Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

1.1 Principle of Arteriovenous Difference Measurement

Adipose tissue (AT) is more than a collection of adipocytes. It is a highly organized tissue in which different cell types interact, and in which a complex mix of hormones and substrates arriving in the plasma, together with neural input and the rate of blood flow (BF), all regulate metabolic activity. These multiple, interacting influences cannot be reproduced in vitro, and, hence, if we wish to understand the integration of AT metabolism in the whole body, it is essential to perform studies of AT metabolism in vivo. There are a number of ways in which this can be done (1). One of the most specific and informative, in a quantitative sense, is the measurement of arteriovenous (A-V) differences across the tissue.

The principle of this technique is simple: Differences in the composition of blood sampled from the arterial supply to the tissue, and from the venous drainage from the tissue, reflect the net metabolic activity of the tissue. In a straightforward example, if the glucose concentration in venous blood is less than that in arterial blood, then the tissue is consuming glucose in a net sense. The net uptake of glucose can be measured quantitatively by estimating the A-V difference for glucose and the rate of BF. The net rate of uptake is then given by (see Notes 1 and 2 ):

In the case of AT we may also be interested in the release of substances such as nonesterified fatty acids (NEFA), glycerol, and other adipocyte products, such as leptin. Then the venous concentration will be greater than the arterial, and the net rate of release is given by:

where V-A difference is the V-A concentration difference.

If a tissue were both utilizing a substrate from blood, and releasing it into the blood, then only the net difference between the rates of those processes would be reflected in a measurement of A-V concentration difference. The addition of a tracer can help to measure absolute rates of substrate uptake or release. One commonly used example is the measurement of NEFA uptake by skeletal muscle. Most muscle beds also contain AT, and net A-V difference measurements may show a net release of NEFA from such a tissue bed. If an isotopically labeled fatty acid (FA) is infused intravenously, it will be found that labeled FAs are extracted from blood across the tissue. This is assumed to represent absolute utilization. Once absolute utilization has been measured, it can be added to net release, to give absolute release. Details of the combination of A-V difference and tracer techniques are not covered here; the reader is referred to the book by Wolfe (2) for further information.

The principles of calculation of net substrate uptake or release, and of measurement of absolute rates of uptake or release with tracers, require the maintenance of steady-state conditions (see Note 3 ).

1.2 Choosing a Suitable Site for Blood Sampling

Although the principle is simple, the practice is not easy. The starting point is to identify a vessel that carries the specific venous drainage from AT. Although such vessels must exist in principle, they are probably too small to be identified. Vessels large enough to be identified and to have blood sampled from them probably always carry some drainage from other tissues, such as skin. However, there are some sites where the contribution of AT seems to predominate, probably because other tissues, such as skin, may be less active metabolically, and also because AT may predominate in mass terms. In humans, this applies to the veins draining from the subcutaneous (sc) anterior abdominal AT into the inferior epigastric vein (3), and probably to no other readily accessible vessels ( Fig. 1 ). There are vessels that fulfill this criterion in other species ( Table 1 ), but this chapter is confined to humans.

Fig 1.
figure 1

Anatomy of the venous drainage from the sc AT of the anterior abdominal wall in humans. Reproduced with permission from ref. 8.

Full size image
Table 1 Measurement of White AT Metabolism by A-V Differences in Different Species
Full size table

Blood sampling from the necessary vein in humans is best achieved by introduction of a flexible, indwelling catheter. Although, in principle, it may be possible to take single samples by venipuncture, in practice the technique requires that a flexible catheter is threaded down a small vein, until it is in a vessel of reasonable size for drawing samples. In addition, the presence of an indwelling catheter means that samples can be taken over a reasonable period, for instance, before and for some hours after eating a test meal, or before and during insulin infusion (11,12).

As well as identifying and sampling from a suitable vein, it is necessary to obtain arterial blood for comparison. The assumption always made is that arterial blood is the same all over the body, and so access specifically to the artery supplying the adipose depot to be studied is not necessary. In humans, arterial blood is usually taken from the radial or brachial artery, although the femoral artery may be used, if this were to be cannulated for other reasons.

1.3 Derived Measurements of AT Metabolism

Measurement of A-V or V-A differences will give estimates of net uptake or release of substances by the adipose depot studied. These may be informative in their own right, but often further information can be gained by looking at the relationships between these measurements (13). For instance, lactate release may be expressed relative to glucose uptake, to give an estimate of the proportion of glucose uptake that is released as lactate. The ratio of glycerol to NEFA release is another example. From this ratio, an estimate of FA esterification in the adipose depot can be made. Such ratios have the advantage over raw A-V or V-A differences, because they are independent of BF (13). Hence, useful quantitative metabolic information may be derived without measurement of BF.

If some assumptions are made, then further estimates of AT metabolic function can be made. The removal of plasma triacylglycerol (TAG) during passage through the tissue seems primarily to reflect the action of lipoprotein lipase (LPL) in AT capillaries, although there is some uptake of intact particles from the larger TAG-rich lipoprotein fractions (9). Therefore, net TAG removal from plasma may be taken as a measure of the physiological rate of LPL action in AT. This increases after a meal (14), as would be expected from the known activation of adipose tissue LPL at this time (15).

The hydrolytic action of LPL on TAG in vitro leads to the accumulation of 2-monoacylglycerol (MAG) (16), but, in vivo, no net release of MAG from adipose tissue can be detected, even during high rates of TAG hydrolysis (10). The assumption can then be made that glycerol is released mole for mole with hydrolysis of TAG by LPL. Total net glycerol release can be measured. If it is assumed that this arises from the combined actions of LPL in AT capillaries, and hormone-sensitive lipase (HSL) within adipocytes, then subtraction of the rate of action of LPL from total glycerol release will give an estimate of the rate of action in vivo of HSL (14) (see Fig. 2 and Notes 1 – 4 ).

Fig. 2.
figure 2

Assumptions involved in the calculation of the rates of action in vivo of LPL and hormone-sensitive lipase (HSL) in white adipose tissue. LPL acts on circulating triacylglycerol- (TAG)-rich lipoproteins; HSL acts upon TAG stored in adipocytes. Both are assumed to release glycerol mole for mole with TAG hydrolysis. It is assumed that there is no other removal of TAG from plasma (known not to be entirely true [9]). Partial hydrolysis of TAG (generation of mono- or diacylglcyerols) is assumed to be insignificant, for which there is good evidence in plasma (10). Then:

Full size image

2 Materials

2.1 Instrumentation

Little instrumentation is needed for a basic study. However, it is helpful to use a clinical infusion pump to perfuse the catheter with saline (0.9% NaCl) between blood samples, at a rate of approx 35 mL/h. This rate is not critical, and almost any clinical infusion pump will suffice. An electric warming pad of the sort available from a high-street pharmacy is useful. A microhematocrit centrifuge is required for some measurements (see Notes 2 and 4 ).

where TAGa is the concentration of TAG measured in arterial plasma, TAGv that in AT venous plasma; BF is AT BF; H is hematocrit expressed as a fraction (see Note 4 ).

where glycerol concentrations in artery (a) and vein (v) are in whole blood. Rate of HSL action = Net glycerol release Rate of LPL action.

Measurement of AT BF by the most common method, the washout of 133Xe, necessitates an external radioactivity monitoring device. We have used a lightweight detector based on a CsI crystal (Mediscint, Oakfield Instruments, Eynsham, UK) (17). See Chapter 23 for further details on the measurement of AT BF.

2.2 Catheters and Other Clinical Disposables

Ohmeda (Swindon, UK) Secalon Hydrocath 22-gage×10 cm, or similar catheter with introducer and guidewire; sterile gloves; sterile saline; local anesthetic (e.g., lignocaine 1%; [lidocaine in US]); sterile forceps (plastic are satisfactory).

3 Methods

3.1 Obtaining Informed Consent

The procedure must be approved by the local research ethics committee. Obtain informed consent from the subject/patient in advance. One requirement of research ethics committees is that the procedure is fully explained to the subject. In making this explanation, it is reasonable to say that the procedure should not hurt (other than the injection of a small amount of local anesthetic), but that it may require more than one attempt.

3.2 Identification of Site

  1. 1.

    Relax the subject in a warm room (30°C if possible), and cover the abdomen with a warming pad.

  2. 2.

    Inspect the abdomen visually (subdued light is best) for veins draining the sc abdominal region. Identify veins that are straight rather than tortuous, medial rather than lateral, and low on the abdomen rather than supraumbilical.

  3. 3.

    Mark the course of the veins with a water-soluble pen.

3.3 Catheterization

  1. 1.

    Have an assistant standing ready, with the guidewire supplied with the catheter, held in sterile gloves.

  2. 2.

    Introduce a small bleb of intradermal local anesthetic at the proximal end of a straight segment of vein.

  3. 3.

    Using the introducing needle provided with the catheter, advance very slowly through the skin (toward the groin), and move the needle slowly forward, watching carefully for flashback of blood into the needle hub. Flashback will be slow, hence the need for very slow movement of the needle.

  4. 4.

    When flashback is obtained, the assistant should thread the guidewire through the needle (straight end first) until >10 cm is inside. If resistance is felt, catheter-ization may not have been successful, and another attempt may be necessary.

  5. 5.

    If the guidewire threads in without obvious resistance, then the assistant should use sterile forceps to hold the guidewire in place, while the introducing needle is withdrawn over the guidewire. Take care when pulling back over the J-end of the guidewire, that the guidewire is not pulled back.

  6. 6.

    Thread the catheter over the guidewire until it is about to enter the skin; about 1 cm of guidewire should protrude from the catheter hub. Lubricate entry into the skin with drops of sterile saline. While holding the exposed end of the guidewire, thread the catheter through the skin, until as much of its length as will easily go has disappeared under the skin. The guidewire can be withdrawn.

  7. 7.

    Flush the catheter with a small amount (1 mL) of sterile saline, and connect the infusion pump (35 mL/h). Tape the catheter firmly in place. Do not attempt to withdraw blood samples for 10–15 min while potential venous spasm subsides. During this time, return the room temperature to that desired for the study to be performed (e.g., 23°C).

  8. 8.

    Possible complications of siting the venous cannula include bruising and superficial infection. However, these are unusual (<3%) with good technique, and such a cannula can be left in situ without problem for 24 h.

3.4 Taking Blood Samples

  1. 1.

    Disconnect the infusion pump.

  2. 2.

    Connect a small syringe (1 mL is preferred) to the catheter, and gently withdraw the plunger until approximately 100–200 μL mixed blood and saline has entered the syringe, to clear the catheter of saline (dead volume approx 100 μL). Remove and discard this syringe.

  3. 3.

    Connect the syringe in which the sample is to be collected. Small syringes (up to 5 mL) are preferred to larger ones, because the tendency is less to pull too hard and collapse the vein. Withdraw blood samples slowly, over a period of a few minutes, if necessary (see Note 5 ).

  4. 4.

    After collecting the desired volume of sample (see Note 6 ), flush the catheter gently with 5 mL saline, and reconnect the saline infusion.

  5. 5.

    Measure hematocrit, if required (see Notes 2 and 4 ).

3.5 Collecting Arterial Blood

Methods for arterial cannulation are outside the scope of this review. There is a small but significant morbidity associated with arterial puncture (18), and this should only be attempted by a skilled operator (see Note 7 ).

3.6 Analytical Techniques

Details of analytical technique are outside the scope of this review, but there are two major points to be borne in mind.

  1. 1.

    Some substances may be labile. This applies, for instance, to plasma TAG, which hydrolyzes on standing (presumably through the action of LPL in plasma), so that the plasma TAG concentration will decrease, but the plasma NEFA concentration will increase. For this reason, draw blood samples into chilled syringes, keep on ice, and separate plasma as rapidly as possible (see Note 5 ).

  2. 2.

    When measuring A-V differences, analytical precision may be crucial. If both arterial and venous concentrations are measured with a standard deviation (SD) of 100 μmol/L (as an example), then the SD of the A-V difference will be √2×100 μmol/L or 141 μmol/L. Because the A-V difference may be small, this SD may render its precise measurement impossible. Potential solutions are to take multiple samples, or to make replicate analyses to improve precision.

4 Notes

  1. 1.

    Units of measurement. If concentrations are measured in μmol/L, and BF in mL blood 100 g tissue/min, then the product of A-V difference and BF gives the net rate of uptake or release in nmol/100 g tissue/min. Removal of substances may be expressed simply as fractional extraction (= A-V difference/arterial concentration), as absolute extraction (= A-V difference×BF) or as clearance (= fractional extraction×BF). If units for A-V difference and BF are as above, units will be: fractional extraction, dimensionless (often expressed as percentage); absolute extraction, nmol/100 g tissue/min; clearance, mL/100 g tissue/min.

  2. 2.

    For substrates that are essentially confined to the plasma compartment (e.g., NEFA, lipids), absolute exchange rates are obtained by multiplication of the plasma A-V difference by the rate of plasma flow. The latter is calculated from the BF by multiplication by (1—hematocrit). For substances carried partly in plasma and partly in red blood cells, it is best to measure total concentrations in whole blood, e.g., after deproteinization with perchloric acid; misleading results from A-V difference studies have been reported when this is not done (19). For AT glucose uptake, there may be special considerations. Glucose uptake is small in comparison with arterial concentration (typical fractional extraction is 3–4%; [20]). Accurate measurement of the A-V difference demands high analytical precision. We have found that glucose A-V differences measured on whole blood are less precise than those measured in plasma (probably because additional dilution factors are involved), and measuring glucose in plasma is preferred. Conversion to whole-blood concentrations is then necessary to estimate net uptake, and should be done using the formula proposed by Dillon (21): B=P (1−0.3 H), where B is glucose concentration in whole blood, P concentration in plasma, and H is hematocrit expressed as a fraction.

  3. 3.

    The requirement for steady-state conditions has been discussed in the literature (22,23). It is most easily seen if the arterial concentration of a substance is rising. Then, if simultaneous samples are taken from artery and vein, the venous sample will match the arterial concentration at some earlier time, depending on the time for transit of blood through the tissue. In addition, small molecules will equilibrate throughout the interstitial space, so, as the arterial concentration rises, there will be apparent uptake of substance, which does not represent true metabolic utilization. If the arterial concentration falls, the reverse will be true. One way of mostly overcoming nonsteady-state problems is to analyze the areas-under-curves for arterial and venous concentrations over reasonably long periods of time, rather than concentrations at individual time-points. For metabolic responses following a meal, for instance, this might be the whole postprandial period (e.g., 6 h), during which concentrations of many substances will rise, then fall again.

  4. 4.

    If concentrations of substances are to be compared, only some of which are confined to the plasma compartment (e.g., NEFA, TAG, and glycerol, when glyc-erol is distributed throughout whole blood), then it is necessary to convert all concentrations to those in whole blood. For NEFA and TAG, this is done as follows: B=P(1−H), where B is glucose concentration in whole blood, P concentration in plasma, and H is hematocrit expressed as a fraction. Measurement of hematocrit in AT venous blood is not required, because it is not measurably different from that in arterial blood (9).

  5. 5.

    Blood flow from the catheter may be slow. Samples of up to 10 mL may be collected from a freely flowing vein, but may require several minutes for withdrawal. A smaller vein, or one that has a tendency to spasm, may yield smaller samples. If the substances to be measured in the blood are labile, consider using a series of smaller syringes, and keep them on ice when not connected to the catheter.

  6. 6.

    If the sample volume desired cannot be collected, the following tips may help. Use finger pressure on the skin near the estimated catheter tip position to attempt to change the geometry of the vein-catheter relationship: This may allow blood to flow again. Ask the subject to move the leg (e.g., flex the hip), or even to cough, for the same reasons. If this fails, but the catheter still appears to be in the vein, the vein may be in spasm, and it is best to wait 60 min or so, with the warming pad in place over the abdomen, and the subject relaxed, to see if the situation has improved. As a last resort, a new catheterization may have to be made.

  7. 7.

    The potential problems associated with arterial cannulation have led some to use arterialized venous blood from a vein draining a warmed hand, as an alternative. Concentrations of many substances are indistinguishable in arterial and arterial-ized venous blood (24). However, this is not true for CO2 (25), and may not be for catecholamines (26); it should be verified for each substance to be measured in this way. The degree of arterialization may vary according the position of the catheter, the method of warming the hand, the state of the subject, and the ambient temperature, and it should be checked by measurement of blood O2 saturation. The best method in our experience is to cannulate the cephalic vein in a retrograde direction, so that the tip of the cannula lies above or distal to the wrist joint, then place the hand in a box with still air warmed to 60°C. Keep the cannula patent between samples, by slow infusion (50 mL/h) of saline.