Perfusion Cookbook 2005

Version 3.1
A Guide to Myocardial Perfusion Analysis During Adenosine Mediated Coronary Vasodilatation for Assessment of Myocardial Perfusion Using Cardiac MR Imaging Release 9 and higher
Exam cards can be found at: http://netforum.medical.philips.com/ German Heart Institute and www.cmr-academy. com Sebastian Kelle Bernhard Schnackenburg Rolf Gebker Ingo Paetsch Axel Bornstedt Eckart Fleck Eike Nagel

Purpose
Magnetic resonance perfusion imaging has shown promising results in preliminary studies and first larger patient trials. In principle the occurrence of myocardial perfusion deficits is a very sensitive indicator of ischemia in the presence of significant coronary artery stenoses. Most perfusion defects only occur during stress, such as pharmacological vasodilation. This can be optimally achieved using adenosine as pharmacological stress agents, which is proved to be safe and well tolerated. Techniques currently used in clinical routine such as SPECT and PET have the disadvantage of radiation and low spatial resolution, which prohibits the assessment of subendocardial ischemia. This cookbook provides instructions for the application and performance of perfusion measurements with cardiac magnetic resonance imaging.

pharmacons. However, coronary arteries with significant stenosis are already maximally dilated at rest (to allow maximal blood flow and compensate for the stenosis) and cannot be dilated any further. Thus, vasodilation with adenosine induces an increase of blood flow in myocardial areas supplied by normal coronary arteries ("coronary steal"), whereas no (or only minimal) change is found in areas supplied by stenotic coronary arteries. With adenosine a maximal coronary vasodilation can be achieved safely with an intravenous infusion at a rate of 140 g/kg/min.

Side Effects
The vasodilatory effect of adenosine may result in a mild-to-moderate reduction in systolic, diastolic and mean arterial blood pressure (< 10 mmHg) with a reflex increase in heart rate. Most patients complain about anginal chest pain. These effects, however, are transient and usually do not require medical intervention. Since adenosine exerts a direct depressant effect on the SA and AV nodes transient first-, second- and third-degree AV block and sinus bradycardia have been reported in 2.9%, 2.6% and 0.8% of patients. Also, adenosine can cause significant hypotension. Patients with intact baroreceptor reflex are able to maintain blood pressure in response to adenosine by increasing cardiac output and heart rate. Adenosine can also cause a paradoxical increase in systolic and diastolic blood pressure, which mostly develops in individuals with significant left ventricular hypertrophy. These increases are transient and resolve spontaneously. Because adenosine is a respiratory stimulant primarily through activation of carotid body chemoreceptors, intravenous administration showed increases in minute ventilation, reduction in arterial PCO2 and respiratory alkalosis. Approximately 14% of patients complain of dyspnea and an urge to breathe deeply during adenosine infusion.

Hard- and Software Requirements

The methodology has been developed for a Gyroscan ACS NT or INTERA 1.5 Tesla (T) whole body system equipped with a 30 mT Master gradient system (slew rate: 150 mT/m/s), using a 5-element cardiac synergy coil. Software: Release 9 and higher

Stress Agent
Pharmacon: concentration: administration: Adenosine preferably 5 mg/ml intravenous

Mode of Action
Adenosine, an endogenous nucleotide, is a potent vasodilator of most vascular beds, except for hepatic and renal arterioles. It exerts its pharmacological effect through the activation of purine A1 and A2 cell-surface adenosine receptors. The essence of the pharmacological mechanism lies in the inhibition of the slow inward Ca2+ current, thereby reducing calcium uptake, and in the activation of adenylate cyclase through A2 receptors in smooth muscle cells. Both, normal and stenotic coronary arteries are dilated to their maximum using these
German Heart Institute Berlin & cmr-academy.com

Safety Studies
Because of the above reported adverse effects, a number of studies have been carried out investigating the safety of intravenous adenosine infusions in different diagnostic modalities of cardiac imaging. So far, there is evidence accumulated in over

10,500 patients studied in thallium radionuclide imaging, echocardiography, SPECT and MRI which shows that pharmacological stress with adenosine presents a safe method of acquiring stress imaging data. Safety of an adenosine infusion at 140 mcg/kg/min was evaluated during radionuclide imaging of 9,256 consecutive patients. The infusion protocol was completed in 80% of patients, required dose reduction in 13% and was terminated early in 7%. Interpretable imaging studies were obtained in 98.7% of patients, and 0.8% of patients received aminophylline. Minor and well tolerated side effects were reported in 81.1% of patients. There were no deaths, one myocardial infarction, seven episodes of severe bronchospasm and one episode of pulmonary edema. Transient AV node block occurred in 706 patients (first-degree in 256, seconddegree in 378 and third-degree in 72) and resolved spontaneously in most patients (n = 508) without alteration in the adenosine infusion. There were no sustained episodes of AV block.

with high-grade AV block or sinus node dysfunction. Adenosine should be used with caution if a patient is receiving any medications that already depress the sinus node and/or AV node (e.g. beta-blockers, calcium channel blockers, cardiac glycosides). Adenosine should be discontinued in patients who develop persistent or symptomatic highgrade block or significant drop in systolic blood pressure (>20 mmHg). The drug should be discontinued in case of persistent or symptomatic hypotension. Also, adenosine should be used with caution in patients with autonomic dysfunction, stenotic valvular heart disease, pericarditis and pericardial effusion, stenotic carotid artery disease, cerebrovascular insufficiency and uncorrected hypovolemia. Adenosine infusion should be exercised with caution in patients with obstructive lung disease not associated with bronchoconstriction (emphysema, bronchitis, etc), and should be avoided in patients with bronchoconstriction or bronchospasm (e.g. asthma). If a patient develops severe respiratory difficulties, adenosine should be immediately discontinued.

Contraindications
Adenosine should be used with caution in patients with pre-existing AV block or bundle branch block and should be avoided in patients

Table 1: Adenosine stress protocol

Pharmacon Patient instruction Adenosine No caffeine (tea, coffee, chocolate, (perfusion) etc.) or medications such as
aminophylline or nitrates 24 hours prior to the examination

Stress protocol

Antidote

140 g/kg/min for a Stop infusion ! maximum of 6 minutes. (if necessary: aminophylline (half-life 4-10 seconds) 250 mg i.v. slowly injected under ECG monitoring)

German Heart Institute Berlin & cmr-academy.com

Contrast Media for Perfusion

Drugs:
A gadolinium derivative is used (e.g. Gd-DTPA; gadobenate dimeglumine (Gd-BOPTA); gadodiamide (Gd-DTPA-BMA) etc.) applied as an intravenous bolus. The cookbook protocol is based on extensive experience with Gd-DTPA; correction of dosage regimen is necessary for Gd-derivatives with different relaxivity.

the radiofrequency cage, or by using special CMR compatible equipment. A defibrillator and all medications for emergency treatment must be available at the CMR site.

Table 3: Contraindications for adenosine

Myocardial infarction <3 days Unstable angina pectoris Severe arterial hypertension Asthma or severe obstructive pulmonary disease requiring treatment AV-block >IIa autonomic nerve dysfunction stenotic valvular disease cerebrovascular insufficiency any obstructive lung disease comedication with beta-blockers, Caantagonists or cardiac glycosides (due to AV / sinus node depression)

Dosage for Gd-DTPA:

0.05 mmol/kg bodyweight injected with 4 ml per second. The bolus is followed by a 20 ml saline flush (infusion rate: 4 ml per second) to facilitate a compact bolus passage. We recommend the use of an automatic infusion system ( e.g. Medrad, Spectris MR injector) for exact dosage and timing.

Caution:

Monitoring and Safety

During stress examinations monitoring of the patient within the magnet is mandatory. In general, monitoring during a MR examination requires the same precautions and emergency equipment as any other stress examination. Specific recommendations are listed in table 2. Apart from the known specific contraindications for MR, the drug related contraindications for adenosine infusion are listed in table 3.

Table 4: Termination criteria

persistent or symptomatic AV block significant drop in systolic blood pressure (> 20 mmHg) persistent or symptomatic hypotension severe respiratory difficulty

Image Interpretation
Visual Assessment
Currently, only limited data is available regarding the accuracy of visual assessment and extensive experience is required to reach an acceptable standard. The alteration of the upslope of the myocardial response curve from stress to rest yields the highest difference between ischemic and normal myocardium. This parameter is superior to maximal signal intensity, which is mainly assessed visually. In our experience a relatively high dose of contrast agent results in improved visual assessment, but often renders semiquantification difficult. Thus, several aspects for visual interpretation need to be taken into account: Imaging artefacts can obscure (e.g. wraparound) or misleadingly be interpreted (e.g. susceptibility) as perfusion deficits. Thus, training in MR image interpretation together with the interplay of the visual criteria given in table 5 will result in a sufficiently high diagnostic accuracy (own unpublished data showed: sensitivity 89%, specificity 80%).

Table 2: Monitoring requirements adenosine stress MR imaging

Heart rate and rhythm Blood pressure Pulse oximetry*
(for rhythm monitoring)

for

Continuously Every minute Continuously Continuously

Symptoms *When the Vector-ECG

required.

is used, pulse oximetry is not

Although adverse events are rare, preparation and practice for rapid removal of the patient from the magnet needs to be practiced in addition to a close compliance with the termination criteria (Table 4). The monitoring of blood pressure, cardiac rhythm and patients' symptoms can either be done by placing standard equipment outside the scanner room connected to the patient with special extensions through a waveguide in
German Heart Institute Berlin & cmr-academy.com

The main artefacts occurring during the initial passage of the contrast bolus are due to susceptibility at the endocardium-bloodpool (=contrast media) interface, occasionally rendering diagnosis of subendocardial perfusion deficits difficult. Especially the trabeculae of the papillary muscles reaching into the left ventricular cavity are washed with contrast agent and, thus, show almost always susceptibility artefacts. Such findings should not be interpreted as regional ischemic perfusion abnormality. Visual criteria for left ventricular myocardial perfusion deficits are given in table 5.

Table 5: Criteria for Visual Assessment of Regional Myocardial Perfusion Deficits During First-Pass of Contrast Agent

Signal-intensity Pattern & Location:

Dark regions arising from the subendocardium with usually irregular intramyocardial border or with completely transmural extent are highly suspicious.

Dynamic Myocardial Filling Pattern:

Semiquantification
Most literature has been published on semiquantification, as described briefly: the endo- and epicardial contours of left ventricular myocardium are traced and corrected manually for changes of diaphragmatic position due to breathing or diaphragmatic drift. Care needs to be taken to place the contours on the myocardium and to exclude the left ventricular cavity and the pericardium. The myocardium is then divided into 6 equiangular segments per slice and numbered clockwise beginning with the anterior septal insertion of the right ventricle. An additional region of interest is placed within the cavity of the left ventricle excluding the myocardial segments and the papillary muscles. Images acquired after premature ventricular beats or insufficient cardiac triggering need to be excluded from the analysis to guarantee steady-state conditions. Signal intensity is determined for all dynamics and segments. The upslope of the resulting signal intensity time curve is determined by the use of a linear fit. To correct for possible differences of the input function, the results of the myocardial segments are corrected by dividing the upslope of each myocardial segment by the upslope of the left ventricular signal intensity curve. Perfusion reserve index is calculated by dividing the results of stress by the results at rest. This approach has yielded sensitivities and specificities of > 90% in selected patient populations. Its value in unselected patients remains to be determined.

Initially, slow or missing enhancement persistent over a few (2 to 10) dynamics with a consecutive signal intensity increase starting from the defect`s epicardial border (epicardial "filling up" of the defect) is typical for regional perfusion abnormalities.

Myocardial Distribution of the Defect:

Evaluate the equatorial slice first, then check whether the suspected perfusion defect can be followed in corresponding segments of the apical or basal slice.

Comparison

If a regional defect is found in the stress scan, but not in the rest scan, inducible ischemia is confirmed ("dynamic lesion"). Regional persistence of the perfusion deficit shows myocardial scar ("static lesion").

Stress

vs.

Rest:

Quantification
Quantification depends on a number of prerequisites, not fully fulfilled with the current contrast agents such as: linearity between signal intensity and contrast agent concentration or adequate downslope. Most centers have not been able to reproduce methods proposed for quantification. Problems due to background correction, image inhomogeneities, water exchange and magnetization transfer remain unsolved.

German Heart Institute Berlin & cmr-academy.com

Protocol Overview
To cover 16 segments we use 3 short axis views. With this approach the apical segment (segment 17) is neglected (for left ventricular segmentation see "DSMR cookbook"). The study includes the following scans, except for the multistack survey they are all breathhold bTFE scans (scan duration ranging from 8 to 12 sec.): (1) (2) (3) (4) (5) (6) (7) (8) multistack survey (bTFE) single-angulated view double-angulated view wall motion scan short axis (3 slices) 4-chamber view 2-chamber view 3-chamber view Perfusion scan (3 slices)

Scan Procedure:
Scan 1: Multistack survey, cardiac coil all elements. Look at the images and check if the coil is optimally positioned. Scan 2: Single-angulated view. Define the plane on transversal slices parallel to the septum through the apex of the left ventricle and the coaptation point of the mitral valve. Scan 3: Repeat Scan 2. Flip the orientation (90) and adjust the plane on the first RAO through the apex and the middle of the mitral valve to get a second long axis view (nearly 4 chamber view). This slice orientation helps to prevent any angulation errors while planning the short axis views. Scan 4: Make use of the double-angulated image to define 3 slices perpendicular to the long axis of the heart representing the short axis geometry.
Note: Under stress conditions even the normal heart
experiences a change in its basal-to-apex dimensions. To avoid visualization of the left ventricular outflow tract as well as to ensure sufficient imaging of the left ventricular cavity (esp. critical is the apical slice), we recommend to perform the planning on the endsystolic images: divide the distance from the apical epicardial border to the mitral valve plane in 5 equal parts. Then, distribute the 3 short axes equally within the inner three-fifth of the distance with adaptation of slice gap.

The perfusion scan will be performed during stress and at rest (after an equlibration time 15 min after the first bolus injection).

Step by Step Protocol

Patient Preparation
It is of special importance to explain not only the course of the examination to the patient but also the breathhold procedure. Generally the breathhold should be performed during endexpiration to ensure reproducible slice geometry. The perfusion scan consists of two breathhold periods. The first is a short one (about 10-12 seconds, baseline acquisition of myocardial intensity), then the patient is asked to inhale and exhale once and hold breath as long as possible. Before starting this breathholdcommand the contrast bolus is administered. The patient should stop breathing for at least 25-30 seconds resulting in a fixed slice geometry during the first-pass of the contrast agent; in case the patient cannot hold his breath throughout the whole scan: ask the patient to inhale and exhale once and hold breath again. Put venous line ( 18 gauge) in cubital vein with two connections: one for adenosine, one for contrast media. Monitor blood pressure and heart frequency on contralateral arm.

Scan 5: Plan the 4-chamber view on the equatorial short axis view, the stack should be aligned through the apex of the right ventricle and the papillary muscle. Scan 6: Plan the 2-chamber view on the previously acquired 4-chamber view by just switching the slice orientation and adjust the angulation (through the left ventricular apex and the coaptation point of the mitral valve). Scan 7: Plan the 3-chamber view on the basal short axis view that displays the mitral valve opening, and the LV outflow tract towards the aortic valve. Scan 8: Stress perfusion. The perfusion scan is performed with an orientation identical to the short axis cine previously acquired (scan 4). Wraparound has to be avoided carefully! We recommend to perform a bTFE prior to the start of the adenosine infusion (e.g. 5 dynamics) with a breathhold command If necessary enlarge the field-of-view (results in decreasing resolution). Scan 9: Repeat scan 8 at rest after 10 - 15 min delay.

German Heart Institute Berlin & cmr-academy.com

Flow chart 1: Adenosine Stress MR Perfusion Imaging

Perfusion Flowchart
Survey
(1) (2) (3) Transversal Single angulated Double angulated

Rest Function
(4) (5) (6) (7)

Short axis (apical, mid, basal) 4-chamber 2-chamber 3-chamber

Test Perfusion

Slice geometry identical to (4) Acquisition of 5 dynamic images NO contrast agent, NO adenosine
Carefully exclude wraparound, check for artifacts. Change foldover direction, enlarge FOV if needed. Repeat scan until satisfied

Continuous RR and rhythm monitoring

Start adenosine 140 g/kg/min

Wait 3-4 minutes Adapt heart rate

(8) Stress Perfusion

50 70 dynamics

Stop adenosine
Wait 10 - 15 minutes Adapt heart rate

(9) Rest Perfusion

50 70 dynamics

German Heart Institute Berlin & cmr-academy.com

Survey MST bTFE Geometry Coil selection/ch. connection FOV(mm)/RFOV Matrix Scan/Reconstruction/Scan% Slices/thickness(mm)/gap Stacks/Type Contrast ScanMode/Technique/Conrast/Fast Imaging Shot mode Partial Echo TE(ms)/Flip(deg)/TR(ms) Half Scan Water fat shift Shim Motion Cardiac synchronization/device Flow Comp./NSA Dyn/Angio Angio/Q.Flow/Dyn. Study Manual start Post Processing Recon Mode real time no/no/no yes no no/1 M2D/FFE/balanced/TFE Single shot no shortest/50/shortest no min auto Syn-Cardiac/12345 450/100% 192/256/50 20/10/0 3/parallel

Angulated and double angulated survey Geometry Coil selection/ch. connection FOV(mm)/RFOV Matrix Scan/Reconstruction/Scan% Slices/thickness(mm)/gap Stacks/Type SENSE*/p reduction Contrast Scan Mode/Technique/Contrast/Fast Imaging Shot mode Partial Echo TE(ms)/Flip(deg)/TR(ms) Half Scan Water fat shift Shim Motion Cardiac synchronization/device
German Heart Institute Berlin & cmr-academy.com

Syn-Cardiac/12345 400/90% 176/256/100 1/8/1/parallel yes/2 (optional) M2D/FFE/balanced/TFE default no shortest/60/shortest no min volume retospective/ECG

Card. Freq./RR%/#Phase Phase percentage Arrhythmia rejection Respiratory compensation/slices per bh Flow Comp./NSA Dyn/Angio Angio/Q.Flow/Dyn. Study Post Processing Recon Mode *Alternatively: Half scan = yes

70/15,15/25 67 % yes breath hold/1 no/1 no/no/no real time

Cine SA/4CH/2CH/3CH bFFE Geometry Coil selection/ch. connection FOV(mm)/RFOV Matrix Scan/Reconstruction/Scan% Slices/thickness(mm)/gap Stacks/Type SENSE*/p reduction Contrast Scan Mode/Technique/Contrast/Fast Imaging Shot mode Partial Echo TE(ms)/Flip(deg)/TR(ms) Half Scan Water fat shift Shim Motion Cardiac synchronization/device Card. Freq./RR%/#Phase Phase percentage Arrhythmia rejection Respiratory comp./slices per bh Flow Comp./NSA Dyn/Angio Angio/Q.Flow/Dyn. Study Post Processing Recon Mode Measured Voxel Size/mm *Alternatively: Half scan = yes real time 2x2x8 no/no/no retrospective/ECG 70/15, 15/25 50 yes breath hold/1 no/1 M2D/FFE/balanced/TFE Default No shortest/60/shortest no min volume Syn-Cardiac/12345 380/90% 192/256/110 1-3/8/user def 1/parallel yes/2 (optional)

German Heart Institute Berlin & cmr-academy.com

Perfusion balanced TFE Geometry Coil selection/ch. connection FOV(mm)/RFOV Matrix Scan/Reconstruction/Scan% Slices/thickness(mm)/gap Stacks/Type SENSE/p reduction Contrast Scan Mode/Technique/contrast/Fast Imaging Partial Echo TE(ms)/Flip(deg)/TR(ms) Half Scan Shot mode Water fat shift Shim TFE prepulse/shared/delay (ms) Motion Cardiac synchronization/device Card. Freq./RR%/#Phase Trigger delay (ms) Flow Comp./NSA Dyn/Angio Angio/Q.Flow/Dyn. Study Dyn scans/dyn scan times Post Processing Recon Mode Measured Voxel Size/mm CAVE (!): Check Info Page TFE shot intervall (beats) 1 (!!), (if not, adjust trigger delay !!!) real time 2.8 x 2.9 x 10 no/no/yes 70/shortest trigger/ECG 70/10,10/1 longest no/1 MS/FFE/balanced/TFE no shortest/50/shortest no single shot min volume saturate/no/100 Syn-Cardiac/12345 360/95 128/256/95 3/10/user def 1/parallel Yes/ 2.3

This cookbook has been assembled from the knowledge available at the time of writing. The authors cannot take liability for dose regimen, infusion schemes, etc. If you find any errors or would like to suggest any improvements, please let us know at [email protected] or [email protected].

German Heart Institute Berlin & cmr-academy.com

10