Pancreatic perfusion and subsequent response to glucose in healthy individuals and patients with type 1 diabetes
- 922 Downloads
The aim of this study was to investigate pancreatic perfusion and its response to a glucose load in patients with type 1 diabetes mellitus compared with non-diabetic (‘healthy’) individuals.
Eight individuals with longstanding type 1 diabetes and ten sex-, age- and BMI-matched healthy controls underwent dynamic positron emission tomography scanning with 15O-labelled water before and after intravenous administration of glucose. Perfusion in the pancreas was measured. Portal and arterial hepatic perfusion were recorded as references.
Under fasting conditions, total pancreatic perfusion was on average 23% lower in the individuals with diabetes compared with healthy individuals. Glucose increased total pancreatic and portal hepatic blood perfusion in healthy individuals by 48% and 38%, respectively. In individuals with diabetes there was no significant increase in either total pancreatic or portal hepatic perfusion.
Individuals with type 1 diabetes have reduced basal pancreatic perfusion and a severely impaired pancreatic and splanchnic perfusion response to intravenous glucose stimulation.
KeywordsBlood flow Glucose Pancreas Pancreatic islets Perfusion Type 1 diabetes
Positron emission tomography
Regions of interest
Volumes of interest
Positron emission tomography (PET) with radioactive labelled water ([15O]H2O) as a tracer is a well-established tool to study changes in perfusion in human organs and has previously been used to study pancreatic perfusion under basal conditions [1, 2]. Pancreatic perfusion is normally closely related to tissue function, for example secretin, a hormone which stimulates exocrine secretion, also increases pancreatic perfusion in humans . The perfusion difference between insulin-deficient and healthy animals mainly reflects the islet vascular component, i.e. the islet blood flow contribution [4, 5, 6]. Moreover, a 2–3 fold increase in islet blood flow after intravenous glucose administration, which facilitates the release and dispersal of insulin, is consistently seen in healthy rodents [4, 5]. In this study we set out to study pancreatic perfusion under basal conditions and its responsiveness to glucose in non-diabetic (‘healthy’) and type 1 diabetic individuals.
All procedures were approved by the regional ethical board and performed in accordance with the Helsinki declaration. Written informed consent was obtained from all participants prior to their inclusion.
Descriptive data of study participants
Male (n, %)
26.7 ± 1.3
25.9 ± 0.9
22.9 ± 1.4
23.6 ± 1.0
HbA1c % DCCT
5.2 ± 0.3
7.1 ± 0.3***
(33.3 ± 0.82)
(53.6 ± 1.3)***
5.4 ± 0.2
6.7 ± 1.0
P-glucose (mmol/l) following glucose administration
12.1 ± 0.3
14.3 ± 1.01
0.6 ± 0.07
0.01 ± 0.006***
C-peptide (nmol/l) following glucose administration
1.6 ± 0.3
0.01 ± 0.006***
All participants were fasting for >4 h prior to the PET examinations. Eight of the ten healthy individuals and all (eight) participants with diabetes underwent one 10 min baseline and one 10 min [15O]H2O dynamic PET examination (400 MBq each) directly after an intravenous injection of 30% (wt/vol.) glucose solution (300 mg/kg). The remaining two healthy individuals were investigated with dynamic PET prior to and 10–20 min after glucose administration.
The pancreas was positioned in the centre of the 15 cm axial field of view of a Discovery ST PET/computed tomography (CT) scanner (GE Healthcare, Milwaukee, MI, USA) by assistance of a low-dose CT scout view (140 kV, 10 mAs). Image acquisition was performed in 3D (26 frames). Images were reconstructed using an iterative Ordered Subsets Expectation Maximisation (OSEM) VUEpoint algorithm (GE Healthcare). Data were analysed using the VOIager 4.0.7 software (GE Healthcare, Uppsala, Sweden). ROIs corresponding to the pancreas and its parts were delineated on transaxial sequential CT slices and combined into volumes of interest (VOIs). The border of the tail (cauda) of the pancreas was assigned to the medial border of the left kidney and the head (caput) of the pancreas was defined as the pancreatic region lateral to the superior mesenteric artery. In order to reduce possible partial volume effect, the ROIs were delineated slightly inside the pancreatic contour. CT VOIs were transferred to co-registered [15O]H2O-PET images. Perfusion was assessed by a standard single-tissue compartment model from dynamic pancreatic [15O]H2O-PET data, using an aortic VOI as the input function .
Portal and arterial hepatic perfusion were calculated and used as comparators . The data of one healthy individual could not be fitted accurately to this model and was therefore excluded.
Differences between two groups of parametric data were assessed by two-sided t tests. Normal distribution of data was assessed by Shapiro–Wilk’s test and equal variance was assessed by Levene’s test. Pearson’s Product Moment Correlation Coefficient was calculated to assess linear correlations. p < 0.05 was considered statistically significant. All values are given as means ± SEM.
As expected, HbA1c levels were increased in individuals with type 1 diabetes. However, after glucose administration, both mean fasting glucose and stimulated glucose concentrations were similar to healthy individuals during the PET examination (Table 1).
In healthy individuals, a 48 ± 8% increase (p = 0.0006, compared with baseline measurements) in total pancreatic perfusion was observed during the first 10 min after intravenous glucose administration (the change in absolute values is shown in Fig. 1e,g) and similar increases were observed in all subregions of the pancreas (the change in absolute values is shown in Fig. 1e–h). The pancreatic perfusion increase in response to glucose was much less pronounced in type 1 diabetic participants irrespective of their fasting glucose levels (the change in absolute values is shown in Fig. 1e,g). There was only a mean 10–12% increase when compared with baseline measurements in the pancreatic head (caput; p = 0.048) and tail (cauda; p = 0.004), but no statistically significant change in total pancreatic perfusion (p = 0.065) in these individuals. Neither portal nor arterial hepatic perfusion increased in diabetic individuals after intravenous glucose injection, while there was a 38 ± 10% (p = 0.0083) increase when compared with baseline measurements in the portal, but not arterial hepatic, perfusion in healthy individuals (the change in absolute values is shown in Fig. 1g). There was no correlation between change in plasma insulin concentrations and pancreatic blood flow in healthy individuals after glucose administration (Fig. 1i–j). Additionally, the magnitude of pancreatic and portal hepatic perfusion increase in healthy individuals did not show any correlation (Fig. 1k). In the two healthy individuals re-examined 10–20 min after glucose administration, there was no longer any pancreatic blood flow increase (1.6 ± 0.3 vs 1.7 ± 0.3 ml × min−1 × g−1), despite blood glucose concentrations remaining increased 20 min after glucose administration (11.3 ± 0.2 mmol/l) (data not shown).
We report a lower baseline perfusion of the pancreas in individuals with long-standing type 1 diabetes when compared with healthy individuals. In previous investigations, pancreatic perfusion in individuals with type 1 diabetes has been shown either not to change or to have a tendency to decrease [2, 7]. Experience from animal studies [4, 5, 6] is suggestive that the perfusion difference between insulin-deficient and healthy animals mainly reflects the islet vascular component. Direct studies of islet blood flow in humans cannot be performed, since the techniques currently available for use in animal studies are invasive and/or terminal [4, 5]. Pancreatic islets are dependent on the effective transport of oxygen to the tissue for glucose metabolism, and a high perfusion for their dispersal of insulin [5, 8]. It is of interest to note that in the healthy individuals an inverse relation between fasting plasma glucose concentrations and pancreatic perfusion was observed, suggesting that higher perfusion may have increased insulin dispersal into the portal circulation. Importantly, in the present study pancreas perfusion was determined from the clearance rate constant of water, which is dependent on the shape of the time-activity curve rather than its amplitude. Thus, it may be assumed that the partial volume effects associated with the smaller pancreas volumes in individuals with type 1 diabetes would not have contributed to the difference in perfusion observed .
We found that all diabetic individuals had an impaired perfusion response to glucose, with a negligible increase in pancreatic perfusion compared with that observed in healthy individuals. An increased islet perfusion in healthy individuals would facilitate insulin dispersal into the systemic circulation, but may not necessarily be linked to an increment in glucose-dependent insulin secretion . Accordingly, we did not observe any correlation between the increase in plasma insulin concentrations and perfusion after glucose administration in healthy individuals.
However, although not shown to correlate on an individual basis, the increase in pancreatic perfusion after glucose infusion in healthy individuals was accompanied by an increase in portal hepatic perfusion, when compared with baseline measurements, to which the pancreatic contribution is only a minor fraction . It is therefore likely that the pancreatic perfusion increase observed in this study did not only reflect an islet blood flow increase, as seen in rodents [4, 5], but may also have occurred as a result of general blood flow increase, caused by parasympathetic signalling through the vagal nerve . In line with this, the absence of an increase in pancreatic perfusion in diabetic individuals may be explained by autonomic neuropathy or endothelial dysfunction , albeit the type 1 diabetic participants involved in this study were chosen based on good metabolic control and absence of known microvascular complications. Additionally, the lack of an increase in arterial hepatic perfusion could be explained by the hepatic arterial buffer response to increased portal flow. Future studies in large animals may be of value to increase our understanding of the partial contribution of islet blood flow increase and general increase in blood perfusion of splanchnic organs.
In contrast to our results, a previous study of human pancreatic blood flow did not show any difference between individuals with type 1 diabetes and healthy individuals, neither at baseline nor after glucose administration . This disparity in findings could be due to the limitations of the technique that was used; arterial spin labelling MRI is a technique that has a high noise to signal ratio, where artefacts are induced by surrounding organ movements and where there are an insufficient number of voxels for sensitive analysis. In comparison, the technique used in the present study demonstrates the possibility of assessing both inter-individual differences and dynamic changes in pancreatic perfusion in a small sample size. This study, therefore, provides evidence for a potentially valuable tool for further studies of pancreatic perfusion in diabetes to gain additional understanding of the possible consequences of dysfunctional blood flow regulation as part of the disease.
This study received financial support from the Swedish Research Council (K2013-55-X-15043-10-5; K2015-54X-12219-19-4; K2013-64X-08268-26-3), the Swedish Diabetes Foundation, the Swedish Juvenile Diabetes Foundation, Diabetes Wellness Sverige, The European Foundation for the Study of Diabetes (EFSD)/Novo Nordisk and the Novo Nordisk Foundation.
Duality of interest
The authors declare that there is no duality of interest associated with this manuscript.
All authors conceived and designed the study and participated in the analysis and interpretation of the data. LC drafted the manuscript and the other authors revised it critically for intellectual content. All authors approved the final version of the paper. POC is the guarantor of this work.