Cardiopulmonary resuscitation does not cause left ventricular rupture of the heart with acute myocardial infarction: a pathological analysis of 77 autopsy cases

1. Introduction 

Cardiac rupture during acute myocardial infarction (AMI) is one of the most frequent causes of sudden cardiac death [1], [2], [3], [4]. It has been suggested that external cardiac massage of infarcted areas may lead to left ventricular rupture. In fact, a few reports of left ventricular rupture with AMI following cardiopulmonary resuscitation (CPR) have been reported [5], [6], [7], although generally rupture occurs spontaneously. To our knowledge, there has been no pathological study comparing cases of cardiac rupture during AMI with CPR to those without CPR. In the present study, therefore, we pathologically compared cases of cardiac rupture during AMI with and those without CPR, and morphologically evaluated whether cardiac rupture in fact occurred as a result of cardiopulmonary resuscitation in these cases.

2. Materials and methods 

3. Results 

No significant differences were found between the CPR and the NCPR group in any of the investigated items (see Table 1, Fig. 1, Fig. 2).

Table 1. Comparison of the autopsy findings between cases with CPR and without CPR

	Autopsy finding	With CPR (n=44)	Without CPR (n=33)	P value
	Location of AMI
	Anteroseptal	19 (43%)	18 (55%)	NS
	Anterior	4 (9%)	3 (9%)	NS
	Inferior	10 (23%)	5 (15%)	NS
	Lateral	11 (25%)	7 (21%)	NS
	Site of rupture
	Anterior	20 (45%)	21 (64%)	NS
	Inferior	10 (23%)	5 (15%)	NS
	Lateral	14 (32%)	7 (21%)	NS
	Type of rupture (Becker's classification)
	Type I	20 (46%)	7 (21%)	NS
	Type II	23 (52%)	24 (73%)	NS
	Type III	1 (2%)	2 (6%)	NS
	Wall thickness of rupture site (%)	54.8±21.4	46.8±19.0	NS
	Number of epicardial tear
	1	31 (71%)	24 (73%)	NS
	2	5 (11%)	5 (15%)	NS
	3	5 (11%)	3 (9%)	NS
	4	3 (7%)	1 (3%)	NS
	Length of epicardial tear (cm)	1.53±1.17	1.48±0.85	NS
	Pericardial hemorrhage (ml)	319.0±114.7	343.6±117.5	NS
	Clotted blood in hemopericardium
	(+)	41	32	NS
	(−)	0	0	NS
	Unknown	3	1	NS
	OMI or fibrosis
	(+)	11 (25%)	6 (18%)	NS
	(−)	33 (75%)	27 (82%)	NS
	Hemorrhagic infarction
	(+)	3 (7%)	2 (6%)	NS
	(−)	41 (93%)	31 (94%)	NS

NS, not significant.

View Large Image Download to PowerPoint

Fig. 1. Histopathologic age of the infarcted myocardium.

View Large Image Download to PowerPoint

Fig. 2. Age of thrombi in the rupture tract.

4. Discussion 

Previous reports have suggested that CPR, particularly external cardiac massage, in patients with AMI can result in rupture of the left ventricle [5], [6], [7]. In the present study, we show that CPR for patients with AMI did not cause left ventricular rupture. The evidence that CPR did not produce left ventricular rupture through the infarcted area were as follows.

First, there were no morphological differences in the left ventricular free wall rupture through the infarcted area between the group that received CPR and the group that did not. This strongly suggests that the two groups were virtually identical with respect to morphology. In particular, the factors that could be amplified by aggressive chest compression, i.e. the number and the length of the epicardial tears and the volume of the pericardial hemorrhage, were virtually equal in both groups. It also reveals that a strong chest compression during CPR does not produce massive injuries in the infracted hearts.

The second, more definitive evidence that CPR did not cause left ventricular rupture is the existence of thrombi along the rupture tract prior to the cardiac rupture. In addition, chronological differences in the appearance of thrombi throughout the rupture tract confirmed that cardiac rupture through the infarcted area precedes external cardiac massage even in cases with CPR. In early thrombus formation in the heart and arterial circulation, once endothelial injury occurs, platelet adhesion (to the culprit lesion) and aggregation (platelet thrombus) begin, and fibrin subsequently enmeshes the platelets and stabilizes the thrombus (MFT) [12], [13]. This stabilizing process of the thrombus takes at least several hours [14]. Our study shows that mural thrombi in the rupture tracts at the subendocardial region (mature fresh thrombi) were more matured than those at the subepicardial region (immature or platelet thrombi), irrespective of whether or not CPR was performed. The existence of stabilized thrombi at the subendocardial region indicates that the rupture of the myocardium begins with laceration of the endocardial tear. In addition, the chronological differences of the appearance of thrombi throughout the rupture tract confirmed that rupture of the infarcted myocardium occurred before CPR.

The other but less definitive evidence that cardiac rupture preceded CPR was the existence of clotted blood in the pericardial hemorrhage. In the present study, clotted blood was present in the hemopericardium of almost all cases in both groups, with the exception of a few cases in which the autopsy records gave no information on clotted blood. Klintschar et al. [15], in their report on cardiac injuries from active compression–decompression cardiopulmonary resuscitation, assumed that a rupture of the infracted left ventricle occurred before the resuscitation based on the fact that some of the hemorrhaged blood was coagulated. We too have experienced in our forensic practices that there was no clotted blood in the pericardial, pleural, or peritoneal hemorrhaging from the CPR-related injury after cardiopulmonary arrest.

Although we have shown in the present study that CPR did not cause the left ventricular rupture associated AMI, a few reports have suggested that external cardiac massage might cause cardiac rupture associated with myocardial infarction [5], [6], [7]. In these cases, however, the authors did not distinguish between spontaneous myocardial rupture and CPR-related injury. Yamada and Fukunaga [5] conjectured that cardiac rupture resulting from external cardiac massage might have occurred based on the higher incidence among the cases having undergone CPR, but the possibility that the rupture occurred prior to massage could not be excluded. Other reports have shown that the jagged edges of fractured ribs [7] or vertebral osteophytes [6] may penetrate the myocardium through the infarcted area and may lead to lethal cardiac tamponade during CPR, but we believe there is insufficient pathological evidence to support this conclusion. For example, neither the conformity of the shape of the rupture tract to that of the object causing injury (i.e. the edge of the fractured ribs) nor the thrombi through the rupture tracts were examined in these cases. Furthermore, our pathological finding that the cardiac rupture of the infarcted myocardium begins with the endocardial portion, would appear to rule out penetration of the object into the heart from the outside.

While a few authors have suggested the possibility of cardiac rupture with myocardial infarction by external cardiac massage, other authors have questioned the causal relationship between cardiac massage and cardiac rupture [16], [17]. In particular, Bodily and Fischer [17], in their report on cardiovascular injuries due to CPR, excluded 11 cases of the left ventricular rupture associated AMI. Baldwin and Edwards [16] also doubted that the cardiac massage caused the left ventricular rupture of the infarcted hearts. The present study supports their conclusions [16], [17].

Furthermore, susceptible sites of cardiac rupture by CPR also support our opinion that CPR does not cause left ventricular rupture with myocardial infarction. Although many CPR-related cardiac injuries have been reported, a rupture of the left ventricle, especially through a non-infarcted area, is extremely rare [17], [18], [19], [20], [21], [22], [23]. In the study by Baldwin and Edward [16], almost all the cardiac ruptures attributed to external cardiac massage occurred in the thinner-wall portions, including the right ventricle, the auricles and the atria, probably as a result of trapping of blood in the chamber due to systemic arterial pressure. Because of this fact, we suppose that complete rupture of the left ventricle due to external cardiac massage does not occur except in unusual cases, such as in patients with a prosthetic mitral valve [24]. More notably, in our subjects, no right ventricular rupture occurred even in those with right ventricular myocardial infarction who had received CPR.

Although none of the parameters investigated in the present study were significantly different between the two groups, the NCPR group had a greater incidence of myocardial infarctions that were less than 1 day old than did the CPR group (NCPR: 13/33 cases, 39.4%; CPR: 11/44 cases, 25%). The reason for this phenomenon is unclear, but it may be related to the fact that the patients with superacute myocardial infarction did not have as much time to get to an emergency room after the attack.

In conclusion, our pathological study revealed that the rupture begins with endocardial laceration during life, and also proved that CPR does not cause left ventricular rupture of the infarcted myocardium.