The Postprandial-to-Fasting Serum C-Peptide Ratio is a Predictor of Response to Basal Insulin-Supported Oral Antidiabetic Drug(s) Therapy: A Retrospective Analysis

The Postprandial-to-Fasting Serum C-Peptide Ratio is a Predictor of Response to Basal... Diabetes Ther (2018) 9:963–971 https://doi.org/10.1007/s13300-018-0404-6 ORIGINAL RESEARCH The Postprandial-to-Fasting Serum C-Peptide Ratio is a Predictor of Response to Basal Insulin-Supported Oral Antidiabetic Drug(s) Therapy: A Retrospective Analysis . . . . Pan-wei Mu De-zhao Liu Ying Lin Dong Liu . . . . Fan Zhang Yong-jun Zhang Shuo Lin Lin-qin Wang . . . Man-man Wang Jiong Shu Long-yi Zeng Yan-ming Chen Received: January 23, 2018 / Published online: March 21, 2018 The Author(s) 2018 drug(s) (OADs). However, some patients are still ABSTRACT unable to control their blood glucose levels even when on basal insulin-supported Introduction: Basal insulin is widely recom- OAD(s) therapy (BOT). The aim of this study mended for the treatment of type 2 diabetes was to investigate the factor(s) predicting mellitus (T2DM) patients who are unable to patient response to BOT. achieve glycemic control with oral antidiabetic Methods: A total of 212 patients with T2DM, ranging in age from 18 to 65 years, admitted to Pan-wei Mu and De-zhao Liu contributed equally to this the university hospital of Sun Yat-sen Univer- work. sity, Guangzhou, China, were enrolled in the Enhanced content To view enhanced content for this study between January 2013 and July 2016. All article go to https://doi.org/10.6084/m9.figshare. patients had fasting blood glucose levels of C 10.0 mmol/L despite receiving OAD(s) treat- ment. According to study design, these patients P. Mu  F. Zhang  S. Lin  M. Wang  J. Shu first received intensive insulin therapy for 2 L. Zeng  Y. Chen (&) weeks to attain and maintain their glycemic Department of Endocrinology, The Third Affiliated Hospital of Sun Yat-sen University, Guangzhou goals and then were switched to BOT. Respon- 510630, People’s Republic of China ders were defined as subjects who maintained e-mail: yanmingch@qq.com their glycemic targets with BOT for at least 3 D. Liu months; all others were considered to be non- Department of Anesthesia, The Third Affiliated responders. The characteristics between Hospital of Sun Yat-sen University, Guangzhou responders and non-responders were compared. 510630, People’s Republic of China Results: Compared with non-responders, Y. Lin responders had a shorter duration of diabetes Department of Gastroenterology, The Third (5.1 ± 5.0 vs. and 10.1 ± 3.2 years; P \ 0.001) Affiliated Hospital of Sun Yat-sen University, and a higher 2-h postprandial C-peptide-to- Guangzhou 510630, People’s Republic of China fasting C-peptide ratio (2 h-PCP/FCP: D. Liu  Y. Zhang 1.95 ± 0.51 vs. 1.67 ± 0.32; P \ 0.01). Department of Endocrinology, The Fifth Affiliated Responders showed a lower proportion of pre- Hospital of ZUNYI Medical University, Zhuhai vious treatment with insulin (69/100 vs 40/3; 519170, People’s Republic of China P \0.001) and sulfonlureas or glinides (116/50 L. Wang vs 40/0; P \0.001) than non-responders. Multi- Nansha Hospital of Traditional Chinese Medicine, variate logistic regression analysis showed that Guangzhou 511462, People’s Republic of China 964 Diabetes Ther (2018) 9:963–971 previous insulin treatment (odds ratio [OR] healthcare costs and save time. The aim of this 17.677, 95% confidence interval [CI] study is to determine which factor(s) may pre- 5.205–60.027; P \ 0.001) and the 2 h-PCP/FCP dict patient response to BOT. ratio (OR 0.241, 95% CI 0.058–0.679; P = 0.007) had predictive value. METHODS Conclusions: A higher 2 h-PCP/FCP ratio and a lack of previous insulin treatment increase the Subjects likelihood of BOT success. All of the participants in this study were being Keywords: Basal insulin; C-peptide; Predictor; treated for T2DM as inpatients in the Depart- Type 2 diabetes mellitus ment of Endocrinology, the 3rd Affiliated Hospital, Sun Yat-sen University (Guangzhou, Peoples’ Republic of China) at the time of INTRODUCTION enrollment (January 2013 to July 2016). The age of the participants at enrollment ranged from Basal insulin is the recommended treatment for 18 to 65 years. All had fasting blood glucose patients with type 2 diabetes mellitus (T2DM) levels of C 10.0 mmol/L, even though they were who are unable to achieve their glycemic targets receiving OAD(s) therapy without insulin. The with oral antidiabetic drugs (OADs) [1–3]. exclusion criteria were: (1) newly diagnosed or However, in clinical practice, some patients still treated-naıve T2DM, (2) type 1 or secondary fail to achieve glycemic control after being diabetes mellitus, (3) positive anti-glutamic acid started on basal insulin as an add-on therapy, decarboxylase antibody, (4) liver cirrhosis, (5) necessitating a switch to other treatment liver or renal dysfunction (i.e. serum alanine schemes, such as a basal-bolus insulin regimen aminotransferase or aspartate aminotransferase or injections with premixed insulin twice or levels were 2.5-fold higher than normal, and three times per day. Such evidence implies the serum creatinine level was [ 177 lmol/L), (6) basal insulin-supported OAD(s) therapy (BOT) systemic infection, (7) the use of corticosteroids fails to effectively control glycemia in some and (8) pregnancy. T2DM patients. It has been reported that vari- ables such as age, body mass index (BMI) and bedtime or post-breakfast plasma glucose levels Study Design may influence patient response to the treatment regimen [4]. For example, in insulin-naı¨ve All therapeutic regimens involving OAD(s) were patients, BOT may be more appropriate for discontinued at admission to the study. The subjects with a lower BMI and higher post- patients were then placed on an intensive breakfast plasma glucose levels [5]. In one study insulin therapeutic regimen (multiple daily population of T2DM patients who switched insulin injections or continuous subcutaneous from a premixed insulin regimen to BOT, the insulin infusion) to attain the pre-defined gly- key factors most closely related to the efficacy of cemic goal of a fasting capillary blood glucose BOT were pre-treatment glycated hemoglobin level of B 6.1 mmol/L and a capillary blood (HbA1c) levels, duration of diabetes and post- glucose level at 2 h after each of three meals of prandial C-peptide levels [6]. However, the B 8.0 mmol/L. This treatment regimen was above-mentioned studies focused on individu- maintained for 2 weeks after the glycemic goal als who plan to switch from treatment with (an) was reached, at which time the patients were OAD(s) alone to BOT. To date, it is unknown switched to BOT. Either glargine insulin or whether one or more of these factors may pre- detemir insulin was chosen as the basal insulin. dict the response to BOT. However, the ability All OAD(s) currently available could be pre- of the treating physician to identify those scribed. Adjustments in therapy were deter- patients who will not respond to BOT before a mined by the treating endocrinologists based new treatment is decided upon would reduce Diabetes Ther (2018) 9:963–971 965 on their experiences. The patients were required as the median and interquartile range. Differ- to self-monitor their capillary blood glucose ences between two groups were assessed by an levels at least four times per week, which had to independent t test or non-parametric test. include the fasting glucose level and the glucose A Chi-square test was performed to analyze the levels at 2 h after each of three meals. Patients differences in rates between the two groups. A were able to reach the specialists in charge at forward LR variable selection-based multivariate any time for consultations during the trial per- logistic regression was utilized to determine iod. Participants who were able to maintain the independent predictors. Pearson’s correlation glycemic target on the BOT regimen for at least and Spearman correlation were performed to 3 months were considered to be responders. analyze the relationships between variables. Those participants who were unable to achieve Statistical significance was set at P \ 0.05. or maintain the glycemic target on the BOT regimen were taken off this treatment by the Compliance with Ethics Guidelines treating physician and switched to a treatment of twice or multiple daily insulin injections All procedures followed in this study were in with or without OAD(s); these individuals were accordance with the ethical standards of the considered to be non-responders. responsible committee on human experimen- tation (institutional and national) and with the Data Collection and Measurements Helsinki Declaration of 1964, as revised in 2013. The protocol and informed consent document A medical history and physical examination were approved by the research ethics board of were conducted at admission. On the second the 3rd Affiliated Hospital of Sun Yat-sen day, fasting blood samples were collected for University. All patients gave written informed measurement of fasting plasma C-peptide (FCP), consents. The trial was not registered since it HbA1c and fasting plasma glucose (FPG) levels was performed in one hospital and was a retro- and for other routine biochemical tests. Break- spective analysis. fast consisted of mixed meals for all patients. At 2 h after breakfast, blood samples were taken to RESULTS measure the 2 h postprandial C-peptide (2 h- PCP) and postprandial plasma glucose (2 h-PPG) Of the 232 subjects screened, 221 patients levels. received the intensive insulin treatment (11 Routine clinical laboratory tests were per- patients withdrew before starting the intensive formed using the Olympus AU640 auto-bio- insulin treatment). Of these 221 patients, seven chemistry analyzer (Olympus Corp., Tokyo, patients declined to go on the BOT therapeutic Japan). A magnetic antibody immunoassay was regimen and two patients were lost to follow- used to measure C-peptide levels (Beijing Bio- up. The final study population for analysis Ekon Biotechnology Co., Ltd., Beijing, China). therefore consisted of 212 T2DM patients HbA1c level was assayed using the Bio-Rad TM (Fig. 1). The characteristics of these 212 study D-10 high-pressure liquid chromatography subjects are shown in Table 1. A total of 169 system (Bio-Rad Laboratories Inc., Hercules, CA, (79.7%) patients achieved and maintained the USA). target (i.e. responders) and 43 (20.3%) patients did not achieve or maintain the target (i.e. non- Statistical Analysis responders). Although non-responders were able to achieve and maintain the target of a Data were analyzed using the PASW statistics fasting capillary blood glucose level 18.0 program (IBM Corp., Armonk, NY, USA). of \ 6.1 mmol/L, they were unable to achieve Continuous variables with a normal distribu- or maintain the target postprandial capillary tion were presented as the mean ± standard blood glucose level of B 8.0 mmol/L. The dura- deviation; in all other cases, data were presented tion of diabetes, the percentage of patients with 966 Diabetes Ther (2018) 9:963–971 Fig. 1 Flow chart of patient selection and treatment. BOT Basal insulin-supported oral antidiabetic drug(s) therapy previous insulin, sulfonylurea or glinide treat- Clinical experience suggests that individuals ments and the 2 h-PCP/FCP ratio were signifi- with the same disease can differ in their cantly different between responders and non- responses to the same treatments [7]. Therefore, responders. These four variables were entered predictors of therapeutic response are of great into the multivariate logistic regression analy- help to the treating physician in deciding upon sis, and the results suggested that previous a treatment program. The UK Prospective Dia- insulin treatment and the 2 h-PCP/FCP ratio betes Study has shown that most patients with had predictive value in terms of the success of T2DM will require treatment with exogenous BOT (Table 2). Spearman correlation analysis insulin at some point during their lifetime determined the coefficient between the dura- [8, 9]. However, insulin is withheld from many tion of diabetes and the 2 h-PCP/FCP ratio to be patients due to a variety of reasons, including - 0.223 (P \ 0.001). concerns about injection pain [10] and lifestyle restrictions [11]. Studies has shown that patients significantly value reducing the num- DISCUSSION ber of insulin injections [12]. In addition, increased injection frequency has been associ- The results of our study indicate that the 2 h- ated with poor adherence to the protocol, with PCP/FCP ratio and previous insulin treatment one study showing that adherence was 78.3% in can predict the success of BOT. More specifi- patients requiring one injection per day and cally, a high 2 h-PCP/FCP ratio and the absence 60.8% in patients requiring four injections per of previous insulin treatment in a patient’s day (P \ 0.0001) [13]. BOT requires only one medical history suggest a higher success rate of single injection per day, which reduces patient BOT. In addition, the duration of diabetes had a reluctance to insulin therapy and increases close correlation with 2 h-PCP/FCP. Conse- compliance to the regimen. Thus, it is of great quently, given the convenience of this param- practical value to identify the predictor(s) of eter, the duration of diabetes may be a patient response to BOT. promising surrogate predictor. Diabetes Ther (2018) 9:963–971 967 Table 1 Characteristics of responders and non-responders to basal insulin-supported oral antidiabetic drug(s) therapy Variables Responders (n = 169) Non-responders (n = P value 43) Male/female 78/91 19/24 – Age (years) 50.0 ± 11.5 51.1 ± 8.3 – Body mass index (kg/m ) 25.78 ± 2.31 25.64 ± 2.61 – Duration (years) 5.1 ± 5.0 10.1 ± 3.2 0.001 Smoking (yes/no) 61/108 18/25 – Triglycerides (mmol/L) 2.51 (0.60) 2.34 (0.52) – Total cholesterol (mmol/L) 5.30 ± 0.93 5.13 ± 0.90 – High-density lipoprotein (mmol/L) 1.17 ± 0.38 1.11 ± 0.36 – Low-density lipoprotein (mmol/L) 3.15 ± 0.61 3.63 ± 0.76 – Urine acid (umol/L) 360.84 ± 90.67 356.07 ± 110.38 – Previous insulin treatment (yes/no) 69/100 40/3 0.001 Previous treatment with sulfonylureas or glinides (yes/no) 116/53 43/0 0.001 Previous treatment with biguanides (yes/no) 173/39 36/7 – Previous treatment with alpha-GI (yes/no) 148/64 28/15 – Previous treatment with thiazolidinediones (yes/no) 76/136 14/29 – Previous treatment with DPP4 inhibitor (yes/no) 35/177 7/36 – FPG (mmol/L) 13.51 ± 2.63 14.26 ± 2.98 – 2 h-PPG (mmol/L) 18.19 ± 3.13 18.65 ± 3.38 – HbA1c (%) 11.19 ± 2.06 11.79 ± 2.58 – (mmol/L) (99.9 ± 1.0) (105.4 ± 4.7) FCP (nmol/L) 0.45 ± 0.39 0.39 ± 0.18 – 2 h-PCP (nmol/L) 0.86 ± 0.71 0.66 ± 0.32 – Ratio of 2 h-PCP/FCP 1.95 ± 0.51 1.67 ± 0.32 0.01 Ratio of FCP/FPG 0.026 (0.018) 0.022 (0.013) – Ratio of 2 h-PCP/2 h-PPG 0.035 (0.030) 0.031 (0.024) – Intensive insulin therapy (MDI/CSII) 66/103 13/30 – f-FBG after intensive insulin therapy 5.16 ± 0.56 5.20 ± 0.73 – f-2 h-PBG after intensive insulin therapy 7.02 ± 0.88 6.89 ± 1.05 – Total dose of basal insulin in MDI in intensive insulin 31.6 ± 6.4 (n = 66) 28.9 ± 8.6 (n = 13) – therapy Total dose of bolus insulin in MDI in intensive insulin 44.7 ± 12.4(n = 66) 49.8 ± 15.6 (n = 13) – therapy 968 Diabetes Ther (2018) 9:963–971 Table 1 continued Variables Responders (n = 169) Non-responders (n = P value 43) Total dose of basal insulin in CSII in intensive insulin 28.2 ± 6.8 (n = 103) 26.1 ± 8.5 (n = 30) – therapy Total dose of bolus insulin in CSII in intensive insulin 38.3 ± 10.8 (n = 103) 40.2 ± 13.1 (n = 30) – therapy Data are presented as an absolute number or the mean ± standard deviation, unless indicated otherwise alpha-GI a-glucosidase Inhibitor, DPP4 Dipeptidyl Peptidase-4, FPG Fasting plasma glucose, 2 h-PPG 2-h postprandial glucose, HbA1c glycated hemoglobin, FCP fasting plasma C-peptide, 2 h-PCP 2-h postprandial C-peptide, MDI multiple daily insulin, CSII continuous subcutaneous insulin infusion, f-FBG finger fasting blood glucose, f-2 h-PBG finger 2-h postprandial blood glucose Data presented as the median with the interquartile range in parenthesis Table 2 Multivariate logistic regression analysis on influencing factors Variable B Standard error Wald test Odds ratio (95% confidence interval) P value Previous insulin treatment 0.287 0.624 21.203 17.677 (5.205–60.027) \0.001 Ratio of 2 h-PCP/FCP - 1.425 0.529 7.252 0.241 (0.058–0.679) 0.007 Intercept - 0.886 1.087 0.665 0.412 0.415 Both BOT and other therapeutic regimens, and fasting and 2 h-PCP level alone or adjusted which in this study included twice or multiple by corresponding glucose levels [6, 15–19]. The daily insulin injections with or without OAD(s), serum C-peptide level reflects endogenous provide basal insulin to the patient. Therefore, insulin secretion more directly than does the as expected, both the FCP levels and the FCP/ serum insulin level. Moreover, the serum FPG ratios were similar between two groups C-peptide level can be used to assess beta-cell (responders and non-responders) in our study. function even in patients undergoing insulin The main distinction between BOT and other therapy. Thus, for our study, we selected serum regimens is that BOT cannot imitate endoge- C-peptide rather than insulin as the relevant parameter. In a number of previous studies, the nous stimulated insulin secretion, while other regimens can. Stimulated insulin secretion ratio of C-peptide adjusted by glucose was contributes significantly to the postprandial superior to C-peptide alone [15–18]. We found, blood glucose level, as confirmed in our study in however, that neither 2 h-PCP alone nor the which all non-responders had to discontinue ratio of 2 h-PCP adjusted by glucose differed BOT because the target postprandial blood glu- between the two groups. This discrepancy cose level could not be achieved. This result between our study and previous ones [15–18]is suggests that the indictor(s) which presumably due to sampling error and target reflect(s) stimulated insulin secretion would be differences. The distributions of both the FCP/ predictor(s) of BOT response. FPG and 2 h-PCP/2 h-PPG ratios were skewed in Various indices have been reported to reflect our study, whereas they showed a normal dis- stimulated insulin secretion and to act as pre- tribution in the above-mentioned studies dictors of insulin therapy, such as first-phase [15–18]. One explanation for this difference in and second-phase insulin secretion [14], gluca- distribution may be the different sampling gon loading serum insulin and C-peptide level, times in these studies. In previous studies, blood Diabetes Ther (2018) 9:963–971 969 samples were collected after or during a period the patients had received insulin treatment of intensive insulin therapy, while in our study, prior to the study. In general, more previous they were collected prior to the patients starting insulin treatment in patients’ history corre- on intensive insulin therapy. A second expla- sponds with a longer duration of diabetes and nation may be differences in the target. The worse beta-cell function. previous studies targeted the requirement for The difference between the responders and insulin therapy, whereas our study targeted the non-responders in terms of previous sulfony- response to BOT. Notably, we found a signifi- lurea or glinide treatments was also significant. cant difference in the 2 h-PCP/FCP ratio However, this variable was removed in the between the responders and non-responders, multivariate logistic regression. Like previous and its predictive value was further supported insulin treatment, as discussed above, the by the results of the multivariate logistic results did not suggest that treatment with sul- regression. Why was the 2 h-PCP/FCP ratio fonylureas or glinides deteriorated beta-cell superior to 2 h-PCP alone or adjusted by glu- function; rather, previous intake of sulfonylurea cose? The answer may be partly attributed to or glinide drugs corresponded with a longer the C-peptide measurement. It is possible that duration of diabetes. factors are present in the C-peptide measure- Pre-treatment HbA1c level has been reported ment which may be very influential [20] and to be one of the key factors closely related to the that this interference is attenuated by adjusting efficacy of BOT [6]. However, in our study it was the 2 h-PCP level by the fasting C-peptide level. similar in both responders and non-responders. The duration of diabetes in responders was This discrepancy is attributed to the differences significantly shorter than that in non-respon- in patients enrolled in these two studies and to ders, but the multivariate logistic regression did differences in the protocols. In the earlier study, not suggest that this variable was a predictor of the patients were treated twice daily with pre- BOT success. The explanation for this difference mixed 30R insulin with or without OAD(s) prior is that beta-cell function was more direct than to initiation of the study and then randomized the duration of diabetes. Therefore, the dura- into either BOT or premixed 30R insulin twice tion of diabetes was removed in the multivari- daily plus OAD(s) treatment group. In contrast, ate logistic regression. It has been proven that in our study these pre-treatment options were beta-cell function progressively deteriorates not required, and all patients received intensive over time [8, 21, 22]. In line with those previous insulin therapy for 2 weeks, followed by a studies, we also found that the 2 h-PCP/FCP switch to BOT. ratio had a tight correlation with the duration In our study, several factors possibly limit of diabetes. Thus, considering the convenience the extent to which these results can be gener- of the duration of diabetes as a predictor of alized. First, the range of FPG level treatment success, this variable is worth further (C 10.0 mmol/L) and age (18–65 years), regard- study. less of pre-treatment options and duration of Compared with the non-responder group, diabetes, were quite wide. Hence, individual the proportion of patients who had received variation may be large, and some characteristics insulin treatment prior to this study was lower may be obscured. However, this scenario is in the responder group. The multivariate logis- closer to reality and is more valuable to clinical tic regression also suggested that a longer pre- practice. Second, modification of the therapeu- vious treatment with insulin reduced the tic regimen was at the discretion of the treating response to BOT. This finding seems contrary to physician, based on personal experiences, earlier reported results showing insulin therapy which could lead to bias. Third, there were far to be beneficial to the protection and preserva- fewer non-responders than responders, which tion of beta-cell function. In fact, these results may have obscured some characteristics or are not contradictory. These previous studies effects of the treatment. focused on the subsequent effects of the various treatment options while we focused on whether 970 Diabetes Ther (2018) 9:963–971 Zeng and Yan-ming Chen) declare that they CONCLUSIONS have nothing to disclose. The results of our study suggest that both the Compliance with Ethics Guidelines. All 2 h-PCP/FCP ratio and previous insulin treat- procedures followed were in accordance with ment are predictors of response to BOT. A the ethical standards of the responsible com- higher 2 h-PCP/FCP ratio and the lack of previ- mittee on human experimentation (institu- ous insulin treatment correspond with a higher tional and national) and with the Helsinki likelihood of BOT success. The duration of dia- Declaration of 1964, as revised in 2013. The betes is a promising predictor that requires fur- protocol and informed consent document were ther study. approved by the research ethics board of the 3rd Affiliated Hospital of Sun Yat-sen University. All patients gave written informed consent. The ACKNOWLEDGEMENTS trial was not registered since it was performed in one hospital and was a retrospective analysis. Funding. This study and article processing Data Availability. The datasets generated charges for this publication were funded by and/or analyzed during the current study are National Key R&D Program of China available from the corresponding author on (2017YFA0105803), the General Program of reasonable request. National Natural Science Foundation of China (81770826), the 5010 Clinical Research Projects Open Access. This article is distributed of Sun Yat-sen University (2015015), the Sci- under the terms of the Creative Commons ence and Technology Plan Projects of Guang- Attribution-NonCommercial 4.0 International dong Province (2016A050502010), the Key License (http://creativecommons.org/licenses/ Special Projects of Medical and Health Collab- by-nc/4.0/), which permits any non- orative Innovation of Guangzhou City commercial use, distribution, and reproduction (201604020016), and the Special Scientific in any medium, provided you give appropriate Research Project of Guangzhou City (2060404). credit to the original author(s) and the source, provide a link to the Creative Commons license, Editorial Assistance. We thank statistician and indicate if changes were made. Lian-xiong Yuan for his advice and help in the statistical analysis. We also thank the American Journal Experts Co. for language revision during the writing of this article. REFERENCES Authorship. All named authors meet the 1. Rodbard HW, Blonde L, Braithwaite SS, Brett EM, International Committee of Medical Journal Cobin RH, Handelsman Y, et al. American Associa- Editors (ICMJE) criteria for authorship for this tion of Clinical Endocrinologists medical guidelines for clinical practice for the management of diabetes manuscript, take responsibility for the integrity mellitus. Endocr Pract. 2007;13[Suppl 1]:1–68. of the work as a whole, and have given final approval for the version to be published. 2. Nathan DM, Buse JB, Davidson MB, Ferrannini E, Holman RR, Sherwin R, et al. Medical management of hyperglycemia in type 2 diabetes: a consensus Giving Thanks. We would like to thank all algorithm for the initiation and adjustment of of the doctors, nurses, technicians and patients therapy: a consensus statement of the American involved for their dedication to this study. Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care. Disclosures. All named authors (Pan-wei 2009;32:193–203. Mu, De-zhao Liu, Ying Lin, Dong Liu, Fan 3. Garber AJ, Abrahamson MJ, Barzilay JI, Blonde L, Zhang, Yong-jun Zhang, Shuo Lin, Lin-qin Bloomgarden ZT, Bush MA, et al. American Wang, Man-man Wang, Jiong Shu, Long-yi Diabetes Ther (2018) 9:963–971 971 Association of Clinical Endocrinologists’ compre- in patients with type 2 diabetes. QJM. hensive diabetes management algorithm 2013 2007;100:345–50. consensus statement–executive summary. Endocr Pract. 2013;19:536–57. 14. Lin JD. Levels of the first-phase insulin secretion deficiency as a predictor for type 2 diabetes onset by 4. Lavernia F. What options are available when con- using clinical-metabolic models. Ann Saudi Med. sidering starting insulin: premix or basal? Diabetes 2015;35:138–45. Technol Ther. 2011;13[Suppl 1]:S85–92. 15. Fujiwara D, Takahashi K, Suzuki T, Shii M, Naka- 5. Fonseca V, Davidson J, Home P, Snyder J, Jellinger shima Y, Takekawa S, et al. Postprandial serum P, Dyhr TA, et al. Starting insulin therapy with basal C-peptide value is the optimal index to identify insulin analog or premix insulin analog in T2DM: a patients with non-obese type 2 diabetes who require pooled analysis of treat-to-target trials. Curr Med multiple daily insulin injection: analysis of C-pep- Res Opin. 2010;26:1621–8. tide values before and after short-term intensive insulin therapy. J Diabetes Investig. 2013;4:618–25. 6. Bu S, Guo XH, Yang WY, Lu GZ, Yang ZJ, Ren TT, et al. Post-hoc analyses of type 2 diabetes patients 16. Saisho Y, Kou K, Tanaka K, Abe T, Shimada A, Kawai switch from premixed insulin regimen to basal T, et al. Postprandial serum C-peptide to plasma insulin plus oral hypoglycemic agents regimen. glucose ratio predicts future insulin therapy in Zhonghua Yi Xue Za Zhi. 2007;87:3115–8. Japanese patients with type 2 diabetes. Acta Dia- betol. 2013;50:987–8. 7. Brewster LM, Seedat YK. Why do hypertensive patients of African ancestry respond better to cal- 17. Saisho Y, Kou K, Tanaka K, Abe T, Kurosawa H, cium blockers and diuretics than to ACE inhibitors Shimada A, et al. Postprandial serum C-peptide to and beta-adrenergic blockers?. A systematic review. plasma glucose ratio as a predictor of subsequent BMC Med. 2013;11:141. insulin treatment in patients with type 2 diabetes. Endocr J. 2011;58:315–22. 8. [No authors listed] Intensive blood-glucose control with sulphonylureas or insulin compared with 18. Funakoshi S, Fujimoto S, Hamasaki A, Fujiwara H, conventional treatment and risk of complications Fujita Y, Ikeda K, et al. Utility of indices using in patients with type 2 diabetes (UKPDS 33). UK C-peptide levels for indication of insulin therapy to Prospective Diabetes Study (UKPDS) Group. Lancet achieve good glycemic control in Japanese patients 1998;352:837–53 with type 2 diabetes. J Diabetes Investig. 2011;2:297–303. 9. Wright A, Burden AC, Paisey RB, Cull CA, Holman RR. Sulfonylurea inadequacy: efficacy of addition of 19. Goto A, Takaichi M, Kishimoto M, Takahashi Y, insulin over 6 years in patients with type 2 diabetes Kajio H, Shimbo T, et al. Body mass index, fasting in the U.K. Prospective Diabetes Study (UKPDS 57). plasma glucose levels, and C-peptide levels as pre- Diabetes Care. 2002;25:330–6. dictors of the future insulin use in Japanese type 2 diabetic patients. Endocr J. 2010;57:237–44. 10. Cefalu WT. Evaluation of alternative strategies for optimizing glycemia: progress to date. Am J Med. 20. Jones AG, Hattersley AT. The clinical utility of 2002;113[Suppl 6A]:23S–35S. C-peptide measurement in the care of patients with diabetes. Diabet Med. 2013;30:803–17. 11. Polonsky WH, Fisher L, Guzman S, Villa-Caballero L, Edelman SV. Psychological insulin resistance in 21. Matthews DR, Cull CA, Stratton IM, Holman RR, patients with type 2 diabetes: the scope of the Turner RC. UKPDS 26: sulphonylurea failure in problem. Diabetes Care. 2005;28:2543–5. non-insulin-dependent diabetic patients over 6 years. UK Prospective Diabetes Study (UKPDS) 12. Hauber AB, Johnson FR, Sauriol L, Lescrauwaet B. Group. Diabet Med. 1998;15:297–303. Risking health to avoid injections: preferences of Canadians with type 2 diabetes. Diabetes Care. 22. [No authors listed] Effect of intensive blood-glucose 2005;28:2243–5. control with metformin on complications in over- weight patients with type 2 diabetes (UKPDS 34). 13. Donnelly LA, Morris AD, Evans JM. Adherence to UK Prospective Diabetes Study (UKPDS) Group. insulin and its association with glycaemic control Lancet 1998;352:854–65. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Diabetes Therapy Springer Journals

The Postprandial-to-Fasting Serum C-Peptide Ratio is a Predictor of Response to Basal Insulin-Supported Oral Antidiabetic Drug(s) Therapy: A Retrospective Analysis

Free
9 pages
Loading next page...
 
/lp/springer_journal/the-postprandial-to-fasting-serum-c-peptide-ratio-is-a-predictor-of-BJ2VB6oav5
Publisher
Springer Healthcare
Copyright
Copyright © 2018 by The Author(s)
Subject
Medicine & Public Health; Internal Medicine; Diabetes; Cardiology; Endocrinology
ISSN
1869-6953
eISSN
1869-6961
D.O.I.
10.1007/s13300-018-0404-6
Publisher site
See Article on Publisher Site

Abstract

Diabetes Ther (2018) 9:963–971 https://doi.org/10.1007/s13300-018-0404-6 ORIGINAL RESEARCH The Postprandial-to-Fasting Serum C-Peptide Ratio is a Predictor of Response to Basal Insulin-Supported Oral Antidiabetic Drug(s) Therapy: A Retrospective Analysis . . . . Pan-wei Mu De-zhao Liu Ying Lin Dong Liu . . . . Fan Zhang Yong-jun Zhang Shuo Lin Lin-qin Wang . . . Man-man Wang Jiong Shu Long-yi Zeng Yan-ming Chen Received: January 23, 2018 / Published online: March 21, 2018 The Author(s) 2018 drug(s) (OADs). However, some patients are still ABSTRACT unable to control their blood glucose levels even when on basal insulin-supported Introduction: Basal insulin is widely recom- OAD(s) therapy (BOT). The aim of this study mended for the treatment of type 2 diabetes was to investigate the factor(s) predicting mellitus (T2DM) patients who are unable to patient response to BOT. achieve glycemic control with oral antidiabetic Methods: A total of 212 patients with T2DM, ranging in age from 18 to 65 years, admitted to Pan-wei Mu and De-zhao Liu contributed equally to this the university hospital of Sun Yat-sen Univer- work. sity, Guangzhou, China, were enrolled in the Enhanced content To view enhanced content for this study between January 2013 and July 2016. All article go to https://doi.org/10.6084/m9.figshare. patients had fasting blood glucose levels of C 10.0 mmol/L despite receiving OAD(s) treat- ment. According to study design, these patients P. Mu  F. Zhang  S. Lin  M. Wang  J. Shu first received intensive insulin therapy for 2 L. Zeng  Y. Chen (&) weeks to attain and maintain their glycemic Department of Endocrinology, The Third Affiliated Hospital of Sun Yat-sen University, Guangzhou goals and then were switched to BOT. Respon- 510630, People’s Republic of China ders were defined as subjects who maintained e-mail: yanmingch@qq.com their glycemic targets with BOT for at least 3 D. Liu months; all others were considered to be non- Department of Anesthesia, The Third Affiliated responders. The characteristics between Hospital of Sun Yat-sen University, Guangzhou responders and non-responders were compared. 510630, People’s Republic of China Results: Compared with non-responders, Y. Lin responders had a shorter duration of diabetes Department of Gastroenterology, The Third (5.1 ± 5.0 vs. and 10.1 ± 3.2 years; P \ 0.001) Affiliated Hospital of Sun Yat-sen University, and a higher 2-h postprandial C-peptide-to- Guangzhou 510630, People’s Republic of China fasting C-peptide ratio (2 h-PCP/FCP: D. Liu  Y. Zhang 1.95 ± 0.51 vs. 1.67 ± 0.32; P \ 0.01). Department of Endocrinology, The Fifth Affiliated Responders showed a lower proportion of pre- Hospital of ZUNYI Medical University, Zhuhai vious treatment with insulin (69/100 vs 40/3; 519170, People’s Republic of China P \0.001) and sulfonlureas or glinides (116/50 L. Wang vs 40/0; P \0.001) than non-responders. Multi- Nansha Hospital of Traditional Chinese Medicine, variate logistic regression analysis showed that Guangzhou 511462, People’s Republic of China 964 Diabetes Ther (2018) 9:963–971 previous insulin treatment (odds ratio [OR] healthcare costs and save time. The aim of this 17.677, 95% confidence interval [CI] study is to determine which factor(s) may pre- 5.205–60.027; P \ 0.001) and the 2 h-PCP/FCP dict patient response to BOT. ratio (OR 0.241, 95% CI 0.058–0.679; P = 0.007) had predictive value. METHODS Conclusions: A higher 2 h-PCP/FCP ratio and a lack of previous insulin treatment increase the Subjects likelihood of BOT success. All of the participants in this study were being Keywords: Basal insulin; C-peptide; Predictor; treated for T2DM as inpatients in the Depart- Type 2 diabetes mellitus ment of Endocrinology, the 3rd Affiliated Hospital, Sun Yat-sen University (Guangzhou, Peoples’ Republic of China) at the time of INTRODUCTION enrollment (January 2013 to July 2016). The age of the participants at enrollment ranged from Basal insulin is the recommended treatment for 18 to 65 years. All had fasting blood glucose patients with type 2 diabetes mellitus (T2DM) levels of C 10.0 mmol/L, even though they were who are unable to achieve their glycemic targets receiving OAD(s) therapy without insulin. The with oral antidiabetic drugs (OADs) [1–3]. exclusion criteria were: (1) newly diagnosed or However, in clinical practice, some patients still treated-naıve T2DM, (2) type 1 or secondary fail to achieve glycemic control after being diabetes mellitus, (3) positive anti-glutamic acid started on basal insulin as an add-on therapy, decarboxylase antibody, (4) liver cirrhosis, (5) necessitating a switch to other treatment liver or renal dysfunction (i.e. serum alanine schemes, such as a basal-bolus insulin regimen aminotransferase or aspartate aminotransferase or injections with premixed insulin twice or levels were 2.5-fold higher than normal, and three times per day. Such evidence implies the serum creatinine level was [ 177 lmol/L), (6) basal insulin-supported OAD(s) therapy (BOT) systemic infection, (7) the use of corticosteroids fails to effectively control glycemia in some and (8) pregnancy. T2DM patients. It has been reported that vari- ables such as age, body mass index (BMI) and bedtime or post-breakfast plasma glucose levels Study Design may influence patient response to the treatment regimen [4]. For example, in insulin-naı¨ve All therapeutic regimens involving OAD(s) were patients, BOT may be more appropriate for discontinued at admission to the study. The subjects with a lower BMI and higher post- patients were then placed on an intensive breakfast plasma glucose levels [5]. In one study insulin therapeutic regimen (multiple daily population of T2DM patients who switched insulin injections or continuous subcutaneous from a premixed insulin regimen to BOT, the insulin infusion) to attain the pre-defined gly- key factors most closely related to the efficacy of cemic goal of a fasting capillary blood glucose BOT were pre-treatment glycated hemoglobin level of B 6.1 mmol/L and a capillary blood (HbA1c) levels, duration of diabetes and post- glucose level at 2 h after each of three meals of prandial C-peptide levels [6]. However, the B 8.0 mmol/L. This treatment regimen was above-mentioned studies focused on individu- maintained for 2 weeks after the glycemic goal als who plan to switch from treatment with (an) was reached, at which time the patients were OAD(s) alone to BOT. To date, it is unknown switched to BOT. Either glargine insulin or whether one or more of these factors may pre- detemir insulin was chosen as the basal insulin. dict the response to BOT. However, the ability All OAD(s) currently available could be pre- of the treating physician to identify those scribed. Adjustments in therapy were deter- patients who will not respond to BOT before a mined by the treating endocrinologists based new treatment is decided upon would reduce Diabetes Ther (2018) 9:963–971 965 on their experiences. The patients were required as the median and interquartile range. Differ- to self-monitor their capillary blood glucose ences between two groups were assessed by an levels at least four times per week, which had to independent t test or non-parametric test. include the fasting glucose level and the glucose A Chi-square test was performed to analyze the levels at 2 h after each of three meals. Patients differences in rates between the two groups. A were able to reach the specialists in charge at forward LR variable selection-based multivariate any time for consultations during the trial per- logistic regression was utilized to determine iod. Participants who were able to maintain the independent predictors. Pearson’s correlation glycemic target on the BOT regimen for at least and Spearman correlation were performed to 3 months were considered to be responders. analyze the relationships between variables. Those participants who were unable to achieve Statistical significance was set at P \ 0.05. or maintain the glycemic target on the BOT regimen were taken off this treatment by the Compliance with Ethics Guidelines treating physician and switched to a treatment of twice or multiple daily insulin injections All procedures followed in this study were in with or without OAD(s); these individuals were accordance with the ethical standards of the considered to be non-responders. responsible committee on human experimen- tation (institutional and national) and with the Data Collection and Measurements Helsinki Declaration of 1964, as revised in 2013. The protocol and informed consent document A medical history and physical examination were approved by the research ethics board of were conducted at admission. On the second the 3rd Affiliated Hospital of Sun Yat-sen day, fasting blood samples were collected for University. All patients gave written informed measurement of fasting plasma C-peptide (FCP), consents. The trial was not registered since it HbA1c and fasting plasma glucose (FPG) levels was performed in one hospital and was a retro- and for other routine biochemical tests. Break- spective analysis. fast consisted of mixed meals for all patients. At 2 h after breakfast, blood samples were taken to RESULTS measure the 2 h postprandial C-peptide (2 h- PCP) and postprandial plasma glucose (2 h-PPG) Of the 232 subjects screened, 221 patients levels. received the intensive insulin treatment (11 Routine clinical laboratory tests were per- patients withdrew before starting the intensive formed using the Olympus AU640 auto-bio- insulin treatment). Of these 221 patients, seven chemistry analyzer (Olympus Corp., Tokyo, patients declined to go on the BOT therapeutic Japan). A magnetic antibody immunoassay was regimen and two patients were lost to follow- used to measure C-peptide levels (Beijing Bio- up. The final study population for analysis Ekon Biotechnology Co., Ltd., Beijing, China). therefore consisted of 212 T2DM patients HbA1c level was assayed using the Bio-Rad TM (Fig. 1). The characteristics of these 212 study D-10 high-pressure liquid chromatography subjects are shown in Table 1. A total of 169 system (Bio-Rad Laboratories Inc., Hercules, CA, (79.7%) patients achieved and maintained the USA). target (i.e. responders) and 43 (20.3%) patients did not achieve or maintain the target (i.e. non- Statistical Analysis responders). Although non-responders were able to achieve and maintain the target of a Data were analyzed using the PASW statistics fasting capillary blood glucose level 18.0 program (IBM Corp., Armonk, NY, USA). of \ 6.1 mmol/L, they were unable to achieve Continuous variables with a normal distribu- or maintain the target postprandial capillary tion were presented as the mean ± standard blood glucose level of B 8.0 mmol/L. The dura- deviation; in all other cases, data were presented tion of diabetes, the percentage of patients with 966 Diabetes Ther (2018) 9:963–971 Fig. 1 Flow chart of patient selection and treatment. BOT Basal insulin-supported oral antidiabetic drug(s) therapy previous insulin, sulfonylurea or glinide treat- Clinical experience suggests that individuals ments and the 2 h-PCP/FCP ratio were signifi- with the same disease can differ in their cantly different between responders and non- responses to the same treatments [7]. Therefore, responders. These four variables were entered predictors of therapeutic response are of great into the multivariate logistic regression analy- help to the treating physician in deciding upon sis, and the results suggested that previous a treatment program. The UK Prospective Dia- insulin treatment and the 2 h-PCP/FCP ratio betes Study has shown that most patients with had predictive value in terms of the success of T2DM will require treatment with exogenous BOT (Table 2). Spearman correlation analysis insulin at some point during their lifetime determined the coefficient between the dura- [8, 9]. However, insulin is withheld from many tion of diabetes and the 2 h-PCP/FCP ratio to be patients due to a variety of reasons, including - 0.223 (P \ 0.001). concerns about injection pain [10] and lifestyle restrictions [11]. Studies has shown that patients significantly value reducing the num- DISCUSSION ber of insulin injections [12]. In addition, increased injection frequency has been associ- The results of our study indicate that the 2 h- ated with poor adherence to the protocol, with PCP/FCP ratio and previous insulin treatment one study showing that adherence was 78.3% in can predict the success of BOT. More specifi- patients requiring one injection per day and cally, a high 2 h-PCP/FCP ratio and the absence 60.8% in patients requiring four injections per of previous insulin treatment in a patient’s day (P \ 0.0001) [13]. BOT requires only one medical history suggest a higher success rate of single injection per day, which reduces patient BOT. In addition, the duration of diabetes had a reluctance to insulin therapy and increases close correlation with 2 h-PCP/FCP. Conse- compliance to the regimen. Thus, it is of great quently, given the convenience of this param- practical value to identify the predictor(s) of eter, the duration of diabetes may be a patient response to BOT. promising surrogate predictor. Diabetes Ther (2018) 9:963–971 967 Table 1 Characteristics of responders and non-responders to basal insulin-supported oral antidiabetic drug(s) therapy Variables Responders (n = 169) Non-responders (n = P value 43) Male/female 78/91 19/24 – Age (years) 50.0 ± 11.5 51.1 ± 8.3 – Body mass index (kg/m ) 25.78 ± 2.31 25.64 ± 2.61 – Duration (years) 5.1 ± 5.0 10.1 ± 3.2 0.001 Smoking (yes/no) 61/108 18/25 – Triglycerides (mmol/L) 2.51 (0.60) 2.34 (0.52) – Total cholesterol (mmol/L) 5.30 ± 0.93 5.13 ± 0.90 – High-density lipoprotein (mmol/L) 1.17 ± 0.38 1.11 ± 0.36 – Low-density lipoprotein (mmol/L) 3.15 ± 0.61 3.63 ± 0.76 – Urine acid (umol/L) 360.84 ± 90.67 356.07 ± 110.38 – Previous insulin treatment (yes/no) 69/100 40/3 0.001 Previous treatment with sulfonylureas or glinides (yes/no) 116/53 43/0 0.001 Previous treatment with biguanides (yes/no) 173/39 36/7 – Previous treatment with alpha-GI (yes/no) 148/64 28/15 – Previous treatment with thiazolidinediones (yes/no) 76/136 14/29 – Previous treatment with DPP4 inhibitor (yes/no) 35/177 7/36 – FPG (mmol/L) 13.51 ± 2.63 14.26 ± 2.98 – 2 h-PPG (mmol/L) 18.19 ± 3.13 18.65 ± 3.38 – HbA1c (%) 11.19 ± 2.06 11.79 ± 2.58 – (mmol/L) (99.9 ± 1.0) (105.4 ± 4.7) FCP (nmol/L) 0.45 ± 0.39 0.39 ± 0.18 – 2 h-PCP (nmol/L) 0.86 ± 0.71 0.66 ± 0.32 – Ratio of 2 h-PCP/FCP 1.95 ± 0.51 1.67 ± 0.32 0.01 Ratio of FCP/FPG 0.026 (0.018) 0.022 (0.013) – Ratio of 2 h-PCP/2 h-PPG 0.035 (0.030) 0.031 (0.024) – Intensive insulin therapy (MDI/CSII) 66/103 13/30 – f-FBG after intensive insulin therapy 5.16 ± 0.56 5.20 ± 0.73 – f-2 h-PBG after intensive insulin therapy 7.02 ± 0.88 6.89 ± 1.05 – Total dose of basal insulin in MDI in intensive insulin 31.6 ± 6.4 (n = 66) 28.9 ± 8.6 (n = 13) – therapy Total dose of bolus insulin in MDI in intensive insulin 44.7 ± 12.4(n = 66) 49.8 ± 15.6 (n = 13) – therapy 968 Diabetes Ther (2018) 9:963–971 Table 1 continued Variables Responders (n = 169) Non-responders (n = P value 43) Total dose of basal insulin in CSII in intensive insulin 28.2 ± 6.8 (n = 103) 26.1 ± 8.5 (n = 30) – therapy Total dose of bolus insulin in CSII in intensive insulin 38.3 ± 10.8 (n = 103) 40.2 ± 13.1 (n = 30) – therapy Data are presented as an absolute number or the mean ± standard deviation, unless indicated otherwise alpha-GI a-glucosidase Inhibitor, DPP4 Dipeptidyl Peptidase-4, FPG Fasting plasma glucose, 2 h-PPG 2-h postprandial glucose, HbA1c glycated hemoglobin, FCP fasting plasma C-peptide, 2 h-PCP 2-h postprandial C-peptide, MDI multiple daily insulin, CSII continuous subcutaneous insulin infusion, f-FBG finger fasting blood glucose, f-2 h-PBG finger 2-h postprandial blood glucose Data presented as the median with the interquartile range in parenthesis Table 2 Multivariate logistic regression analysis on influencing factors Variable B Standard error Wald test Odds ratio (95% confidence interval) P value Previous insulin treatment 0.287 0.624 21.203 17.677 (5.205–60.027) \0.001 Ratio of 2 h-PCP/FCP - 1.425 0.529 7.252 0.241 (0.058–0.679) 0.007 Intercept - 0.886 1.087 0.665 0.412 0.415 Both BOT and other therapeutic regimens, and fasting and 2 h-PCP level alone or adjusted which in this study included twice or multiple by corresponding glucose levels [6, 15–19]. The daily insulin injections with or without OAD(s), serum C-peptide level reflects endogenous provide basal insulin to the patient. Therefore, insulin secretion more directly than does the as expected, both the FCP levels and the FCP/ serum insulin level. Moreover, the serum FPG ratios were similar between two groups C-peptide level can be used to assess beta-cell (responders and non-responders) in our study. function even in patients undergoing insulin The main distinction between BOT and other therapy. Thus, for our study, we selected serum regimens is that BOT cannot imitate endoge- C-peptide rather than insulin as the relevant parameter. In a number of previous studies, the nous stimulated insulin secretion, while other regimens can. Stimulated insulin secretion ratio of C-peptide adjusted by glucose was contributes significantly to the postprandial superior to C-peptide alone [15–18]. We found, blood glucose level, as confirmed in our study in however, that neither 2 h-PCP alone nor the which all non-responders had to discontinue ratio of 2 h-PCP adjusted by glucose differed BOT because the target postprandial blood glu- between the two groups. This discrepancy cose level could not be achieved. This result between our study and previous ones [15–18]is suggests that the indictor(s) which presumably due to sampling error and target reflect(s) stimulated insulin secretion would be differences. The distributions of both the FCP/ predictor(s) of BOT response. FPG and 2 h-PCP/2 h-PPG ratios were skewed in Various indices have been reported to reflect our study, whereas they showed a normal dis- stimulated insulin secretion and to act as pre- tribution in the above-mentioned studies dictors of insulin therapy, such as first-phase [15–18]. One explanation for this difference in and second-phase insulin secretion [14], gluca- distribution may be the different sampling gon loading serum insulin and C-peptide level, times in these studies. In previous studies, blood Diabetes Ther (2018) 9:963–971 969 samples were collected after or during a period the patients had received insulin treatment of intensive insulin therapy, while in our study, prior to the study. In general, more previous they were collected prior to the patients starting insulin treatment in patients’ history corre- on intensive insulin therapy. A second expla- sponds with a longer duration of diabetes and nation may be differences in the target. The worse beta-cell function. previous studies targeted the requirement for The difference between the responders and insulin therapy, whereas our study targeted the non-responders in terms of previous sulfony- response to BOT. Notably, we found a signifi- lurea or glinide treatments was also significant. cant difference in the 2 h-PCP/FCP ratio However, this variable was removed in the between the responders and non-responders, multivariate logistic regression. Like previous and its predictive value was further supported insulin treatment, as discussed above, the by the results of the multivariate logistic results did not suggest that treatment with sul- regression. Why was the 2 h-PCP/FCP ratio fonylureas or glinides deteriorated beta-cell superior to 2 h-PCP alone or adjusted by glu- function; rather, previous intake of sulfonylurea cose? The answer may be partly attributed to or glinide drugs corresponded with a longer the C-peptide measurement. It is possible that duration of diabetes. factors are present in the C-peptide measure- Pre-treatment HbA1c level has been reported ment which may be very influential [20] and to be one of the key factors closely related to the that this interference is attenuated by adjusting efficacy of BOT [6]. However, in our study it was the 2 h-PCP level by the fasting C-peptide level. similar in both responders and non-responders. The duration of diabetes in responders was This discrepancy is attributed to the differences significantly shorter than that in non-respon- in patients enrolled in these two studies and to ders, but the multivariate logistic regression did differences in the protocols. In the earlier study, not suggest that this variable was a predictor of the patients were treated twice daily with pre- BOT success. The explanation for this difference mixed 30R insulin with or without OAD(s) prior is that beta-cell function was more direct than to initiation of the study and then randomized the duration of diabetes. Therefore, the dura- into either BOT or premixed 30R insulin twice tion of diabetes was removed in the multivari- daily plus OAD(s) treatment group. In contrast, ate logistic regression. It has been proven that in our study these pre-treatment options were beta-cell function progressively deteriorates not required, and all patients received intensive over time [8, 21, 22]. In line with those previous insulin therapy for 2 weeks, followed by a studies, we also found that the 2 h-PCP/FCP switch to BOT. ratio had a tight correlation with the duration In our study, several factors possibly limit of diabetes. Thus, considering the convenience the extent to which these results can be gener- of the duration of diabetes as a predictor of alized. First, the range of FPG level treatment success, this variable is worth further (C 10.0 mmol/L) and age (18–65 years), regard- study. less of pre-treatment options and duration of Compared with the non-responder group, diabetes, were quite wide. Hence, individual the proportion of patients who had received variation may be large, and some characteristics insulin treatment prior to this study was lower may be obscured. However, this scenario is in the responder group. The multivariate logis- closer to reality and is more valuable to clinical tic regression also suggested that a longer pre- practice. Second, modification of the therapeu- vious treatment with insulin reduced the tic regimen was at the discretion of the treating response to BOT. This finding seems contrary to physician, based on personal experiences, earlier reported results showing insulin therapy which could lead to bias. Third, there were far to be beneficial to the protection and preserva- fewer non-responders than responders, which tion of beta-cell function. In fact, these results may have obscured some characteristics or are not contradictory. These previous studies effects of the treatment. focused on the subsequent effects of the various treatment options while we focused on whether 970 Diabetes Ther (2018) 9:963–971 Zeng and Yan-ming Chen) declare that they CONCLUSIONS have nothing to disclose. The results of our study suggest that both the Compliance with Ethics Guidelines. All 2 h-PCP/FCP ratio and previous insulin treat- procedures followed were in accordance with ment are predictors of response to BOT. A the ethical standards of the responsible com- higher 2 h-PCP/FCP ratio and the lack of previ- mittee on human experimentation (institu- ous insulin treatment correspond with a higher tional and national) and with the Helsinki likelihood of BOT success. The duration of dia- Declaration of 1964, as revised in 2013. The betes is a promising predictor that requires fur- protocol and informed consent document were ther study. approved by the research ethics board of the 3rd Affiliated Hospital of Sun Yat-sen University. All patients gave written informed consent. The ACKNOWLEDGEMENTS trial was not registered since it was performed in one hospital and was a retrospective analysis. Funding. This study and article processing Data Availability. The datasets generated charges for this publication were funded by and/or analyzed during the current study are National Key R&D Program of China available from the corresponding author on (2017YFA0105803), the General Program of reasonable request. National Natural Science Foundation of China (81770826), the 5010 Clinical Research Projects Open Access. This article is distributed of Sun Yat-sen University (2015015), the Sci- under the terms of the Creative Commons ence and Technology Plan Projects of Guang- Attribution-NonCommercial 4.0 International dong Province (2016A050502010), the Key License (http://creativecommons.org/licenses/ Special Projects of Medical and Health Collab- by-nc/4.0/), which permits any non- orative Innovation of Guangzhou City commercial use, distribution, and reproduction (201604020016), and the Special Scientific in any medium, provided you give appropriate Research Project of Guangzhou City (2060404). credit to the original author(s) and the source, provide a link to the Creative Commons license, Editorial Assistance. We thank statistician and indicate if changes were made. Lian-xiong Yuan for his advice and help in the statistical analysis. We also thank the American Journal Experts Co. for language revision during the writing of this article. REFERENCES Authorship. All named authors meet the 1. Rodbard HW, Blonde L, Braithwaite SS, Brett EM, International Committee of Medical Journal Cobin RH, Handelsman Y, et al. American Associa- Editors (ICMJE) criteria for authorship for this tion of Clinical Endocrinologists medical guidelines for clinical practice for the management of diabetes manuscript, take responsibility for the integrity mellitus. Endocr Pract. 2007;13[Suppl 1]:1–68. of the work as a whole, and have given final approval for the version to be published. 2. Nathan DM, Buse JB, Davidson MB, Ferrannini E, Holman RR, Sherwin R, et al. Medical management of hyperglycemia in type 2 diabetes: a consensus Giving Thanks. We would like to thank all algorithm for the initiation and adjustment of of the doctors, nurses, technicians and patients therapy: a consensus statement of the American involved for their dedication to this study. Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care. Disclosures. All named authors (Pan-wei 2009;32:193–203. Mu, De-zhao Liu, Ying Lin, Dong Liu, Fan 3. Garber AJ, Abrahamson MJ, Barzilay JI, Blonde L, Zhang, Yong-jun Zhang, Shuo Lin, Lin-qin Bloomgarden ZT, Bush MA, et al. American Wang, Man-man Wang, Jiong Shu, Long-yi Diabetes Ther (2018) 9:963–971 971 Association of Clinical Endocrinologists’ compre- in patients with type 2 diabetes. QJM. hensive diabetes management algorithm 2013 2007;100:345–50. consensus statement–executive summary. Endocr Pract. 2013;19:536–57. 14. Lin JD. Levels of the first-phase insulin secretion deficiency as a predictor for type 2 diabetes onset by 4. Lavernia F. What options are available when con- using clinical-metabolic models. Ann Saudi Med. sidering starting insulin: premix or basal? Diabetes 2015;35:138–45. Technol Ther. 2011;13[Suppl 1]:S85–92. 15. Fujiwara D, Takahashi K, Suzuki T, Shii M, Naka- 5. Fonseca V, Davidson J, Home P, Snyder J, Jellinger shima Y, Takekawa S, et al. Postprandial serum P, Dyhr TA, et al. Starting insulin therapy with basal C-peptide value is the optimal index to identify insulin analog or premix insulin analog in T2DM: a patients with non-obese type 2 diabetes who require pooled analysis of treat-to-target trials. Curr Med multiple daily insulin injection: analysis of C-pep- Res Opin. 2010;26:1621–8. tide values before and after short-term intensive insulin therapy. J Diabetes Investig. 2013;4:618–25. 6. Bu S, Guo XH, Yang WY, Lu GZ, Yang ZJ, Ren TT, et al. Post-hoc analyses of type 2 diabetes patients 16. Saisho Y, Kou K, Tanaka K, Abe T, Shimada A, Kawai switch from premixed insulin regimen to basal T, et al. Postprandial serum C-peptide to plasma insulin plus oral hypoglycemic agents regimen. glucose ratio predicts future insulin therapy in Zhonghua Yi Xue Za Zhi. 2007;87:3115–8. Japanese patients with type 2 diabetes. Acta Dia- betol. 2013;50:987–8. 7. Brewster LM, Seedat YK. Why do hypertensive patients of African ancestry respond better to cal- 17. Saisho Y, Kou K, Tanaka K, Abe T, Kurosawa H, cium blockers and diuretics than to ACE inhibitors Shimada A, et al. Postprandial serum C-peptide to and beta-adrenergic blockers?. A systematic review. plasma glucose ratio as a predictor of subsequent BMC Med. 2013;11:141. insulin treatment in patients with type 2 diabetes. Endocr J. 2011;58:315–22. 8. [No authors listed] Intensive blood-glucose control with sulphonylureas or insulin compared with 18. Funakoshi S, Fujimoto S, Hamasaki A, Fujiwara H, conventional treatment and risk of complications Fujita Y, Ikeda K, et al. Utility of indices using in patients with type 2 diabetes (UKPDS 33). UK C-peptide levels for indication of insulin therapy to Prospective Diabetes Study (UKPDS) Group. Lancet achieve good glycemic control in Japanese patients 1998;352:837–53 with type 2 diabetes. J Diabetes Investig. 2011;2:297–303. 9. Wright A, Burden AC, Paisey RB, Cull CA, Holman RR. Sulfonylurea inadequacy: efficacy of addition of 19. Goto A, Takaichi M, Kishimoto M, Takahashi Y, insulin over 6 years in patients with type 2 diabetes Kajio H, Shimbo T, et al. Body mass index, fasting in the U.K. Prospective Diabetes Study (UKPDS 57). plasma glucose levels, and C-peptide levels as pre- Diabetes Care. 2002;25:330–6. dictors of the future insulin use in Japanese type 2 diabetic patients. Endocr J. 2010;57:237–44. 10. Cefalu WT. Evaluation of alternative strategies for optimizing glycemia: progress to date. Am J Med. 20. Jones AG, Hattersley AT. The clinical utility of 2002;113[Suppl 6A]:23S–35S. C-peptide measurement in the care of patients with diabetes. Diabet Med. 2013;30:803–17. 11. Polonsky WH, Fisher L, Guzman S, Villa-Caballero L, Edelman SV. Psychological insulin resistance in 21. Matthews DR, Cull CA, Stratton IM, Holman RR, patients with type 2 diabetes: the scope of the Turner RC. UKPDS 26: sulphonylurea failure in problem. Diabetes Care. 2005;28:2543–5. non-insulin-dependent diabetic patients over 6 years. UK Prospective Diabetes Study (UKPDS) 12. Hauber AB, Johnson FR, Sauriol L, Lescrauwaet B. Group. Diabet Med. 1998;15:297–303. Risking health to avoid injections: preferences of Canadians with type 2 diabetes. Diabetes Care. 22. [No authors listed] Effect of intensive blood-glucose 2005;28:2243–5. control with metformin on complications in over- weight patients with type 2 diabetes (UKPDS 34). 13. Donnelly LA, Morris AD, Evans JM. Adherence to UK Prospective Diabetes Study (UKPDS) Group. insulin and its association with glycaemic control Lancet 1998;352:854–65.

Journal

Diabetes TherapySpringer Journals

Published: Mar 21, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off