol/l) Glucose (mg/dl) Baseline Mean Ϯ SD 1.9 Ϯ 1.9 9.5 Ϯ 1.7 410.6 Ϯ 108.6 225.0 Ϯ 105.6 Median 1.4 9.1 375.0 218.0 Minimum to maximum 0.0–10.7 7.0–14.2 269.0–908.0 64.0–574.0 Week 2* Mean Ϯ SD 3.0 Ϯ 2.2† 9.1 Ϯ 1.5 362.4 Ϯ 76.5† 187.4 Ϯ 91.0† Median 2.7 8.7 348.5 190.0 Minimum to maximum 0.0–12.6 6.8–14.0 232.0–572.0 27.0–448.0 Mean percent change vs. baseline 57.9 4.2 11.7 16.7 Median percent change vs. baseline 92.9 4.4 7.1 12.8 Week 4 Mean Ϯ SD 3.7 Ϯ 2.5† 8.8 Ϯ 1.4† 340.0 Ϯ 79.1† 181.4 Ϯ 102.4† Median 3.5 8.7 321.0 157.0 Minimum to maximum 0.0–13.3 6.6–13.2 206.0–559.0 29.0–701.0 Mean percent change vs. baseline 94.7 7.4 17.2 19.4 Median percent change vs. baseline 150.0 4.4 14.4 28.0 Week 8 Mean Ϯ SD 5.0 Ϯ 3.6† 8.2 Ϯ 1.2† 317.5 Ϯ 75.4† 172.6 Ϯ 100.5† Median 4.8 8.0 295.0 141.0 Minimum to maximum 0.0–15.7 6.3–12.3 197.0–575.0 16.0–545.0 Mean percent change vs. baseline 163.2 13.7 22.7 23.3 Median percent change vs. baseline 242.9 12.1 21.3 35.3 ␮ ␮ Ͻ ␮ Reference ranges: 1,5AG, 10.7–32.0 g/ml (men) and 6.8–29.3 g/ml (women); HbA1c, 6.0%; fructosamine, 205–285 mol/l; glucose, 55–115 mg/dl. *One male patient did not provide a blood sample at week 2, thus the n for this time point is 76. †P Ͻ 0.05 vs. baseline.

from baseline to week 8 between 1,5AG change appearing at 2 weeks of treatment. newly diagnosed with type 2 diabetes and fructosamine (67 of 77 patients, In contrast, HbA1c responded more slow- were monitored for 4 weeks after initia- 87.0%) and HbA1c and fructosamine (75 ly, and both HbA1c and fructosamine dis- tion of oral antihyperglycemic medica- of 77 patients, 97.4%) was also assessed. played more modest changes in value. tions. At the end of the 4-week period, Those patients demonstrating discor- These results are in agreement with the one-half of the patients continued on dance between 1,5AG and HbA1c dis- literature. For instance, in a 1996 study treatment while the rest discontinued played concordance between HbA1c and by Yamanouchi et al. (17), 56 patients treatment. The results showed that 1,5AG fructosamine and vice versa (i.e., the two patients found to be discordant by assessments of HbA1c and fructosamine were concordant when comparing changes of 1,5AG with changes of HbA1c).

CONCLUSIONS — The present study has examined the ability of the GlycoMark assay for 1,5AG to reflect changes of glycemia in a cohort of patients with type 1 and type 2 diabetes who are being treated aggressively with various regimens and who are being monitored with established biochemical markers of glycemia. Over the 8-week monitoring period, values for 1,5AG, HbA1c, and fructosamine were highly correlated, and the results were in good agreement with previous Japanese studies (17,18). Fur- thermore, 1,5AG responded rapidly and significantly to population-based changes Figure 1—Percentage change from baseline by time point for serum 1,5AG (●–●), whole-blood ■ ■ { { ‚ ‚ ϭ in glycemia, with the first significant HbA1c ( – ), serum fructosamine ( – ), or random serum glucose ( – ). Error bars SE.

1862 DIABETES CARE, VOLUME 27, NUMBER 8, AUGUST 2004 McGill and Associates

Figure 2—Longitudinal changes in 1,5AG (A), HbA1c (B), fructosamine (FA; C), and glucose (D) in individual patients.

increased rapidly and demonstrated a sig- dence in Japan that demonstrates a rapid the eight patients with discordant 1,5AG nificant change versus baseline at 2 weeks and significant response of 1,5AG to an- and HbA1c changes revealed that 1,5AG of monitoring. After discontinuation of tihyperglycemic treatments. values either slightly diminished or re- treatment, 1,5AG values sharply de- The present study also demonstrates mained unchanged over the monitoring creased, and the values at the 6-week time that changes in 1,5AG values reflect period, whereas HbA1c values diminished point were significantly different from changes in glycemic control in good ac- slightly, indicating some minimal im- those in the subgroup who continued on cord with HbA1c on an individual patient provement in glycemic control. The rela- therapy. Results for fructosamine and basis. Overall, 89.6% displayed concor- tive insensitivity of 1,5AG in these HbA1c demonstrated a lesser response to dant changes of 1,5AG and HbA1c from patients may be due to excessive deple- therapy in the initial 4 weeks. Further- baseline to week 8. Percent concordance tion of the body pool of 1,5AG. Such de- more, whereas fructosamine responded was reduced at intermediate time points pletion has been observed to occur in to the discontinuation of therapy between due to the relative insensitivity of HbA1c severe hyperglycemia as a result of persis- Ͻ weeks 4 and 6, HbA1c did not. Thus, the to changes in glycemia at time periods 8 tent glucosuria (19). Additionally, a dy- present study adds to the body of evi- weeks. Examination of the results from namic mass balance two-compartment

DIABETES CARE, VOLUME 27, NUMBER 8, AUGUST 2004 1863 1,5-Anhydroglucitol in diabetes model has been developed for 1,5AG treatment regimens. The results are in ex- vanovic L, McGill JB: Serum 1,5-anhy- (20), and it demonstrates that the tissue cellent accord with Japanese studies over droglucitol (GlycoMark): a short-term compartment is two to three times the the previous two decades and serve to in- glycemic marker. Diabetes Technol Ther 5: mass of the plasma compartment. Model dicate that the ethnically and culturally 355–363, 2003 estimates suggest that accelerated deple- diverse population of the U.S. does not 8. Pitkanen E: The serum protein pattern and the urinary polyol excretion in dia- tion of 1,5AG is associated with slow re- negatively impact the perceived use of the betic and in uremic patients. Clin Chim covery upon the initial improvement of marker. Furthermore, the recent clear- Acta 38:211–230, 1972 glycemia, because the body pool of 1,5AG ance granted by the FDA makes the Gly- 9. Pitkanen E: Occurrence of 1,5-anhydro- represents ϳ5 weeks of normal dietary coMark assay available for intermediate- glucitol in human cerebrospinal fluid. intake. If the model is accurate, it suggests term monitoring, which, it is hoped, will Clin Chim Acta 48:159–166, 1973 that the use of 1,5AG is optimal in the allow patients to seek medical interven- 10. Yamanouchi T, Tachibana Y, Akanuma H, range of modest hyperglycemia to near tion in a timely manner, such as when Minoda S, Shinohara T, Moromizato H, euglycemia. In this range, the GlycoMark initiating or altering therapy. This may Miyashita H, Akaoka I: Origin and dis- assay may represent a valuable means for empower patients to achieve and main- posal of 1,5-anhydroglucitol, a major the maintenance of near-normal glucose tain better control of their disease. Of polyol in the human body. Am J Physiol 236:E268–E273, 1992 levels, as it responds rapidly and sensi- course, further studies will be useful to 11. Kametani S, Hashimoto Y, Yamanouchi T, tively to even transiently ascending glu- better clarify the role of 1,5AG measure- Akanuma Y, Akanuma H: Reduced renal cose concentrations. ments in clinical diabetes management. reabsorption of 1,5-anhydroglucitol in di- Clearly the overall clinical manage- abetic rats and mice. J Biochem 102:1599– ment of diabetes, whether type 1 or type 2 1607, 1987 diabetes, requires continual multifactorial Acknowledgments— This study was funded 12. Akanuma Y, Ogawa K, Yamanouchi T, evaluation of glycemic status. The Glyco- by Tomen America and was previously pub- Mashiko S, Oka Y, Kosaka K, Akanuma H: Mark assay for 1,5AG may be a useful tool lished in abstract form. Decreased plasma 1,5-anhydroglucitol in as an adjunct to existing methods. For ex- We express appreciation to the volunteers diabetic patients (Abstract). Diabetes 30 ample, observations of good correlation who participated in the study and to the staff of (Suppl. 1):124A, 1981 13. Yoshioka S, Saitoh S, Negishi C, Fujisawa between circulating 1,5AG and the mag- the Washington University Clinical Research Center. T, Fujimori A, Takatani O, Imura M, Fun- nitude of glycemic excursions within a abashi M: Variations of 1-deoxyglucose day (21) indicate some potential for (1,5-anhydroglucitol) content in plasma 1,5AG to substitute for frequent glucose from patients with insulin-dependent di- measurements in type 2 diabetes. Fur- References abetes mellitus. Clin Chem 29:1396– thermore, circulating 1,5AG concentra- 1. Mokdad AH, Ford ES, Bowman BA, Nel- 1398, 1983 tions may also reflect transient glycemic son DE, Engelgau MM, Vinicor F, Marks 14. Yamanouchi T, Minoda S, Yabuuchi M, excursions due to postprandial hypergly- JS: Diabetes trends in the U.S.: 1990– Akanuma Y, Akanuma H, Miyashita H, cemia and, as such, may provide signifi- 1998. Diabetes Care 23:1278–1283, 2000 Akaoka I: Plasma 1,5-anhydroglucitol as cant benefit for improving long-term 2. Mokdad AH, Bowman BA, Ford ES, Vini- new clinical marker of glycemic control in cor F, Marks JS, Koplan JP: The continu- NIDDM patients. Diabetes 38:723–729, outcomes. Additionally, 1,5AG as an in- ing epidemics of obesity and diabetes in 1989 termediate-term marker could be evalu- the United States. JAMA 286:1195–1200, 15. Fukumura Y, Tajima S, Oshitani S, Ushi- ated between standard 3-month 2001 jima Y, Kobayashi I, Hara F, Yamamoto S, assessments of HbA1c to indicate worsen- 3. Diabetes Control and Complications Trial Yabuuchi M: Fully enzymatic method for ing of glycemic control and to incite be- Research Group: The effect of intensive determining 1,5-anhydro-D-glucitol in havioral change before induction of treatment of diabetes on the development serum. Clin Chem 40:2013–2016, 1994 significant vascular damage. There have and progression of the long-term compli- 16. Expert Committee on the Diagnosis and also been indications of the prognostic ca- cations of insulin-dependent diabetes Classification of Diabetes Mellitus: Report pability of 1,5AG measurements. In a mellitus. N Engl J Med 329:977–986, of the Expert Committee on the Diagno- study by Sone et al. (22), 1,5AG levels 1993 sis and Classification of Diabetes Melli- 4. U.K. Prospective Diabetes Study (UKPDS) tus. Diabetes Care 25 (Suppl. 1):5–20, were measured for 5 consecutive days in a Group: Intensive blood glucose control 2002 cohort of 22 patients with type 2 diabetes. with sulfonylureas or insulin compared 17. Yamanouchi T, Ogata N, Tagaya T, Ka- Changes in 1,5AG during the 5 days were with conventional treatment and risk of wasaki T, Sekino N, Funato H, Akaoka I, significantly correlated (R ϭ 0.70, P Ͻ complications in patients with type 2 di- Miyashita H: Clinical usefulness of serum 0.01) with changes in HbA1c over the sub- abetes (UKPDS 33). Lancet 352:837–853, 1,5-anhydroglucitol in monitoring gly- sequent 3 months. The authors con- 1998 cemic control. Lancet 347:1514–1518, cluded that 1,5AG could be used to 5. U.K. Prospective Diabetes Study (UKPDS) 1996 identify patients at high risk for poor gly- Group: Intensive blood glucose control 18. Yamanouchi T, Akanuma Y, Toyota T, cemic control in the future. with metformin on complications in over- Kuzuya T, Kawai T, Kawazu S, Yoshioka In conclusion, this first U.S. clinical weight patients with type 2 diabetes S, Kanazawa Y, Ohta M, Baba S, Kosaka K: (UKPDS 34). Lancet 352:854–865, 1998 Comparison of 1,5-anhydroglucitol, A1C, trial of the GlycoMark assay for the meta- 6. Peterson CM, Jovanovic L: Glycosylated and fructosamine for detection of diabetes bolically stable analyte, 1,5AG, has dem- proteins in normal and diabetic preg- mellitus. Diabetes 40:52–57, 1991 onstrated the capability of this simple, nancy. Acta Endocrinol 277 (Suppl.):107– 19. Yamanouchi T, Akanuma Y: Serum 1,5- automated method to report changes in 111, 1986 anhydroglucitol (1,5-AG): new clinical glycemia induced by antihyperglycemic 7. Buse JB, Freeman JLR, Edelman SV, Jo- marker for glycemic control. Diabetes Res

1864 DIABETES CARE, VOLUME 27, NUMBER 8, AUGUST 2004 McGill and Associates

Clin Pract 24 (Suppl.):S261–S268, 1994 K, Morishima T, Kawamori R, Kamata T: Y, Odawara M, Matsushima T, Kawai K, 20. Stickle D, Turk J: A kinetic mass balance 1,5-anhydro-D-glucitol evaluates daily Yamashita K: Predicting long-term glyce- model for 1,5-anhydroglucitol: applica- glycemic excursions in well-controlled mic control of post-educational type II di- tions to monitoring of glycemic control. NIDDM. Diabetes Care 18:1156–1159, abetic patients by evaluating serum 1,5- Am J Physiol 273:E821–E830, 1997 1995 anhydroglucitol levels. Diabetes Res Clin 21. Kishimoto M, Yamasaki Y, Kubota M, Arai 22. Sone H, Okuda Y, Yamaoka T, Kawakami Pract 34:83–88, 1996

DIABETES CARE, VOLUME 27, NUMBER 8, AUGUST 2004 1865