 |
|
 |
 |
School
of Medicine, Chiba University Chiba University Hospital |
|
|
|
| Chiba Medical Journal | |
| Journal Index | |
| 〔Original Paper〕 | ||
| Measurement of insulin clearance and factors affecting insulin clearance in septic patients with glucose intolerance
－analysis under strict blood glucose control by means of bedside-type artificial pancreas－ |
||
|
|
||
| Masami Hoshino1), Yoshikura Haraguchi3), Hiroyuki Hirasawa4), Iwanori Mizushima2) Chie Tanaka2), Yasumasa Morita1), Takehito Yokoi1) and Motohiro Sakai2) | ||
| （Received January 27, 2006, Accepted January 30, 2006） | ||
|
|
||
SUMMARY
|
||
Key Words
|
||
| Ⅰ．Introduction Insulin production in acute phase is usually increased because the stress, caused by acute illness, stimulates insulin secretion from pancreas through autonomic nervous system and endocrine system （stress hormones）［1］. However, insulin sensitivity is often decreased and hyperglycemia occurs at the same time, owing to increased cytokines, stress hormones, some metabolic substances （ex. free fatty acid （FFA））, and toxins （ex. endotoxin）［2-8］. Hyperglycemia is well known undesirable pathophysiological state that causes decreased immune function［9］, delayed wound healing［9］, and hypercoagulable state［10］, all of which are the risk factors for the development of organ dysfunction［9,11,12］. Therefore, control of the hyperglycemia by insulin infusion is the essential therapy. We have been using bedside-type artificial pancreas （AP） （STG-22: manufactured by NIKKISOH Corporation, Tokyo Japan） on those patients since 1985［12］. Intensive insulin therapy on critically ill patients was recently reported to result in reduced morbidity and mortality［9,13］. On the other hand, beneficial effect of insulin itself is also known. Insulin has anti-inflammatory and anti-thrombotic action［14］. Therefore, analysis of insulin metabolism is considered to be important for establishing the appropriate treatment for acutely ill patients with glucose intolerance. Among the many aspects of insulin metabolism, insulin clearance （IC） seems to be clinically most important because of the following reasons: 1） IC directly influences blood insulin concentration and therapeutic effect of insulin, 2） Increased IC may be one the important factors causing glucose intolerance, 3） Intensive insulin therapy can be evaluated or may be justified in terms of IC, and 4） Clinicians can perform insulin therapy safely, effectively, and confidently with the knowledge of IC. IC and factors related to IC in chronic disease has been well investigated as follows. Factors related to decreased IC: aged［15,16,19,20］, fattish （ with increased body mass index （BMI）［15,17-19］） people, hypertension［19,20］, fatty liver［18］, liver cirrhosis［19］, Cushing’s disease［19,21］, uremia［22］, patients taking sulfonylurea［23］, increased blood glucose （BG）［18,19,24］, FFA［17-19］, triglyceride （TG）［17］, growth hormone （GH）［15］, and testosterone［25］levels. Factors related to increased IC: exercise［19,26］, pregnancy［27］, muscular people［18,19］, Graves’ disease［19,21］, cystic fibrosis［28］, and increased blood lactate level［29］. However, studies on IC in acutely ill patients or septic patients have been rarely reported［30-33］, partly because the measurement of IC was not easy to perform in acute phase, and partly because the importance of intensive insulin therapy and insulin metabolism was not recognized as nowadays. The purpose of this study is, 1） to clarify IC by establishing the method of the measurement of IC, 2） to detect factors affecting IC, and from the results of 1）, and 2）, 3） to consider and evaluate intensive insulin therapy and glucose intolerance in terms of IC, on acutely ill septic patients with glucose intolerance. Ⅱ．Materials and Methods BG levels were strictly controlled by means of AP in our intensive care unit （ICU） on 198 acutely ill patients with glucose intolerance from December 1985 to November 2002. Both hemodynamic monitoring with Swan-Ganz catheter （Swan-Ganz CCOmbo Edwards Lifesciences Corporation USA） and measurement of IC by the modified glucose clamp method （GC） by using AP （as described below） was applied for all those patients since February 1997 （41 patients）. We performed GC at intervals of seven days after the initial trial of GC basically. The measurement of whole steps in initial trial of GC was successful in 27 patients （success rate in initial trial of GC: 66% （27/41））. Among those 27 patients, we investigated 22 patients in whom the measurement of IC was performed in early phase or within 3 days after admission （mean±standard deviation （SD）: 2.2±1.0 days after admission）. Those 22 patients were all in septic state. Fourteen patients in whom the measurement was unsuccessful （panperitonitis 9 〔acute pancreatitis 4, ileus 2, postoperative 2, cholecystitis 1〕, pneumonia 4, fulminant hepatitis 1） were all in septic state with severe glucose intolerance and had multiple organ dysfunction syndrome （MODS）. Seven of them had diabetes mellitus （type Ⅰ 1, type Ⅱ 6）. Daily mean blood glucose level （BGm） and the amount of the daily administered insulin of the 14 patients were 208±32 mg/dL and 2.7±1.5 U/kg/day, respectively. Nine of them did not survive （mortality: 64% 〔9/14〕）. As the control to those 22 septic patients, we investigated also non-septic 7 patients in stable clinical condition in whom primary diseases improved and IC was considered to indicate normal value. Normality of IC on those 7 patients was supported from the low level of SOFA score （Sequential Organ Failure Assessment score）［34］, improved daily mean blood glucose level （BGm）, and normal and almost normal M value and I （blood insulin concentration during GC） as indicated in （Table 1） and （Table 3）. Measurement of IC on those 7 patients, in whom Swan-Ganz catheter was already removed, was performed on 17.3±5.3 hospital day after admission. In this study, patients with the following conditions were excluded: 1） acute cardiac diseases, 2） bleeding tendency, 3） frequent convulsion, 4） pregnancy, 5） under age of eighteen. This study was approved by the review board of Tokyo Police Hospital and informed consent of performing GC was obtained from the patient or from the family members. Demographic data of the patients are listed in （Table 1）. Sepsis was diagnosed based on the American College of Chest Physicians/Society of Critical Care Medicine （ACCP/SCCM） criteria for sepsis［35］. Organ dysfunction and multiple organ dysfunction syndrome （MODS） were diagnosed using SOFA score. Thirteen septic patients （13/22: 59%） had renal dysfunction （in whom renal score was one or more） and 10 patients （10/22: 45%） had liver dysfunction （in whom liver score was one or more）. Diabetic patients were not excluded because there is no definite relationship reported between the diabetes or insulin sensitivity and IC［15,19,36］. There was no patient with past medical histories or underlying diseases that might significantly affect hemodynamic parameters. Seven septic patients were under the administration of dopamine and dobutamine when GC was performed. Five patients were under continuous infusion of dopamine and two were receiving dobutamine. However, the amount of dopamine used was less than 3μg/kg/min for all five patients and that of dobutamine was 4μg/kg/min for one patient. Dopamine and dobutamine for these 6 patients was used for the purpose of increasing renal blood flow. The amount of dobutamine was 20μg/kg/min for one patient, who was severely ill septic patient with SOFA score of 19 and with cardiac index （CI） of 2.0L/min/m2. Nutritional support was achieved with total parenteral nutrition （TPN） in all patients. |
||
|
||
|
Glucose Clamp Method （GC） Modified part of our GC for the measurement of IC as compared to original GC by DeFronzo［37］is two points as follows. First, clamped blood glucose level is set to 80mg/dL instead of fasting blood glucose level, and second, two-step insulin infusion rate is performed. The details of our modified GC are indicated in （Table 2）. First step: Initial insulin infusion rate （IIR1） is 1.12 mU/kg/min （40mU/㎡/min）. After confirmation of both BG level being settled to 80mg/dL and infusion rate of glucose from AP being stabilized for one hour, M value （glucose metabolized: the amount of glucose infused from AP for maintaining BG level to 80mg/dL） is measured （M1）, and blood sample for insulin （I1） and C-peptide （C1） measurement is obtained. Second step: The insulin infusion rate （IIR） is increased to three times or 3.36 mU/kg/min （IIR3） and M value （M3）, insulin （I3）, and C-peptide （C3） are measured again. Regarding normal value of M value and blood insulin concentration （I） during original GC by DeFronzo, when IIR is set to 1.12 mU/kg/min and clamped BG level is set to fasting BG level, normal value of M value and I is reported to be 6-8 mg/kg/min［37,38］and around 100 mU/L［37］, respectively. Measurement of Insulin Clearance （IC） IC （Insulin Clearance） is calculated by dividing “the amount of insulin metabolized” by “blood insulin concentration”. IC is usually measured by GC （glucose clamp method） with clamped BG （blood glucose） at the “fasting BG level”. When steady state of BG level （clamped BG level） and blood insulin concentration （I） is achieved by GC, IC is calculated by the following formula［39］. IC＝ （the amount of insulin metabolized）/（blood insulin concentration） ＝ （IIR+the amount of endogenous insulin production）/（blood insulin concentration） ＝IIR/（blood concentration of exogenous insulin） ＝IIR/（I （clamp）－（blood concentration of endogenous insulin）） ＝IIR/（I （clamp）－I （basal）×C （clamp）/C （basal）） I （clamp） and C （clamp）: I and C （C-peptide） in the measurement of M value. I （basal） and C （basal）: I and C just before the beginning of GC. Normal range of IC is considered to be about 10－15 mL/kg/min［19,20,39］. However, “fasting BG level” usually can not be obtained in acutely ill patients with glucose intolerance, because insulin is already used and insulin therapy should not be stopped. Therefore, IC has to be calculated at certain clamped BG （our clamped BG is 80mg/dL） instead of “fasting BG level”. Besides, by using our modified GC with two step insulin infusion, IC is considered to be calculated by dividing “the difference of the amount of exogenous insulin infusion” by “the difference of the blood concentration of exogenous insulin”. Blood concentration of endogenous insulin is calculated by k （constant） × C （C-peptide）. Accordingly, IC is calculated by the following formula. IC （mL/kg/min）＝⊿ IIR/⊿ I ＝ （IIR3－IIR1）/⊿ I ＝1000×（3.36－1.12）/{（I3－kC3）－（I1－kC1）} ＝2240/{（I3－I1） + k （C1－C3）} ⊿ IIR （mU/kg/min）: difference of the amount of exogenous insulin infusion; （IIR3－IIR1） ⊿ I （mU/L）: difference of the blood concentration of exogenous insulin k: constant （7－14）, C: C-peptide （μg/L） kC: blood concentration of endogenous insulin （mU/L） Therefore, considering from the formula described above, If k ―（C1－C3）is small enough as compared to （I3－I1）, then IC can be approximated as the following formula. IC＝2240/（I3－I1） --- A Investigation of the factors affecting IC Septic patients were classified into three groups according to IC calculated by using formula A; N group （patients with normal IC （10≦IC≦15）, n＝8）, H group （patients with increased （high） IC （15＜IC）, n＝11）, and L group （patients with decreased （low） IC （IC＜10）, n＝3）. Mean values of the aforementioned parameters which have been reported to influence IC in chronic disease, and parameters which are related to organ dysfunction, were compared among the groups. Daily mean BG level （BGm） was calculated from the 24 BG data, which were measured by AP hourly before 24 hours of the beginning of GC. Blood sampling and the measurement of the parameters except M value were performed just before the beginning of GC. The data are shown as mean±SD. The unpaired Student’s t test and χ2 test were used for the comparison of mean values and the ratio of the patients with concerned parameter, respectively. Strength of the relationships between the data is indicated by correlation coefficient r, and the correlations between the data are shown by a regression line. P＜0.05 was considered significant. Ⅲ．Results Accuracy of the measurement of IC by GC with two step insulin infusion method There was no complication associated with this GC. Parameters related to glucose tolerance and IC are listed in （Table 3）. Mean percentage of k（C1－C3）to （I3－I1） in non-septic and septic patients were within the range from 0.2 to 0.4% and from 2.1 to 4.2%, respectively. Therefore, IC was considered to be approximated by the aforementioned formula A in both non-septic and septic patients. | ||
|
||
|
Measured value of IC and factors affecting IC Average values of the IC calculated from the above mentioned formula A in non-septic and septic patients were 11.9±1.4 mL/kg/min （minimum: 10.2, maximum: 14.6）, 15.6±6.3 mL/kg/min （minimum: 3.9, maximum: 30.7）, respectively. Distribution of the IC of the patients is shown in （Fig. 1）. Because only three patients had decreased IC （IC of the three patients were 3.9, 8.8, and 9.1 mL/kg/min, respectively）, analysis of L group was excluded. Average value of the IC in N group and H group was 11.5±1.4 and 20.9±4.2, respectively. There was not significant difference in FFA and TG between non-septic and septic patients （Average value of FFA and TG in non-septic patients was 0.22±0.22mEq/L and 1410±640mg/L, respectively）. Tendency of higher level of GH and cortisol in septic patients as compared to those in non-septic patients was found although the difference was not significant （average value of GH and cortisol in non-septic patients was 4.5±1.3μg/L and 164±57μg/L , respectively）. Comparison of the parameters between N group and H group are listed in （Table 4） and （Table 5）. BMI, systemic vascular resistance index （SVRI）, and total bilirubin in H group were significantly lower than those in N group, and CI in H group was significantly higher than that in N group. Tendency of negative correlation was found between SVRI and IC （r＝－0.41, P＜0.057）, and between blood total bilirubin and IC （r＝－0.38, P＜0.08）, respectively. Significant positive correlation was found between CI （cardiac index） and IC （r＝0.63, P＜0.002） （（Table 6）, （Fig. 2））. The amount of daily administered insulin in L group, N group and H group were 0.35±0.24, 0.90±0.55, and 1.20±0.24 U/kg/day, respectively. Tendency of increased amount of insulin administration was found in the group with increased IC. Number of the diabetics in N group and in H group was two （type Ⅱ 2, 2/8; 25%）, four （type Ⅰ 2, type Ⅱ 2, 4/11; 36%）, respectively. Two patients in N group and two in H group did not survive. | ||
|
||
|
||
|
Ⅳ．Discussion Measurement of IC by GC with two step insulin infusion method IC （Insulin Clearance） in non-septic and septic patients with glucose intolerance was considered to be reliably measured by our modified GC （Glucose Clamp method）. However, there are several points which should be taken into account for the measurement of IC and for the assurance of the reliability of our method. Therefore, those points are discussed below （（1）～（4））. （1） Regarding the reliability of our method, our method is considered to be reliable because IC of all non-septic, improved patients, in whom IC seemed to be normal as described before, was within the range from 10 to 15ml/kg/min, which was reported to be the normal value［19,20,39］. Besides, normality of the IC in non-septic patients in whom five out of seven patients were the diabetics may indicate that diabetic patients usually have normal IC in stable condition. （2） When performing our GC, clamped BG （blood glucose） level must be determined to a certain level as described before. High clamped BG level will cause endogenous insulin secretion on one hand, and low clamped BG level will trigger circulatory and hormonal counter-regulatory responses on the other. The reason why we determined the clamped BG level to 80mg/dL is that the level is neither hypoglycemic nor stimulative to endogenous insulin secretion［40］. （3） Regarding IIR （insulin infusion rate） of GC, we adopted IIR of 1.12mU/kg/min, because that rate is usually believed to increase I （blood insulin concentration） to around 100mU/L, which leads to a 60% reduction in basal insulin secretion and decreases hepatic glucose production to less than 10－15% （0.2－0.4mg/kg/min）［37］. Both conditions are essential for the reliable measurement of M value. However, mean value of I1 in our septic patients （68mU/L） was lower than the reported value （around 100mU/L）, which suggests increased IC due to acute illness with sepsis. Our success rate of the measurement of IC in initial trial of GC was 66% （27/41）, and the unsuccessful measurement was caused by the failure or retardation of decreasing and maintaining BG to clamped BG until the time limit of the examination （until closing time of the central laboratory: 17:00） because of the severe glucose intolerance. Therefore, considering relatively low I1 （68mU/L）, increasing IIR1 to more than 1.12mU/kg/min seemed to be one of the methods in order to increase the success rate. （4） IC was approximated by the formula A. However, relationship between IC and I （blood insulin concentration） must be considered in order to assure the agreement of the IC measured by usual GC and that measured by ours. IC measured by usual GC is IC under IIR of 1.12mU/kg/min, while IC measured by our GC is IC under IIR of 3.36mU/kg/min. Regarding the relationship between IC and I, there are some articles reporting that IC is constant over the physiological range of I （lower than 500mU/L） and decreases in the extra-high supraphysiological range［19,27,39］. By contrast, other studies indicate that IC decreases as I increases even within the physiological I （lower than 90mU/L）［19］. Therefore, considering I3 （244±127mU/L）, IC measured by our GC might be lower in some degree than IC measured by usual GC. Factors related to IC Mean value of IC in septic patients was 15.6mL/kg/min and half of the septic patients （11/22） had increased IC （H group）. Only three patients had decreased IC （IC＝9.1,8.8,3.9） among which one patient had definitely decreased IC （IC＝3.9）. This was severely ill septic patient with SOFA score of 19. These results indicate that IC in septic patients is usually preserved within normal limit or rather increased except in patients with extremely severe terminal state. Therefore, insulin therapy for BG control in septic patients except terminal state can be achieved safely, should be performed sufficiently, and should not be reduced. SVRI （systemic vascular resistance index） in H group was significantly lower than that in N group. Significant positive correlation was found between CI and IC. These results indicate close relationship between hyperdynamic state caused by sepsis and increased IC. Our results support the following four articles which reported: 1） half life of insulin in trauma patients was significantly shorter than that in control patients［30］, 2） IC in the burn patients was higher than that in the normal volunteers［31］, 3） IC in the multiple trauma patients was almost twice normal［32］, 4） insulin turnover in postoperative septic patients was significantly increased compared to control［33］. Besides, the result that tendency of increased amount of insulin administration was found in the group with increased IC indicate that increased IC is one of the important factors which cause increased demand of insulin in septic patients. Our results, showing increased IC in septic patients with hyperdynamic state, and patients with increased IC having the tendency of increased requirement of insulin infusion, indicate the possibility of increased demand of insulin, and support positive insulin supply, or intensive insulin therapy, in hyperdynamic septic patients. On the other hand, regarding the increased IC in septic patients with hyperdynamic state, the following items seem to be important: 1） mechanism of the increased IC, 2） relationship between increased IC and glucose intolerance, and 3） relationship between increased IC and outcome on septic patients. Proportion of the insulin degradation in certain organ or tissue to total insulin degradation in healthy subjects is reported to be from about 50% to 80% in the liver, from 24% to 30% in the kidney, and from 13% to 25% in the peripheral tissues［19,41］. Therefore, increased insulin degradation by the liver with increased hepatic blood flow due to increased CI may be the main cause of increased IC in septic patients. However, this hypothesis may not be persuasive enough to be stated true, because proportion of the insulin degradation in specific organ or tissue to total insulin degradation in septic patients may different from that in healthy subjects. Regarding relationship between increased IC and glucose intolerance, there was no significant difference between N group and H group in BGm, M1, and M3, as indicated in （Table 4）. The reason why H group had the same glucose intolerance as N group is unclear. H group might have some other factors which improved glucose tolerance. Regarding relationship between increased IC and outcome on septic patients, there was no significant difference between N group and H group in SOFA score and mortality. However, considering the lowest IC was found in patient with extremely severe terminal state and two out of three patients with low IC （L group） did not survive, further investigation concerning relationship between IC and outcome is considered to be important. Accordingly, as mentioned above, clinical usefulness and importance of measuring IC in septic patients with glucose intolerance is summarized as follows. First, intensive insulin therapy is supported by clarifying IC, as our results indicated that IC in septic patients was usually preserved within normal limit or rather increased especially in hyperdynamic state except in patients with extremely severe terminal state. Second, further investigation concerning relationships among IC, glucose tolerance, and outcome is considered to be necessary, partly because increased IC is theoretically one of the factors causing glucose intolerance which worsens outcome, and partly because mortality of the patients with low IC was high. There were not significant correlation between IC and other parameters but CI. These results suggest that those parameters were not the main factors which directly influence IC in septic patients. There was not significant difference in FFA and TG between non-septic and septic patients. These values are usually increased in septic patients owing to increased lipolysis. One of the reasons is considered to be the suppression of lipolysis in both groups, caused by the sufficient administration of glucose and insulin under strict blood glucose control by means of AP. Further investigation of IC on septic patients is considered to be necessary in terms of the following points of view, owing to the limitations of this study. First, the number of the patients to be investigated should be increased because of the limited number of the patients in this study. Second, measuring IC by our method in normal volunteers, and ideally, comparing the IC obtained by conventional method to that by our method in the same subject, are considered to be necessary in order to give the further assurance of the reliability of our method. Third, several blood sampling for the measurement of blood concentration of insulin and C-peptide during GC might be better for the strict analysis of IC. Forth, the investigation performed on the patients with and without diabetes mellitus separately, may be meaningful because this study included the diabetics. Finally, influence of IC to glucose intolerance and outcome should be further investigated and elucidated. | ||
|
|
||
Conclusions
|
||
要旨
|
||
References
|
||
Others
|
||
|
|
|
|
|
|
|
