| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J Thorac Cardiovasc Surg 1998;115:94-98
© 1998 Mosby, Inc.
SURGERY FOR CONGENITAL HEART DISEASE |
Departments of Neurosciencesa and Cardiothoracic Surgery,b Institute of Child Health (UCL) and Great Ormond Street Hospital forChildren, Department of Paediatrics and Neonatal Medicine, Royal PostgraduateMedical School,c Department of Medical Physics and Bioengineering,University College London, London, United Kingdom, d and Departmentof Surgery Duke University Medical Center, Durham, N.C.e C.E.C.is a Medical Research Council Training Fellow. This work was funded by theBritish Heart Foundation.
Received for publication June 23, 1997; revisions requested August12, 1997; revisions received Sept. 15, 1997; accepted for publication Sept.15, 1997. Address for reprints: Idris Roberts, BSc, Biomedical EngineeringDepartment, Great Ormond Street Hospital for Children, Great Ormond Street,London WC1 3JH, United Kingdom.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
100 gm–1 min–1. In an additional 10 children in whom repeated measurementswere made, the coefficient of variation was 11% ± 7% (mean ±standard deviation). In a further study, the method was compared with microsphereinjection in five piglets undergoing cardiopulmonary bypass. The comparisonwithin each animal with the linear least squares method gave values for R2 in the range 0.91 to 0.99. The gradientsof the fits ranged from 0.5 to 1.8 (median 1.0). The mean difference betweenthe two techniques was 5.7 ml 100 gm–1 min–1 or 7%. The coefficient of variation for the piglets was 14%± 9% (mean ± standard deviation). |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
1 healthy adults, 2 and children undergoing cardiopulmonarybypass (CPB). 3 However, theuse of oxyhemoglobin makes assumptions about cerebral oxygen uptake that maynot be appropriate during CPB. This article reports the use of an improvedmethod that uses the optical dye indocyanine green (ICG) as an intravasculartracer. |
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
4 These optical properties allow transmissionspectroscopy to be performed in vivo. 5
For linear transillumination of tissue, the modified Beer-Lambert lawcan be expressed as 6:
A =(a[c]LB)+G (EQ1) where A is the attenuationof light in optical densities; a is the extinction coefficient of the chromophore(mmol–1·L·cm–1); [c] is the concentration of chromophore (mmol·L-1); L is the distance between pointsof light entry and exit (cm); B is a "pathlengthfactor" that takes account of the increased optical path caused by thescattering of light in the tissue; and G isa factor related to tissue and optode geometry.
The attenuation caused by scattering and the geometric factor (G) isusually unknown and thus [c] cannot be quantified absolutely. However, ina given tissue with fixed optical geometry it is likely to remain constantin the short term; thus changes in attenuation can be attributed solely tochanges in the concentration of chromophore: 
[c] = A/(aLB)(EQ2) Scattering causes the distance traveled by the light to begreater than L by a factor B, and the distance traveled is given by the productof L and B. Provided these factors and the extinction coefficients of thechromophores present are known, changes in chromophore concentration can bequantified using equation 2. This study used a value of B of 4.39 for infantsless than 2 years old and 5.93 for older children. 4,7
In brain, the major endogenous chromophores that absorb light in thespectral range 700 to 1000 nm and whose concentrations change over the shortterm are oxyhemoglobin and deoxyhemoglobin and cytochrome oxidase. 8 ICG is an exogenous chromophorethat can be administered by injection. It is a tricarbocyanine dye of molecularweight 924.9 that absorbs near-infrared light maximally between 795 and 805nm. It has a low acute toxicity with a median lethal dose (LD50)in mice of 650 mg kg–1. The maximum recommended dose is5 mg kg–1. The dye is strongly protein bound to albuminin plasma and rapidly and completely eliminated by the liver, although excretionis slowed during cardiac operations. Full pharmacologic and toxicologic dataare available, 9 and a productlicense for the use of ICG is in force in the United States and the UnitedKingdom. Changes in the concentrations of these four chromophores can be calculatedfrom observed changes in the attenuation of light at four wavelengths.
Theory 2: Measurement of CBF using NIRS.
The Fick principle states that the rate of accumulation of a tracersubstance in an organ (Q) is equal to the difference between the rate of arrivaland the rate of departure of that substance. If a tracer is suddenly introducedinto the arterial blood, a measurement of the amount accumulated in the organcan be made at a time (t) later. While t is less than the minimum transittime of blood through the organ, the tracer will not appear in the venousefflux, and flow (f) can be measured as the ratio of tracer accumulated tothe quantity of tracer introduced over the time period (t). The quantity oftracer introduced by time t is equal to the integral of the change in thearterial concentration of tracer at time t (Pa[t]) with respect to time. Thus: f = Q(t)/ 
Pa(t)dt (EQ3)
ICG can be injected into the CPB circuit and the concentration in bloodentering the aorta measured by an optical device situated in the efferentlimb of the bypass circuit. Because no further dilution of [ICG] occurs betweenthis point and the brain, the concentration measured is the cerebral arterialconcentration ([ICG]blood), and the quantity of tracer introduced(Pa(t)) is given by the integral of [ICG]blood with respectto time. The increase in ICG in the brain ([ICG]brain) canbe measured by NIRS; this quantifies the tracer accumulation (Q(t)). Thus: CBF = k([ICG]brain)/**[ICG]blooddtml100gm–1min–1 (EQ4)where k is a constant reflecting the molecular weight of ICG,as well as tissue density and decimal conversions that allow blood flow tobe expressed per 100 gm of tissue.
Theory 3: Estimation of mean cerebral transit time.
Mean transit time (MTT) can be calculated after a rapid injection ofnondiffusable tracer by the "height over area" method. 10 The MTT is related to CBF andcerebral blood volume (CBV) by the Stewart-Hamilton equation: MTT= CBV/CBF (EQ5)
Measurement of CBF with NIRS.
Fig. 1 shows the apparatus used to measure CBF during CPB. Near-infrared light was carried fromthe NIRS spectrometer (NIRO 500; Hamamatsu Photonics KK, Hamamatsu City, Japan)to the head by a fiberoptic bundle (optode) placed against the skin overlyingthe parietal region of the brain at least 3.5 cm apart and held by an elasticbandage underneath a cover to reduce background light. The light transmittedacross the head was collected by a second optode and carried back to the spectrometer,which used pulsed laser diodes at 4 wavelengths (776, 819, 843, and 913 nm)as light sources; changes in light attenuation were measured every half second.The algorithm used to calculate the changes in concentrations of the chromophoresincorporated a linear least squares curve fitting technique. 11,12 The extinctioncoefficients used in the algorithm were corrected for the wavelength dependenceof pathlength using the method of Essenpreis. 12 The correction factors for the wavelengths used were 0.9995, 0.9550,0.9162, and 0.7805 (M. Cope, personal communication, October 1994).
|
Study protocols.
Studies were approved by the Research Ethics Committees of Great OrmondStreet Hospital for Children (November 1991) and of Duke University (March1994), and consent was obtained from the parents of the children studied.
All animals received humane care in compliance with the "Guidefor the Care and Use of Laboratory Animals" published by the NationalInstitutes of Health.
Measurement of CBF and MTT in children.
Nine children undergoing CPB surgery for congenital heart disease werestudied. The diagnosis and physiologic details are given in Table I. The pump flow was variable but was maintained at2.4 L · min-1 · m-2 whenever possible. CBF and MTT weremeasured in all patients.
|
Comparison with microspheres in an animal model.
Five newborn piglets were premedicated with ketamine (20 mg kg–1) and acepromazine (1 mg kg–1) and ventilatedmechanically (Sechrist infant ventilator, model IV-100B, Sechrist IndustriesInc., Anaheim, Calif.). They were then anesthetized with intravenous fentanyl(100 µg kg–1 bolus and 50 µg kg–1 per hour infusion), paralyzed with pancuronium (0.3 mg kg–1), and placed on CPB. Systematically, to induce measurablechanges in CBF, pump flow and body temperature were changed. Injections of15 µm radioactive microspheres were made at (1) pump flow of 100 ml kg–1 min–1 and temperature of 37°C; (2) 100 ml kg–1 min–1,22° C; (3) 50 ml kg–1 min–1, 22° C; (4) 50 ml kg–1 min–1, 17° C; and (5) 25 ml kg1 min–1, 17° C. Microspheres (3 x 106 in1 ml 10% dextran and 0.01% Tween, 153Gd, 113Sn, 103Ru, 95Nb, and 46Sc in random order; NEN ResearchProducts, DuPont, Wilmington, Del.) were injected into a left atrial catheter.Blood was withdrawn from the femoral arterial catheter at a rate of 7 ml min–1 over 120 seconds starting 10 seconds before the injection.Measurements of CBF using ICG as a tracer were made simultaneously with themicrosphere injection as described above; one optode was placed over the frontalregion and the other over the parietal region, both on the right side. Ateach of five pump flow and temperature settings, a microsphere CBF was comparedwith the mean of 3 to 7 (median 4) ICG CBF measurements.
At the end of the study the animals were killed, and the brain was harvested.The brain was subdivided into the cortex, basal ganglia, cerebellum, and brainstem. All the samples were counted in a gamma counter (Minaxi Auto-gamma Counter,5000 Series, Packard Instrument Co., Meriden, Conn.), and individual nuclideactivity was calculated (Compusphere Microsphere Multinuclide Analysis Software,Packard Instrument Co.).
The mean CBF value of the repeated NIRS measurements (t = 4 seconds)was compared with the CBF value calculated for cortex flow by the microspheremethod. Cortex flow was chosen for comparison because it is believed thatthe near-infrared technique probably reflects cortical flow because of thepenetrance of light. The mean, median, and range of the differences were calculated.Reproducibility was assessed from the coefficient of variation and by performingan analysis of variance with pig as a factor and time as a covariate.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
|
|
100 gm–1 min–1 for calculations where t = 2 seconds. When a greater proportionof flow from longer transit time compartments was included by increasing tto 4 seconds (see discussion), CBF ranged from 8.2 to 32.4 (median 15.9) ml 100 gm–1·min–1.
|
100 gm–1 min–1. The time betweenmeasurements within a group ranged from 0.5 to 7.4 (median 2.0) minutes. Thetime between groups ranged from 16.7 to 78.7 (median 30.3) minutes. The groupmean CBFs ranged from 3.2 to 70.7 (median 11.9) ml 100 gm–1 · min–1. For the repeated measurements, the coefficientof variation was 11% ± 7%. Analysis of variance showed that the interindividualdifferences and the effects of different pump flows and temperatures accountedfor 73% of the variability; the residual variability (27%) includes the reproducibilityof the technique and any other factors not measured.
Comparison with microspheres in piglets.
In three pigs, five measurements of CBF were made using radiolabeledmicrospheres; two pigs had four measurements. The relation between the twomethods is shown in Fig. 5. The fit within each animal as estimatedby R2 was good (R2 range 0.91 to 0.99). The gradientof the fits ranged from 0.5 to 1.8 (median 1.0). The mean difference was 5.7ml 100 gm–1 min–1 (median7.1, range –30.1 to 21.9).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Near-infrared spectra for hemoglobin and cytochrome oxidase have beenmeasured under carefully controlled conditions by several groups. 13–15 Spectrameasured in vivo and in vitro are identical and confirmed that the brain normallycontains only three mobile chromophores. 8,11 Because the exogenous tracermolecule ICG is strongly protein-bound in vivo, we have now measured absorptionspectra for ICG in the presence of plasma protein (see Table III).
|
6
The scattering properties of adult brain in vivo and postmortem neonatalbrain have been measured, and the extinction coefficients used in the conversionalgorithm are corrected for the measured variation in the scattering of lightby cerebral tissue at different wavelengths. 4 We have used the DPF measured in infants for children up to 2 yearsof age and for adults above that age based on an estimation of the amountof myelin in the developing brain. This will introduce some inaccuracy intoour results that will be corrected in the future when routine measurementsof the pathlength of light become available. Technology has recently beendeveloped that will allow pathlength to be measured routinely by phase modulationduring all NIRS procedures. 16
Measurements of changes in the attenuation of light at four wavelengthswere analyzed using a linear least squares technique. Other workers have demonstratedthat the errors induced by using this technique are very small. 17 System noise and systematic error are a small fractionof the total signal. 18
Measurement of CBF by NIRS.
The use of ICG as an intravascular tracer rests on several assumptions.It is assumed (1) that the tracer is not consumed by the organ, (2) that theflow to the organ remains constant, (3) that the total volume of blood inthe organ remains constant, (4) that there is no further admixture of bloodbeyond the point of [ICG]blood measurement in the bypass circuit,and (5) that the measurement is made over a time period less than the minimumtransit time of the tracer through the organ. These will be addressed in turn.
ICG is an inert substance that is not consumed by the brain.
The assumption of constant blood flow is required for all methods ofCBF measurement, but it poses less of a problem for the present method, wherea measurement requires only 4 to 6 seconds, than for conventional methodssuch as the Kety-Schmidt technique, or measurements of 133Xe clearance,which assume constant flow for 10 minutes or more, a condition unlikely tobe met during CPB.
Constant CBV is also assumed by other measurement techniques. NIRS measureschanges in total cerebral hemoglobin concentration, which is directly relatedto CBV. Thus any change in CBV during measurement of CBF can be observed andthe measurement disregarded. This is not possible with other methods for estimationof CBF.
Any admixture of blood beyond the point of [ICG]blood measurementwould cause CBF to be underestimated. However, this is extremely unlikelyto occur given the placement of the efferent catheter from the pump-oxygenator.
CBF measurements should be completed in less than the minimum transittime of blood through the organ for quantification of the total amount oftracer accumulated. Although minimum transit time was not measured in oursubjects, values obtained for MTT were much higher than the chosen valuesfor t, suggesting that only trivial volumes of blood would have passed throughthe brain in less than the measurement period.
Inhomogeneity of CBF.
The period of integration t is chosen with respect to the distributionof transit times within the organ and should be explicit because any specifict weights the measured value to a particular distribution of blood flow. 19 Low values for t will measurethe flow of blood that passes through the brain with a short transit time,whereas higher values for t will include components of total CBF with longertransit times. Comparison of values for CBF calculated for values of t between2 and 6 seconds 1,2 revealed predictable inhomogeneity of flow withinthe brain (
).
CBFs calculated over shorter periods were generally higher than flowcompartments of longer transit time. The value of CBF obtained will thus bemodulated by the value chosen for t. Measurements that use fewer points onthe tracer accumulation curve have larger variability, 19 and studies of adult volunteers using oxyhemoglobinas a tracer, 19 as well asinspection of , suggest that integrationmore than 4 seconds provides a stable and consistent measurement of globalCBF.
Cerebral blood flow and MTT.
Measurement of CBF requires only the first few seconds of tracer accumulation,but if the full period of tracer passage through the brain is observed, MTTcan be calculated by conventional methods. 20 The inverse relation between CBF and MTT predicted by the Stewart-Hamiltonequation was found and is shown in .
Comparison studies.
The coefficient of variation for the studies in children and in pigletswere comparable with values for other methods of measuring CBF. The fit withineach piglet as estimated by R2is good. However, the slopes of the fits are variable. It is known from measurementsin human beings that the DPF varies considerably between subjects. 4,16 However, to account for all the variation in slope found here, a standarddeviation of 50% of the mean DPF would be required. The R2 for each animal implies good agreement on changefor percentage flow change between the two methods across a wide range oftemperature and pump flow values. This would suggest that for one subjectthe DPF value remains constant across the range of CPB conditions even thoughit may be subject to intersubject variation.
Technical difficulties and assumptions exist for all CBF techniques,including those using microspheres, and these may become even more importantduring CPB. The microsphere technique measures regional CBF as the NIRS methodis also considered to do. An inherent variability exists in both the microspheretechnique (of 10% to 20%) and in the NIRS method that may account for someof the discrepancy between them. In addition, the NIRS technique may interrogatea variable contribution from tissues not accounted for in any of the microspheremeasurements, such as skull and skin.
Clinical value of CBF measurements during CPB.
A significant number of children and adults who undergo CPB operationssurvive with neurodevelopmental or intellectual impairment, probably causedby perioperative derangements of CBF. 21 The method described here should allow multiple measurements of CBFto be made during a single operating session and permit CBF, as well as othervariables such as CBV, MTT, cerebral oxygen, and glucose delivery, to be routinelymeasured. It may be possible to define critical levels of CBF that shouldtrigger appropriate responses from physicians caring for patients during surgeryto prevent ischemic injury from occurring.
Future developments.
The technique detailed here requires validation, perhaps comparing withthe Kety-Schmidt technique, under the special conditions of CPB surgery, bothin children and in adult patients. Improvements in the accuracy of the techniquecan be expected as NIRS technology improves, for example with the introductionof continuous measurement of the optical pathlength. 22 Use of NIRS allows the acquisition of other dataon cerebral oxygenation simultaneously during CBF measurements, 3 and there is also the prospect of regional flow estimationswhen NIRS localization becomes available. 23
We thank Dr. R. Ungerleider, Department of Surgery, Duke UniversityMedical Center, Durham, North Carolina, for his collaborative support; AngieWade, Department of Biostatistics, Institute of Child Health (UCL), London,United Kingdom, for statistical advice; and Pulsion Medizintechnik for theloan of a model IVH4 intravascular fibreoptic dye densitometer.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  W. M. Kuebler How NIR is the future in blood flow monitoring? J Appl Physiol, April 1, 2008; 104(4): 905 - 906. [Full Text] [PDF]
|
|
|
|
|
|
J. A. Guenette, I. Vogiatzis, S. Zakynthinos, D. Athanasopoulos, M. Koskolou, S. Golemati, M. Vasilopoulou, H. E. Wagner, C. Roussos, P. D. Wagner, et al. Human respiratory muscle blood flow measured by near-infrared spectroscopy and indocyanine green J Appl Physiol, April 1, 2008; 104(4): 1202 - 1210. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Meybohm, G. Hoffmann, J. Renner, A. Boening, E. Cavus, M. Steinfath, J. Scholz, and B. Bein Measurement of Blood Flow Index During Antegrade Selective Cerebral Perfusion with Near-Infrared Spectroscopy in Newborn Piglets Anesth. Analg., March 1, 2008; 106(3): 795 - 803. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. M. Hemmerling, J.-F. Olivier, F. Basile, N. Le, and I. Prieto Bispectral Index as an Indicator of Cerebral Hypoperfusion During Off-Pump Coronary Artery Bypass Grafting Anesth. Analg., February 1, 2005; 100(2): 354 - 356. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C Terborg, S Bramer, S Harscher, M Simon, and O W Witte Bedside assessment of cerebral perfusion reductions in patients with acute ischaemic stroke by near-infrared spectroscopy and indocyanine green J. Neurol. Neurosurg. Psychiatry, January 1, 2004; 75(1): 38 - 42. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Boushel, H. Langberg, J. Olesen, M. Nowak, L. Simonsen, J. Bulow, and M. Kjar Regional blood flow during exercise in humans measured by near-infrared spectroscopy and indocyanine green J Appl Physiol, November 1, 2000; 89(5): 1868 - 1878. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Hopton, T. S. Walsh, and A. Lee Measurement of cerebral blood volume using near-infrared spectroscopy and indocyanine green elimination J Appl Physiol, November 1, 1999; 87(5): 1981 - 1987. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
