- Email: [email protected]
Morphologic Validation of Reperfused Hemorrhagic Myocardial Infarction by Cardiovascular Magnetic Resonance
Morphologic Validation of Reperfused Hemorrhagic Myocardial Infarction by Cardiovascular Magnetic Resonance Cristina Basso, MD, PhDa, Francesco Corbetti, MDb,*, Caterina Silva, MD, PhDc, Aierken Abudureheman, MD, PhDa, Carmelo Lacognata, MDb, Luisa Cacciavillani, MDc, Giuseppe Tarantini, MD, PhDc, Martina Perazzolo Marra, MDc, Angelo Ramondo, MDc, Gaetano Thiene, MDa, and Sabino Iliceto, MDa The purposes of this study were to assess the ex vivo cardiovascular magnetic resonance (CMR) signals of pathologically proved hemorrhagic myocardial infarction (MI) and to correlate these with in vivo CMR findings. Late gadolinium hypoenhancement within a hyperenhanced area in reperfused acute MI is ascribed to severe microvascular obstruction. The hearts of 2 patients, who died from cardiogenic shock after acute MIs and who had undergone coronary recanalization and in vivo CMR, were examined by T2 and T1 late enhancement sequences as well as by gross and histologic investigation. Four corresponding short-axis slices of each cardiac specimen from the base to the left ventricular apex were selected to assess the extent of MI and hemorrhage and were compared with the in vivo T2 and late enhancement CMR scans. On pathologic examination, the extent of MI was 57 ⴞ 30% and 44 ⴞ 24%, and the extent of hemorrhage was 23 ⴞ 13% and 19 ⴞ 8% of the left ventricular area, respectively, showing progressive increases from the base to the apex. The low-signal intensity areas observed by ex vivo T2 CMR strongly correlated with the hemorrhage quantified on histology (R ⴝ 0.93, p ⴝ 0.0007). Using ex vivo late gadolinium sequences, bright areas surrounded by thin dark rims, consistent with magnetic susceptibility effects, were detected, corresponding with hemorrhage. On in vivo CMR images, low-signal intensity and hyperintense areas with peripheral susceptibility artifacts were observed within the MI core on T2 and late gadolinium sequences, respectively. In conclusion, in reperfused MI, CMR hypointense T2 signal and susceptibility effects within the late gadolinium hypoenhanced areas are consistent with interstitial hemorrhage due to irreversible vascular injury, as proved by pathologic study. © 2007 Elsevier Inc. All rights reserved. (Am J Cardiol 2007;100:1322–1327) Reperfused myocardial infarction (MI) frequently appears reddish on morphologic examination because of intramyocardial hemorrhage, which is believed to be caused by vascular cell damage with the leakage of blood from the injured vessels.1– 6 The clinical implications of hemorrhagic versus white infarcts remain undetermined, and this is also related to the lack of reliable and reproducible imaging modalities to assess its presence
in vivo. Cardiovascular magnetic resonance (CMR) can provide a wide range of information in acute MI by detecting infarct size, microvascular obstruction, and edema.7–12 However, to date, CMR findings in acute hemorrhagic MI have not been fully elucidated, and no correlations have been reported among late gadolinium images, T2 images, and histologic findings in humans. We present the first such correlations from the cardiac specimens of 2 patients who underwent CMR and died of cardiogenic shock after acute MIs.
a
Department of Medic-Diagnostic Sciences and Special Therapies, bService of Radiology, and c Department of Cardio, Thoracic, and Vascular Sciences, University of Padua Medical School, Padua, Italy. Manuscript received April 4, 2007; revised manuscript received and accepted May 22, 2007. This work was supported by the Registry of Cardio-Cerebrovascular Pathology, Veneto Region, Venice, Italy. *Corresponding author: Tel: 0039-049-8211982; fax: 0039-049-8272284. E-mail address: [email protected] (F. Corbetti).
Case Descriptions: CMR and Pathology Correlates Patient 1, a 68-year-old man, presented to the emergency room 5 hours after the onset of acute chest pain. Electrocardiography showed ST-segment elevation in leads V2 to V6 and I to aVL. Thrombolysis with recombinant tissue plasminogen activator was performed (time to treatment 330 min-
0002-9149/07/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2007.05.062
utes). The peak troponin I level was 971 g/L at hour 13. Echocardiography revealed akinesia of the septum and the anterior midapical and lateroapical walls, with an end-diastolic volume of 60 ml/m2 and an ejection fraction of 42%. Coronary angiography performed 1 day later showed a 90% stenosis of the mid left anterior descending coronary artery, and stenting was successfully performed (baseline Thrombolysis In Myocardial Infarction [TIMI] flow grade 2, baseline blush grade 0; final TIMI flow grade 3, final blush grade 0). CMR was performed on day 9, and the patient died on day 12 from cardiogenic shock. Patient 2, a 60-year-old woman, presented to the emergency room with new onset of acute chest pain. Electrocardiography showed ST-segment elevation in leads V1 to V4 and aVR and ST-segment depression in leads I, II, III, and aVF. Coronary angiography revealed a 100% occlusion of the mid left anterior descending coronary www.AJConline.org
Case Report/Hemorrhagic Myocardial Infarction by Magnetic Resonance
1323
Figure 1. Patient 1: transmural anteroseptal and lateral hemorrhagic MI. Gross (A, B, C, D), histologic (E, F, G, H) (Heidenhain trichrome stain), and ex vivo T2 CMR (I, L, M, N) short-axis sections of the heart, from the base to the apex. Note the perfect overlap between the hemorrhage seen on pathologic examination and the low-signal intensity areas observed on ex vivo T2 CMR.
artery and a 90% stenosis of the left circumflex, and stenting was successfully performed in the 2 arteries (time to treatment 5 hours; baseline TIMI flow grade 0, baseline blush grade 0; final TIMI flow grade 3, final blush grade 0). The peak troponin I level was 369.43 g/L on the first day. Echocardiography revealed akinesia of the apex, septum, and anterior wall, with an end-diastolic volume of 59 ml/m2 and an ejection fraction of 46%. The patient underwent CMR on day 7 and was discharged home on day 11, but on day 24 she died from cardiogenic shock. CMR in vivo images were acquired during repeated breath-holds on a 1.0-T body magnet (Harmony; Siemens Medical Systems, Erlangen, Germany) using cardiologic software (MRease SYNGO 2002B; Siemens Medical Systems) and a 4-channel phased-array receiver coil. A standard imaging protocol was used, including cine steady-state free precession sequences in 4 short-axis (from the base
to the apex) and 3 long-axis (2, 3, and 4 chamber) views for functional evaluation, T2-weighted fat saturation black-blood turbo spin-echo sequences (repetition time 2 to 3 RR intervals, echo time 92 ms, mean voxel size 1.6 ⫻ 1.4 ⫻ 7 mm), and myocardial delayed enhancement (MDE) sequences (segmented inversion recovery [IR] turbo fast low-angle shot, repetition time 1 to 2 RR intervals, echo time 6 ms, flip angle 30°, mean voxel size 1.8 ⫻ 1.3 ⫻ 8 mm, inversion time optimized in each patient, range 190 to 220 ms) performed 10 to 15 minutes after the administration of 0.2 mmol/kg of a gadolinium-based contrast agent (MultiHance; Bracco Diagnostics, Inc., Milan, Italy). The excised hearts were enveloped in a surface flexible coil (Flex Large; Siemens Medical Systems) and serial short-axis views from the base to the apex plus 3 long-axis views, as in vivo, were acquired using T2- and T1weighted sequences. For T2 study, a standard turbo spin-echo sequence
without blood preparation was used (repetition time 2,400 ms, echo time 104 ms, field of view 160 mm, mean voxel size 0.7 ⫻ 0.6 ⫻ 3 mm). For T1-weighting purposes, the same IR turbo fast low-angle shot sequence used for MDE images in vivo was applied, adapting the parameters to achieve better resolution (repetition time 1,200 ms, echo time 7 ms, flip angle 30°, field of view 180 mm, mean voxel size 0.8 ⫻ 0.6 ⫻ 3 mm, inversion time 150 ms [a value commonly used in conventional IR sequences to suppress fat signal]). As previously reported, hyperintense and dark areas on ex vivo T2 images were assumed to indicate infarct size and hemorrhage, respectively.13,14 CMR images were analyzed quantitatively by the user-interactive program Cine-Tool (GE Medical Systems, Waukesha, Wisconsin), manually tracing the endocardial and epicardial borders for left ventricular mass computation. MI and hemorrhage extent were then quantified by manually tracing the borders of the hyperintense
1324
The American Journal of Cardiology (www.AJConline.org)
Figure 2. Patient 2: transmural anteroseptal and lateral hemorrhagic MI. Gross (A, B, C, D), histologic (E, F, G, H) (Heidenhain trichrome stain), and ex vivo T2 CMR (I, L, M, N) short-axis sections of the heart, from the base to the apex. Note the perfect overlap between the hemorrhage seen on pathologic examination and the low-signal intensity areas observed on ex vivo T2 CMR.
and dark areas, respectively, and were expressed as percentages of total left ventricular area in each slice. Four short-axis myocardial slices of the excised formalin-fixed hearts (base, midbase, midapical, and apical) correlating with CMR slices were obtained, photographed, and embedded with paraffin wax. Three tissue sections 7 m thick were cut and stained with hematoxylin-eosin, Perl’s Prussian blue (to visualize ferric iron in hemosiderin) and Heidenhain trichrome, respectively. Short-axis slices of the heart stained with Heidenhain trichrome were analyzed by quantitative computerized morphometry (Image-Pro Plus version 4.0; Media Cybernetics, Inc., Bethesda, Maryland). Left ventricular area was obtained by manually tracing the endocardial and epicardial borders; hemorrhage and MI were quantified by manual tracing and were expressed as percentages of total left ventricular area in each slice. Analysis was performed independently for the histopathologic (CB, GT) and CMR (FC, CS) studies, with the
researchers for histopathology blinded to the CMR results and vice versa. Linear regression was used to compare the extent of MI and hemorrhage on T2 ex vivo CMR and pathology. Cardiogenic shock with acute pulmonary edema and peripheral congestion was found to be the cause of death at autopsy. Pathologic examination of the 2 hearts revealed hemorrhagic transmural septal and anterolateral MIs, in the setting of patent coronary artery stents (Figures 1 and 2). Hemorrhage was confined to the zone of necrosis and was massive at the core of MI, with diffuse and confluent packed red blood cells, interspersed between myocytes with coagulation necrosis in patient 1, whereas an almost complete removal of dead myocytes was seen in patient 2. Moving toward the periphery, hemorrhage was slight to moderate, with patchy distribution. In patient 2, at the border between the hemorrhagic core and the healing organizing tissue, patchy hemosiderin deposits were detected (Figure 3). In both instances, the total
destruction of the microvasculature within the MI core was evidenced. The MI extent was 57 ⫾ 30% and 44 ⫾ 24% of the left ventricular area, with progressive increases from the base to the apex (ranges 29% to 88% and 16% to 68%, respectively). Hemorrhagic extent was 23 ⫾ 13% and 19 ⫾ 8% of the left ventricular area, also showing progressive increases from the base to the apex (ranges 12% to 40% and 9% to 27%, respectively). T2-weighted images showed areas of signal hypointensity inside the hyperintense areas (Figures 1 and 2). The MI extent and the area of hemorrhage (hypointense regions or “dark areas”) measured by ex vivo T2 CMR strongly correlated with the pathologic quantification (Figure 4). Using ex vivo MDE sequences, bright–signal intensity areas surrounded by dark rims, consistent with a magnetic susceptibility effect, were visible within the MI regions. Compared with histopathologic findings, the bright signal and the dark border were found to correspond to areas of massive hem-
Case Report/Hemorrhagic Myocardial Infarction by Magnetic Resonance
1325
Figure 3. Patient 2: on ex vivo T2 (A) and IR MDE (B) images, a perfect correspondence with histologic findings is visible (C), showing 2 areas of massive intramyocardial hemorrhage in the septum (hematoxylin-eosin ⫻5); (D) close-up of the upper boxed area showing packed extravasated intact red blood cells (hematoxylin-eosin ⫻31); (E) neovascularized organizing tissue with patchy hemosiderin deposits at the border close to hemorrhage (Perl’s Prussian blue ⫻31).
orrhage with red blood cell extravasation, with additional hemosiderin deposits at the periphery in patient 2 (Figure 3). As expected, high-signal intensity on in vivo T2-weighted sequences was found in the MI area, reflecting tissue edema. However, low-signal areas were also detectable within the edematous myocardial wall and correlated well in location and extent with the hemorrhage seen on the ex vivo T2 images and proved by histology (Figures 1 to 3 and 5). On in vivo MDE images, a dark hypoenhanced core resembling persistent microvascular obstruction was present within the hyperenhanced MI. However, on closer evaluation, a dark rim surrounding tiny central areas of higher signal intensity was visible, with susceptibility
effects similar to those observed on ex vivo images (Figure 5). Comments By clinicopathologic study in 2 patients who died of cardiogenic shock after reperfused acute MIs, we provide evidence that hemorrhagic MI accounts for CMR late gadolinium hypoenhancement associated with magnetic susceptibility effects and hypointense T2 signal, thus casting some doubt on the belief that low-signal intensity areas surrounded by bright zones on late enhancement CMR always represent noreflow zones. Herein, we demonstrate that an alternative explanation for this phenomenon may be hemorrhage within the infarct core. In the heart, the ability of CMR to
detect hemorrhage has been reported by Lotan et al13,14 using ex vivo T2weighted spin-echo sequences in a canine model of acute MI and by Ochiai et al15 using CMR with a T2 star gradient-echo sequence in humans. In the latter study, ex vivo T2 star CMR was performed in 1 patient who died 3 days after the procedure, showing an overlap between the hypointense zone and the myocardial hemorrhage visible in the pathologic specimen. Our data agree with these previous findings, showing a perfect overlap among the location, spatial extent, and shape of the hypointense regions by ex vivo T2 CMR images and the hemorrhage by pathologic investigation. The possibility of visualizing myocardial hemorrhage as hypointense signal by T2 sequences in vivo is also pro-
1326
The American Journal of Cardiology (www.AJConline.org)
Figure 4. Comparison of CMR T2 sequences and pathology by linear regression analysis: MI size (top) (R ⫽ 0.93, p ⫽ 0.0009) and hemorrhagic extent (bottom) (R ⫽ 0.93, p ⫽ 0.0007) showed a positive correlation. LV ⫽ left ventricular.
Figure 5. Patient 2: in vivo versus ex vivo CMR. (A) In vivo T2-weighted CMR, 4-chamber views: a hypointense rim (asterisk) is clearly seen within the hyperintense edematous septum. (B) In vivo IR MDE sequence, long-axis view: a complex signal pattern is seen within the MI core of infarcted septum, with tiny bright areas surrounded by a dark rim (arrows), a feature typical of the susceptibility effect. (C) Ex vivo IR MDE sequence, long-axis view: the same susceptibility effect as in (B) is clearly seen around the septal hemorrhage but is more evident because of the higher resolution of the image and the additional presence of hemosiderin, which normally occurs only at this stage of hemorrhage (i.e., 24 days after MI onset).
vided, given the correspondence of T2 in vivo and ex vivo images. In this respect, Asanuma et al16 considered markers of myocardial hemorrhage not only the low-signal intensity area within the risk area by gradient-echo acquisition but also the hypoenhanced
area within a high-signal intensity zone by gadolinium-enhanced spin echo CMR. However, their definition of hemorrhage was based only on the CMR findings, without any histopathologic validation. Our CMR findings and their corre-
lation with histology indicate for the first time that peculiar signal features of late gadolinium hypoenhancement within the MI core can be related to hemorrhage. In our patients, the highsignal intensity within the hemorrhage on ex vivo T1 IR MDE sequence was typical of the paramagnetic effect of intracellular metahemoglobin17–19 and was observed also in vivo on close evaluation of the late hypoenhanced area. Furthermore, on ex vivo T1 IR MDE study, the bright signal within the hemorrhagic areas was surrounded by a thin dark rim, a finding that is observed at the periphery of hematomas not only on T2 sequences17 but also on T1 sequences,20 especially by using gradient-echo and low-bandwidth sequences,19 as we did in our cases. This feature can be related to the compartmentalization of fresh hemorrhage and the presence of metahemoglobin within intact packed red cells, resulting in a susceptibility effect at the periphery of hematoma with signal loss; in patient 2, who died 24 days after MI, this effect on ex vivo images was enhanced by the peripheral deposition of hemosiderin, the latest breakdown product of hemoglobin that appears first at the periphery of hemorrhage in the subacute or chronic stage and induces strong susceptibility effects due to its superparamagnetic properties.17–20 Of course, these findings must be interpreted as preliminary data, because they are the results of case reports of only 2 patients. However, this study was not an experimental but an observational investigation of human hearts, supported by detailed postmortem ex vivo CMR and pathologic studies in 2 patients who underwent in vivo T2 and late enhancement CMR and died from cardiogenic shock. Experimental studies in animals subjected to coronary occlusion and reperfusion as well as further human correlative studies are needed to assess the relative role of severe microvascular obstruction compared with hemorrhage on persistent late gadolinium hypoenhancement in reperfused acute MI. Acknowledgment: We are deeply indebted to Mauro Pagetta for technical assistance.
Case Report/Hemorrhagic Myocardial Infarction by Magnetic Resonance 1. Reimer KA, Lowe JE, Rasmussen MM, Jennings RB. The wavefront phenomenon of ischemic cell death. 1. Myocardial infarct size vs duration of coronary occlusion in dogs. Circulation 1977;56:786 –194. 2. Fishbein MC, Rit J, Lando U, Kanmatsuse K, Mercier JC, Ganz W. The relationship of vascular injury and myocardial hemorrhage to necrosis after reperfusion. Circulation 1980; 62:1274 –1279. 3. Kloner RA, Rude RE, Carlson N, Maroko PR, DeBoer LW, Braunwald E. Ultrastructural evidence of microvascular damage and myocardial cell injury after coronary artery occlusion: which comes first? Circulation 1980;62:945–952. 4. Roberts CS, Schoen FJ, Kloner RA. Effect of coronary reperfusion on myocardial hemorrhage and infarct healing. Am J Cardiol 1983; 52:610 – 614. 5. Garcia-Dorado D, Théroux P, Solares J, Alonso J, Fernandez-Avilés F, Elizaga J, Soriano J, Botas J, Munoz R. Determinants of hemorrhagic infarcts. Histologic observations from experiments involving coronary occlusion, coronary reperfusion, and reocclusion. Am J Pathol 1990;137:301–311. 6. Basso C, Thiene G. The pathophysiology of myocardial reperfusion: a pathologist’s perspective. Heart 2006;92:1559 –1562. 7. Kim RJ, Fieno DS, Parrish TB, Harris K, Chen EL, Simonetti O, Bundy J, Finn JP, Klocke FJ, Judd RM. Relationship of MRI delayed contrast enhancement to irreversible injury, infarct age, and contractile function. Circulation 1999;100:1992–2002. 8. Wu KC, Zerhouni EA, Judd RM, LugoOlivieri CH, Barouch LA, Schulman SP, Blumenthal RS, Lima JA. Prognostic signifi-
9.
10.
11.
12.
13.
cance of microvascular obstruction by magnetic resonance imaging in patients with acute myocardial infarction. Circulation 1998;97:765–772. Simonetti OP, Kim RJ, Fieno DS, Hillenbrand HB, Wu E, Bundy JM, Finn JP, Judd RM. An improved MR imaging technique for the visualization of myocardial infarction. Radiology 2001;218:215–223. Lund GK, Stork A, Saeed M, Bansmann MP, Gerken JH, Muller V, Mester J, Higgins CB, Adam G, Meinertz T. Acute myocardial infarction: evaluation with first-pass enhancement and delayed enhancement MR imaging compared with 201Tl SPECT imaging. Radiology 2004;232:49 –57. Hombach V, Grebe O, Merkle N, Waldenmaier S, Hoher M, Kochs M, Wohrle J, Kestler HA. Sequelae of acute myocardial infarction regarding cardiac structure and function and their prognostic significance as assessed by magnetic resonance imaging. Eur Heart J 2005;26:549 –557. Tarantini G, Cacciavillani L, Corbetti F, Ramondo A, Marra Perazzolo M, Bacchiega E, Napodano M, Bilato C, Razzolini R, Iliceto S. Duration of ischemia is a major determinant of transmurality and severe microvascular obstruction after primary angioplasty: a study performed with contrastenhanced magnetic resonance. J Am Coll Cardiol 2005;46:1229 –1235. Lotan CS, Miller SK, Bouchard A, Cranney GB, Reeves RC, Bishop SP, Elgavish GA, Pohost GM. Detection of intramyocardial hemorrhage using high-field proton (1H) nuclear magnetic resonance imaging. Catheter Cardiovasc Diagn 1990;20:205–211.
1327
14. Lotan CS, Bouchard A, Cranney GB, Bishop SP, Pohost GM. Assessment of postreperfusion myocardial hemorrhage using proton NMR imaging at 1.5 T. Circulation 1992;86: 1018 –1025. 15. Ochiai K, Shimada T, Murakami Y, Ishibashi Y, Sano K, Kitamura J, Inoue S, Murakami R, Kawamitsu H, Sugimura KL. Hemorrhagic myocardial infarction after coronary reperfusion detected in vivo by magnetic resonance imaging in humans: prevalence and clinical implications. J Cardiovasc Magn Reson 1999;1:247–256. 16. Asanuma T, Tanabe K, Ochiai K, Yoshitomi H, Nakamura K, Murakami Y, Sano K, Shimada T, Murakami R, Morioka S, Beppu S. Relationship between progressive microvascular damage and intramyocardial hemorrhage in patients with reperfused anterior myocardial infarction: myocardial contrast echocardiographic study. Circulation 1997; 96:448 – 453. 17. Thulborn KR. Biochemical basis of the MRI appearance of cerebral hemorrhage. In: Edelman RR, Hesselink JR, Zlatkin MB, Crues JV, eds. Clinical Magnetic Resonance Imaging. 3rd ed. Philadelphia, PA: Saunders, 2006:174 –186. 18. Parizel PM, Makkat S, Van Miert E, Van Gorthem JW, van den Hauve L, De Schepper AM. Intracranial hemorrhage: principles of CT and MRI interpretation. Eur Radiol 2001; 11:1770 –1783. 19. Bradley WG Jr. MR appearance of hemorrhage in the brain. Radiology 1993;189: 15–26. 20. Atlas SW, Thulborn KR. MR detection of hyperacute parenchymal hemorrhage of the brain. Am J Neuroradiol 1998;19:1471–1477.
in vivo. Cardiovascular magnetic resonance (CMR) can provide a wide range of information in acute MI by detecting infarct size, microvascular obstruction, and edema.7–12 However, to date, CMR findings in acute hemorrhagic MI have not been fully elucidated, and no correlations have been reported among late gadolinium images, T2 images, and histologic findings in humans. We present the first such correlations from the cardiac specimens of 2 patients who underwent CMR and died of cardiogenic shock after acute MIs.
a
Department of Medic-Diagnostic Sciences and Special Therapies, bService of Radiology, and c Department of Cardio, Thoracic, and Vascular Sciences, University of Padua Medical School, Padua, Italy. Manuscript received April 4, 2007; revised manuscript received and accepted May 22, 2007. This work was supported by the Registry of Cardio-Cerebrovascular Pathology, Veneto Region, Venice, Italy. *Corresponding author: Tel: 0039-049-8211982; fax: 0039-049-8272284. E-mail address: [email protected] (F. Corbetti).
Case Descriptions: CMR and Pathology Correlates Patient 1, a 68-year-old man, presented to the emergency room 5 hours after the onset of acute chest pain. Electrocardiography showed ST-segment elevation in leads V2 to V6 and I to aVL. Thrombolysis with recombinant tissue plasminogen activator was performed (time to treatment 330 min-
0002-9149/07/$ – see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2007.05.062
utes). The peak troponin I level was 971 g/L at hour 13. Echocardiography revealed akinesia of the septum and the anterior midapical and lateroapical walls, with an end-diastolic volume of 60 ml/m2 and an ejection fraction of 42%. Coronary angiography performed 1 day later showed a 90% stenosis of the mid left anterior descending coronary artery, and stenting was successfully performed (baseline Thrombolysis In Myocardial Infarction [TIMI] flow grade 2, baseline blush grade 0; final TIMI flow grade 3, final blush grade 0). CMR was performed on day 9, and the patient died on day 12 from cardiogenic shock. Patient 2, a 60-year-old woman, presented to the emergency room with new onset of acute chest pain. Electrocardiography showed ST-segment elevation in leads V1 to V4 and aVR and ST-segment depression in leads I, II, III, and aVF. Coronary angiography revealed a 100% occlusion of the mid left anterior descending coronary www.AJConline.org
Case Report/Hemorrhagic Myocardial Infarction by Magnetic Resonance
1323
Figure 1. Patient 1: transmural anteroseptal and lateral hemorrhagic MI. Gross (A, B, C, D), histologic (E, F, G, H) (Heidenhain trichrome stain), and ex vivo T2 CMR (I, L, M, N) short-axis sections of the heart, from the base to the apex. Note the perfect overlap between the hemorrhage seen on pathologic examination and the low-signal intensity areas observed on ex vivo T2 CMR.
artery and a 90% stenosis of the left circumflex, and stenting was successfully performed in the 2 arteries (time to treatment 5 hours; baseline TIMI flow grade 0, baseline blush grade 0; final TIMI flow grade 3, final blush grade 0). The peak troponin I level was 369.43 g/L on the first day. Echocardiography revealed akinesia of the apex, septum, and anterior wall, with an end-diastolic volume of 59 ml/m2 and an ejection fraction of 46%. The patient underwent CMR on day 7 and was discharged home on day 11, but on day 24 she died from cardiogenic shock. CMR in vivo images were acquired during repeated breath-holds on a 1.0-T body magnet (Harmony; Siemens Medical Systems, Erlangen, Germany) using cardiologic software (MRease SYNGO 2002B; Siemens Medical Systems) and a 4-channel phased-array receiver coil. A standard imaging protocol was used, including cine steady-state free precession sequences in 4 short-axis (from the base
to the apex) and 3 long-axis (2, 3, and 4 chamber) views for functional evaluation, T2-weighted fat saturation black-blood turbo spin-echo sequences (repetition time 2 to 3 RR intervals, echo time 92 ms, mean voxel size 1.6 ⫻ 1.4 ⫻ 7 mm), and myocardial delayed enhancement (MDE) sequences (segmented inversion recovery [IR] turbo fast low-angle shot, repetition time 1 to 2 RR intervals, echo time 6 ms, flip angle 30°, mean voxel size 1.8 ⫻ 1.3 ⫻ 8 mm, inversion time optimized in each patient, range 190 to 220 ms) performed 10 to 15 minutes after the administration of 0.2 mmol/kg of a gadolinium-based contrast agent (MultiHance; Bracco Diagnostics, Inc., Milan, Italy). The excised hearts were enveloped in a surface flexible coil (Flex Large; Siemens Medical Systems) and serial short-axis views from the base to the apex plus 3 long-axis views, as in vivo, were acquired using T2- and T1weighted sequences. For T2 study, a standard turbo spin-echo sequence
without blood preparation was used (repetition time 2,400 ms, echo time 104 ms, field of view 160 mm, mean voxel size 0.7 ⫻ 0.6 ⫻ 3 mm). For T1-weighting purposes, the same IR turbo fast low-angle shot sequence used for MDE images in vivo was applied, adapting the parameters to achieve better resolution (repetition time 1,200 ms, echo time 7 ms, flip angle 30°, field of view 180 mm, mean voxel size 0.8 ⫻ 0.6 ⫻ 3 mm, inversion time 150 ms [a value commonly used in conventional IR sequences to suppress fat signal]). As previously reported, hyperintense and dark areas on ex vivo T2 images were assumed to indicate infarct size and hemorrhage, respectively.13,14 CMR images were analyzed quantitatively by the user-interactive program Cine-Tool (GE Medical Systems, Waukesha, Wisconsin), manually tracing the endocardial and epicardial borders for left ventricular mass computation. MI and hemorrhage extent were then quantified by manually tracing the borders of the hyperintense
1324
The American Journal of Cardiology (www.AJConline.org)
Figure 2. Patient 2: transmural anteroseptal and lateral hemorrhagic MI. Gross (A, B, C, D), histologic (E, F, G, H) (Heidenhain trichrome stain), and ex vivo T2 CMR (I, L, M, N) short-axis sections of the heart, from the base to the apex. Note the perfect overlap between the hemorrhage seen on pathologic examination and the low-signal intensity areas observed on ex vivo T2 CMR.
and dark areas, respectively, and were expressed as percentages of total left ventricular area in each slice. Four short-axis myocardial slices of the excised formalin-fixed hearts (base, midbase, midapical, and apical) correlating with CMR slices were obtained, photographed, and embedded with paraffin wax. Three tissue sections 7 m thick were cut and stained with hematoxylin-eosin, Perl’s Prussian blue (to visualize ferric iron in hemosiderin) and Heidenhain trichrome, respectively. Short-axis slices of the heart stained with Heidenhain trichrome were analyzed by quantitative computerized morphometry (Image-Pro Plus version 4.0; Media Cybernetics, Inc., Bethesda, Maryland). Left ventricular area was obtained by manually tracing the endocardial and epicardial borders; hemorrhage and MI were quantified by manual tracing and were expressed as percentages of total left ventricular area in each slice. Analysis was performed independently for the histopathologic (CB, GT) and CMR (FC, CS) studies, with the
researchers for histopathology blinded to the CMR results and vice versa. Linear regression was used to compare the extent of MI and hemorrhage on T2 ex vivo CMR and pathology. Cardiogenic shock with acute pulmonary edema and peripheral congestion was found to be the cause of death at autopsy. Pathologic examination of the 2 hearts revealed hemorrhagic transmural septal and anterolateral MIs, in the setting of patent coronary artery stents (Figures 1 and 2). Hemorrhage was confined to the zone of necrosis and was massive at the core of MI, with diffuse and confluent packed red blood cells, interspersed between myocytes with coagulation necrosis in patient 1, whereas an almost complete removal of dead myocytes was seen in patient 2. Moving toward the periphery, hemorrhage was slight to moderate, with patchy distribution. In patient 2, at the border between the hemorrhagic core and the healing organizing tissue, patchy hemosiderin deposits were detected (Figure 3). In both instances, the total
destruction of the microvasculature within the MI core was evidenced. The MI extent was 57 ⫾ 30% and 44 ⫾ 24% of the left ventricular area, with progressive increases from the base to the apex (ranges 29% to 88% and 16% to 68%, respectively). Hemorrhagic extent was 23 ⫾ 13% and 19 ⫾ 8% of the left ventricular area, also showing progressive increases from the base to the apex (ranges 12% to 40% and 9% to 27%, respectively). T2-weighted images showed areas of signal hypointensity inside the hyperintense areas (Figures 1 and 2). The MI extent and the area of hemorrhage (hypointense regions or “dark areas”) measured by ex vivo T2 CMR strongly correlated with the pathologic quantification (Figure 4). Using ex vivo MDE sequences, bright–signal intensity areas surrounded by dark rims, consistent with a magnetic susceptibility effect, were visible within the MI regions. Compared with histopathologic findings, the bright signal and the dark border were found to correspond to areas of massive hem-
Case Report/Hemorrhagic Myocardial Infarction by Magnetic Resonance
1325
Figure 3. Patient 2: on ex vivo T2 (A) and IR MDE (B) images, a perfect correspondence with histologic findings is visible (C), showing 2 areas of massive intramyocardial hemorrhage in the septum (hematoxylin-eosin ⫻5); (D) close-up of the upper boxed area showing packed extravasated intact red blood cells (hematoxylin-eosin ⫻31); (E) neovascularized organizing tissue with patchy hemosiderin deposits at the border close to hemorrhage (Perl’s Prussian blue ⫻31).
orrhage with red blood cell extravasation, with additional hemosiderin deposits at the periphery in patient 2 (Figure 3). As expected, high-signal intensity on in vivo T2-weighted sequences was found in the MI area, reflecting tissue edema. However, low-signal areas were also detectable within the edematous myocardial wall and correlated well in location and extent with the hemorrhage seen on the ex vivo T2 images and proved by histology (Figures 1 to 3 and 5). On in vivo MDE images, a dark hypoenhanced core resembling persistent microvascular obstruction was present within the hyperenhanced MI. However, on closer evaluation, a dark rim surrounding tiny central areas of higher signal intensity was visible, with susceptibility
effects similar to those observed on ex vivo images (Figure 5). Comments By clinicopathologic study in 2 patients who died of cardiogenic shock after reperfused acute MIs, we provide evidence that hemorrhagic MI accounts for CMR late gadolinium hypoenhancement associated with magnetic susceptibility effects and hypointense T2 signal, thus casting some doubt on the belief that low-signal intensity areas surrounded by bright zones on late enhancement CMR always represent noreflow zones. Herein, we demonstrate that an alternative explanation for this phenomenon may be hemorrhage within the infarct core. In the heart, the ability of CMR to
detect hemorrhage has been reported by Lotan et al13,14 using ex vivo T2weighted spin-echo sequences in a canine model of acute MI and by Ochiai et al15 using CMR with a T2 star gradient-echo sequence in humans. In the latter study, ex vivo T2 star CMR was performed in 1 patient who died 3 days after the procedure, showing an overlap between the hypointense zone and the myocardial hemorrhage visible in the pathologic specimen. Our data agree with these previous findings, showing a perfect overlap among the location, spatial extent, and shape of the hypointense regions by ex vivo T2 CMR images and the hemorrhage by pathologic investigation. The possibility of visualizing myocardial hemorrhage as hypointense signal by T2 sequences in vivo is also pro-
1326
The American Journal of Cardiology (www.AJConline.org)
Figure 4. Comparison of CMR T2 sequences and pathology by linear regression analysis: MI size (top) (R ⫽ 0.93, p ⫽ 0.0009) and hemorrhagic extent (bottom) (R ⫽ 0.93, p ⫽ 0.0007) showed a positive correlation. LV ⫽ left ventricular.
Figure 5. Patient 2: in vivo versus ex vivo CMR. (A) In vivo T2-weighted CMR, 4-chamber views: a hypointense rim (asterisk) is clearly seen within the hyperintense edematous septum. (B) In vivo IR MDE sequence, long-axis view: a complex signal pattern is seen within the MI core of infarcted septum, with tiny bright areas surrounded by a dark rim (arrows), a feature typical of the susceptibility effect. (C) Ex vivo IR MDE sequence, long-axis view: the same susceptibility effect as in (B) is clearly seen around the septal hemorrhage but is more evident because of the higher resolution of the image and the additional presence of hemosiderin, which normally occurs only at this stage of hemorrhage (i.e., 24 days after MI onset).
vided, given the correspondence of T2 in vivo and ex vivo images. In this respect, Asanuma et al16 considered markers of myocardial hemorrhage not only the low-signal intensity area within the risk area by gradient-echo acquisition but also the hypoenhanced
area within a high-signal intensity zone by gadolinium-enhanced spin echo CMR. However, their definition of hemorrhage was based only on the CMR findings, without any histopathologic validation. Our CMR findings and their corre-
lation with histology indicate for the first time that peculiar signal features of late gadolinium hypoenhancement within the MI core can be related to hemorrhage. In our patients, the highsignal intensity within the hemorrhage on ex vivo T1 IR MDE sequence was typical of the paramagnetic effect of intracellular metahemoglobin17–19 and was observed also in vivo on close evaluation of the late hypoenhanced area. Furthermore, on ex vivo T1 IR MDE study, the bright signal within the hemorrhagic areas was surrounded by a thin dark rim, a finding that is observed at the periphery of hematomas not only on T2 sequences17 but also on T1 sequences,20 especially by using gradient-echo and low-bandwidth sequences,19 as we did in our cases. This feature can be related to the compartmentalization of fresh hemorrhage and the presence of metahemoglobin within intact packed red cells, resulting in a susceptibility effect at the periphery of hematoma with signal loss; in patient 2, who died 24 days after MI, this effect on ex vivo images was enhanced by the peripheral deposition of hemosiderin, the latest breakdown product of hemoglobin that appears first at the periphery of hemorrhage in the subacute or chronic stage and induces strong susceptibility effects due to its superparamagnetic properties.17–20 Of course, these findings must be interpreted as preliminary data, because they are the results of case reports of only 2 patients. However, this study was not an experimental but an observational investigation of human hearts, supported by detailed postmortem ex vivo CMR and pathologic studies in 2 patients who underwent in vivo T2 and late enhancement CMR and died from cardiogenic shock. Experimental studies in animals subjected to coronary occlusion and reperfusion as well as further human correlative studies are needed to assess the relative role of severe microvascular obstruction compared with hemorrhage on persistent late gadolinium hypoenhancement in reperfused acute MI. Acknowledgment: We are deeply indebted to Mauro Pagetta for technical assistance.
Case Report/Hemorrhagic Myocardial Infarction by Magnetic Resonance 1. Reimer KA, Lowe JE, Rasmussen MM, Jennings RB. The wavefront phenomenon of ischemic cell death. 1. Myocardial infarct size vs duration of coronary occlusion in dogs. Circulation 1977;56:786 –194. 2. Fishbein MC, Rit J, Lando U, Kanmatsuse K, Mercier JC, Ganz W. The relationship of vascular injury and myocardial hemorrhage to necrosis after reperfusion. Circulation 1980; 62:1274 –1279. 3. Kloner RA, Rude RE, Carlson N, Maroko PR, DeBoer LW, Braunwald E. Ultrastructural evidence of microvascular damage and myocardial cell injury after coronary artery occlusion: which comes first? Circulation 1980;62:945–952. 4. Roberts CS, Schoen FJ, Kloner RA. Effect of coronary reperfusion on myocardial hemorrhage and infarct healing. Am J Cardiol 1983; 52:610 – 614. 5. Garcia-Dorado D, Théroux P, Solares J, Alonso J, Fernandez-Avilés F, Elizaga J, Soriano J, Botas J, Munoz R. Determinants of hemorrhagic infarcts. Histologic observations from experiments involving coronary occlusion, coronary reperfusion, and reocclusion. Am J Pathol 1990;137:301–311. 6. Basso C, Thiene G. The pathophysiology of myocardial reperfusion: a pathologist’s perspective. Heart 2006;92:1559 –1562. 7. Kim RJ, Fieno DS, Parrish TB, Harris K, Chen EL, Simonetti O, Bundy J, Finn JP, Klocke FJ, Judd RM. Relationship of MRI delayed contrast enhancement to irreversible injury, infarct age, and contractile function. Circulation 1999;100:1992–2002. 8. Wu KC, Zerhouni EA, Judd RM, LugoOlivieri CH, Barouch LA, Schulman SP, Blumenthal RS, Lima JA. Prognostic signifi-
9.
10.
11.
12.
13.
cance of microvascular obstruction by magnetic resonance imaging in patients with acute myocardial infarction. Circulation 1998;97:765–772. Simonetti OP, Kim RJ, Fieno DS, Hillenbrand HB, Wu E, Bundy JM, Finn JP, Judd RM. An improved MR imaging technique for the visualization of myocardial infarction. Radiology 2001;218:215–223. Lund GK, Stork A, Saeed M, Bansmann MP, Gerken JH, Muller V, Mester J, Higgins CB, Adam G, Meinertz T. Acute myocardial infarction: evaluation with first-pass enhancement and delayed enhancement MR imaging compared with 201Tl SPECT imaging. Radiology 2004;232:49 –57. Hombach V, Grebe O, Merkle N, Waldenmaier S, Hoher M, Kochs M, Wohrle J, Kestler HA. Sequelae of acute myocardial infarction regarding cardiac structure and function and their prognostic significance as assessed by magnetic resonance imaging. Eur Heart J 2005;26:549 –557. Tarantini G, Cacciavillani L, Corbetti F, Ramondo A, Marra Perazzolo M, Bacchiega E, Napodano M, Bilato C, Razzolini R, Iliceto S. Duration of ischemia is a major determinant of transmurality and severe microvascular obstruction after primary angioplasty: a study performed with contrastenhanced magnetic resonance. J Am Coll Cardiol 2005;46:1229 –1235. Lotan CS, Miller SK, Bouchard A, Cranney GB, Reeves RC, Bishop SP, Elgavish GA, Pohost GM. Detection of intramyocardial hemorrhage using high-field proton (1H) nuclear magnetic resonance imaging. Catheter Cardiovasc Diagn 1990;20:205–211.
1327
14. Lotan CS, Bouchard A, Cranney GB, Bishop SP, Pohost GM. Assessment of postreperfusion myocardial hemorrhage using proton NMR imaging at 1.5 T. Circulation 1992;86: 1018 –1025. 15. Ochiai K, Shimada T, Murakami Y, Ishibashi Y, Sano K, Kitamura J, Inoue S, Murakami R, Kawamitsu H, Sugimura KL. Hemorrhagic myocardial infarction after coronary reperfusion detected in vivo by magnetic resonance imaging in humans: prevalence and clinical implications. J Cardiovasc Magn Reson 1999;1:247–256. 16. Asanuma T, Tanabe K, Ochiai K, Yoshitomi H, Nakamura K, Murakami Y, Sano K, Shimada T, Murakami R, Morioka S, Beppu S. Relationship between progressive microvascular damage and intramyocardial hemorrhage in patients with reperfused anterior myocardial infarction: myocardial contrast echocardiographic study. Circulation 1997; 96:448 – 453. 17. Thulborn KR. Biochemical basis of the MRI appearance of cerebral hemorrhage. In: Edelman RR, Hesselink JR, Zlatkin MB, Crues JV, eds. Clinical Magnetic Resonance Imaging. 3rd ed. Philadelphia, PA: Saunders, 2006:174 –186. 18. Parizel PM, Makkat S, Van Miert E, Van Gorthem JW, van den Hauve L, De Schepper AM. Intracranial hemorrhage: principles of CT and MRI interpretation. Eur Radiol 2001; 11:1770 –1783. 19. Bradley WG Jr. MR appearance of hemorrhage in the brain. Radiology 1993;189: 15–26. 20. Atlas SW, Thulborn KR. MR detection of hyperacute parenchymal hemorrhage of the brain. Am J Neuroradiol 1998;19:1471–1477.