Nuclear cardiology involves the use of radionuclides, which emit electromagnetic radiation. The radiation is in the form of gamma or x-rays. The quantum particle of this radiation is called a photon, which becomes visible by photoelectric absorption and Compton scatter, causing light flashes (or scintillations).

These photons are detected by a gamma (scintillation) camera. The basic camera is called the Anger camera, which consist of a single sodium iodide crystal, over which there is a flat piece of lead with numerous holes (a collimator).

This collimator allows photons emanating from the patient to strike the crystal. This collimator absorbs scattered photons and those coming from outside the field of view. The photons passing through the holes interact with the crystal to produce scintillations (flashes), which are detected by photomultiplier tubes.

These tubes convert the scintillations into an electrical signal, which are fed into an electronic processor to determine the position of scintillation from the outputs of the photomultiplier tubes. A pulse analyzer allows only scintillations with a preset range to pass through (thus, within the photopeaks of the particular radionuclide used such as 99mTc (technetium- 99m) and 201TL (thallium- 201).

The photomultiplier tubes are directly connected to an imaging computer, which allows acquisition and display of multiple-framed gated blood pool imaging studies to assess ventricular function, ejection fraction (how much blood is ejected), optimizing contrast and intensity of images, quantitation of distribution and redistribution of radionuclides in myocardial perfusion imaging studies and tomographic acquisition and display (known as single-photon-emission computed tomography (or SPECT).

The SPECT camera, with one to three heads, rotates around the patient to acquire the images, using 201Tl or 99mTc teboroxime.

Reference: Johnson, L.L., MD, Pohost, G.M., MD, Nuclear Cardiology, Hurst's The Heart, 8th edition, p 2281-2295.

Thallium-201:

Thallium-201 is a cyclotron-produced radionuclide with a half life of 73 hours. It is used as a myocardial-imaging agent because of its primarily intracellular distribution, similar to ionic potassium (K+).

Reference: Johnson, L.L., MD, Pohost, G.M., MD, Nuclear Cardiology, Hurst's The Heart, 8th edition, p 2281-2295.

Initial distribution of 201Tl:

After injection of a bolus of 201Tl, the heart blood pool activity rises rapidly to a peak and falls more gradually as the thallium is extracted by the tissues of the body.

Afterwards, the blood pool activity declines slowly due to excretion. Following injection of an intravenous bolus of thallium, about 5 percent of the dose distributes to the myocardium, which extracts 80 to 90 percent of the thallium as it passes through the coronary circulation (see fig.104f, fig.104g, fig.104h).

If thallium is given at rest, the time to peak myocardial thallium activity is delayed. But at maximal exercise (as in nuclear stress testing) or with vasodilators (like dipridamole or adenosine used in nuclear stress testing), the time to peak activity is early.

Peak myocardial thallium activity distal to a coronary stenosis will be delayed relative to myocardium supplied by a nonstenotic coronary artery.

Regions supplied by normal (see fig.109g, fig.109d) and abnormal stenotic arteries account for the defect(s) observed on myocardial images (see fig.109c, fig.109h, fig.109i, fig.109k).

Myocardial extraction of thallium is related to blood flow. The longer the exposure, the more effective the extraction and conversely. At very high flow rates (like those induced by dipridamole or adenosine) extraction may decrease. The extraction occurs only if the myocardium is viable.

Reference: Johnson, L.L., MD, Pohost, G.M., MD, Nuclear Cardiology, Hurst's The Heart, 8th edition, p 2281-2295.

Redistribution of 201Thallium:

There is an increase in myocardial thallium to peak and then a slow clearance. But in myocardial zones jeopardized by important stenosis, peak activity is delayed and the concentration of thallium is lower compared to nonjeopardized myocardium (see fig.109c, fig.109h, fig.109i, fig.109k).

Complete thallium redistribution can occur only in the absence of myocardial scar. But the persistent thallium defect is not specific for nonviable myocardium.

Up to 50 percent of viable myocardial segments will demonstrate persistent defects with the conventional 3-to 4- hour followup imaging interval. The basis for this appears to be more rapid clearance of thallium from blood pool with less residual thallium available for re-uptake.

There are two methods to get thallium into these persistent defects in viable myocardium.

One is to get follow-up images at 24 to 48 hours after the initial injection. Another method is to give a second injection of thallium to boost the blood pool activity 3 to 4 hours after the initial injection (see fig.109i, fig.109k).

Reference:Johnson, L.L., MD, Pohost, G.M., MD, Nuclear Cardiology, Hurst's The Heart, 8th edition, p 2281-2295.

Clinical Approaches to Myocardial Perfusion Imaging (Nuclear stress testing):

Both thallium and the technetium based agents can be administered during exercise stress, infusion of the vasodilators like dipyridamole or adenosine, or in unstable patients at rest.

When the patient has achieved his/her highest level of exercise, a bolus of 201Th, 99mTc-MIBI or 99mTc-teboroxime is given intravenously.

The patient is encouraged to continue exercise for an additional minute. The EKG, blood pressure and pulse rate are continuosly monitored throughout the test.

FALSE POSITIVE TESTS

False positive tests do occur for a variety of reasons. When using thallium-201 as the radioactive tracer, the timing of the aquisition of the the pictures poststress is important, since excessive delay can result in decreased sensitivity for the detection of coronary disease, owing to to early redistribution of the thallium. But the aquisition of Thallium-201 or Technetium-MIBI based myocardial agents should begin in 10 minutes following exercise injection because of the frequent observation of an artifactual perfusion defect due the "upward creep of the heart". This phenomenon is related to the increased depth of respiration very early postexercise, which is associated with an average lower position of the diaphram in the chest compared with normal ventilatory state. This causes the heart to gradually move cephalad during the early part of the camera aquisition of data for pictures, resulting in a form of motion artifact after reconstruction. By delaying acquisition until 10-15 min. after stress, this 'upward creep" artifact is avoided.
Several cameras provide hardware and software implementation of attentuation correction protocols, but they are imperfect reducing but not eliminating the apparent perfusion defects due to soft tissue attentuation in normal patients. At times, true perfusion defects might be obscured or eliminated by the application of these approaches.

Reference:Johnson, L.L., MD, Pohost, G.M., MD, Nuclear Cardiology, Hurst's The Heart, 8th edition, p 2281-2295.

POSITRON EMISSION TOMOGRAPHY FOR THE NONINVASIVE STUDY AND QUANTIFICATION OF BLOOD FLOW AND METABOLISM IN HUMANCARDIAC DISEASE

Heinrich R. Schelbert

TOOLS FOR PROBING -----------------------Tracer Kinetic Principles / 630--------------- CLINICAL APPLICATIONS
MYOCARDIAL TISSUE FUNCTION 629------Positron-Emitting Tracers of Myocar ------- Identification --------------Imaging withPositron-Emitting Radiopharaceuticals-/629--------------------------ial Tissue Function/630 and charcterization of CAD / 632
-------------------------------------------------------------------------------------------------------Assessment ofMyocardiial viability/637

The study of the human heart with conventional radionuclide techniques remains confined to primarily ventricular function and relative distributions of regional myocardial blood flow (MBF). Positron emission tomography (PET) exceeds these capabilities. It offers the probing and defining regional functional processes in absolute units in the human heart spanning from MBF to biochemical reaction rates, substrate fluxes, and neuronal activity. The many positron-emitting, biologically active tracers, the quantitative imaging capability, and the in vivo application of tracer kinetic principles are unique to PET and account for this capability. The human heart's physiology and pathophysiology thus can be characterized more comprehensively. Also, novel insights into the function of the human heart can be gained, while, at the same time, PET can have a decisive impact on patient diagnosis and management. This chapter describes the key ingredients of PET and the tools for the evaluation and/or quantification of local functional processes in the human heart. It then examines how these tools can be applied to the diagnosis and characterization of coronary artery disease (CAD) and its consequences on regional myocardial function and discusses the impact of PET findings on patient management.

TOOLS FOR PROBING MYOCARDIAL TISSUE FUNCTION

Fundamental to the uniqueness of PET are
(1) the quantitative imaging and high temporal resolution capability,
(2) the in vivo application of tracer kinetic principles, and
(3) the large number of physiologically active radiotracers.

Imaging with Positron-Emitting Radiopharmaceuticats

DEDICATED PET SYSTEMS

The quantitative imaging capability results from the physical properties unique to positrons. After losing their kinetic energy, they combine with an electron and "annihilate." The annihilation represents a conversion of mass into energy; i.e., the combined mass of the positron and the electron converts into two 511-keV photons that leave the site of the annihilation in diametrically opposed directions. If both strike two scintillation detectors connected by a coincidence circuitry at the same time, an annihilation event is registered. Its location in space can be defined by circular arrays of scintillation detectors. The near simultaneous arrival of two 511-keV photons at the two scintillation detectors positioned in opposite directions allows the use of tomographic reconstruction algorithms analogous to those used with x-ray computed tomography. Accordingly, the spatial resolution throughout the image plane is rather homogeneous, unlike that obtainable with single-photon-emission computed tomography (SPECT), where the spatial resolution declines as a function of the distance between the imaged object and the scintillation detectors. Further, by acquiring "transmission" images with external rotating or circular sources of positronemitting isotopes, the images of the tracer tissue concentrations ("emission" images) can be corrected for photon attenuation so that the resulting tomographic images represent accurately the true regional radioactivity concentrations (mCi or MBq/ cm3). Current PET systems offer spatial resolutions of as high as 4 to 5 mm full-width half-maximum (FWHM). Further, because modern tomographs are stationary circular devices, images can be acquired at sampling rates in the range of seconds. It is therefore possible with PET to rapidly measure changing radiotracer concentrations in tissues.

COMBINED SPECT AND PET

Several institutions use SPECT and PET for the evaluation of cardiovascular disease. For example, the distribution of MBF is determined with 201T1- or 99°Tc-labeled tracers of MBF and SPECT and then compared with the distribution of myocardial glucose use by imaging [1RF]deoxyglucose with dedicated PET systems. While this approach yields a diagnostic accuracy comparable with that achieved with dedicated PET systems, diag-nostic difficulties can arise due to differences in the geometry of the heart and spatial resolution on SPECT and PET images, as well as artifacts on the SPECT images due to photon attenuation, especially of the inferior wall and the interventricular septum .

MULTIPURPOSE IMAGING SYSTEMS

Most clinical applications of PET do not require assessment of functional processes in absolute units. This is particularly true for the identification of myocardial viability and, to some extent, for the detection of CAD. In view of the high cost of dedicated PET systems, lower-cost "hybrid" or "multipurpose" imaging systems are now available. Generally, two types of systems have emerged (Fig. 216-1). One is a SPECT-like system equipped with ultra-high-energy general-purpose collimators to accommodate the 511-keV photons of positrons instead of the 70- to 160keV photon energies of conventional radiotracers. When used together with 201TI- or 9'mTc-labeled flow tracers, the ['SF]deoxyglucose images provide diagnostic information comparable with that available with dedicated PET systems. The second type of system entails a SPECT-like dual-head device with coincidence detection. While highly promising because of its superior spatial resolution, its use is still evolving, and clinical studies are scarce.Initial studies demonstrate a spatial and contrast resolution that is superior to the high-energy-collimator SPECT system and approaches that of dedicated PET systems. Critical, however, is appropriate correction for photon attenuation that now seems feasible with 511-keV photon sources. Both camera types also can be used for conventional single-photon-emitting tracers, hence the name hybrid or multipurpose systems.

Tracer Kinetic Principles

Positron-emitting isotopes of elements that constitute major parts of living matter such as carbon-11 ("C), nitrogen-13 (13N), and oxygen-15 (150) are inserted into biomolecules without disturbing their very physiologic properties. Their high specific activity (radioactivity per mass) permits administration of true tracer quantities without exerting a mass effect and perturbing the very process to be studied. Since their physical half-life is short, functional processes can be measured repeatedly or different aspects of the myocardial tissue function can be explored within the same study session. The radioactivity concentrations of these tracers in tissues such as arterial blood and myocardium and their changes over time can be determined noninvasively. The time-activity curves derived from serially acquired tomographic images at sampling rates of 1 to 10 s are fitted with operational equations that are derived from tracer kinetic models and yield quantitative estimates of regional functional processes. Tracer compartment models describe the distribution of the tracer radiolabel in tissue and its time-dependent changes. Because only the activity concentration of the tracer radiolabel can be measured externally, these models relate the externally derived signal to the metabolic fate of the tracer label and its relationship to the functional process under study. Such tracer kinetic models typically consist of functional rather than anatomic pools or compartments that contain the radiotracer or its metabolites. Exchange of radiotracers between compartments is described typically by first-order rate constants. Flux of a radiotracer through a given compartment depends on the flux rate of tracer or of its metabolite and on the size of the compartment. Tracer compartment models provide the basis for developing operational equations; applied to the externally derived radioactivity signal as, for example, tissue time-activity curves, estimates of regional functional processes are derived in absolute units.

Positron-Emitting Tracers of Myocardial Tissue Function

BLOOD VOLUME AND TISSUE
CHARACTERIZATION

Blood can readily be radiolabeled with minute quantities of 150 or 11C carbon monoxide (CO). Once inhaled, the radiolabeled CO binds to hemoglobin, thereby tagging red blood cells. The latter serve to define the components of the myocardium in terms of vascular space, viable and normal myocytes, and scar tissue. One such characterization assumes that only living myocytes exchange water rapidly.Transmission images represent the densities of the various tissues in the chest. They resemble low-spatial-resolution x-ray computed tomographic (CT) images and delineate the volume of the myocardium together with the blood in its cavities. The true extravascular volume is obtained by subtracting blood pool images from the transmission images. The fraction of the extravascular volume that exchanges water rapidly is then estimated with 15O-labeled water and is referred to as the water perfusable tissue index (PTI). If all the extravascular volume does indeed rapidly exchange water, then the PTI approaches unity." If a portion of the myocardium is injured irreversibly and scar tissue has formed, this fraction becomes less than unity . Further, the PTI also will be reduced in diffuse interstitial fibrosis. Initial clinical investigations demonstrated that the fraction of irreversibly injured myocardium or of regional scar tissue formation does indeed indicate whether an impairment in contractile function is irreversible or whether a postrevascularization improvement is likely. If functionally compromised but viable myocardium exchanges water as rapidly as normal myocardium, this may limit the predictive value of the PTI. In recent observations, areduced PTI had a high negative predictive value, but a nearnormal PTI predicted less accurately than [1&F]deoxyglucose an improvement in contractile dysfunction. Moreover, the sum of viable and normal myocytes in a given myocardial segment also serves as a reference to which transmural estimates of MBF or substrate metabolism can be related.

MYOCARDIAL BLOOD FLOW

Several approaches exist for measurements of regional MBF in absolute units. Tracers such as 82Rb or [13N]ammonia are retained in myocardium in proportion to MBF. Images of their regional activity concentrations in the myocardium depict the relative distribution of MBF at the time of tracer injection. Each tracer offers advantages and disadvantages. For example, 8Rb is available through a generator based pushbuttonoperated infusion system and hence is easy to use clinically. Its physical half-life of only 75 s affords repeat studies at only 10-min time intervals, enabling evaluation of changes in regional MBF in response to physiologic or pharmacologic interventions. The short physical half-life, however, can result in low-count and, thus, statistically noisy images. The longer physical halflife of [13N]ammonia (10 min), by contrast, produces images of higher count rates and higher diagnostic quality but requires 40- to 50-min time intervals between studies.

The various approaches yield comparable estimates of MBF in the human myocardium during rest and during pharmacologically induced hyperemia Some variability between studies probably derives from methodologic differences but also from intergroup differences in the hemodynamic state. Importantly, MBF in the normal myocardium depends largely on oxygen demand and thus on cardiac work as estimated from the rate pressure product. Thus individual flow measurements should be interpreted within the context of the rate-pressure product.Finally, gender-related differences in MBF have been reported. Compared with an age-matched group of males, women demonstrated higher MBFs both at rest and during hyperemia, which the authors attributed to higher high-density lipoprotein (HDL) cholesterol and lower triglyceride plasma levels in females.

Repeat studies in the same normal volunteers report a 10 ± 11 percent reproducibility (average percentage difference of flows normalized to the rate pressure product) for rest MBF and a 12 ± 9 percent reproducibility for hyperemic MBFs Other studies report similar values for both, [13N]ammonia and for [11O]water Furthermore, the validity of the noninvasive measurements of MBF has been extensively established in ani
mal experiments as well as in humans, using intracoronary flow velocity probes .

MYOCARDIAL SUBSTRATE METABOLISM

Figure 216-2 depicts the major aspects of the myocardial substrate metabolism. According to this simplified depiction, the myocardium chooses between various substrates; foremost are free fatty acid (FFA), glucose, lactate, and ketone bodies. Selection of a given fuel substrate depends largely on its concentration in plasma and the overall hormonal milieu. These in turn are governed by the dietary state, the level of physical activity, and the plasma concentrations of catecholamines, insulin, and glucagon. In the fasting state, circulating FFA levels are high and insulin levels are low so that as much as 70 to 80 percent of the myocardium's oxygen consumption can be accounted for by oxidation of FFA. Conversely, oral glucoseby invasive techniques, where frequently the predominantly vascular smooth muscle-mediated vasodilator response to, for example, intracoronary papavarine or adenosine was preserved despite the presence of coronary risk factors, while primarily endothelial-mediated responses were markedly abnormal.As Bache suggests, however, the major resistance to flow through the coronary circulation resides at vessels in the diameter range of 100 to 400 aitm. If increases in flow exert shear stresses on the endothelium of the 400-sm vessels, then primarily endothelialdependent mechanisms augment the flow response to predominantly vascular smooth muscle vasodilators. Conversely, as forearm blood flow measurements have shown, pharmacologic impairments of endothelial function reduce the maximal flow response to intravascular adenosine by about 25 to 35 percent.112 Consequently, coronary risk factors such as low-density lipoprotein (LDL) cholesterol, triglycerides, and diabetes interfere with endothelial function and account for the diminished hyperemic response to adenosine or dipyridamole.

Another potentially important new concept for identifying diffuse coronary artery narrowing without discrete stenoses has been introduced recently. Diffuse luminal narrowing causes a greater decline in pressure along the coronary artery. This then is associated with a progressive decrease in myocardial perfusion from the proximal to the distal portion of the coronary arterial system and hence in a longitudinal base-to-apex perfusion gradient. A similar longitudinal perfusion gradient likely exists during hyperemia in patients without CAD but with coronary risk factors.

The predominantly endothelial-mediated coronary vasomotion can be assessed by the cold pressor test. In invasive studies, cold pressor testing evoked paradoxical changes in the diameter of the conduit vessels comparable with those evoked by intracoronary acetyicholine and, further, alterations at the level of the resistance vessels that correlated with those produced by intracoronary acetylcholine. Observations with PET-based measurements of MBF in long-term smokers support the use of the cold pressor test for this purpose.In these young smokers, vascular smooth muscle-mediated hyperemic flow responses to intravenous dipyridamole were normal, but cold pressor testing produced only modest, nonsignificant increases in MBF as compared with nonsmokers, where MBF rose in direct proportion to the cold-induced increase in cardiac work, as estimated by the rate-pressure product (Fig. 216-7). Similarly, abnormal flow responses to cold also were noted in postmenopausal women with and without CAD risk factors.

MONITORING RESPONSES TO RISK FACTOR MODIFICATION

Given their high degree of reproducibility, PET-based measurements of MBF are equally suited for monitoring responses in coronary vasomotion to pharmacologic interventions and risk factor modification. Early studies in only 13 participants in a 6-week cardiovascular conditioning program demonstrated this capability.'2' Dietary changes, regular exercise, and lifestyle modifications were associated with weight loss, decreases in heart rate and blood pressure at rest, and significant decreases in plasma total and LDL cholesterol. A 12 percent decline in MBF at rest was proportionate to the decrease in resting cardiac work. Cardiovascular conditioning also produced a 9 percent increase in intake elevates the plasma glucose level and thus insulin levels while lowering FFA levels so that myocardium shifts its fuel selection to glucose." Strenuous physical exercise increases plasma levels of lactate, which then becomes the major fuel substrate .4142 In fact, as much as 60 percent of the 02 consumption can be accounted for by oxidation of lactate. On the other hand, catecholamines accelerate lipolysis so that circulating FFA levels increase, shifting the heart's substrate selection to FFA.

Glucose enters the cell via facilitated transport systems, the largely insulin-independent glucose transporter GLUTI and the largely insulin-dependent glucose transporter GLUT4. The hexokinase reaction phosphorylates glucose to glucose-6-phosphate. This compound then may be synthesized to glycogen or, alternatively, enter glycolysis with pyruvate as its end product. Converted to lactate, it may leave the myocardium or, if activated to acetyl-CoA, enters the tricarboxylic acid (TCA) cycle as the final oxidative pathway shared by most fuel substrates. Exogenous lactate can be converted via NAD+ to pyruvate, which then again after esterification to acyl-CoA enters the TCA cycle. FFA also may enter two different metabolic pathways. On entering the cells, it is esterified by the thiokinase reaction to acetyl-CoA. This compound then enters an endogenous lipid pool, consisting mostly of glycerides and phospholipids, and/or proceeds via the carnitine shuttle to the inner mitochondrial membrane. It is there where (3-oxidation cleaves of the long-chain acyl-CoA units' two-carbon fragments, which then engage in the TCA cycle. The TCA cycle metabolizes the two-carbon units into CO2 and H2O. The rate of flux through the TCA cycle is coupled closely with oxidative phosphorylation, where the energy resulting from the synthesis of oxygen and hydrogen ions is stored in the high-energy phosphate bonds of adenosine triphosphate (ATP). The latter is shuttled into the cytosol with transfer of energy to the high-energy phosphate bond of creatine phosphate. Other sites of high-energy production include glycolysis. The energy yields relative to oxygen differ between the various substrates; e.g., for 1 mol of oxygen, glucose yields 6.3, lactate 6, and FFA 5.7 mol ATP."

Myocardial Glucose Utilization

The initial metabolic step of exogenous glucose metabolism can be evaluated and quantitated with [18F]deoxyglucose (see Fig. 216-2). This radiolabeled glucose analog exchanges across the capillary and sarcolemmal membranes in proportion to glucose with which it then competes for hexokinase for phosphorylation to ['"F]deoxyglucose6-phosphate .44.45 The phosphorylated glucose analog is a poor substrate for glycogen formation, glycolysis, and the fructosepentose shunt; its rate of dephosphorylation is low in myocardium, and it is relatively impermeable to the cell membrane. The phosphorylated tracer thus becomes trapped in the cell so that images of the myocardial `F concentrations at 40 to 60 min after tracer injection reflect the relative distribution of glucose utilization rates. Because the compound traces only the initial steps of glucose utilization (up to the branch point between glycogen synthesis and glycolysis; see Fig. 216-2), it offers no direct information on glycolytic rates, glucose oxidation, or glycogen synthesis. Yet, in states of glycogen depletion, such as, for example, during ischemia, exogenous glucose serves as the major source of glycolytic flux so that [18F]deoxyglucose may offer an estimate of the rate of glycolysis.

Myocardial Fatty Acid Metabolism

This aspect of the substrate metabolism can be evaluated with 1-["C]palmitate. The labeled long-chain FFA participates fully in the metabolic fate of its natural counterpart (see Fig. 216-2). Once esterified to acyl-CoA, a fraction of tracer label proceeds via the carnitine shuttle into mitochondria, where /3-oxidation catabolizes the long-chain fatty acid into two-carbon fragments that are oxidized via the TCA cycle. The label is released from the myocardium in the form of "CO2. The remaining fraction of the initially extracted and activated tracer enters intracellular lipid pools, mostly those of di- and triglycerides and phospholipids. The biexponential morphology of the myocardial time-activity curve reflects the metabolic fate of the tracer. The slow turnover rate of the intracellular lipid pools accounts for the slow clearance phase, whereas the rapid clearance curve component corresponds to the fraction of tracer that enters /3-oxidation and its rate of oxidation. Ischemia reduces the rate of FFA oxidation and of TCA cycle activity. The relative size and rate of the rapid clearance curve component on the "C myocardial timeactivity curve typically decline during acute myocardial ischemia46'47 A disproportionately greater fraction of tracer label then enters the slower-turnover endogenous lipid pool. Used mostly as a tracer for the qualitative evaluation of regional myocardial fatty acid metabolism, recent studies suggest the possibility of quantitating myocardial fatty acid oxidation in milliequivalents of FFA per gram of myocardium per minute."

Preferential use of a given fuel substrate (e.g., glucose, lactate, or FFA) depends on its concentration in arterial blood, which, in turn, depends on dietary state, serum levels or insulin resistance, and physical stress. A change in the myocardium's preferential substrate use can be demonstrated with either [C]palmitate and [18F]deoxyglucose or both. In the presence of high FFA and low glucose and insulin levels, use of FFA as the preferred substrate is reflected on the [C]palmitate curve by the large relative size of the rapid clearance phase and its steep slope (both corresponding to increased fatty acid oxidation) and the low or even undetectable [18F]deoxyglucose uptake. Ingestion of carbohydrates raises plasma glucose levels, stimulates insulin secretion, and depresses FFA levels. The shift to glucose use is reflected by a decline in the size and slope of the rapid-clearance phase of ["C]palmitate and by an increase in myocardial [18F]deoxyglucose uptake.

Myocardial Oxygen Consumption (MVO2)

While molecular 0 oxygen is available for measurements of the (MVO2, the
more widely applied approach entails rapid serial imaging with ["C] acetate. The radiotracer clears rapidly from blood into the myocardium and produces high signal-to-background images It directly traces the rate of substrate flux through the TCA cycle as the final oxidative pathway common to most fuel substrates. The rate of clearance of "C activity from the myocardium on serially acquired images corresponds to the TCA cycle activity and, because of its close coupling to oxidative phosphorylation, to oxidative metabolism and MVO,. Of note, the tracer yields rate constants only, which can be converted into units of 02 per minute per gram. Unlike ["C]palmitate or ['"F]deoxyglucose, the clearance rate of ["C]acetate from myocardium is relatively insensitive to changes in myocardial substrate utilization.'' A tracer compartment model, based on biochemical assays of the tracer tissue kinetics of [14C]acetate in isolated rat hearts=" forms the base for estimating MVO2 in absolute units in the human heart and, at the same time, of regional MBFs.

CLINICAL APPLICATIONS

Clinical imaging of positron-emitting radionuclides with PET or PET-like devices is gaining momentum because of research showing the considerable potential for contributing to diagnosis, characterization, treatment, and monitoring of disease. New, lower-cost positron imaging devices and the availability of positron-emitting ;acers through regional distribution centers have accelerated the pace of dissemination. Foremost in cardiology have been (1) the identification and characterization of CAD and (2) the detection of myocardial viability.

Identification and Characterization of CAD

GENERAL CONSIDERATIONS

Most studies with PET, such as, for example, those performed with [13N]ammonia or "Rb, evaluate the relative distribution of MBF from the retention of tracer in the myocardium. More recent investigations use PET's quantitative capability for estimating regional MBF in milliliters of blood per minute per gram of myocardium in order to demonstrate abnormalities in vasomotion of the human coronary circulation during the early stages of coronary atherosclerosis.

Unlike other radionuclide approaches, PET employs almost exclusively pharmacologic stress for the detection of CAD and determination of its extent and functional significance. The transmission images, essential for correction of photon attenuation, must be acquired with the patient in exactly the same position as during the emission images. Both dipyridamole and adenosine afford the determination of the myocardial flow reserve as the ratio of hyperemic to rest MBFs. The now classic studies by Gould et al.demonstrated a curvilinear, inverse correlation between stenosis severity and hyperemic flows or flow reserve. Thus the attenuation of the MBF response to dipyridamole induced hyperemia depends on the functional stenosis severity. As demonstrated by flow measurements with either [15O]water or [13N]ammonia, dipyridamole and adenosine as direct vascular smooth muscle dilators evoke interindividually variable hyperemic responses but induce on average four
to fivefold increases in MBF .The magnitude of the hyperemic flow response is similar for dipyridamole (at a dose of 0.56 mg/kg over 4 min) and for adenosine (140 µg/kg/min) . Increases in the dipyridamole dose by 50 percent do not produce higher flows, nor do they reduce the interpatient variability in flow responses. Additionally, the values of the normal flow reserve were derived from studies in young normal volunteers with an average age of 34 ± 16 years. This is important as evidence accumulates that flow reserve declines progressively with age (Fig. 216-3). Contributing factors include an agedependent decline in the vasodilator capacity and an agedependent increase in baseline MBF due to higher rate-pressure products as a major determinant of MBF. A progressive decline in vascular compliance is another possible explanation. Surprisingly, increases in the mean arterial blood pressure due to either isometric handgrip exercise or supine bicycle exercise attenuated the maximum flow response, most likely because of increased vascular resistance due to greater extravascular resistive forces . These factors also may contribute to lesser flow increases during physical exercise when flow increases in proportion to MVO2. Thus pharmacologically induced hyperemia may not necessarily prove to be more accurate in identifying functionally significant coronary stenoses. Even though flows in remote myocardium may rise less with exercise depending on the level of cardiac work, higher intracavitary left ventricular (LV) pressures and regional wall stresses in ischemic or dysfunctional myocardium may enhance extravascular resistive forces so that flow responses in stenosis-dependent myocardium in fact may be even more attenuated. Because of differences between pharmacologically and physically stressed-induced ischemia, the vasodilator reserve as determined pharmacologically may not necessarily reflect truly the myocardium's ability to raise flow during physical exercise. An example is hypertrophic cardiomyopathy, where MBF during exercise failed to increase despite some residual flow reserve demonstrated with dipyridamole.

Another important consideration with regard to pharmacologic stress is the variability of the hyperemic response. In normal individuals, responses range from about two- to sixfold increases in MBF. Several factors may account for this variability. Among these are (1) the coronary driving pressure, best reflected by the mean arterial blood pressure, (2) extravascular resistive forces as a function of wall tension and tension development, which in turn depend on the diastolic volume and the myocardium's contractile state, (3) /3- and especially a-adrenergic control of the basic vasomotor tone,(4) endothelial-dependent vasomotion, and (5) pharmacologic effects on vascular smooth muscle relaxation. The latter may be altered by antagonists of dipyridamole and adenosine, such as, for example, caffeine or theophylline-containing agents. It thus is imperative that patients refrain from these substances for at least 24 h prior to a pharmacologic stress study.

Positive inotropic agents also are used for stress interventions. Dobutamine raises MBF in proportion to increases in cardiac work, as evidenced by increases in the rate pressure product. In one study, intravenous infusion of dobutamine in normal volunteers at a rate of 40 µg/kg of body weight per minute increased the rate-pressure product by about 200 percent, which was paralleled by a 225 percent increase in MBF.71 Lower infusion rates produced lesser increases in the ratepressure product and thus in MBF.

ASSESSMENT OF HEMODYNAMICALLY SIGNIFICANT CAD

For the detection of CAD, the relative distribution of MBF is examined at rest and during pharmacologic vasodilation. Either 82Rb or [13N]ammonia is used. Both are retained in myocardium in proportion to MBF so that the resulting images depict the distribution of MBF at rest and during hyperemia. The approach identifies flow defects at rest as well as attenuated responses of regional MBF to hyperemia as a consequence of a coronary stenosis (Fig. 216-4). The baseline and hyperemia flow images are analyzed by visual inspection combined with circumferential activity profile techniques or polar map approaches. While most studies rely on visual analysis, several laboratories employ quantitative image analysis. The regional tracer activity concentrations in a patient are compared with databases of normal displayed graphically in various cartographic forms, such as polar (or azimuthal) or cylindrical (Mercator-like) projections or surface rendered three-dimensional displays of the LV myocardium.

Clinical investigations confirmed PET's high diagnostic performance for the detection of CAD. Sensitivities range from 87 to 97 percent; and specificities from 78 to 100 percent. Most studies compared rest or stress-induced flow defects to arteriographic findings by visual analysis, and most defined a 50 to 70 percent diameter luminal narrowing as significant stenosis. Given the well-known limitation of visual analysis, Gould et al.' and, subsequently, Demer et al . graded stenosis severity by estimates of coronary flow reserve by quantitative arteriography. Coronary arteries were classified as moderately to severely stenosed if the predicted coronary flow reserve was less than 3, as intermediate if the coronary flow reserve ranged from 3 to 4, and as minimal for coronary flow reserve values of greater (FIGURE 216-4) than 4. According to this classification, 94 percent of vessels with moderate to severe, 49 percent of vessels with intermediate, and 5 percent of vessels with minimal stenosis were accurately identified with PET and pharmacologic vasodilator stress.

COMPARISON OF PET WITH CONVENTIONAL TECHNIQUES

The diagnostic accuracy of PET must be directly compared with that of more conventional approaches in order to define the diagnostic gain. Demer et al. indirectly compared their findings with those by another laboratory using 201T1 SPECT but an identical angiographic approach for defining stenosis severity. In this comparison, PET outperformed SPECT. Both studies defined stenosis severity by the angiographically predicted coronary flow reserve. Moderate to severe coronary stenoses were detected with a 95 percent sensitivity by PET and a 72 percent sensitivity by 201T1 SPECT; intermediate stenoses were detected with a 49 percent sensitivity by PET, whereas none were detected by SPECT.

Other studies compared the PET with the SPECT approach in the same patients. An early study used supine bicycle stress and [13N]ammonia in 48 patients with CAD and reported comparable diagnostic performances for PET and SPECT.78 In another investigation of 202 patients, MBF was evaluated with 82Rb at rest and again 4 min after the dipyridamole infusion." About 8 to 9 min later, or a total of 12 to 13 min after the end of the dipyridamole infusion, 201T1 was injected and SPECT imaging performed within 10 min. PET and SPECT exhibited comparable specificities, while PET demonstrated a significantly higher sensitivity than SPECT. The results were similar when only 132 of the 202 patients without prior cardiac events were analyzed. A third study reported somewhat different findings in 81 patients." Again, all patients underwent rest and dipyridamole stress imaging with S2Rb and PET; for the 201T1 SPECT study, 38 (or 47 percent) of the patients underwent treadmill testing, and the remaining 43 (or 53 percent) underwent pharmacologic stress with dipyridamole. In that study, PET and SPECT exhibited comparable sensitivities; however, the specificity was higher for PET than for SPECT. The diagnostic accuracies were similar for patients submitted to treadmill stress testing and patients with pharmacologically induced hyperemia for SPECT imaging with 201T1.

Thus both studies demonstrate the high diagnostic accuracy for PET but differ in terms of higher sensitivities and specificities. The average decay half-time of 33 min for the hyperemic
response amounts to an only 10 percent decline in the hyperemic response over a 4-min period" that is unlikely to fully explain the lower sensitivity of 201T1 SPECT. The gain in specificity in the study by Stewart et al." most likely resulted from the adequate correction of photon attenuation and thus a reduction of falsely positive findings. Although the reasons for the observed differences between both studies remain unclear, image analysis at different points of the receiver operating curve may be one possible explanation.

On balance, the reported studies demonstrate a statistically significant gain in diagnostic accuracy for the detection of CAD by PET. Although larger clinical trials are needed, especially in previously undiagnosed patients with normal MBF and normal wall motion at baseline, current information indicates an improved diagnostic accuracy that may eliminate additional diagnostic procedures. A recent report compared the effect of PET and of SPECT on the subsequent referral to coronary angiography in 1490 and 102 patients, respectively." Pretest likelihoods for CAD were similar for both patient groups. However, the rate of angiography was significantly less (16.7 percent) after PET than after SPECT (31.4 percent), which produced an approximately 23 percent cost saving per patient.

EFFECT OF CORONARY STENOSES ON REGIONAL MBF

Recent investigations took advantage of PET's ability for measurements of regional MBF with [15O]water or [13N]ammonia in order to define the relationships between the angiographic stenosis severity, hyperemic flow responses, and vasodilator capacity These studies noted significant correlations between the anatomic stenosis severity and an attenuation of the hyperemic response to pharmacologic vasodilation. A similar correlation between MBF and coronary stenosis severity was observed when MBF was increased with dobutamine4r (Fig. 216-5).11 One study describes an inverse, nonlinear correlation between the cross-sectional area reduction of the stenosis and the flow reserve in the stenosis-dependent myocardium" (Fig. 216-6) that resembles the nonlinear correlation observed in animals." In exploring the existence of such correlation in human CAD, the latter study excluded confounding factors such as stenoses in series or collateral vessels to stenosis-dependent myocardium." While such correlation had in fact been expected in human CAD, the considerable scatter of the data about the regression line was not. Factors accounting for this scatter include possible inaccuracies in regional MBF measurements, the variability of the hyperemic response to pharmacologic vasodilation, age differences, and different baseline hemodynamic states. The scatter of the data may further point to a disparity between the anatomic and functional properties of human coronary artery stenoses. Different from the controlled and idealized coronary artery stenoses in the experimental setting, human coronary artery stenoses are of remarkably greater morphologic complexity, including eccentricity, variable stenosis inflow and outflow angles, and different lengths and irregular surfaces, that
( FIGURE 216-5 )
may not be fully appreciated by angiography nor be adequately accounted for by assumptions underlying model-based estimates of stenosis severity. It thus seems probable that the evaluation of flow, either semiquantitatively or quantitatively, renders more accurate functional information on the stenosis severity and, more broadly, on CAD. Moreover, estimates of an attenuated flow reserve obtained from static images of the relative distribution of MBF during hyperemic stress clearly offer invaluable information on the functional significance of coronary artery stenosis. Yet, in view of the nonlinear response in flow tracer uptake to increases in blood flow, such "semiquantitative" estimates would tend to be less accurate than those available through true measurements of MBF.

ASSESSMENT OF CORONARY VASOMOTION AND PRECLINICAL CAD

Noninvasive measurements of regional MBF offer the intriguing possibility to uncover vasomotor abnormalities of the human coronary circulation. If such abnormalities exist already during the early stages of CAD, it then may become possible to detect the disease during its evolutionary and preclinical stages. Such measurements further offer the prospect of monitoring disease progression as well as the responses to interventions aiming at regression of disease or slowing or halting its progression. Several lines of evidence support such possibility.

The now well-established beneficial effects of cholesterol lowering and especially of HMG-CoA reductase inhibitors have shifted the emphasis to the assessment of function rather than to the morphology of the human coronary circulation. Dietary and/or pharmacologic cholesterol lowering affect the anatomic stenosis severity only little, if at all, at least over the time periods studied, but strikingly reduced cardiac morbidity and mortality. Hence the beneficial effects are attributed to plaque stabilization and improvements in endothelial function.

Invasive studies of the human coronary circulation, performed during cardiac catheterization with intracoronary administration of direct vascular smooth muscle dilator agents such as adenosine or papavarine and of acetylcholine as a pharmacologic probe of predominantly endothelial-mediated coronary vasomotion, emphasize the importance of endothelial dysfunction early during the development of atherosclerosis . For example, human coronary arteries with minimal atherosclerotic changes but no flow-limiting stenoses or even without any structural changes but in the presence of coronary risk factors alone revealed normal, predominantly vascular smooth muscle-mediated vasodilator capacities but attenuated or even highly abnormal endothelial-mediated flow responses. These invasive studies test endothelial function at two sites of the coronary circulation, the large epicardial conduit and the coronary resistance vessel . Measurements of regional MBF by PET offer the opportunity to probe the function of the human coronary circulation entirely noninvasively and mostly at the level of the resistance vessels.

PET-based measurement of MBF in asymptomatic patients with hypercholesterolemia revealed an approximately 32 percent reduction in myocardial flow reserve or an approximately 18 percent reduction in hyperemic flow during adenosine adminstration. In fact,the myocardial flow reserve was correlated with the ratio of plasma total cholesterol over HDL cholesterol .Subsequent investigations confirmed these observations but also noted that elevated triglycerides or,in young individuals, a family history of CAD alone or of hypertennsion was associated with diminished vasodilator capacities and myocardial flow reserves.Other studies again observed diminished hyperemic responses in patients with diabetes,and one study found a correlation between the hyperemic MBF response and the therapuetic control of the diabetic state.
( FIGURE 216-6 )
To some extent, these observations differ from those made by invasive techniques, where frequently the predominantly vascular smooth muscle-mediated vasodilator response to, for example, intracoronary papavarine or adenosine was preserved despite the presence of coronary risk factors, while primarily endothelial-mediated responses were markedly abnormal. As Bache suggests, however, the major resistance to flow through the coronary circulation resides at vessels in the diameter range of 100 to 400 microm. If increases in flow exert shear stresses on the endothelium of the 400-microm vessels, then primarily endothelialdependent mechanisms augment the flow response to predominantly vascular smooth muscle vasodilators. Conversely, as forearm blood flow measurements have shown, pharmacologic impairments of endothelial function reduce the maximal flow response to intravascular adenosine by about 25 to 35 percent. Consequently, coronary risk factors such as low-density lipoprotein (LDL) cholesterol, triglycerides, and diabetes interfere with endothelial function and account for the diminished hyperemic response to adenosine or dipyridamole.

Another potentially important new concept for identifying diffuse coronary artery narrowing without discrete stenoses has been introduced recently. Diffuse luminal narrowing causes a greater decline in pressure along the coronary artery. This then is associated with a progressive decrease in myocardial perfusion from the proximal to the distal portion of the coronary arterial system and hence in a longitudinal base-to-apex perfusion gradient. A similar longitudinal perfusion gradient likely exists during hyperemia in patients without CAD but with coronary risk factors.

The predominantly endothelial-mediated coronary vasomotion can be assessed by the cold pressor test. In invasive studies, cold pressor testing evoked paradoxical changes in the diameter of the conduit vessels comparable with those evoked by intracoronary acetyicholine and, further, alterations at the level of the resistance vessels that correlated with those produced by intracoronary acetylcholine. Observations with PET-based measurements of MBF in long-term smokers support the use of the cold pressor test for this purpose. In these young smokers, vascular smooth muscle-mediated hyperemic flow responses to intravenous dipyridamole were normal, but cold pressor testing produced only modest, nonsignificant increases in MBF as compared with nonsmokers, where MBF rose in direct proportion to the cold-induced increase in cardiac work, as estimated by the rate-pressure product (Fig. 216-7). Similarly, abnormal flow responses to cold also were noted in postmenopausal women with and without CAD risk factors.

MONITORING RESPONSES TO RISK FACTOR MODIFICATION

Given their high degree of reproducibility, PET-based measurements of MBF are equally suited for monitoring responses in coronary vasomotion to pharmacologic interventions and risk factor modification. Early studies in only 13 participants in a 6-week cardiovascular conditioning program demonstrated this capability. Dietary changes, regular exercise, and lifestyle modifications were associated with weight loss, decreases in heart rate and blood pressure at rest, and significant decreases in plasma total and LDL cholesterol. A 12 percent decline in MBF at rest was proportionate to the decrease in resting cardiac work. Cardiovascular conditioning also produced a 9 percent increase in
( FIGURE 216-7 )
hyperemic flow, and MBF reserve increased by a total of 20 percent. Rigorous lifestyle and risk factor modification had been shown previously with PET to result in smaller and less severe stress-induced perfusion defects. More recent studies with PET-based measurements of MBF have demonstrated beneficial effects of cholesterol lowering by HMG-CoA reductase inhibitors. In one study, a 6-month course of fiuvastatin treatment produced a 26 percent increase in hyperemic MBFs (at 6 months but not at 2 months) and thus in vasodilator capacity. Of interest was the delayed improvement in vasodilator capacity ( Fig. 216-8 ). In these patients with CAD, the cumulative coronary function improved in myocardial territories subtended by both diseased and nondiseased coronary arteries. These observations differ with those of another study that demonstrated a significant improvement in vasodilator capacity only in territories with stress-induced perfusion defects but not in apparently normal myocardium, whereas
( FIGURE 216-8 )
a third study demonstrated again a 20 percent improvement of hyperemic flows in remote myocardium. Another study reported immediate (within 24 h) improvements in hyperemic MBFs following LDL cholesterol plasma apheresis. In the latter study, however, plasma LDL cholesterol apheresis reduced total cholesterol by 42 percent and LDL cholesterol by 58 percent, which was greater than in the fluvastatin study with total and LDL cholesterol reductions of 29 and of 37 percent, respectively.

Other investigations explored pharmacologic effects on predominantly endothelial-dependent coronary vasomotion. For example, intravenous L-arginine (30 g) as the substrate of nitric oxide synthase (NOS) in long-term smokers normalized the MBF response to cold pressor testing, suggesting that endothehal function or, at least, the bioactivity of nitric oxide had norma1ized. Whether increases in the substrate for NOS accelerate production of nitric oxide remains uncertain, especially in view of the low Km, which renders the reaction relatively substrate-independent. One possibility could be a nonspecific effect, perhaps on the oxidative stress, as recently demonstrated with cold pressor testing in response to acute administration of vitamin C. Other possible mechanisms include competitive displacement of asymmetric dimethylarginine, an inhibitor of NOS with elevated plasma levels in hypercholesteremic patients. An insulin-dependent mechanism is also possible, especially because L-arginine infusions prompted three- to fourfold increases in plasma insulin concentrations. Similarly, hormone-replacement therapy in postmenopausal women without coronary risk factors can normalize the MBF response to cold, while the responses remain abnormal in postmenopausal women with coronary risk factors despite hormone-replacement therapy)

Assessment of MyocardiaL Viability

Myocardial viability pertains to an impairment of myocardial contractile function that is potentially reversible. Distinction of such potentially reversible from irreversible impairment of contractile function often is of considerable clinical importance but remains diagnostically challenging. Both types of tissue injury share several features, including similar degrees of abnormal systolic wall motion, of reduced MBF, and of electrocardiographic abnormalities. Persistence of metabolic activity for sustaining vital, energy-requiring processes, however, including cellular homeostasis, depends on some residual MBF for removal of inhibitory metabolites as well as for supply of fuel substrates. Hence key features of viable myocardium include

o Impairment of systolic wall motion at rest
o Normal or reduced, but not absent, blood flow
o Preservation of cellular homeostasis
o Persistent metabolic activity for high-energy phosphate production
o Recruitable contractile reserve

GENERAL CONSIDERATIONS

Research studies in animals provided the base for the detection of myocardial viability. Known alterations in substrate metabolism during acute myocardial ischemia were demonstrated noninvasively with positron-emitting tracers of myocardial substrate metabolism. Consistent with an impaired FFA oxidation was the diminished initial uptake of delayed clearance from the myocardium. Additionally, the known increase in glucose extraction and use was reflected by a regional increase in 8F]deoxyglucose uptake. Initial studies in patients with acute myocardial ischemia revealed blood flow and glucose metabolism patterns that were virtually identical to those in animals, e.g., enhanced deoxyglucose uptake in hypoperfused dysfunctional myocardial regions. Unexpectedly, the same pattern existed in patients with chronic CAD but no signs of acute ischemia ( Fig. 216-9 ). This raised the question of whether the observed blood flow-metabolism pattern was unique to acute ischemia or represented a more general metabolic pattern in chronically dysfunctional and hypoperfused myocardium. Also intriguing were observations in other CAD patients with regionally reduced deoxyglucose uptake that paralleled the reduction in regional MBFIM (see Fig. 216-9 ). A more systematic study in patients scheduled for surgical revascularization confirmed the hypothesis that the regionally enhanced deoxyglucose uptake, in contrast to a reduction, reflected metabolic activity as evidence of viability in myocardium with complete or partial loss of contractile function. Restoration of tissue perfusion was followed by improved contractile function in myocardium with but not in myocardium without persistent glucose metabolic activity.

Possible Mechanisms of the Blood Flow Metabolism Pattern

The preceding observations established the clinical utility of these PET findings, but the underlying mechanisms remained uncertain. Patients with CAD revealed after supine bicycle exercise in stress-induced flow defects an augmented deoxyglucose uptake when the radiotracer was administered 20 to 30 mill after exercise and after the stress-induced flow defect had already resolved. This implicated myocardial stunning as one possibility, subsequently supported by observations in animal experiments and in patients with either collaterized myocardium or unstable angina.These studies demonstrated the evolution of a blood flow-metabolism pattern in chronically reperfused myocardium: An immediate postreperfusion decrease in glucose uptake was followed by an increase that subsequently declined to normal as contractile function returned. The enhanced 8F]deoxyglucose uptake was attributed to increased lactate release and thus anaerobic glycolysis that persisted even after blood flow had been restored. The evolution of such metabolic pattern also may pertain to early postinfarction patients but does not fully explain all observations in patients with chronic CAD. Another possibility includes repetitive stunning as the reason for the persistent increase in cose uptake in dysfunctional myocardium. An impairment in contractile function associated with enhanced glucose use was noted in collateral-dependent myocardium only if the flow reserve was markedly restricted. It limits the coronary circulation's ability to respond appropriately to transient and frequent increases in oxygen demand during daily life, leading to transient ischemic episodes, each followed by stunning and preventing recovery of contractile function.

Myocardial hibernation serves as another explanation.The postulated downregulation of contractile function in response to diminished rest MBF is thought to be associated with an alteration of the myocardium's substrate metabolism with a dominant role for the more oxygen-efficient glucose. Hibernation in its truest sense then implies that the downregulated
( click for FIGURE 216-9 )

( click for FIGURE 216-10 )
energy requirements match the available energy supply. A new supply-demand imbalance is established, but at a lower level. Such a new balance, however, will be a precarious one because even moderate increases in demand or decreases in supply disturb the steady state and cause ischemia. It is thus possible and likely that both hibernation and stunning coexist to varying extents in many patients. Observations in experimental animals suggest that sustained reductions in both blood flow and contractile function can be maintained for some time without significant necrosis, but development of structural alterations resembling those in patients with chronic CAD supports the concept of hibernation.

Both concepts, repetitive stunning and hibernation, may, in their purest form, represent the two ends of a spectrum. As Fig. 216-10 illustrates, the spectrum begins with a reduction in myocardial flow reserve, where increased demand can no longer be matched by an appropriate increase in supply and which ends with a loss of the flow reserve and a decline in regional MBF at rest, associated with a downregulation of contractile function and adaptation of substrate metabolism. Such a spectrum could represent a temporal progression in coronary artery stenosis severity. Recent findings in chronically instrumented animals with a progressive decline in and ultimately loss of regional flow reserve associated with a decrease in rest blood flow support such a scenario. Reductions in flow or flow reserve also may occur suddenly in view of the high incidence of blood flow metabolism mismatches in early postinfarction patients. In acute animal studies, sudden moderate reductions in regional MBF are associated initially with evidence of acute ischemia (e.g., release of lactate and enhanced glucose uptake). An apparent resetting or adjustment of demand follows, and lactate release converts to uptake, high-energy phosphate stores are replenished, anda new supply-demand balance seems to have returned.

Some debate focused on the issue of whether MBF at rest can indeed be chronically reduced. Nevertheless, findings in chronic animal experiments, as well as substantial improvements in resting MBF following surgical revascularization, argue in favor of the possibility of a true chronic regional hypoperfusion.

Ultrastructural and Histochemical Observations Other attempts to gain mechanistic insights into the enhanced [18F]deoxyglucose uptake include morphometric and histochemical analyses of biopsy specimens harvested from dysfunctional human myocardium during surgical revascularization. Prior autopsy studies indicated a general correlation between the degree of regional myocardial fibrosis and the severity of the impairment of regional contractile function. Yet there were exceptions. In some instances, dyskinetic myocardium was free of fibrosis at autopsy, or conversely, some normally contracting myocardium contained as much as 40 percent fibrosis. It also was known that "abnormal" myocytes ( Fig. 216-11 ) existed in chronically dysfunctional myocardium. More recent investigations noted correlations between the externally determined relative blood flows and relative [18F]deoxyglucose concentrations with the morphometrically determined fractions of fibrosis, abnormal myocytes, and normal myocardium. The various studies agree on a general correlation between relative blood flow and the percentage of tissue fibrosis ( Fig. 216-12 ) but differ on the fraction of abnormal myocytes. In one study, this fraction is the same in reversibly and irreversibly dysfunctional myocardium, 161 whereas a second study notes a significantly greater fraction in reversibly than in irreversibly dysfunctional myocardium. 161 Because the centrally located glycogen granules are key features of such abnormal myocytes and a significant correlation exists

( click for FIGURE 216-11 )

( click for FIGURE 216-12 )

between the fraction of such abnormal myocytes and the relative [18F]deoxyglucose uptake, these abnormal myocytes have been considered the ultrastructural correlate of enhanced ['8F]deoxyglucose uptake in chronically dysfunctional myocardium. Other observations argue against this notion. Again, electron microscopy and histochemistry of biopsy samples from the center of the dysfunctional myocardial wall demonstrate highly different degrees of severity of morphologic alterations in myocardial regions with blood flow-metabolism mismatches . Despite identical flow and glucose metabolism findings on PET, nearly half the patients in this study revealed only minimal, if any, morphologic changes, whereas the other half demonstrated severe structural abnormalities. Such variability in morphologic alterations argues against the structurally abnormal myocyte and especially the glycogen granules as an explanation of the enhanced [18F]deoxyglucose uptake. More likely explanations include translocation and possibly upregulation of GLUT1 as a flux-generating step, uncoupling of glycolysis from glucose oxidation, regulated probably by malonyl-CoA and carnitine palmitate transferase and possibly an ischemia-related loss of adrenergic innervation or function associated with increased exogenous glucose use. In cardiac allografts, glucose use was about 70 percent higher in denervated than in reinnervated myocardium.

Myocytes in Chronically Dysfunctional Myocardium

Whether abnormal myocytes as described initially by Flameng et al. and subsequently observed in biopsy material from mismatched myocardium point specifically in the direction of or are ingredients unique to any particular pathophysiologic mechanism underlying the chronic, though potentially reversible, impairment of contractile function remains uncertain. Two schools of thought exist. One holds that the morphologic alterations result from
(1) contractile unloading,
(2) increased wall stress (stretch), and
(3) a metabolic substrate switch to preferential glucose use.
In fact, contractile unloading recently has been demonstrated to result in virtually identical structural changes. The expression and distribution patterns of other features such as of a-smooth muscle actin, cardiotin, and titin,as well as an increased expression of glucose transporter 1 (GLUT1) mRNA, features that resemble those in embryonic and/or neonatal myocytes, suggested that the changes of abnormal myocytes may represent "dedifferentiation." Histochemical analysis further uncovered alterations in the extracellular matrix, with increased amounts of collagen and fibronectin surrounding the abnormal myocytes."' Finally, similar to neonatal myocytes, these abnormal myocytes have been found to be relatively tolerant of ischemia. The absence of true degenerative changes further has been claimed to support this possibility.

The other school of thought emphasizes a progressive deterioration rather than a stable state of the cell's morphology and therefore referred to hibernation as "incomplete adaptation to ischemia. The process begins with few structural changes but a switch in substrate selection to glucose, either because of its greater oxygen efficiency or, alternatively, because of loss of enzymes essential for fatty acid oxidation, followed by loss of contractile protein and accumulation of glycogen and mitochondrial and nuclear alterations, ultimately leading to cell death and scar tissue formation ( Fig. 216-13 ). Other studies again report reduced expression of contractile and cytoskeletal proteins associated with increased expression of extracellular matrix proteins, implying a progressive loss of contractile protein and of the cell structure that is paralleled by accelerated formation of tissue fibrosis and hence a progressive loss of viability that was further found to be associated with apoptosis and replacement fibrosis. Biopsies from patients with preoperatively viable myocardium but without a postrevascularization improvement in contractile dysfunction demonstrated an about threefold increase in mRNA of caspase-3, a promoter of apoptosis, together with an about 50 percent reduction in the expression of the antideath genes Bcl-2 and p53, again consistent with continued cell death and replacement fibrosis. Chronic animal experimental studies similarly have demonstrated significant increases in apoptotic myocytes in hibernating myocardium with reduced rest MBF and critically reduced or absent flow reserve . The fact that myocyte apoptosis in these studies occurred scattered and not in clusters raises the question of whether apoptosis is indeed the end point of the progressively deteriorating abnormal myocyte or such apoptosis represents a process that occurs in parallel. To some extent this may depend on the duration and severity of the ischemic compromise. For instance, other animal studies with more sudden reductions in flow and shorter time periods report higher rates of myocyte apoptosis occurring in clusters.

A progressive deterioration of reversibly dysfunctional myocardium is also consistent with clinical observations that point to the high prevalence of mismatch patterns in patients with prior myocardial infarctions ,but note a declining incidence of blood flow-metabolism mismatches as a function of time after an acute myocardial infarction. Moreover, the loss of the capability to improve global LV function if revascularization was delayed by more than 6 months"" or an increase in fibrosis and loss of functional recovery after revascularization as a function of the duration of clinical symptoms seem to support such progression. The blood flow-metabolism mismatch may represent a transient rather than a permanent state of reversibly dysfunctional myocardium. It is possible that reversibility can be maintained up to a certain point. Once this critical point has been reached, myocytes become committed to irreversibility and cell death.

In the clinical setting, prompt restoration of adequate tissue perfusion through interventional revascularization therefore will be essential, regardless of whether abnormal myocytes represent dedifferentiation or degeneration. It would seem that ultimately the return of contractile function will depend on the amount of connective tissue. Once fibrosis and scar tissue occupy more than 35 to 40 percent of the myocardium, dysfunction has been shown to be irreversible. Also, the presence of structural changes in viable myocardium, as demonstrated with blood flow-metabolism imaging, implies that if the contractile machinery in abnormal or dedifferentiated myocytes can be reconstructed, the recovery of contractile function will not be immediate but slow, as animal experimental and clinical investigations have indeed demonstrated. The delay in cell repair also may explain the persistence of

( click for FIGURE 216-13 )

artifactual reductions in tracer concentrations due to photon attenuation .This can limit the ability to accurately estimate the extent of a blood flow-metabolism mismatch.
More recent approaches rely solely on the use of multipurpose SPECT-like systems, either equipped with ultrahigh-photon-energy general-purpose collimators or with coincidence-detection systems (see also Fig. 216-1). Little information thus far has become available on the clinical performance of coincidence-detection systems, whereas systematic studies with ultra-high-photon-energy general-purpose collimator SPECT systems using 2207T1- or 99mTc-labeled flow tracers and [18F]deoxyglucose report predictive accuracies that are comparable with those reported with dedicated PET systems.
Further, 201T1 rest-redistribution imaging has been useful for identifying myocardial viability and for predicting the postoperative outcome of ischemic cardiomyopathy, although with a somewhat lower predictive accuracy. This approach suffers from instrumentation-related shortcomings, especially in patients with poor LV function, and consequently, poor signal-to-noise ratios. Thallium-201 offers a negative signal (reduced tracer uptake) as compared with [18F]deoxyglucose, with a positive signal (enhanced tracer uptake) that is more readily accessible to visual analysis. One study reporting that [18F]deoxyglucose and PET identified myocardial viability in 18 of 20 patients with an average LV ejection fraction of 23 percent and only fixed 201T1 defects on SPECT` is consistent with a more recent report of viability by [18F]deoxyglucose in 17 of 33 patients (LV ejection fraction <35 percent) with fixed or minimally redistributing 201T1 defects. Further, a comparison study of 201T1 and [18F]deoxyglucose SPECT reports a generally excellent agreement between both approaches but observed disparities in patients with severely depressed LV ejection fractions, where [18F]deoxyglucose revealed more viable myocardial segments than 201T1 SPECT.

In synthesizing the currently available information, it appears that ultimately the total fraction of scar tissue in a given myocardial segment determines largely whether or not contractile function will improve. Because of the linear correlation between scar tissue and relative MBF,- evaluation or even quantitation of regional MBF offers information on potential reversibility. On the other hand, if in viable though functionally compromised myocardium MBF is also reduced, then the augmented glucose use, as evidenced by the enhanced [18F]deoxy-glucose uptake, offers additional and critical information. This has prompted most investigators to predict the ultimate functional outcome from a combined assessment of blood flow and [18F]dcoxyglucose uptake. Further, the temporal recovery of contractile function after revascularization appears to depend on the degree of ultrastructural changes of myocytes as well as the fractional distribution between myocytes with only mild and those with severe ultrastructural changes. If, as postulated, only mild structural changes are associated with a full functional recovery within 3 months, more severe structural changes may require substantially longer time periods and, further, may account for the persistence of increased [18F]deoxyglucose uptake even for many months following revascularization.

CLINICAL ROLE OF PET VIABILITY ASSESSMENT

Among the various PET approaches. the blood flow and glucose metabolism approach has gained the greatest clinical acceptance. Viability assessments with PET can decisively affect therapeutic strategies in patients with advanced CAD and ischemic cardiomyopathy. The therapeutic options in these patients range from aggressive medical management to surgical revascularization and cardiac transplantation. While conservative pharmacologic approaches to the management of such patients has improved markedly over the past decade, the long-term survival of medically treated patients remains relatively poor 201 Cardiac transplantation as another approach offers a better long-term survival and an improvement in the quality of life, but the supply of donor hearts has not kept pace with the increasing demand, so this therapeutic option remains limited . At present, the prevalence of ischemic cardiomyopathy in the United States alone amounts to about 2.5 million cases, thus affecting roughly 1 percent of the U.S. population. The decision to revascularize frequently depends on the answers to several questions. First, what is the leading cause of poor LV function? Second, if CAD has been identified as the culprit, is there enough viable myocardium so that surgical revascularization produces an improvement in LV performance and/or congestive heart failure symptoms? Third, will revascularization avert future catastrophic cardiac events and prolong survival? And finally, can the surgical risk be predicted, since this will influence the preoperative risk-benefit ratio?

Ischemic versus Idiopathic Dilated Cardiomyopathy

In addition to heart failure symptoms, ischemic cardiomyopathy shares several other features with idiopathic dilated cardiomyopathy, such as, for example, the LV enlargement, the often diffuse hypokinesis, the markedly depressed LV ejection fraction, and frequently, mitral regurgitation. Biventricular enlargement has been thought of as a feature characteristic of idiopathic dilated cardiomyopathy but also can be present in ischemic cardiomyopathy. Conduction abnormalities often limit the accuracy of electrocardiographic criteria to distinguish between both entities. Additionally, an intrinsic myopathic process including LV remodeling also may exist in a number of patients with CAD so that the major cause of the poor LV function may remain unknown or difficult to elucidate. Importantly, however, the therapeutic approach to both disease entities will differ strikingly .

Both disease entities, however, reveal remarkably different patterns of blood flow and substrate metabolism on PET. A comparative study in patients with ischemic cardiomyopathy and idiopathic dilated cardiomyopathy found the distribution of MBF to be characteristically homogeneous in idiopathic cardiomyopathy as compared with distinct flow reductions clearly corresponding in ischemic cardiomyopathy to the coronary vascular territories. Similarly, uptake of [18F]deoxyglucose was noted to be homogeneous in dilated cardiomyopathy, whereas matches and/or mismatches between blood flow and [18F]deoxuglucose uptake were present in ischemic cardiomyopathy ( Fig. 216-15 ). Combined imaging of blood flow and glucose metabolism distinguished with an overall accuracy of 85 percent between both disease entities.

Prediction of the Outcome in Global LV Function

Numerous clinical investigations have reported the high accuracy of of increased [18F]deoxyglucose uptake after successful revascularization.

Viability Assessment in the Clinical Setting

The classic and now most widely applied approach entails evaluation of the relative distribution of blood flow and exogenous glucose use with [18F]deoxyglucose. Initial studies uncovered three distinct patterns:

• Normal blood flow and normal or enhanced glucose uptake
• Reduced blood flow but normal glucose uptake in excess of blood flow (mismatch)
• Reduced blood flow and proportionately reduced glucose uptake (match)

While these terms are purely operational, they infer, at least to some extent, the underlying pathophysiology accounting for the contractile dysfunction. Normal flow and/or metabolism may represent stunned myocardium, whereas the classic mismatch may be consistent with hibernating myocardium. Both patterns predict a postrevascularization improvement in contractile function, whereas the concordant reduction in blood flow and metabolism predicts that function will not improve. '35,182.183 It should be emphasized that the reduction in regional flow for both matches and mismatches may vary considerably between patients.
Because of the observed correlation between tissue fibrosis and relative flow tracer uptake, the evaluation of regional MBF alone can provide information on the presence of reversible contractile dysfunction ( Fig. 216-12 ). Severe reductions to less than 25 percent of normal or complete absence of blood flow reflects complete or nearly complete transmural scar tissue formation and hence nonreversiblity. In another study, flow reductions of more than 60 percent were highly accurate in predicting nonreversiblity of contractile dysfunction. Conversely, completely normal or only mild reductions (<20 percent) of MBF in dysfunctional myocardium argue against the presence of significant amounts of tissue fibrosis; it possibly reflects myocardial stunning and thus indicates functional reversibility. Mild to moderate flow reductions are less reliable discriminators. If combined with a metabolic study, the [18F]deoxyglucose uptake in the case of a small nontransmural/infarction with otherwise normal myocardium would be reduced in proportion to blood flow. Conversely, an increase in glucose uptake would indicate the coexistence of reversibly dysfunctional myocardium with scar tissue and predict an improvement in contractile function.

Another, again limited approach for identifying reversible contractile dysfunction is the use of [18F]deoxyglucose alone. This approach assumes that regional reductions in [18F]deoxyglucose greater than 50 percent relative to remote myocardium represent irreversible contractile function, whereas mildly reduced or normal uptake indicates the presence of reversible dysfunction. While used for some time as a benchmark for defining the accuracy of 201T1-based techniques for assessing myocardial viability, only recently has the validity of this particular approach been tested against the postrevascularization outcome in regional contractile function. Electrocardiographic gated image acquisition affords simultaneous evaluation of regional function and metabolism and thu- can further augment the predictive accuracy of the [18F]deoxyglucose standalone approach. A more recent report emphasizes the utility of measurements of exogenous glucose use in absolute units.Using a threshold value of 0.25 micromol/g per minute offered a 93 percent positive and a 95 percent negative predictive accuracy for the improvement of contractile dysfunction. Nevertheless, this approach is severely limited when glucose use and hence [18F]deoxyglucose uptake cannot be controlled sufficiently. In such instances, it may be difficult to distinguish between scar tissue, normal myocardium, and reversibly contractile dysfunction, which then could be readily clarified by evaluating the distribution of regional MBF.

The pattern of normal blood flow and glucose metabolism in mildly to severely hypokinetic myocardium of severely depressed left ventricles is a difficult clinical problem. One study reports that of 32 such myocardial regions, only 8 regions (or 25 percent) improved following surgical revascularization.194 Such regions therefore may represent remodeled LV myocardium. Conversely, an improvement in wall motion may be consistent with myocardial stunning. If suspected, careful evaluation of the coronary anatomy or, if unavailable, the addition of a pharmacologic stress study can aid in distinguishing between stunned and remodeled LV myocardium ( Fig. 216-14 ).

Alternate Approaches to Blood Flow and Glucose Metabolism Imaging

Several institutions use SPECT myocardial perfusion imaging with either 201T1 or 99mTc-sestamibi and metabolic imaging with [18F]deoxyglucose with dedicated PET systems.

The reported predictive accuracies for segmental and global LV function approach those obtained with dedicated PET systems. Nevertheless, such combined PET/SPECT approaches present at times with diagnostic limitations, especially because of considerable differences in contrast and spatial resolutions as well as artifactual reductions in tracer concentrations due to photon attenuation. This can limit the ability to accurately estimate the extent of a blood flow-metabolism mismatch.

More recent approaches rely solely on the use of multipurpose SPECT-like systems, either equipped with ultrahigh-photon-energy general-purpose collimators or with coincidence-detection systems (see also Fig. 216-1 ). Little information thus far has become available on the clinical performance of coincidence-detection systems, whereas systematic studies with ultra-high-photon-energy general-purpose collimator SPECT systems using 201T1- or 99mTc-labeled flow tracers and [18F]deoxyglucose report predictive accuracies that are comparable with those reported with dedicated PET systems.

Further, 201T1 rest-redistribution imaging has been useful for identifying myocardial viability and for predicting the postoperative outcome of ischemic cardiomyopathy, although with a somewhat lower predictive accuracy . This approach suffers from instrumentation-related shortcomings, especially in patients with poor LV function, and consequently, poor signal-to-noise ratios. Thallium-201 offers a negative signal (reduced tracer uptake) as compared with [18F]deoxyglucose, with a positive signal (enhanced tracer uptake) that is more readily accessible to visual analysis. One study reporting that [18F]deoxyglucose and PET identified myocardial viability in 18 of 20 patients with an average LV ejection fraction of 23 percent and only fixed 201TI defects on SPECT197 is consistent with a more recent report of viability by [8F]deoxyglucose in 17 of 33 patients (LV ejection fraction <35 percent) with fixed or minimally redistributing 201T1 defects. Further, a comparison study of 201T1 and [18F]deoxyglucose SPECT reports a generally excellent agreement between both approaches but observed disparities in patients with severely depressed LV ejection fractions, where [18F]deoxyglucose revealed more viable myocardial segments than 201T1 SPECT.

In synthesizing the currently available information, it appears that ultimately the total fraction of scar tissue in a given myocardial segment determines largely whether or not contractile function will improve. Because of the linear correlation between scar tissue and relative MBF , evaluation or even quantitation of regional MBF offers information on potential reversibility. On the other hand, if in viable though functionally compromised myocardium MBF is also reduced, then the augmented glucose use, as evidenced by the enhanced [18F]deoxyglucose uptake, offers additional and critical information. This has prompted most investigators to predict the ultimate functional outcome from a combined assessment of blood flow and [18F]deoxyglucose uptake. Further, the temporal recovery of contractile function after revascularization appears to depend on the degree of ultrastructural changes of myocytes as well as the fractional distribution between myocytes with only mild and those with severe ultrastructural changes.If, as postulated, only mild structural changes are associated with a full functional recovery within 3 months, more severe structural changes may require substantially longer time periods and, further, may account for the persistence of increased [18F]deoxyglucose uptake even for many months following revascularization.

CLINICAL ROLE OF PET VIABILITY ASSESSMENT

Among the various PET approaches, the blood flow and glucose metabolism approach has gained the greatest clinical acceptance.Viability assessments with PET can decisively affect therapeutic strategies in patients with advanced CAD and ischemic cardiomyopathy.

( click for FIGURE 216-14 )

The therapeutic options in these patients range from aggressive medical management to surgical revascularization and cardiac transplantation. While conservative pharmacologic approaches to the management of such patients has improved markedly over the past decade, the long-term survival of medically treated patients remains relatively poor. Cardiac transplantation as another approach offers a better long-term survival and an improvement in the quality of life, but the supply of donor hearts has not kept pace with the increasing demand, so this therapeutic option remains limited ). At present, the prevalence of ischemic cardiomyopathy in the United States alone amounts to about 2.5 million cases, thus affecting roughly 1 percent of the U.S. population. The decision to revascularize frequently depends on the answers to several questions. First, what is the leading cause of poor LV function? Second, if CAD has been identified as the culprit, is there enough viable myocardium so that surgical revascularization produces an improvement in LV performance and/or congestive heart failure symptoms? Third, will revascularization avert future catastrophic cardiac events and prolong survival? And finally, can the surgical risk be predicted, since this will influence the preoperative risk-benefit ratio?

Ischemic versus Idiopathic Dilated Cardiomyopathy

In addition to heart failure symptoms, ischemic cardiomyopathy shares several other features with idiopathic dilated cardiomyopathy, such as, for example, the LV enlargement, the often diffuse hypokinesis, the markedly depressed LV ejection fraction, and frequently, mitral regurgitation. Biventricular enlargement has been thought of as a feature characteristic of idiopathic dilated cardiomyopathy but also can be present in ischemic cardiomyopathy. Conduction abnormalities often limit the accuracy of electrocardiographic criteria to distinguish between both entities. Additionally, an intrinsic myopathic process including LV remodeling also may exist in a number of patients with CAD so that the major cause of the poor LV function may remain unknown or difficult to elucidate. Importantly, however, the therapeutic approach to both disease entities will differ strikingly .

Both disease entities, however, reveal remarkably different patterns of blood flow and substrate metabolism on PET. A comparative study in patients with ischemic cardiomyopathy and idiopathic dilated cardiomyopathy found the distribution of MBF to be characteristically homogeneous in idiopathic cardiomyopathy as compared with distinct flow reductions clearly corresponding in ischemic cardiomyopathy to the coronary vascular territories. Similarly, uptake of [18F]deoxyglucose was noted to be homogeneous in dilated cardiomyopathy, whereas matches and/or mismatches between blood flow and [18F]deoxuglucose uptake were present in ischemic cardiomyopathy ( Fig. 216-15 ). Combined imaging of blood flow and glucose metabolism distinguished with an overall accuracy of 85 percent between both disease entities.

Prediction of the Outcome in Global LV Function

Numerous clinical investigations have reported the high accuracy of [18F]deoxyglucose imaging with PET in predicting the postrevascularization outcome in regional LV wall mo-tion. Even though some of these investigations employed permutations of the initially described blood flow-metabolism approach or relied only on the evaluation of regional [18F]deoxyglucose uptake in dysfunctional myocardium, the predictive accuracy, both positive and negative, continued to be high. Such studies have been important because they prove the concept of blood flow-metabo- FDG lism patterns as accurate predictors of the outcome of regional wall motion after restoration of MBF. More relevant in the clinical setting is whether blood flowmetabolism patterns can predict the postrevascularization outcome in global LV function.

Initial semiquantitative studies demonstrated some correlation between the extent of the blood flow-metabolism mismatch and the postrevascularization gain in LV ejection fraction. Patients with blood flow-metabolism mismatches that occupied at least two or more of seven total myocardial segments revealed a statistically significant increase in the LV ejection fraction following coronary bypass grafting. No such improvement was observed in patients with only one mismatch segment or with only matches. Subsequent studies reported significant gains in LV function in patients with blood flow-metabolism mismatches as compared with no improvement in those patients without metabolic evidence of viability. Fig. 216-16 summarizes the findings in 19 investigations including a total of 570 patients. The gain in global LV function is most striking in patients with an LV ejection fraction of less than 35 percent, who had a 34 percent postoperative increase in the LV ejection fraction as compared with a 19 percent (p < 0.02) improvement in patients with an LV ejection fraction of more than 35 percent. Additionally, recent studies reported significant correlations between the percentage of the left ventricle with a blood flow-metabolism mismatch and the postrevascularization increase in the LV ejection fraction Thus the extent of a blood flow-metabolism mis match has some predictive value on the postoperative gain in global LV performance ( Fig. 216-17 ). The absence of such improvement in one specific laboratory may be attributable to differences in the imaging and analysis approach used. MBF is evaluated with 82Rb at rest and during pharmacologic stress,and the distributionof MBF during stress is compared with the myocardial glucose uptake at rest. Thie approach identifies both stress induced ischemia and "viable myocardium" at rest. Hence blood flow and possibly wall motion at rest may be normal in some patients so that revascularization predominantly improves the capcity of the left ventricle to respond to exercise.

The improvement in regional and especially global LV function may not occur immediately but slowly though progressively following revascularization. In a highly selected patient group, blood flow had been shown to recover promptly following revascularization by angioplasty, whereas contractile function remained initially unchanged. On reexamination 67+-19 days later.no further improvements inregionalMBF had occurred, but systolic wall motion had now signicantly improved. The disparity between recovery ofMBF and contractile function may be attributed to stunningg and/or to rebuilding o the con-

( click for FIGURE 216-15 )

tractile machinery that had been lost in abnormal myocytes. Preliminary observations suggest a correlation between severity of the mismatch and the rate of recovery of contractile function. The rate of recovery appears to be faster when flow is relatively well preserved as compared with segments with more severe flow reductions and possibly more severe ultrastructural changes. Segments with largely preserved flows recovered faster than segments with more severe flow reductions. The contractile function appears to recover or improve more promptly in myocardial regions without marked ultrastructural abnormalities or a lesser fraction of abnormal myocytes.Finally, in addition to a slow recovery of contractile function in reversibly dysfunctional myocardium, other studies describe an associated decline in end-diastolic and end-systolic volumes, suggesting the possibility of a reversal of LV remodeling.

Effect on Congestive Heart Failure-Related Symptoms A related clinical question is whether such improvement is also

( click for FIGURE 216-17 )

associated with relief or amelioration of congestive heart failure symptoms. Several retrospective studies indicate such possible symptomatic improvement. Two investigations concluded that patients with blood flow-metabolism mismatches undergoing surgical revascularization demonstrated a significantly higher incidence of improvement in NYHA functional class than patients without mismatches or patients with matches but not submitted to regional revascularization 2 Among the 52 patients with mismatches and congestive heart failure class III or IV, 81 percent of the 26 patients undergoing revascularization revealed a significant improvement in congestive heart failure class as compared with only 23 percent of 26 patients treated
conservatively.

The amount of viable myocardium on blood flow and [18F]deoxyglucose imaging contains information on the magnitude of the postrevascularization improvement in congestive heart failure symptoms. The level of physical activity patients were able to perform prior to and 24 i- 14 months following coronary artery bypass grafting was graded on a specific activity scale and expressed in metabolic equivalents .220 Among the 36 patients in this study with an average LV ejection fraction of only 28 ± 6 percent prior to revascularization, the extent of the blood flow-metabolism mismatch ranged from 0 to 74 percent (mean 23 ± 22 percent) on polar map analysis. When patients were grouped according to the extent of the mismatch, 11 patients with a mismatch occupying less than 5 percent of the LV myocardium revealed a statistically significant but only mild improvement in functional status (34 percent increase in metabolic equivalents) (see Chap. 17). Intermediate-sized mismatches (5 to 17 percent) in 8 patients were associated with a 42 percent increase in metabolic equivalents, whereas large mismatches, i.e., greater than 18 percent, in 17 patients were followed after revascularization by an average increase of 107 percent in metabolic equivalents. Furthermore, the improvement in functional status was linearly correlated with the anatomic extent of the blood flowmetabolism mismatch. Lastly, blood flow-metabolism mismatches of 18 percent or more were 70 percent sensitive and 78 percent specific in predicting an improvement in physical activity or functional status following successful surgical revascularization.

Impact on Long-Term Survival

Several studies examined the long-term fate of patients after being evaluated for MBF and metabolism with PET. These studies presented compelling evidence for an increased prevalence of cardiac events in patients with blood flow-metabolism mismatches not submitted to interventional revascularization. They also implied that revascularization of blood flow-metabolism mismatches may avert future cardiac events.

Despite this general agreement, important differences emerged from these studies. One study in 129 chronic CAD patients followed for a time period of 17 ± 19 months found the presence of mismatches in the absence of revascularization to be independent predictors of the 17 nonfatal ischemic events . Nevertheless, the LV ejection fraction and the patient's age contained the highest predictive values for the 13 cardiac deaths in this patient group. In patient series with more homogeneously depressed LV function, the predictive value of a low LV ejection fraction applied equally to all groups. As shown in Fig. 216-18, the cumulative long term survival was lowest in the patient subgroup with blood flow-metabolism mismatches, who were on medical treatment. Of note, all four subgroups were similar with regard to age and clinical and hemodynamic findings. There were no significant intergroup differences in the LV ejection fraction, which for the whole group averaged only 25+-7 percent. Of note, patients with mismatches who underwent revascularization revealed a significantly better cumulative survival that no longer differed significantly from that of groups without mismatches (Fig. 216-18). In this study, the LV ejection fraction was without significant predictive valve, whereas by Cox model analysis the extent of a mismatch had a significant negative effect on survival (p<0.02), and revascularization of mismatch patients had a significant positive effect on survival (p<0.04). A second study in patients with similar uniform depression of LV ejection fraction reached similar conclusions.

( click for FIGURE 216-16 )

( click for FIGURE 216-18 )