Cocaine and the heart

Cocaine and the Heart HENRY D. TAZELAAR, MD,*t STEVEN B. KARCH, MD,* BOYD G. STEPHENSMD,~ AND MARGARET E. BILLINGHAM, MB, BS, MRCPath *w and Simpson and Edwards TM reported the presence of acute and chronic obstructive coronary lesions, contraction bands, myocardial fibrosis, and lymphocytic myocarditis in association with cocaine use. In an attempt to clarify the pathophysiology o f cocaineassociated cardiotoxicity and to search for pathologic changes that might be useful in a forensic setting, we reviewed microscopic sections of hearts from 30 cases of cocaine-associated death seen by the San Francisco Medical Examiner.

T h e r e is increasing evidence that cocaine can have serious ad-

verse effects on the heart. Angina, myocardial infarction, coronary artery spasm, arrhythmia, a n d sudden death have been rep o r t e d in association with its use. There have been only two reports of actual myocardial pathology. In an attempt to clarify the p a t h o p h y s i o l o g y of cocaine-associated cardiotoxiclty and to search for pathologic changes that might be useful forensically, w e r e v i e w e d r a n d o m m i c r o s c o p i c sections of h e a r t s from 30 cases of cocaine-associated death seen by the San Francisco Medical Examiner. T h e age of the patients at death ranged from 25 to 74 years (mean 33.9 years). Pathologic f i n d i n g s i n c l u d e d the p r e s e n c e o f mild atherosclerotic coronary artery disease without evidence of thrombosis in three cases, associated with mild interstitial fibrosis in one case as well as mild focal myocardial fibrosis without coronary disease in four other cases. The most notable abnormality was the presence of myocardial contraction b a n d s in 28 (93 per cent) cases. I n comparison to 20 control cases o f d e a t h secondary to s e d a t i v e - h y p n o t i c overdose, the hearts f r o m the cases of cocalne-associated contained significantly m o r e myocardial contraction bands (P < .001; two-sided). T h e diffuseness of the contraction b a n d s correlated directly with the level of cocaine f o u n d in the u r i n e and blood at autopsy d u r i n g r o u t i n e screening. T h e p r e s e n c e a n d n u m b e r of c o n t r a c t i o n bands in these cases was i n d e p e n d e n t o f o t h e r drugs found in the u r i n e a n d blood, the n u m b e r of sections of myocardinm examined, and a history of attempted resuscitation. Contraction bands may act to supply the anatomic substrate for the arrhythmias associated with cocaine use. They may also provide a morphologic marker that can be sought in suspected cases of lethal cocaine o v e r d o s e . T h e i r presence may also suggest a cause of death in cases of sudden and unexpected death in which autopsy reveals no other pathology, a n d a d r u g screen is positive for cocaine. HUM PATHOL 18:195--199, 1987.

Numerous reports have documented the association of cocaine use with angina, myocardial infarction, coronary artery spasm, arrhythmia, and sudden death. T M Because the d r u g predictably increases pulse rate, blood pressure, and oxygen consumption, its deleterious effects on those with underlying coronary disease are not surprising. There appears to be a common acceptance that the drug is directly toxic to the myocardium, which might account for the occurrence of typically ischemic cardiac sequelae in young individuals. However, specific myocardial pathology has been reported only twice. Isner et al. l~ reported the presence o f eosinophilic myocarditis, Received from the *Department of Pathology, Stanford University Medical Center, Stanford, and the ~Medical Examiner's Office, City and County of San Francisco, San Francisco, California. Accepted for publication September 15, 1986. Supported by tNational Research Service Individual Postdoctoral Fellowship 1 F32 HL07189-01 and w grant HL 13108 from the National Heart, Lung and Blood Institute. Address correspondence and reprint requests to Dr. Tazelaar: Department of Pathology, Stanford University Medical Center, Stanford, CA 94305. 0046-8177~7 $0.00 + .25

MATERIALSAND METHODS

All deaths attributed to cocaine toxicity seen by the Office o f Chief Medical E x a m i n e r - C o r o n e r of San Francisco from July 1, 1983 (the date when such records for cocaine were initiated), through J u n e 30, 1985, were reviewed. Demographic and historical data were obtained f r o m t h e Coroner's registry and included age, sex, race, history of resuscitation, the presence of intravenous needle tracks, and all drugs - - i n c l u d i n g c o c a i n e - - f o u n d in a r o u t i n e d r u g screen. T h e control group was comprised o f all cases from J a n u a r y 1984 to J u l y 1985 who died secondary to sedative-hypnotic overdose without cocaine. All cases in which tricyclic antidepressants were found in either blood or urine were excluded, because these substances have been shown to produce ventricular arrhythmias, is T h e Medical Examiner's Office submits random sections of the heart for light microscopy in every case. All sections were fixed in 10 per cent buffered formalin, processed routinely, paraffin e m b e d d e d for light microscopy, and stained with hematoxylineosin. For the purposes of this study, slides of other organs were not studied specifically, although general autopsy findings were reviewed. The number of sections of myocardium examined for each case and control was recorded. All sections o f myocardium were examined simultaneously by three of the authors without knowledge of blood or urine d r u g levels. Sections o f myocardium were examined for the presence of pathologic features that might account for the cocaine-associated deaths. Sections were evaluated specifically for the presence of myocardial contraction bands, coronary artery disease, inflammatory infiltrates, and myocardial fibrosis. In an initial review o f the material, the most notable abnormality was the presence of focal myocyte contraction bands. In hematoxylin-eosin-stained sections, these bands appear as hypereosinophilic bands traversing the myofiber cytoplasm formed by clumped sarcomeres. 195

HUMAN PATHOLOGY

Volume 18, No. 2 (February 1987]

FIGURE 1. Photomicrograph showing hypereosinophilic contraction bands [arrows] in the myocardium of a 35-year-old man dying of cocaine toxicity. (Hematoxylin-eosin stain, x 400.]

T h e presence o f contraction bands, therefore, was graded as 0 when absent; 1 + when present focally, and up to 3 + when widespread and abundant. T h e contraction band score o f the cases was compared to that o f the control g r o u p using the Mann-Whitney Wilcoxon two-tailed test. 16 Routine blood and urine toxicologic examination was p e r f o r m e d on both study cases and controls. This involved screening by standard techniques for the following substances: volatile substances such as acetone, barbiturates, sedative-hypnotics, benzodiazepines, m e t h a q u a l o n e , a m p h e t a m i n e s , phencyclidine, antihistamines, cocaine, m o r p h i n e - t y p e alkaloids, tricycyclic antidepressants and analytically related compounds, phenothiazines, and salcylates, and ethyl alcohol. T h e d r u g levels found in the blood and urine were then correlated with the contraction band score.

RESULTS T h i r t y - f o u r cases o f cocaine-associated death were identified. All were cases in which the coroner ruled cocaine toxicity as the cause of death, based on both the presence or level of cocaine found in body fluids at the time o f autopsy and a lack of other significant findings. T w o cases were excluded because slides were unavailable; two others were excluded because o f tissue autolysis. T h e remaining 30 cases included 26 men (86 per cent) and four women (14 per cent). T h e r e were 21 whites (70 per cent), seven blacks (23 per cent), and two orientals (7 per cent). T h e mean age was 33.9 years (range, 25 to 74 years). T h e r e was attempted resuscitation in 15 (50 per cent)

o f the cases. O ne patient survived two days following resuscitation, and one survived three days. Twentyone (70 per cent) were intravenous drug users as evidenced by the presence o f needle tracks unrelated to resuscitation. In two cases cocaine was the only d r u g identified in the d r u g screen. In six cases cocaine was associated only with ethyl alcohol and/or aspirin in either the blood or urine. T h e levels o f cocaine found in the blood ranged for 0 to 20.8 mcg/dl and in the urine from 0 to 218.3 mcg/dl. T h e range of lethal blood levels o f cocaine reported previously 17 are 0.1 to 20.9 mcg/dl, although what constitutes a lethal blood level o f cocaine at autopsy is difficult to determine, because there continues to be metabolism o f cocaine after death and susceptibility to cocaine's toxic effects may vary between individuals based on previous exposure to the drug. Other drugs found in the blood or urine o f the cases included metamphetamine, methadone, amphetamine, morphine, diazepam, nordiazepam, lidocaine, theophylline, prop o x y p h e n e , n o r p r o p o x y p h e n e , librium, and codeine. In no case was the c o n c e n t r a t i o n o f o t h e r d ru g s found in the blood or urine at the time o f death considered to be lethal or responsible for death. A mean o f 2.2 sections (range one to five) was e x a m i n e d f o r each case. Cont ract i on bands were present in 28 (93 per cent) of the cocaine cases (fig. 1) and were j u d g e d moderately diffuse to abundant and widespread (grade 2 + to 3 + ) in more than half o f the cases (table 1). O t h e r findings include a sparse and focal neutrophilic infiltrate associated with contraction bands in two cases, mild atherosclerotic coronary disease without evidence of thrombosis in three cases, associated with mild focal myocardial fibrosis in one case, and mild focal interstitial fibrosis without coronary disease in four cases. T h e contraction band score correlated directly with the serum and urine concentrations o f cocaine (fig. 2). Twenty-three control cases were identified, o f which three were excluded: one because o f a history o f mediastinal irradiation, one because of autolysis o f the heart sections, and one because no longitudinal myocardial fibers were present in the tissue sections, and these are necessary for the identification o f contraction bands. T h e control g r o u p c o n t a i n e d 19 whites (95 per cent) and one black (5 per cent). T h e m ean age o f the control g r o u p (42.5) was slightly higher than that o f the cocaine group (33.9);ears), but had almost exactly the same range, 22 to 75

TABLE 1,

Distribution of Contraction Band Scores for Study Cases a n d Controls

Cocaine Cases (n = 30)

Control Cases (n = 20)

Contraction Band Score

2 8 7 13

11 6 I 2

0 1+ 2+ 3+

196

COCAINE AND THE HEART[Tazelaar et aL]

20

230

18

210

E -.~

16

_~ 160 E -.~

z

14

z 140 O

O ICE

i,-

a: 120

12

I.Z

Z UJ

Z

o

r

t.tJ

z 100

10

O

uJ

Z

~

gJ

z ,,r

s

8

O

~ 0

80

6

w Z

0

..I

m

4

~

40

B

$

20 9

9

I

IP

a

0 1 2 3 CONTRACTION BAND SCORE

b

I-

Z l.u

60-

n" uJ

..,...~

40..=.; ~:o

2O

0

,.,,,,,o"-

~,t.~l-v

~,P""

~,Ox' DRUG

197

9

0 I 2 3 CONTRACTION BAND SCORE

100

F I G U R E 2 (top]. a, S c a t t e r g r a m showing the relationship of the contraction band score to the blood cocaine concentration found at autopsy in 30 cases of cocaine associated death, b, Scattergram showing the relationship of the contraction band score to the urine cocaine concentration found at autopsy in 30 cases of cocaine associated death [data available in 27 cases]. FIGURE 3 [bottom). Histogram showing frequency of drugs found in the blood and urine at autopsy in the 20 sedativehypnotic overdose controls. "Other" includes salicylate (one case], thioridazine [one case], antihistamine [one case], and digoxin [one case].

9

,-

HUMAN PATHOLOGY

Volume 18, No. 2 (February 1987]

years. There was attempted resuscitation in five (25 per cent) of the controls with no short-term survivors. T h e frequency of drugs found in the blood and urine screen in the controls is found in figure 3. A mean of 3.2 sections of heart (range one to six sections) was examined for each control. Contraction bands were identified in nine o f 20 (45 per cent) control cases; the distribution based on our scoring scheme is found in table 1. T h e contraction band score was 3 + in only two (10 per cent ) of cases. Overall, the study cases had significantly higher scores for the presence o f contraction bands (P < .001; two-sided) than the controls. T h e individual contraction band scores for both the cases and controls were independent of the number of sections of myocardium examined and whether or not resuscitation was attempted (data not shown). For the study cases there also was no statistical difference between the scores o f cases with only cocaine, alcohol, or aspirin identified in either blood or urine and those with other drugs in addition. DISCUSSION T h e euphoria and central nervous system stimulation p r o d u c e d by cocaine have been known in South America for thousands of years) s Until the nineteeth century, however, cocaine was virtually ign o r e d in Western E u r o p e , a l t h o u g h it had been known there for at least 300 years. But by the late 1800s, its use had become widespread, and it was touted as a "cure-all, ''[9 particularly by Freud. Cocaine has been considered a relatively safe drug. The recent upswing in its popularity and use, 2~ however, has highlighted that not only can cocaine result in addiction and a withdrawal syndrome, 23 but it also can have catastrophic effects on the cardiovascular system. 2-]4,24 While cocaine has secondary effects on the cardiovascular system via the central nervous system, it also has direct actions on the heart, although these effects may vary depending on the dose route and chronicity of administration. Acute cocaine administration has two distinct dose-related effects on the heart. At low plasma concentrations, it is a local anesthetic with membrane potential stabilizing effects not unlike those of quinidine. ~5 At higher concentrations, cocaine prevents the reuptake of n o r e p i n e p h r i n e by preganglionic sympathetic nerve endings, resulting in a local excess of norepinephr!ne at the synaptic cleft.26 This results m an increase m automaticity and heart rate, a decrease in atrioventricular nodal refractoriness, and an increase in the conduction velocity of the His-Purkinje system. 27 The action of norepinephrine is terminated largely by the mechanism of reuptake. Cocaine prevents this reuptake and thereby potentiates the effects of norepinephrine, the main cardiac catecholamine. C h r o n i c a d m i n i s t r a t i o n o f cocaine has been shown to markedly increase the norepinephrine content o f rat left ventricle, although the rate of cate198

cholamine synthesis, as reflected by tissue levels of tyrosine hydroxylase, is decreased. 28 This suggests that the body responds to chronic cocaine stimulation by attempting to decrease sympathetic tone. Catecholamine-induced myocardial damage has been recognized for many years. For example, the so-called pheochromocytoma heart is characterized by the widespread presence of myocardial contraction bands frequently associated with neutrophils early and lymphocytes later; this subsequently heals by scarring. 29 Due to the intimate relationship between cocaine and catecholamines, our observation o f a direct correlation between tile presence o f contraction bands and the level of cocaine found in body fluids at autopsy has several implications regarding the possible mechanism of death in these cocaine users. It is known that the occurrence of contraction bands represents a catecholamine-induced disruption o f i n t r a c e l l u l a r calcium homeostatis. ~~ T h e pathogenesis is complex, but the end result is a recognizable lesion that was present in 93 per cent of our cases. It has been postulated that this type of myocardial damage provides the anatomic pathways necessary to produce potentially lethal reentrant arrhythmias 31 as contraction bands can be identified in more than 80 per cent of patients dying of sudden cardiac death but in only 15 per cent of those dying o f thrombotic infarctions. This may have been the mechanism o f death in many o f our cases. That acute and chronic cocaine use has different effects in the heart suggests that the mechanism of death may differ in chronic as opposed to occasional recreational users and that chronic users may actually be primed for a fatal arrhythmia by accumulating excess norepinephrine. More importantly, if tile areas o f interstitial fibrosis present in those hearts without c o r o n a r y artery atherosclerosis represent healed zones of contraction band necrosis in chronic users (as is known to happen in animals with pheochromocytomas), the chronic user may also have a fixed anatomic basis for being more susceptible to arrhythmias with continued cocaine use. In those users with obstructive coronary lesions as in the case of Simpson and Edwards, 14 the fibrosis may also be the result of ischemia. None of our cases, however, had the type of coronary lesions described by Simpson and Edwards, 14 nor did they show evidence of eosinophilic or lymphocytic myocarditis, 13A4 although foci of lymphocytic myocarditis have been identified in endomyocardial biopsy specimens in the setting of cocaine abuse (H. D. Tazelaar, MD, M. E. Billingham, MB, BS, MRCPath, unpublished observations). Contraction bands were also present in the myocardia o f o u r control group (9/20), although to a much lesser extent than in the cases. In the control g r o u p , the contraction bands were probably secondary to other factors known to contribute to their development (e.g., shock, defibrillation, or anoxia). 3~ T h e contribution of these other factors in producing contraction bands in our study cases, including that

COCAINE AND THE HEART[Tazelaar et al.]

of other drugs known to cause them (e.g., anapltetamines), probably was outweighed by the effects of cocaine. We would not have expected the direct association between the blood a n d urine levels of cocaine and the contraction bands if this were not the case. T h e c o n t r a c t i o n b a n d s i d e n t i f i e d in the case o f Simpson and Edwards 14 may have been due to ischemia per se, a direct effect of cocaine toxicity or resuscitation. Contraction bands are a nonspecific finding in the myocardium, and their presence is not pathognomonic of cocaine cardiotoxicity. T h e y do indicate that a complex set o f metabolic d e r a n g e m e n t s , chiefly involving calcium ions and catecholamines, has occurred. 3~ T h e correlation between cocaine and contraction bands as shown in this study and that of Simpson and Edwards 14 is provocative and potentially clinically important. Contraction bands may supply the anatomic substrate for tire arrhythmias and sudden deaths associated wittt cocaine use in the absence of coronary disease, suggesting a pathophysiologic mechanism of cocaine-associated cardiac toxicity. The ability to demonstrate this morphologic lesion in the myocardium at autopsy with routine histologic stains may provide confirmatory evidence that cocaine toxicity was the cause of death in a known user; it might also suggest cocaine as a cause for sudden or unexpected death if no other cause for contraction bands is identified. A d r u g or urine screen showing the presence of cocaine would be necessary to c o n f i r m this, however. With the increasing use o f cocaine, 2~ particularly a m o n g young people, 2z a rise in the number of cocaine-associated cardiac events and deaths is to be expected. Clinicians and pathologists must maintain a high index of suspicion for such cases so that appropriate drug tests and pathologic examinations can be performed. If the diagnosis is missed, the real incidence of cocaine toxicity will be grossly underestimated.

Acknowledgments. T h e authors tlmnk the staff o f the San Francisco Coroner's Office, particularly J o s e p h Surdyka, for help in obtaining the necessary information and material for this study. T h e y also thank Dr. Lincoln Moses for statistical analysis, Dr. Danna J o h n s o n and Dr. Brian H o f f m a n for helpful comments, Phil Verzola for photographic assistance, and Margaret Beers for secretarial assistance. REFERENCES 1. Pollin W: The danger of cocaine. JAMA 254:98, 1985 2. Bednarczyk LR, Gressmann EA, Wymer RL: Two cocaine-induced fatalities. J Anal Toxicol 4:263, 1980 3. Coleman DL, Ross TF, Naughton JL: Myocardial iscbemia and infarction related to recreational cocaine use. West J Med 136:444, 1982 4. Wetli CV, Wright RK: Death caused by recreational cocaine use.JAMA 241:2519, 1979 5. Benchimol A, Bartall H, Dresser KB: Accelerated ventricular rhythm and cocaine abuse. Ann Intern Med 88:519, 1978 6. Pasternack PR, Colvin SB, Baumann FG: Cocaine-induced an-

gina pectoris and acute myocardial infarction in patients younger than 40 )'ears. Am J Cardiol 55:847, 1985 7. Howard RE, Hueter DC, Davis GJ: Acute.myocardial infarction following cocaine abuse in a young woman with normal coronary arteries. JAMA 254:95, 1985 8. Kossowsky WA, Lyon AF: Cocaine and acute myocardial infarction, a probable connection. Chest 86:729, 1984 9. Young D, Glauber J J: Electrocardiographic changes resuhing from acute cocaine intoxication. Am tteartJ 34:272, I947 10. Schaclme JS, Roberts BH, Tbompson PD: Coronary artery spasm and myocardial infarction associated with cocaine use. N Englj Med 310:1665, 1984 I I. Cregler LL, Mark H: Relation of acute myocardial infarction to cocaine abuse. A m J Cardiol 56:794, 1985 12. Wilkins CE, Mathur V, Ty RC, et al: Myocardial infarction associated with cocaine abuse. Texas Heart Inst J 12:385, 1985 13. Isner JM, Estes NAM, Thompson PD, et al: Cardiac consequences of cocaine: prenmture myocardial infarction, ventricular tachyarrythmias, myocarditis and sudden death. Circulation 72(Suppl 3):415, 1985 14. Simpson RW, Edwards WD: Pathogenesis of cocaine-induced ischemic heart disease. Autopsy findings in a 21-year-old man. Arch Pathol Lab Med 110:479, 1986 15. Goldberg RJ, Capone RJ, HuntJD: Cardiac complications following tricyclic antidepressant overdose. JAMA 254:1772, 1985 16. Moses LE, Emerson JD, ttosseini H: Analyzing data from ordered categories. N EnglJ Med 311:442, 1984 17. Mittleman RE, Wetli CV; Death cause by recreational cocaine use: an update. JAMA 252:1889, 1984 18. Plowman T: Botanical perspectives on coca. ht Jeri FR (ed): Cocaine 1980, Proceedings of the lnteramerican Seminar on Coca and Cocaine. Lima, Peru, Pacific Press, 1980, pp 90-105 19. Aldrich MR, Barker RW: Historical aspects of cocaine nse and abuse. In Muld SJ: Cocaine: Chemical Biological, Clinical, Social and Treatment Aspects. Cleveland, CRC Press, 1976, pp 3-11 20. Jekel JF, Podlewski H, Dean-Patterson S, et al: Epidemic freebase cocaine abuse. Case study from the Bahamas. Lancet 1:459, 1986 21. Castro J: Battling the enemy within. Time 127:52, 1986 22. Morganthau T: Kids and cocaine. Newsweek 107:58, 1986 23. Caldwell J: Pltysiological aspects of cocaine usage. In Mul6 SJ: Cocaine: Chemical Biological, Clinical, Social and Treatment Aspects. Cleveland, CRC Press, 1976, pp 189-199 24. Barth CW III, Bray M, Roberts WC: Rupture of the ascending aorta dr, ring cocaine intoxication. Am J Cardiol 57:496, 1986 25. Ritchie JM, Greene NM: Local anesthetics. In Gilman AG, Goodman LS, Gilman A: The Pharmacological Basis of Therapentics, ed 6. New York, Macmillan, 1980, pp 307-308 26. Dart AM, Dietz R, Kubler W, et al: Effects of cocaine and desipramine on the neurally evoked overflow of endogenous noradrenaline from the rat heart. Br J Pharmacol 79:71, 1983 27. Harrison DC, Mason JW: Effects of catecholamines and adrenergic innervation on cardiac conduction and arrhythmias. In Mezey KC, Caldwell ADS (eds): Catecholamines and the Heart. Royal Society of Medicine Congress Symposium No. 8. London, Grune and Stratton, 1979, pp 17-30 28. Tarizzo V, Rubio MC: Effects of cocaine on several adrenergic system parameters. Geu Pharmacol 16:71, 1985 29. Rosenbanm J, Billingham ME, Ginsberg R, et al: Cardiomyopathy in a rat model of pheochromocytoma: morphological and functional alterations. Submitted 30. Karch SB, Billingham ME: Myocardial contraction bands revisited. HuM PATHOL 17:9, 1986 31. Reichenbach DD, Moss NS: Myocardial necrosis and sudden death in humans. Circulation 51(suppl 3):60, 1975

199