If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Ischemic heart failure mortality is not predicted by cardiac insulin resistance but by diabetes per se and coronary flow reserve: A retrospective dynamic cardiac 18F-FDG PET study
Corresponding author at: Department of Nuclear Medicine & PET-Centre, Aarhus University Hospital, Palle Juul-Jensens Boulevard 165, 8200 Aarhus N, Denmark.
Department of Nuclear Medicine & PET-Centre, Aarhus University Hospital, Palle Juul-Jensens Boulevard 165, 8200 Aarhus N, DenmarkSteno Diabetes Center Aarhus, Aarhus University Hospital, Hedeager 3, 8200 Aarhus N, Denmark
Department of Nuclear Medicine & PET-Centre, Aarhus University Hospital, Palle Juul-Jensens Boulevard 165, 8200 Aarhus N, DenmarkSteno Diabetes Center Aarhus, Aarhus University Hospital, Hedeager 3, 8200 Aarhus N, Denmark
The connection between peripheral insulin resistance (IR) and coronary artery disease is well-established. Both are major risk factors for the development of ischemic cardiomyopathy potentially leading to heart failure (HF). Whether cardiac IR also impacts overall survival and morbidity is still debated. We therefore aimed to test if cardiac IR predicts mortality and major cardiovascular events (MACE) in patients with HF scheduled for cardiac viability testing before revascularization.
Methods
This retrospective study included 131 patients with a clinical diagnosis of ischemic HF (114 (87%) male, 33 (25%) with diabetes) referred to a viability Rubidium-82 (perfusion) and dynamic 18F-Fluorodeoxyglucose (metabolism) positron emission tomography combined with computed tomography prior to a potential revascularization procedure. Cardiac IR was assessed by myocardial glucose uptake (MGU) in a remote (non-scarred) area of the left ventricle during a hyperinsulinemic-euglycemic clamp (1mIE/kg/min).
Results
MGU correlated with skeletal muscle glucose uptake (p < 0.001) and whole-body glucose uptake (M-value) (p < 0.001), whereas no association was observed for individuals with diabetes. MGU did not predict the risk of death or MACE. However, both overt diabetes and reduced coronary flow reserve predicted overall survival.
Conclusion
Even though diabetes and related small-vessel disease is associated with increased mortality, cardiac IR per se does not predict cardiovascular morbidity and mortality.
]. Therefore, it is tempting to speculate that cardiac IR is directly responsible for the poor prognosis in patients with heart failure and diabetes, but the prognostic value of cardiac IR is unknown.
Skeletal muscle is responsible for the majority of the insulin-stimulated GU and therefore is the primary determinant of whole-body insulin sensitivity. Other organs such as the liver, adipose tissue, and the heart accounts for a smaller amount of glucose disposal [
]. Therefore, the consequences of cardiac IR on cardiac function remains to be elucidated.
With the use of dynamic positron emission tomography (PET)/computed tomography (CT), and the glucose radiotracer analogue 18F-Fluorodeoxyglucose (FDG), it is possible to measure organ-specific insulin sensitivity [
]. In this study, we aimed to determine the prognostic impact of cardiac IR assessed by myorcardial glucose uptake (MGU) on overall mortality and morbidity. We hypothesized, that MGU was a predictor for overall death and major cardiovascular events (MACE) in patients with heart failure.
2. Research design and methods
2.1 Study population and follow-up
This retrospective study included 131 patients with a clinical diagnosis of ischemic HF referred to The Department of Nuclear Medicine & PET Centre, Aarhus University Hospital, Denmark, between January 1st, 2015, and December 31st, 2017. Patient demographics and clinical viability parameters have been briefly presented in our recent mini-review [
]. All patients were referred by experienced cardiologists to undergo clinical cardiac viability testing by Rubidium-82 (82Rb) PET/CT and a dynamic FDG cardiac PET/CT prior to a potential revascularization procedure. Baseline characteristics (height, weight, age, and diabetes status) were measured by the time of the PET scans.
Access to patient files was granted by the Danish Patient Safety Authority (31-1521-273) and the institutional review board at Aarhus University Hospital. Due to the retrospective nature of the study, individual patient consent was waived.
Retrospectively, myocardial perfusion imaging data were collected from electronic patient records and clinical information was obtained from the time of the viability cardiac PET scan until end of study June 1st, 2020. All-cause mortality was registered as the primary outcome and secondary outcomes were listed as the first and second MACE defined as the combination of myocardial infarction, stroke, and cardiovascular death. Survival was calculated from the date of the FDG cardiac PET until death or end of study.
2.2 82Rb PET/CT perfusion scans
Patients were instructed to fast for 6 h and avoid caffeine-containing food or beverages and anti-ischemic medication 24 h prior to the procedure. Two cannulas were inserted into a superficial forearm vein for infusion of 82Rb and adenosine. For each patient, a rest and adenosine stress PET/CT scan was performed on a GE Discovery 690 or MI Digital Ready PET/CT (GE Healthcare, Waukesha, WI), which have identical PET-detectors. At rest, infusion of 1110 MBq of 82Rb (CardioGen, Bracco) was administered followed by a 7 min list-mode scanning of the heart and subsequently re-binned into 27 dynamic frames (12 × 5 s; 6 × 10 s; 4 × 20 s; 4 × 40 s; 1 × 60 s), one static frame (2.5–7 min), and 8 gated frames (2.5–7 min). All common corrections were applied including attenuation and scatter. The stress scan was performed 10 min after the end of the rest scan to ensure correct elution of 82Rb, and the same imaging protocol was conducted 2 min into an adenosine infusion (0.14 mg/kg/min over 6 min). Blood pressure, heart rate, and a 12‑lead ECG were recorded throughout the procedure.
2.3 Hyperinsulinemic euglycemic clamp (HEC)
After completion of the perfusion scan, a HEC was started. Patients without known diabetes and patients with type 2 diabetes treated with oral glucose-lowering drugs were examined in the fasting state. Patients treated with metformin were allowed to take their medication on the morning of the scan. Patients with insulin treated diabetes could eat breakfast before the perfusion PET and HEC and were instructed to take insulin and oral glucose-lowering drugs as usual. At time 0 min, a constant infusion of insulin (1 mIE/kg/min insulin aspart, Novo Nordisk, Denmark) mixed with 20 ml NaCl was initiated. After 15 min, an exogenous infusion with 20% glucose was started. Plasma glucose concentration in arterialized blood was measured at baseline and every 10 min during the HEC, and the infusion rate was adjusted continuously to maintain a plasma glucose concentration at 5.0 mmol/l. During steady state, the exogenous glucose infusion rate will equal the total amount of metabolized glucose in the body and is consequently a measurement of whole-body insulin sensitivity commonly known as the M-value (μmol/kg/min). The HEC was performed for 2 h and the M-value was calculated during the last 30 min, where steady state was assured [
Sixty minutes into the HEC, patients were placed in the PET scanner and a bolus injection of 400 MBq FDG was administered. Immediately following radiotracer injection, a 60-min dynamic FDG PET list-mode scanning of the heart was started with data re-binned into 36 dynamic frames (1 × 10 s; 8 × 5 s; 4 × 10 s; 3 × 20 s; 5 × 30 s; 5 × 60 s; 4 × 150 s; 4 × 300 s; 2 × 600 s), one static frame (45–60 min) and 8 gated frames (45–60 min). All common corrections were applied including attenuation and scatter. The HEC continued during image acquisition and regulation of plasma glucose level was managed as described above. Infusion of insulin was terminated after the scan, whereas glucose infusion continued until the plasma glucose stabilized at 5 mmol/l. The radiation dose for each patient was estimated to be 12 mSv in total for the whole procedure including 2 x low-dose CT, 82Rb PET, and FDG PET.
2.5 PET data analysis for cardiac measurements
82Rb and static FDG images were analyzed in QPET (v2015, Cedar-Sinai Medical Center, Los Angeles, CA, USA). For assessment of reversible (transient) ischemia in the left ventricle (LV), a visual method was used to assess tracer uptake as previously described [
]. The LV was divided into the standard areas comprising of the left anterior descending artery, circumflex branch of left coronary artery, and right coronary artery. Of these three areas, a remote area with the highest FDG uptake and the least amount of scarring was selected to represent maximal MGU of participating patients [
Early assessment of regional myocardial blood flow and metabolism in thrombolysis in myocardial infarction flow grade 3 reperfused myocardial infarction using carbon-11-acetate.
] (supplemental fig. 1). The remote area was determined both visually (on static images) and quantitatively expressed by the area with the highest Ki value. MGU values (μmol/min/100 g perfused tissue) were calculated by multiplying the Ki value with the average plasma glucose concentration (mmol/l) during the last 50 min of the HEC. A lumped constant of 1.16 for skeletal muscle was used [
]. All image analyses were performed performed by the same investigator (TVL) blinded to clinical variables.
2.6 PET data analysis for skeletal muscle measurements
The fused dynamic FDG PET/CT images were analyzed using PMOD 4.0 software (PMOD Technologies Ltd., Zürich, Switzerland) with the Patlak model to assess skeletal muscle glucose uptake (SGU). Regions of interests were drawn in patients' erectus spinae muscle (supplemental fig. 2) and tracer activity from the aorta/LV from the aQuant Research analysis was used as input function to calculate SGU values.
2.7 Statistical analysis
Analyses were performed using Stata 14 and 16 (StataCorp LP, College Station, Texas, USA). GraphPad Prism 8 (GraphPad Software Inc., San Diego, USA) was used to perform linear regression of bivariate numerical data. Numerical variables are expressed as mean ± standard deviation, and categorical variables are expressed as frequency with percentage (%). Comparisons were analyzed using Fisher's exact test for categorical variables and t-test for parametric numerical data. A p-value of <0.05 was deemed significant.
We computed cumulative incidences of first and second MACE after PET scan using the Aalen-Johansen estimator to account for death as competing risk. We determined all-cause mortality risk for patients with diabetes compared with patients without diabetes as well as for patients with an MGU above the median compared to those below, using the Kaplan Meier estimator for Patlak assessed MGU. Additionally, we determined all-cause mortality for patients with an MGU above the median compared to those below in an analysis stratified by diabetes status. In univariate analyses we used Cox's proportional hazards regression analysis to compute hazard ratio (HR) of death according to gender (male as reference), age category (33–45, 46–55, 56–65, 66–75 and 75 years of age or older), PET assessed ejection fraction (EF), MGU, total CFR, and revascularization intervention (ITV). The variables significantly associated with death in the univariate analysis (95% confidence interval not including 1) were included in a multivariate regression analysis. Proportional-hazards assumption was checked using Schoenfeld residuals. Five patients had missing values for MGU. We carried out sensitivity analysis assigning missing values the 1st, 50th and 99th percentile of MGU. As this did not change the point estimate (data not shown), all further analyses were carried out on data with complete observations. Twenty-two patients had missing values for CFR. While assigning the 1st and 50th percentile of total CFR did not change the point estimate, the 99th percentile changed the point estimate towards 1 (data now shown). However, in a clinical setting it is unlikely that patients with a missing CFR would have a high CFR as cancellation of an 82Rb adenosine stress scan would most likely be due to a poor cardiac pump function. Thus, it is more likely that these patients would have a low CFR and so all further analyses were carried out on data with complete observations.
3. Results
3.1 Patient characteristics and dynamic FDG cardiac PET values
A total of 131 patients were included in this study, and baseline characteristics are summarized in Table 1. The patients were characterized depending on diabetic status (33/131 = 25%), with only mean blood sugar level (mmol/l) and glucose infusion rate during the HEC differing between the two groups. There were no significant differences in age, gender, BMI, scar %, hibernation %, and viability metrics. In total, 44 out of the 131 patients underwent ITV of which 13/44 = 29% were individuals with diabetes.
Table 1Baseline subject characteristics and dynamic FDG cardiac PET data. Values are shown as mean±standard deviation or no. (%).
Variable
Participants, n = 131
No diabetes, n = 98
Diabetes, n = 33
Two-sided P-value (no diabetes vs. diabetes)
Age (years)
67.0±10.5
66.6±10.6
68.3±10.3
0.4181
Male gender, n (%)
114 (87%)
85 (87%)
29 (88%)
Fisher's exact P = 1.000
BMI
27.7±5.7
27.3±5.7
28.9±5.7
0.1552
Mean plasma glucose during HEC (mmol/l)
5.2±0.8
5.0±0.6
5.8±1.1
<0.001
Mean glucose infusion rate (mg/kg/min)
3.9±1.9
4.2±1.9
3.1±1.8
0.0064
Scar tissue (%)
20.4±13.2
21.4±13.5
17.5±11.9
0.1403
Hibernating tissue (%)
3.6±5.1
3.34±4.8
4.5 ± 5.9
0.2696
EF (%)
32.9±12.5
33.2±12.9
32.1±11.4
0.6630
LV MGU remote area (μmol/min/100 g tissue) (missing data for 5 subjects)
32.2±11.3
33.1±10.4 (n = 93)
29.7±13.2 (n = 33)
0.1387
SGU (μmol/min/100 g tissue) (missing data for 5 subjects)
3.2 Associations between myocardium, skeletal muscle, and whole-body GU rates
For all subjects, whole-body GU was significantly correlated with MGU (R2 = 0.09, p < 0.05) (Fig. 1a ). However, when stratified for the presence of diabetes, no correlation was observed for individuals with diabetes, whereas whole-body GU was correlated with MGU in individuals without diabetes (diabetes: R2 = 0.03, p = 0.37; no diabetes: R2 = 0.11, p < 0.05) (Fig. 1b). Overall, SGU and whole-body GU also correlated (R2 = 0.17, p < 0.05) (Fig. 1c). The correlation was weaker and not significant for patients with diabetes compared to patients without diabetes (diabetes: R2 = 0.09, p = 0.10; no diabetes: R2 = 0.22, p < 0.05) (Fig. 1d). The same pattern was observed for the association between SGU and MGU for the whole group (R2 = 0.09, p < 0.05) (Fig. 1e), and for individuals with diabetes compared to individuals without diabetes in these groups (diabetes: R2 = 0.03, p = 0.36; no diabetes: R2 = 0.13, p < 0.05) (Fig. 1f).
Fig. 1PET measurements of glucose uptake. Correlation of whole-body GU with MGU (A) and differentiated by diabetic status (B), the same two correlations of whole-body GU with SGU (C, D), and for MGU and SGU (E, F).
Patients were followed from the day of PET-scan until death or end of study June 1st, 2020. The median follow-up period was 1260 days (range: 52–1958 days), and 31 patients (31/131, 23.7%) died during the follow-up period. 75 patients (75/131, 57.3%) experienced a first MACE and 37 patients (37/131, 28.2%) experienced a second MACE with the median time of 205 days (range: 8–1471) and 444 days (range: 70–1848), respectively.
The cumulative incidence of first- and second MACE did not differ according to diabetes status (Fig. 2a and b ). Individuals with diabetes had an increased mortality compared to individuals without diabetes (log-rank test 7.53, p < 0.05) (Fig. 2c). After 1 year of follow-up, mortality was 5.1% (95% CI: 2%–12%) in individuals without diabetes and 18%(95% CI: 9%–36%) in individuals with diabetes. After 4 years of follow-up, mortality was 20% (95% CI: 12%–30%) in individuals without diabetes and 40% (95% CI: 26%–59%) in individuals with diabetes. All-cause mortality was similar for individuals with a MGU below the median compared to a MGU above the median (log-rank test 0.30, p = 0.58) (Fig. 2d). After 1 year of follow-up, mortality was 7.9% (95% CI: 3.4%–18%) in patients with MGU below the median and 8.8% (95% CI: 4.1%–18.6%) in patients with MGU above the median. After 4 years of follow-up mortality was 24% (95% CI: 14.3%–38.5%) in patients with MGU below the median and 25.8% (95% CI: 16.9%–28.4%) in patients with MGU above the median. In a stratified analysis of patients with and without diabetes, mortality was similar for individuals with a MGU value below the median compared to a MGU above the median (diabetes: log-rank test 0.04, p = 0.83; no diabetes: log-rank test 1.04, p = 0.31) (Fig. 2e).
Fig. 2Prediction of MACE and overall survival. Cumulative incidence among the non-diabetic and diabetic group for the first MACE event following PET scan (A) and for the second MACE event following PET scan (B). Kaplan-Meier curve adjusted for diabetic status (C). Kaplan-Meier curve adjusted for MGU above and below median (D). Kaplan-Meier curve with MGU above and below median adjusted for diabetic status (E).
Table 2 shows HRs of mortality from uni- and multivariate analyses. In univariate analyses, diabetes and increasing age were associated with an increased mortality HR. Higher CFR was associated with a lower mortality HR both in the univariate (HR: 0.3 (95% CI: 0.1–0.7)) *and in the multivariate analysis (HR: 0.32 (95% CI: 0.12–0.85)).
Table 2HR for univariate and multivariate analysis for diabetes, MGU, EF, age, gender, CFR, and ITV.
Recently, there have been speculations that isolated cardiac insulin resistance is particularly detrimental in patients with ischemic HF and that cardiac insulin resistance could in fact contribute to the development of HF [
]. Therefore, we retrospectively reviewed 131 individuals with heart failure referred for cardiac viability work-up in whom both PET derived cardiac perfusion, absolute MGU values and outcome measures were available.
4.1 Peripheral and myocardial insulin sensitivity is not always closely correlated
Progression of insulin resistance can be seen in every insulin dependent tissue and is in general perceived to move in tandem with changes in whole-body insulin sensitivity [
]. However, this is not always the case. For example, MGU has been demonstrated to be comparable in individuals with diabetes and hypertriglyceridemia and healthy controls despite substantial differences in skeletal and whole-body glucose uptake [
]. Furthermore, the same two studies demonstrated that MGU did not correlate to skeletal muscle or whole-body glucose uptake. This indicates that development of insulin resistance in the myocardium is affected by other factors than those responsible for skeletal muscle insulin resistance. An analogous decoupling of hepatic and whole body glucose uptake has been reported by Honka et al., who measured endogenous glucose production and glucose uptake in skeletal muscle, adipose tissue, and liver using FDG PET/CT and a HEC in patients without diabetes. They found that glucose uptake in skeletal muscle and adipose tissue correlated with whole-body glucose uptake, whereas the correlation between hepatic glucose uptake and whole-body glucose uptake was opposite [
In this study, myocardial, skeletal muscle and whole-body glucose uptake was associated in individuals without diabetes, whereas these were dissociated in individuals with diabetes. As previously described, we observed a tissue-dependent decoupling of myocardial and whole-body glucose uptake in individuals with diabetes. There may be several explanations for this. First, it is conceivable that cardiac and skeletal muscle glucose uptake simply do not fluctuate in a linear fashion regardless of which type of patients are studied. ~80% of energy to the heart is obtained from lipids, lactate and ketones, even under conditions of hyperinsulinemia [
]. Therefore, it is possible that subtle increases in glucose uptake during a two-hour HEC is not detected to the same extent as changes in SGU.
Alternatively, the results may be explained by the composition of our patient cohort. Thus, patients enrolled in this study suffered from ischemic HF, which is associated with increased MGU due to an increase in glycolysis [
]. Moreover, animal studies have indicated that transient ischemia may alter the expression of GLUT isoforms towards the insulin independent isoforms [
]. Although we did measure MGU in remote areas (areas not scarred), it is likely that even these areas of the myocardium could be ischemic at times causing a gradual shift of glucose transport towards insulin-independent GLUT1 transporters. Such a shift would incur an apparent decoupling of insulin stimulated SGU and MGU, particularly during a HEC.
4.2 The prognostic impact of MGU and diabetes
Our second aim was to investigate the impact of cardiac IR as a predictor of coronary artery diease. Few studies have addressed this previously. In the most comprehensive study [
Clinical significance of quantitative assessment of right ventricular glucose metabolism in patients with heart failure with reduced ejection fraction.
], dynamic FDG PET data were used as prognostic indicators for mortality and MACE. It included 75 HF patients characterized by LV EF <40%, where more than half of the subjects had diabetes (53.7%). Somewhat surprisingly, both mortality rate and MACE were increased in individuals with high global MGU values. To explain this counterintuitive finding, Tsai et al. also pointed to the specific metabolic modulations described above that occur in an ischemic heart [
]. Restricted energy production in the ischemic myocardium due to limited oxygen supply leads to a decrease in mitochondrial oxidative metabolism and increase in glycolysis as well as an uncoupling between glycolysis and glucose oxidation [
]. Thus, what would appear to be an insulin sensitive heart with insulin stimulated glucose uptake could in fact be a severely strained myocardium taking in glucose independently of insulin stimulation and with adverse metabolic effects. The study was in our opinion limited by several factors. Firstly, a global MGU was used with a possible inclusion of scar tissue. Secondly, an oral glucose loading was used instead of the recommended HEC [
An abbreviated hyperinsulinemic-euglycemic clamp results in similar myocardial glucose utilization in both diabetic and non-diabetic patients with ischemic cardiomyopathy.
Myocardial glucose utilization and optimization of (18)F-FDG PET imaging in patients with non-insulin-dependent diabetes mellitus, coronary artery disease, and left ventricular dysfunction.
], and lastly the study had a relatively short follow-up time (<2 years). Furthermore, it can also be argued that using ROC based thresholding to determine high and low MGU will by definition increase the likelihood of finding statistically different outcomes.
In our study, no difference in overall survival was observed between MGU below and above the median during the five-year follow-up period, even after adjusting for diabetic status. This was also the case when we analyzed our data using global MGU to predict outcome (shown in supplemental). We believe this difference compared to Tsai et al. may be explained by our use of a proper HEC instead of oral glucose loading. The insulin response after oral glucose loading is affected by factors such as the incretin response and the absorption from the gastrointestinal tract, which may differ between individuals and therefore introduce excess variation not attributed to insulin resistance. As a result, FDG may not be entirely cleared from the lumen of the cardiac ventricles, and spill-in of FDG activity into especially scarred tissue frequently occurs. This in turn may lead to incorrect and higher estimates of MGU. Using only remote myocardium ensures that no such errors are made. The size of our cohort improves our confidence in observation that MGU does not predict mortality or MACE.
A few potential limitations to our study should be considered. First, a longer follow-up period and a wider age range in our study population including younger subjects would have been preferred. However, we believe that we observed a sufficient number of events during the observation period to allow us to assess the prognostic value of MGU with some certainty. Second, SGU was measured in the erectus spinae muscle and not in a larger muscle group (e.g. in the legs) due to the limited field of view of the PET scanner. SGU varies between muscle groups, and the most precise estimate of maximal insulin stimulated glucose uptake is probably obtained using larger muscles with substantial perfusion in the upper or lower extremities [
Altered glucose uptake in muscle, visceral adipose tissue, and brain predict whole-body insulin resistance and may contribute to the development of type 2 diabetes: a combined PET/MR study.
]. Third, the lack of a control group with no known HF and myocardial scar tissue precludes us from evaluating the potential predictive importance of MGU in e.g. patients with type 2 diabetes without ischemic heart disease. Fourth, our overall insulin sensitivity assessment was only based on a non-adjusted M-value since our study population was included as part of a normal clinical evaluation. The use of the M-value alone may be influenced by differences in body composition as recently reported by Ter Horst et al. [
]. It is therefore very likely that an insulin adjusted M-value (M/I), as used in a large number of studies and considered the optimal assessment of insulin sensitivity [
], had resulted in a more precise estimate of the relationship between peripheral and cardiac insulin sensitivity. Fifth, Since this was a retrospective study, we have no knowledge of skewed changes in e.g., smoking cessation, improved blood pressure control, weight loss, exercise, and other relevant measures, which could have impacted outcome. To address the effects of these preventive measures, a prospective study with rigorous follow-up visits would have to be conducted. Finally, it has been established that women with diabetes have a significantly increased risk of developing cardiovascular disease and a higher mortality rate compared to men with diabetes [
]. The causes of death include both all-cause mortality and diabetes-related cause of death. Since our study population only consisted of 13% women, the possible prognostic value of MGU on overall survival could be underestimated. The same notion goes for the lack of stratification of our study population based on EF. HF patients with reduced EF (HFrEF) are more likely to be younger, male, and have an ischemic aetiology, but less likely to have hypertension and to be obese compared to HF patients with preserved EF (HFpEF) [
Prevalence and prognostic impact of non-cardiac co-morbidities in heart failure outpatients with preserved and reduced ejection fraction: a community-based study.
Epidemiology and one-year outcomes in patients with chronic heart failure and preserved, mid-range and reduced ejection fraction: an analysis of the ESC heart failure long-term registry.
Epidemiology and one-year outcomes in patients with chronic heart failure and preserved, mid-range and reduced ejection fraction: an analysis of the ESC heart failure long-term registry.
]. The majority of our study population was most likely HFrEFs, since the mean EF was 32.9%. This could lead to a higher mortality rate and an underestimation of the prognostic value of MGU.
5. Conclusions
In this study, we observed the anticipated decoupling of insulin-stimulated skeletal and myocardial glucose uptake in ischemic HF patients with type 2 diabetes, whereas the association was preserved in ischemic HF patients without diabetes. However, contrary to our preliminary hypothesis, myocardial insulin resistance assessed by MGU was not a predictor of overall survival or outcome in general. Instead, diabetes per se and CFR were predictive of overall survival. Taken together, these observations underscore the notion that diabetes related HF is caused by small vessel disease and not by cardiac IR, and that perturbations in MGU is of minor importance for overall morbidity and mortality.
Data availability
Data collection and analyses are complete. Data will be shared on reasonable request. The study protocol and statistical analysis plan may be obtained from the corresponding author.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Author statement
The corresponding author attests that all listed authors meet the authorship criteria and that no others meeting the criteria have been omitted.
CRediT authorship contribution statement
Thien Vinh Luong: Investigation, Data curation, Visualization, Writing – original draft, Validation. Mette Glavind Bülow Pedersen: Formal analysis, Visualization, Writing – original draft, Validation. Mette Louise Blouner Gram Kjærulff: Investigation, Data curation, Visualization, Writing – original draft, Validation. Simon Madsen: Investigation, Data curation, Visualization, Writing – original draft, Validation. Katrine Meyer Lauritsen: Investigation, Data curation, Visualization, Writing – original draft, Validation. Esben Søndergaard: Conceptualization, Methodology, Visualization, Writing – original draft, Validation. Lars C. Gormsen: Conceptualization, Methodology, Visualization, Writing – original draft, Validation.
Declaration of competing interest
All authors declare no conflict of interests.
Acknowledgements
We thank the technical personnel from the Department of Nuclear Medicine & PET Centre, Aarhus University Hospital for their excellent help in carrying out the clamping procedures.
Early assessment of regional myocardial blood flow and metabolism in thrombolysis in myocardial infarction flow grade 3 reperfused myocardial infarction using carbon-11-acetate.
Clinical significance of quantitative assessment of right ventricular glucose metabolism in patients with heart failure with reduced ejection fraction.
An abbreviated hyperinsulinemic-euglycemic clamp results in similar myocardial glucose utilization in both diabetic and non-diabetic patients with ischemic cardiomyopathy.
Myocardial glucose utilization and optimization of (18)F-FDG PET imaging in patients with non-insulin-dependent diabetes mellitus, coronary artery disease, and left ventricular dysfunction.
Altered glucose uptake in muscle, visceral adipose tissue, and brain predict whole-body insulin resistance and may contribute to the development of type 2 diabetes: a combined PET/MR study.
Prevalence and prognostic impact of non-cardiac co-morbidities in heart failure outpatients with preserved and reduced ejection fraction: a community-based study.
Epidemiology and one-year outcomes in patients with chronic heart failure and preserved, mid-range and reduced ejection fraction: an analysis of the ESC heart failure long-term registry.
Significant value.
00162-1&id=gr1.jpg){kind=link}
00162-1&id=gr2.jpg){kind=link}