Hovorka R. Isec Calculating Insulin Secretion Program Http Www.soi.city.ac.uk Sg331.

Methods for the Assessment of p-Cell Function In Vivo

Andrea Mariand GiovanniPacini

Introduction

The P cell plays a key role in the maintenance of glucose homeostasis and P-cell dysfunction is a characteristic feature of many states of glucose intolerance. Thus, the assessment of P-cell function is of fundamental importance in the study of metabolic disorders, particularly type 2 diabetes, and in the evaluation of drugs to treat P-cell dysfunction. At the present time, P-cell function continues to be very actively studied, as strategies for preventing the decline of P-cell function and design of drugs that may achieve this effect are a promise for the alleviation of the burden of diabetes.

The scope of this chapter is to provide a comparative and critical account of the most important methods for the assessment of P-cell function in vivo. We illustrate the basic technical aspects of the methodologies, referring the reader to the original publications for the details of test protocols. We also try to relate the characteristics of the in vivo tests to the P-cell molecular processes, knowledge of which has been greatly increased in the recent years. We hope that this attempt, though imperfect, will improve critical understanding of the methods. Other useful reviews of P-cell function assessment can be found in several journals (Hovorka & Jones 1994; Kahn 2003; Pacini & Mari 2003; Ferrannini & Mari 2004; Mari 2006) and textbooks (Porte et al. 2003; DeFronzo et al. 2004; LeRoith et al. 2004).

Methods for insulin secretion in vivo

The assessment of P-cell function requires the determination of the insulin secretory response to a given stimulus and, if the stimulus is not standardised, normalisation of the response to the stimulus. Thus, the evaluation of insulin secretion is a prerequisite for the assessment of P-cell function. The insulin secretory response can be simply determined from insulin concentration, or else from C-peptide levels using more complex methodologies.

Clinical Diabetes Research: Methods and Techniques Edited by Michael Roden © 2007 John Wiley & Sons, Ltd ISBN 978-0-470-01728-9

Insulin concentration

Simple measurement of plasma insulin is a classic approach still employed in many studies. However, peripheral insulin concentration reflects pancreatic insulin secretion only partially. Insulin in fact undergoes a first pass hepatic removal of about 50 %, i.e. about half the insulin secretion never reaches the periphery. Most importantly, insulin clearance, of which hepatic extraction is a major determinant, may vary in different metabolic conditions (Duckworth et al. 1998). Differences in peripheral insulin may not only reflect differences in insulin secretion but also differences in insulin clearance (Ferrannini et al. 1997; Camastra et al. 2005). Nevertheless, the observed insulin concentration profile almost parallels that of insulin secretion, as insulin kinetics is fast. Rapid insulin release, such as that seen in first phase secretion, is clearly reflected in insulin concentration.

C-peptide methods

To avoid the problem of non-constant insulin clearance, an alternative approach based on the measurement of C-peptide was developed almost 30 years ago (Eaton et al. 1980). C-peptide

Peptide Image

Figure 2.1 Illustration of deconvolution. The mathematical representation of the relationship between insulin secretion (ISR) and C-peptide concentration is based on convolution. The determinant of this relationship is the C-peptide concentration response to a C-peptide bolus injection, which quantitatively describes C-peptide kinetics. Convolution is the operation with which C-peptide concentration is calculated from ISR. If C-peptide kinetics is known (either by direct assessment or using the allometric formula of Van Cauter et al. (1992)), it is possible to reverse the convolution operator and calculate ISR from C-peptide concentration. This operation is called deconvolution and is illustrated in the graphs on the top (data from Figure 2.4). In this example, ISR is represented as a piecewise constant function over one-minute intervals (top left). For a given ISR time-course, the C-peptide kinetic model allows calculation (by convolution) of the corresponding C-peptide concentration. Thus, the ISR values can be determined by fitting the calculated C-peptide values to the measured ones. This is done using a modified least-squares method that ensures a smooth ISR profile. The graphs on top show measured (dots) and fitted (solid line) C-peptide values obtained with this procedure (right graph) and the calculated ISR (left graph).

Figure 2.1 Illustration of deconvolution. The mathematical representation of the relationship between insulin secretion (ISR) and C-peptide concentration is based on convolution. The determinant of this relationship is the C-peptide concentration response to a C-peptide bolus injection, which quantitatively describes C-peptide kinetics. Convolution is the operation with which C-peptide concentration is calculated from ISR. If C-peptide kinetics is known (either by direct assessment or using the allometric formula of Van Cauter et al. (1992)), it is possible to reverse the convolution operator and calculate ISR from C-peptide concentration. This operation is called deconvolution and is illustrated in the graphs on the top (data from Figure 2.4). In this example, ISR is represented as a piecewise constant function over one-minute intervals (top left). For a given ISR time-course, the C-peptide kinetic model allows calculation (by convolution) of the corresponding C-peptide concentration. Thus, the ISR values can be determined by fitting the calculated C-peptide values to the measured ones. This is done using a modified least-squares method that ensures a smooth ISR profile. The graphs on top show measured (dots) and fitted (solid line) C-peptide values obtained with this procedure (right graph) and the calculated ISR (left graph).

is co-secreted with insulin in equimolar amounts, undergoes negligible hepatic extraction and has linear and relatively constant kinetics. Thus, C-peptide concentration reflects more precisely true pancreatic insulin secretion, although its time-course, compared to that of insulin concentration, is somewhat blunted and delayed with respect to insulin secretion. For this reason, a mathematical operation, called deconvolution, is used to reconstruct insulin secretion from C-peptide concentration. Deconvolution, illustrated in Figure 2.1, is the mathematical operation with which C-peptide (i.e. insulin) secretion is calculated from C-peptide concentration (see Hovorka & Jones 1994 for details).

To perform deconvolution, C-peptide kinetics must be known. In the original approach (Eaton et al. 1980), followed in several successive studies, C-peptide kinetics was determined by a bolus injection of biosynthetic C-peptide in each individual. In a later study (Van Cauter et al. 1992), the difficulty of the assessment of the individual C-peptide kinetics was circumvented by developing a method by which approximate C-peptide kinetic parameters could be derived from anthropometric measurements. This approach has made wide application of the C-peptide deconvolution methodology possible, and is currently one of the most common methods for insulin secretion.

C-peptide deconvolution remains unfortunately a rather specialised approach, as it requires specific software and technical expertise (Hovorka & Jones 1994). A publicly available (though rather old) program for deconvolution (Hovorka et al. 1996) can be found at http://www.soi.city.ac.uk/~sg331/software.html (web search keywords: isec,site:city.ac.uk).

p-cell response characteristics in vivo p-cell response

Figure 2.2 summarises the most relevant characteristics of a normal P-cell response: 1) when glucose is raised gradually, insulin secretion is progressively stimulated (Figure 2.2A) and a dose-response relationship is observed between glucose concentration and insulin secretion (Figure 2.2B). 2) When glucose concentration is briskly increased and maintained at a suprabasal level, insulin secretion shows a biphasic pattern, with an initial burst (first phase insulin secretion) followed by a gradually increasing secretion that approaches a nearly constant level after about 1-2 h (second phase insulin secretion) (Figure 2.2C). The amplitude of both first and second phase response is a function of the glucose increment; the amplitude of the second phase depends on the P-cell dose-response. A biphasic response is also observed with the intravenous glucose tolerance test (IVGTT), in which a strong first phase secretion peak is followed by a slower and more blunted secretion rise. 3) Prolonged exposure to hyperglycaemia produces an increase of the insulin response (both first and second phase). This phenomenon, called potentiation, is clearly visible by repeating the same stimulus after a short rest period (Figure 2.2C). 4) The P-cell response to an oral glucose stimulus, such as an oral glucose tolerance test (OGTT), is considerably higher than that obtained with an intravenous glucose infusion at matched glucose levels (Figure 2.2D). This augmentation of the secretory response is mainly attributed to gut-secreted hormones called incretins, and in particular to glucose-dependent insulinotropic peptide (GIP) and glucagon-like peptide-1 (GLP-1). 5) The P-cell responds also to various non-glucose stimuli (e.g. some aminoacids, sulphonylureas and glucagon). An aminoacid frequently used in P-cell function testing is arginine, which is a powerful secretagogue. Injection of arginine produces a strong first phase insulin response, which is potentiated by hyperglycaemia (Figure 2.2E). 6) In the presence of insulin resistance, P-cell function is increased to cope with the increased insulin demand necessary to set resistant glucose uptake and production processes to adequate levels. This phenomenon is exemplified in Figure 2.2F for the acute insulin response (AIR),

Insulin Demand

Figure 2.2 Characteristics of the p-ceLL response: A) insulin secretion (ISR) obtained with a gradual increase in glucose concentration; B) the corresponding dose-response relating glucose concentration to insulin secretion. Redrawn from Byrne et al. (1995); C) response to a repeated square wave of hyperglycaemia (mean glucose levels shown in the bars). Both glucose stimulations produce the typical biphasic secretory pattern. Potentiation of the secretory response is evident in the second stimulus. Redrawn from Cerasi(1981); D) Enhancement of the secretory response observed when glucose is administered orally. The insulin curves are obtained with an OGTT (closed circles and solid line) and with an intravenous glucose infusion reproducing the same glucose levels of the OGTT (open circles and dashed line). Redrawn from Nauck et al. (1986); E) response to arginine injection at different glucose levels. The four smaller peaks are relative to a basal glucose concentration of ~5 mM - glucose levels for the higher peaks are shown in the bars above the peaks. The response to arginine is potentiated by hyperglycaemia. Redrawn from Ward et al. (1984); F) inverse relationship between the minimal model parameter of insulin sensitivity (SI) and the acute insulin response to an IVGTT (AIR). The solid line represents a hyperbolic interpolation in a group of 93 subjects (the dashed lines represent the dispersion, as percentiles of the disposition index). As a subject becomes insulin resistant (from point S to R), his secretion increases to maintain glucose tolerance normal. Redrawn from Kahn et al. (1993).

Figure 2.2 Continued.

a parameter of first phase secretion of the IVGTT (see below). Fasting insulin secretion exhibits a similar relationship with insulin sensitivity.

While the P-cell characteristics illustrated in Figure 2.2 are well known, the relative importance of the various response modes to glucose regulation and the pathophysiology of type 2 diabetes is still debated. The response modes are often but not always correlated (Ferrannini & Mari 2004). Therefore, although the use of a single P-cell function index may be a practical necessity, complete assessment of P-cell function requires multiple indices.

Cellular processes underlying the p-cell response

Some background on the relevant cellular processes is useful to shed light on the physiological meaning of the P-cell response characteristics observed in the various tests, although our understanding of the molecular aspects of insulin secretion is still largely incomplete. The molecular aspects of P-cell secretion are discussed in depth in several excellent reviews (Henquin 2000; Rorsman et al. 2000; Bratanova-Tochkova et al. 2002; Henquin et al. 2002; Straub & Sharp 2002).

Figure 2.3 is a simplified representation of the P-cell secretory machinery. Two pathways, denoted as triggering and amplifying (Henquin 2000), are shown. In the triggering pathway, glucose activates exocytosis by increasing cytosolic calcium concentration through a chain of events that ends with the opening of the calcium channels and an influx of extracellular calcium (see the legend to Figure 2.3). Through the amplifying pathway, glucose modulates insulin secretion independently of changes in calcium concentration. Both these pathways are important determinants of the P-cell response.

In the P cell, insulin is stored in granules. Only granules in a specific status (usually denoted as immediately releasable) can undergo exocytosis by activation of the triggering pathway. The translocation of granules from one status to another is an additional determinant of the phasic insulin response. According to some viewpoints, an immediately releasable pool of granules is responsible for first phase insulin release (Daniel et al. 1999), while the second phase involves the translocation of new granules to the plasma membrane (see the reviews cited above).

Potassium Glucose Transporter

Figure 2.3 Simplified representation of the p cell. In the triggering pathway, glucose metabolism, which depends on glucose influx through the glucose transporters, modulates the ATP/ADP ratio, which increases (block arrow) when glucose concentration and metabolism increase. The increase in the ATP/ADP ratio closes the ATP-sensitive potassium channels; the closure of these channels produces a membrane depolarisation that opens the voltage-dependent calcium channels. Opening of the calcium channels increases the calcium influx and the cytosolic calcium concentration, which triggers exocytosis of the insulin granules that are ready for release. The amplifying pathway, which encompasses complex and incompletely understood phenomena depicted here very schematically, is responsible for increases in insulin secretion independent of changes in cytosolic calcium concentration. This pathway is also activated by glucose metabolism. The incretin hormones (GLP-1 in the figure) are thought to interact with the amplifying pathway to augment insulin secretion. Another key phenomenon in insulin release is the translocation of granules from pools inside the cell to the plasma membrane, as only granules in a particular state on the plasma membrane can be released by calcium-mediated exocytosis.

Figure 2.3 Simplified representation of the p cell. In the triggering pathway, glucose metabolism, which depends on glucose influx through the glucose transporters, modulates the ATP/ADP ratio, which increases (block arrow) when glucose concentration and metabolism increase. The increase in the ATP/ADP ratio closes the ATP-sensitive potassium channels; the closure of these channels produces a membrane depolarisation that opens the voltage-dependent calcium channels. Opening of the calcium channels increases the calcium influx and the cytosolic calcium concentration, which triggers exocytosis of the insulin granules that are ready for release. The amplifying pathway, which encompasses complex and incompletely understood phenomena depicted here very schematically, is responsible for increases in insulin secretion independent of changes in cytosolic calcium concentration. This pathway is also activated by glucose metabolism. The incretin hormones (GLP-1 in the figure) are thought to interact with the amplifying pathway to augment insulin secretion. Another key phenomenon in insulin release is the translocation of granules from pools inside the cell to the plasma membrane, as only granules in a particular state on the plasma membrane can be released by calcium-mediated exocytosis.

p-cell function tests Intravenous vs. oral tests

An important distinction should be made between intravenous and oral P-cell function tests. In fact, ingestion of glucose stimulates the entero-insular axis, i.e. a complex hormonal and neural response that markedly potentiates insulin secretion (Figure 2.2D) (Unger & Eisentraut 1969; Fehmann et al. 1995; Creutzfeldt 2001). The magnitude of the potentiation response depends not only on the degree of neural activation and secretion of gut incretin hormones but also on intrinsic P-cell function, as incretin hormones bind to specific P-cell receptors and activate signalling for secretion (Figure 2.3). Thus, oral tests give a more comprehensive assessment of P-cell function, but they cannot distinguish the intrinsic P-cell defects from those of the entero-insular axis (e.g. a defective GLP-1 production or impaired neural stimulation). In addition, with oral tests the secretory stimulus (e.g. glucose concentration)

cannot be standardised, and thus assessment of P-cell function requires appropriate methods for normalisation of insulin secretion to the stimulus.

Intravenous glucose tolerance test (IVGTT)

The IVGTT is the typical test for first phase insulin secretion, although a second phase is also present. For first phase assessment, a 10-min IVGTT is sufficient. However, the IVGTT is often also used to evaluate insulin sensitivity with the minimal model and possibly second phase secretion. Here the test format for the minimal model, which is the most widely used, will be described. More details can be found in Chapter 3.

The IVGTT minimal model protocol (Figure 2.4) is as follows: 1) a standardised glucose bolus (0.3 g/kg body weight) is injected after a baseline control period of about 20-30 min; 2) glucose, insulin and often C-peptide concentrations are measured at frequent intervals (12-30 samples) for 3-4 h. Frequent samples (at 1-2 min intervals) are collected in the initial 8-10 min for first phase assessment; 3) the typical first phase secretion index is the acute insulin response (AIR), i.e. the average incremental insulin concentration obtained in the first 5-10 min of the test; 4) second phase insulin secretion is calculated using empirical indices, usually computed from the areas under the insulin and glucose concentration curves, or by modeling (Toffolo et al. 1980; Toffolo et al. 1995; Toffolo et al. 1999), using both insulin and C-peptide; 5) In the insulin-modified IVGTT (shown in Figure 2.4), used to improve the minimal model insulin sensitivity estimate, a standardised insulin dose (0.03-0.05 U/kg) is administered 20 min after the glucose bolus. Exogenous insulin obviously masks the endogenous insulin response. However, if C-peptide concentration is measured, the second phase can be still determined (Toffolo et al. 1999) (with the proviso that exogenous insulin may interfere with endogenous secretion).

The IVGTT, together with the hyperglycaemic clamp, is the typical test for first phase insulin secretion, and AIR is the most widely used index. Assessment of second phase secretion is on the other hand made difficult by the necessity of accounting for glucose levels, which may vary considerably. Compared to AIR, the empirical and the model-based indices of second phase secretion have received limited attention. First phase insulin release depends on the amplitude of the glucose increment after the bolus. However, as the glucose dose is standardised, it is reputed that AIR does not require normalisation to the glucose peak. On the other hand, in normal subjects AIR is dependent on insulin sensitivity (Figure 2.2F). Thus, AIR per se may not be a good index of P-cell function, i.e. comparison of AIR in populations with different insulin sensitivity may lead to inappropriate conclusions. This important problem is discussed in a later section.

One drawback of the IVGTT is that first phase insulin secretion, though important, is only one of the modes of response of the P cell, and is thus insufficient to characterise P-cell function satisfactorily. As known since long time (Seltzer et al. 1967) and re-emphasized recently (Ferrannini & Mari 2004), diabetic subjects may totally lack first phase secretion but still respond to an OGTT.

First phase secretion quite likely represents the discharge of a pool of immediately releasable insulin granules through the activation of the triggering pathway (Daniel et al. 1999; Rorsman et al. 2000; Straub & Sharp 2002). As the magnitude of this pool depends on a complex equilibrium between exocytosis and refilling from a precursor pool, in which several cellular processes are involved, an observed defect of first phase insulin release may result from quite different causes. Therefore, while there is ample evidence that

Ivgtt Peptide

Figure 2.4 Illustration of the IVGTT protocol. The figure shows the insulin-modified frequently sampled IVGTT (23 samples), which is currently in use for assessing both insulin sensitivity (with the minimal model) and secretion. A 0.3 g/kg body weight glucose bolus is given at time 0, and a five-minute insulin infusion (0.03 U/kg) is administered after between 20 and 25 min. The regular IVGTT protocol is similar but does not include insulin infusion. Shorter and less frequently sampled protocols can be employed, in particular if only first phase secretion is needed. Measurement of C-peptide is optional, but allows calculation of insulin secretion by deconvolution during the whole test, avoiding the confounding effect of exogenous insulin (see Figure 2.1). Individual data from Mari(1998).

Figure 2.4 Illustration of the IVGTT protocol. The figure shows the insulin-modified frequently sampled IVGTT (23 samples), which is currently in use for assessing both insulin sensitivity (with the minimal model) and secretion. A 0.3 g/kg body weight glucose bolus is given at time 0, and a five-minute insulin infusion (0.03 U/kg) is administered after between 20 and 25 min. The regular IVGTT protocol is similar but does not include insulin infusion. Shorter and less frequently sampled protocols can be employed, in particular if only first phase secretion is needed. Measurement of C-peptide is optional, but allows calculation of insulin secretion by deconvolution during the whole test, avoiding the confounding effect of exogenous insulin (see Figure 2.1). Individual data from Mari(1998).

first phase secretion is a sensitive marker of (-cell function, impairment of this function may not be a primary (-cell defect (Mari 2006).

Hyperglycaemic glucose clamp

The hyperglycaemic glucose clamp (DeFronzo et al. 1979) assesses both first and second phase secretion. The protocol is as follows (see Figure 2.5): 1) after an initial baseline control period, an intravenous priming dose of glucose, followed by a variable glucose infusion, is administered to sharply raise glucose concentration to the desired hyperglycaemic value; 2) to keep glucose concentration constant in the successive period, glucose infusion rate is frequently adjusted based on quick bedside measurement of glucose concentration, similarly to the euglycemic clamp (see Chapter 4). An approximate equilibrium is reached after about

Glucose Clamp Technique

Figure 2.5 Illustration of the hyperglycemic clamp protocol. A 4 mM step in glucose concentration (top panel) is generated by a primed glucose infusion (bottom panel), as described by DeFronzo et al. (1979). The priming is performed by increasing the glucose infusion rate in the initial 10 min (bottom panel). The glucose infusion rate is frequently adjusted based on quick bedside glucose concentration measurements to keep glucose constant. The typical biphasic insulin response is visible in the second panel. Individual data from Nataliet al. (1998).

Figure 2.5 Illustration of the hyperglycemic clamp protocol. A 4 mM step in glucose concentration (top panel) is generated by a primed glucose infusion (bottom panel), as described by DeFronzo et al. (1979). The priming is performed by increasing the glucose infusion rate in the initial 10 min (bottom panel). The glucose infusion rate is frequently adjusted based on quick bedside glucose concentration measurements to keep glucose constant. The typical biphasic insulin response is visible in the second panel. Individual data from Nataliet al. (1998).

two hours; 3) glucose and insulin (and possibly C-peptide) are sampled in the basal period, frequently in the initial 8-10 min to assess first phase secretion, and successively during the final steady-state period to evaluate the second phase; 4) the secretory response can be evaluated from insulin concentration, using C-peptide deconvolution and also by modeling (Bonadonna et al. 2003). Several indices of first and second phase secretion can be calculated (e.g. AIR, absolute and incremental insulin or secretion values during the second phase); 5) because insulin secretion is available at two glucose levels (basal and hyperglycaemic), the test yields two points of the P-cell dose-response relating insulin secretion to glucose concentration; 6) diabetic subjects are usually not brought to euglycaemia by infusing insulin before the hyperglycaemic clamp is started, as typically done with the euglycemic clamp for insulin sensitivity, as this would confound the endogenous insulin response. Thus, the hyperglycaemic clamp typically evaluates the insulin secretion increment at a fixed glucose concentration increment, rather than absolute insulin secretion at standardised glucose levels (as is done with the graded glucose infusion test, discussed below).

The hyperglycaemic glucose clamp, though cumbersome, is the typical test for both first and second phase secretion. Because glucose levels are standardised, the hyperglycaemic clamp is a test of (-cell function, though comparison between subjects with different glucose levels (e.g. normal and diabetic subjects) may not be possible.

The hyperglycaemic clamp index of first phase secretion is similar to that obtained with the IVGTT and shares the same physiological interpretation. In contrast to the first phase, which is mainly dependent on the activation of the triggering pathway, the second phase response depends on the activation of the amplifying pathway (Figure 2.3).

Graded glucose infusion test

The graded glucose infusion test (Byrne et al. 1995) is an experimental determination of the (-cell dose-response relating insulin secretion to glucose concentration (Figure 2.2A and B). The procedure is as follows: 1) after a basal control period, glucose is infused in multiple 40-min steps at progressively increasing rates, so that glucose levels increase gradually (Figure 2.2A); 2) glucose and C-peptide concentrations are measured and insulin secretion is calculated by deconvolution; 3) the mean insulin secretion values at the end of each step are plotted against the corresponding mean glucose levels, thus obtaining an individual (-cell dose-response (Figure 2.2B); 4) to determine insulin secretion at glucose values below baseline, and in particular to test diabetic subjects at glucose concentrations comparable to normal fasting glucose values, glucose is lowered by an insulin bolus or infusion before the start of the graded glucose infusion. Because insulin secretion is computed from C-peptide, insulin infusion does not interfere with the calculations. Thus, the (-cell dose-response can be determined at standardised glucose levels in both normal and diabetic subjects.

The graded infusion test, though cumbersome (almost six hours), is a direct and standardised evaluation of the (-cell dose-response. The dose-response is an essential feature of the ( cell, also observable at a cellular level (Henquin 2000), and clearly characterises the (-cell function impairment in various pathophysiological states (e.g. Byrne et al. 1996). As for second phase secretion from the hyperglycaemic clamp, the dose-response obtained with this test is in relation with the activation of the amplifying pathway (Figure 2.3). The graded infusion test does not assess first phase secretion.

Arginine tests

The injection of arginine produces a large burst in insulin secretion that is reputed to represent a form of maximal first phase insulin release. The secretory response to arginine is dependent on the prevailing glucose levels and is usually stronger than that elicited by glucose injection such as in the IVGTT. The basic arginine test requires measurement of insulin (and possibly C-peptide) concentration before the arginine injection and successively for about 10 min. The study protocol is similar to that employed with the IVGTT, with injection of arginine instead of glucose and about 10-min sampling. The secretory response is typically expressed as the mean incremental insulin (or C-peptide) level after arginine injection, similarly to the IVGTT AIR.

An important variant of the basic arginine test involves multiple injections of arginine at different glucose levels (Ward et al. 1984) (Figure 2.2E). The rationale of this approach is that exposure to hyperglycaemia for a short time potentiates the insulin response to arginine and the magnitude of the potentiation effect, in addition to the absolute response to arginine, is an index of P-cell function. An outline of this quite complex experimental test is the following (see Ward et al. 1984 for details): 1) arginine is injected in basal conditions and the response analysed as described above; 2) glucose is infused at a variable rate for about 30 min to raise glucose concentration to a fixed level (similarly to a hyperglycaemic clamp) and a second arginine bolus is given. The same procedure is repeated at multiple glucose levels. The maximal response to arginine is obtained at a glucose level of about 30 mmol/l; 3) a dose-response, relating the glucose levels at which arginine is injected and the corresponding secretory responses, is constructed (Figure 2.6). Beside the absolute responses, an index of P-cell function is the slope of this dose-response, evaluated between the first two glucose levels (classically denoted as glucose potentiation slope); 4) to study the arginine response in diabetic subjects at the same glucose levels as normal subjects, insulin is infused before the test to lower glucose concentration. Arginine activates the triggering pathway, though with mechanisms different from glucose (Smith et al. 1997; Henquin 2000). In comparison to the IVGTT, the arginine test releases a larger pool of insulin granules - the physiological significance of this is unclear - and elicits a first phase response in diabetic subjects.

The arginine test at multiple glucose levels is unique in its investigation of the potentiating effects of hyperglycaemia and their impairment. No other test for exploring this P-cell function has received a similarly wide application. Despite how interesting this mode of response of the P cell is, however, the complexity of the test is a serious drawback.

Acute Insulin Response Calculation

Figure 2.6 Dose-response relating the pre-arginine injection glucose levels and the acute insulin response to arginine (AIR, mean incremental insulin levels after injection). The glucose potentiation slope is the initial slope of the curves. The solid line is the dose-response in normal subjects; the broken line is that in type 2 diabetic subjects. Redrawn from Ward et al. (1984).

Figure 2.6 Dose-response relating the pre-arginine injection glucose levels and the acute insulin response to arginine (AIR, mean incremental insulin levels after injection). The glucose potentiation slope is the initial slope of the curves. The solid line is the dose-response in normal subjects; the broken line is that in type 2 diabetic subjects. Redrawn from Ward et al. (1984).

OGTT and meal tests

The OGTT methods are based on a standard 75 g OGTT and on measurement of glucose, insulin and possibly C-peptide in a variable number of blood samples, usually five or more, collected over 2-3 h (see e.g. Reinauer et al. 2003, available online at http://www.who.int/bookorders/anglais/catalog1.jsp?sesslan=1 (web search keywords: reinauer,site:www.who.int), for details on the protocol). A typical sampling schedule is 0 (pre-load), 30, 60, 90 and 120 min post glucose load. A standardised mixed meal is often used in place of the OGTT, with the same purpose and similar characteristics.

With the OGTT and meal tests, insulin concentration or secretion must be normalised to the prevailing glucose levels, which are clearly not standardised. This is achieved using empirical indices or modeling methods. The empirical indices are typically ratios between insulin and glucose levels. Perhaps the most widely used formula is the so-called insulinogenic index, calculated as the ratio between the supra-basal increments at 30 min of insulin and glucose concentration (see e.g. Pacini & Mari 2003). Similar indices based on the areas under the concentration curves have also been proposed. A recent alternative approach uses empirical formulas to provide estimates of the first and second phase insulin secretion as calculated using a hyperglycaemic clamp (Stumvoll et al. 2000, 2001). These formulas have been derived from a regression analysis of the OGTT indices against the hyperglycaemic clamp.

The modeling methods, discussed in the following section, have the conceptual advantage of being based on an explicit mathematical representation of the dynamic relationship between glucose concentration and insulin secretion. They provide an estimate of the P-cell dose-response relating insulin secretion to glucose concentration during the oral test. Some models also provide a parameter quantifying early insulin release during the oral test (Breda et al. 2001; Cretti et al. 2001; Mari et al. 2002a, 2002b), reputed to be a marker of first phase secretion. This parameter, however, may have limited reliability (Mari et al. 2002a; Steil et al. 2004). One model provides a potentiation parameter (Mari et al. 2002a, 2002b).

The OGTT and meal tests elicit a complex P-cell response. Not only the triggering and amplifying pathways involved in the intravenous tests are activated, but also the pathways stimulated by the activation of the entero-insular axis, such as the GLP-1 receptor signalling (Figure 2.3). Thus, oral tests give a more complete assessment of P-cell function. However, the relative roles of the various mechanisms involved in the response cannot be evaluated.

Delicious Diabetic Recipes

Delicious Diabetic Recipes

This brilliant guide will teach you how to cook all those delicious recipes for people who have diabetes.

Get My Free Ebook


Post a comment