congestive heart failure, Study notes of Histology

Download congestive heart failure and more Study notes Histology in PDF only on Docsity! PATHOPHYSIOLOGY AND NATURAL HISTORY CONGESTIVE HEART FAILURE Human myocardial histologic characteristics in congestive heart failure DONALD V. UNVERFERTH, M.D., JULIE K. FETTERS, B.S., BARBARA J. UNVERFERTH, M.S., CARL V. LEIER, M.D., RAYMOND D. MAGORIEN, M.D., ANTHONY R. ARN, B.S., AND PETER B. BAKER, M.D. ABSTRACT The purpose of this study was to identify the histologic characteristics of human myocardium in congestive heart failure (CHF) by cellular hypertrophy, nuclear area, endocardial thickness, and percentage of fibrosis and to correlate histologic findings to cause, severity, and duration of disease. Right ventricular endomyocardial biopsies from 109 patients were quantitatively analyzed. Ten patients with normal cardiac history, physical examination results, and cardiac function served as the control group. The remaining patients were divided into the following groups: those treated with doxorubicin (n = 18), and those with chest pain with normal coronary arteries (n = 8), familial CHF (n = 3), CHF associated with myotonic dystrophy (n = 3), peripartal CHF (n = 2), valvular CHF (n = 9), alcohol-induced CHF (n = 13), postviral CHF (n = 6), or idiopathic CHF (n = 36). Linear regression analyses showed a strong correlation between cell diameter and nuclear area (r = .70, p < .001) and weaker correlations between amount of fibrosis and cell diameter (r = .30, p < .005) and fibrosis and nuclear area (r = .29, p < .005). Results of function studies and histologic measurements (e.g., echocardiographically measured change in the minor-axis dimension of the left ventricle with systole and cell diameter) correlated poorly (r = - .33, p < .005). Duration of dyspnea did not correlate with any histologic factor. Within the normal group there was a direct correlation of cell diameter with age (r = .67, p < .05). Analysis of covariance revealed that the doxorubicin-treated patients had significantly less cellular hypertrophy but more fibrosis than the other patients. In those with alcohol-induced CHF the opposite pattern of more hypertrophy and less fibrosis was observed. The idiopathic and valvular CHF groups did not differ from each other or from the other groups. The results of this study demonstrate that fibrosis and hypertrophy tend to be worse in patients with severe CHF and that some myocardial insults (especially that of doxorubicin) induce a characteristic response to the injury. Circulation 68, No. 6, 1194-1200, 1983. THE MYOCARDIUM has a simple structure and a limited range of responses to a wide variety of insults. These responses include fibrosis and cellular hypertro- phy. The pattern and severity of fibrosis and hypertro- phy have been used with varying success to differenti- ate hypertrophic from congestive cardiomyopathy,' 2 but have not been used to characterize the various types of nonischemic congestive heart failure (CHF). The endocardial biopsy procedure has facilitated the study of cardiomyopathy. Before the advent of heart biopsy procedures, study was confined to that of au- topsy specimens. The results of these studies reflected From the Departments of Medicine and Pathology, Ohio State Uni- versity College of Medicine, Columbus. This study was supported in part by the Samuel J. Roessler Research Foundation, the James D. Casto Research Foundation, and the Central Ohio Heart Chapter of the American Heart Association. Address for correspondence: Donald V. Unverferth, M.D., 657 Means Hall, 1655 Upham Dr., Columbus, OH 43210. Received Feb. 15, 1983; revision accepted Aug. 11, 1983. 1194 agonal and postmortem autolytic changes and all specimens were obtained from patients with severe end-stage disease. Endomyocardial biopsy has enabled pathologists to study cardiomyopathies of different se- verity and duration.3-7 The purpose of this study was to use the endomyocardial biopsy to characterize CHF by cause, severity, and duration of myocardial injury. The hypothesis was that, although the heart has a limit- ed range of responses to injury, the response will vary according to the nature as well as the duration and severity of injury. Our results demonstrated that the light-microscopic characteristics of cellular hyper- trophy, nuclear size, endocardial thickness, and per- cent fibrosis correlate weakly with myocardial func- tion. Doxorubicin-induced CHF had a distinctive pattern of marked fibrosis with little hypertrophy, while alcohol-induced and valvular CHF did not have a distinctive pattern. In addition, age correlated with cell size in normal subjects. Equally important was the CIRCULATION D ow nloaded from http://ahajournals.org by on June 17, 2022 PATHOPHYSIOLOGY AND NATURAL HISTORY-CONGESTIVE HEART FAILURE absence of correlation of histologic changes with dura- tion of dyspnea. It is hoped that this information can help in the understanding of the pathogenesis of CHF and possibly the cause of idiopathic cardiomyopathy. Methods Patient selection. Endomyocardial biopsy samples from 109 patients were evaluated. These patients comprised the 10 fol- lowing groups. (1) Normal subjects (n = 10). These subjects had normal cardiovascular histories and physical examination results. In addition, results of noninvasive tests to determine systolic time intervals and of echocardiographic examinations were normal. Each of these subjects had had a tumor necessitat- ing doxorubicin chemotherapy and had participated in a study in which serial endomyocardial biopsy procedures (including at baseline) were performed.8 Data obtained from the baseline biopsies are the control data in this study. The Human Research Committee of our institution approved the previous study8 and also approved of our use of the tissue for the present study. (2) Doxorubicin-treated patients (n = 18). These patients had re- ceived over 200 mg/mi2 of doxorubicin (mean = 400 mg/mi2). Cardiac function test results at the time of biopsy ranged from normal to severely depressed (table 1). The biopsies were taken from I to 6 months after the last dose of doxorubicin. All patients had had normal cardiac function before undergoing doxorubicin chemotherapy. (3) Patients with chest pain and normal coronary arteries (n = 8). These patients had normal systolic function and normal coronary arteries as determined by cardiac catheterization, but each had chest pain that mimicked angina. (4) Patients with familial CHF (n = 3). These patients had dilated congestive cardiomyopathy of undetermined cause. Each had two or more first-degree relatives with dilated conges- tive cardiomyopathy. (5) Patients with CHF associated with myotonic dystrophy (n = 3). The clinical findings in these patients were consistent with myotonic dystrophy and there was no other apparent cause of their CHF. (6) Patients with peripar- tum CHF (n = 2). These two patients developed heart failure within I month after the delivery of a child and there was no other apparent cause of their CHF. (7) Patients with valvular CHF (n = 9). All patients in this group had severe valvular disease necessitating mitral (in three) or aortic (in four) valve replacement or both (in two). Each of these patients was still in heart failure 3 months or longer after valve replacement surgery. Catheterization revealed normal coronary arteries and normal prosthetic function. (8) Patients with alcohol-induced CHF (n - 13). Each of these patients had a dilated heart with poor systolic and diastolic function (table 1) and a history of heavy alcohol abuse that included the consumption of greater than I pint of whiskey (or equivalent) per day for over 2 years. None of these patients had consumed alcohol in the 2 months before they underwent the biopsy procedure. (9) Patients with postviral CHF (n = 6). Each of these patients had had biopsy-document- ed inflammatory myocarditis (more than five inflammatory cells per high-power field, n = 4) and/or a rise of viral titers during an acute viral-like illness that included arrhythmias and electrocardiographic changes followed by the development of heart failure (n = 3). The biopsied tissue used for this study was obtained at least 3 months after the acute episode. No inflamma- tory cells were present in the biopsy samples at the time of this study. (10) Patients with idiopathic CHF (n = 36). These pa- tients had dilated cardiomyopathies not attributable to alcohol abuse, family history of heart failure, myocarditis, primary valvular disease, cancer chemotherapy, or exposure to other cardiac toxins and onset was not within 3 months of the delivery of a child. They fulfilled the criteria of the World Health Organ- ization for the classification of idiopathic dilated cardiomyopa- thy.9 Ten of these patients had a history of hypertension (> 140/ 90). None of these 10 had had uncontrolled or malignant hyper- tension. All subjects gave written informed consent before un- dergoing biopsy. Noninvasive testing. Noninvasive cardiac function evalua- tion was performed in all 109 patients. Each patient rested for 30 min before testing. Echocardiograms were performed with a Unirad Series C ultrasonoscope and an Electronics for Medicine VR6 recorder or an SK Echoline 20A echocardiograph with an Irex Recorder. The change in the minor-axis dimension of the left ventricle with systole (%AD) was the echocardiographic determinant of ventricular function. Echocardiographic inter- pretation was performed in accordance with the recommenda- tions of the American Society of Echocardiography. '0 Systolic time intervals were determined with an Electronics for Medicine VR6 or an Irex Continutrace Unit with the specifications pre- viously outlined.1' The ratio of the preejection period to left ventricular ejection time was used as the measure of left ventric- ular function. The reproducibility and reliability of these tests have been demonstrated in our laboratory.'2 Hemodynamic evaluation. Diagnostic cardiac catheteriza- tion was performed in all subjects in groups 3 through 10. No coronary artery had an obstruction measuring > 50% of the diameter of the vessel. Left and right heart pressures were mea- sured with an Electronics for Medicine VR16 Multichannel Recorder, which has an M2101 Pressure Amplification Unit, or a Hewlett-Packard Model 8800 Recording System, which has a 4568C pressure unit. The frequency cutoff for both systems was from 0 to 250 Hz. Left ventriculography was performed with 42 ml of sodium meglumine diatrizoate (Renografin 76) injected over 3.5 sec (12 ml/sec). Left ventricular ejection fraction was calculated by the method of Sandler and Dodge.'3 Cardiac out- put was determined by a dye dilution technique in which indo- cyanin green dye is used.'4 Endomyocardial biopsy. A right ventricular endomyocar- dial biopsy procedure was performed via the right internal jugu- lar vein'5 within 1 week from the time of the noninvasive tests and cardiac catheterization. Each sample used for this study measured 1 to 2 mm' and weighed I to 4 mg. The tissue was placed in 2% buffered glutaraldehyde solution (pH 7.35) within 20 sec (mean 11 sec) of the closure of the jaws of the bioptome on the right ventricular septal myocardium. This sam- ple was sent for histologic evaluation. Biopsy procedures were repeated until adequate amounts of tissue (> 1 mm3) were ob- tained. Histologic evaluation. The glutaraldehyde-fixed tissue was processed in a routine manner. Light-microscopic slides, 3 to 5 ,um thick, were prepared from paraffin-embedded sections. The slides were stained by hematoxylin and eosin and Masson's trichrome stain by the methods of Sheehan and Hrapchak.'6 Hematoxylin and eosin-stained myocardial slices were exam- ined with a 43 x objective lens and 1O x eyepiece with an ocular micrometer disc. The diameter of a myocardial cell was determined by measuring the distance across the cell at its most narrow plane across the nucleus. At least 50 cells from each biopsy were measured and the diameters of the 50 cells were averaged to determine the value for each patient. Ninety-five percent of measured cell diameters for a single biopsy were within 4 gm of the mean. A lesser range was noted for normal biopsy tissue or for biopsies with a smaller mean cell diameter. In addition, the average nuclear area was determined in the 50 cells used for cell size measurements. The height (h) and width (w) of each nucleus was measured and nuclear area was calcu- lated from the formula: nuclear area = (h + w/4)2 X 7r. Ninety-five percent of the measured nuclear areas were within 9 ,um2 of the mean. Trichrome-stained sections were examined at a magnification of 400 x on a viewing screen. The viewing Vol. 68, No. 6, December 1983 1195 D ow nloaded from http://ahajournals.org by on June 17, 2022 UNVERFERTH et al. 24 - 22 CELL 20 * DIAMETER ,u 18- 16- 14- rr-.67 *t p<0.05 12 26 34 42 50 58 66 AGE IN YEARS FIGURE 3. In group 1 (10 normal subjects) cell diame directly with the age of the patients. comparing the slopes for the individual gro slopes were inscribed by plotting the cel (figure 2) or percentage of fibrosis (figure the echocardiographic %AD. The linear analyses were forced through a "normal' rived from the data of the 10 normal subjc study. This was done because we assumec patient was normal before myocardial insult line for each group should reflect the path myocardial response to injury. The comparison of slopes for the change with progressive heart failure (as determii echocardiographic %AD) is shown in figur four largest groups. The slope inscribed b from the alcohol-induced CHF group was si different from that inscribed by data from dc treated patients (p < .05). The idiopathic ar CHF group slopes were intermediate and I cantly different from those for either the duced CHF or doxorubicin-treated groups illustrates the slopes derived for fibrosis plot %zcD. The doxorubicin-treated group was s ly different from the alcohol-induced and CHF groups (both p < .05), but not from tI CHF group. The valvular CHF group did from any of the others. Discussion Methodologic considerations. The use of q histologic procedures in endomyocardial plagued by the problems of sampling error a artifact. The biopsied tissue used in this stu taken from the right side of the interventri tum. However, Baandrup and Olsen2 have right ventricular endomyocardial biopsy st equivalent to those removed from the left. II our laboratory demonstrated a consistent pattern of fibrosis, vacuolization, and hypertrophy in samples /0 from the right and left.'8 Although there are varying degrees of degeneration at different levels of the myo- cardium, the patterns are consistent in CHF. 18 Also, in a study of myocardial biopsy samples taken close to the time of death, it was demonstrated that these sam- ples were representative of overall cardiac pathol- ogy.3 19 Thus, right-sided endomyocardial biopsies can be used to determine cardiac morphologic alter- ations and can be compared with biopsies of the same -,------,---. area of the hearts of normal subjects and other CHF 74 82 patients. Fixation artifact is primarily a problem in electron microscopy but can cause some vacuolization in light- microscopic specimens.20 This problem has been obvi- ups. These ated by the use of tissue samples of similar size (I to 2 11 diameter mm'), eliminating cells with contraction-band arti- 3) against facts, and the use of the same techniques for the pro- regression cessing of all tissues. point de- Histologic and clinical correlations. For the total patient ects in this population (n = 109), in which there was a spectrum I that each of normal-to-severe CHF and disparate causes of dis- Thus, the ease, several interesting correlations were noted. As way of the expected, cell diameter correlated closely with nuclear area (r = .70, p < .0001). Increased nuclear activity in cell size (and therefore size) is a primary requisite for cellular ned by the hypertrophy.21 22 There were also weak correlations ^e 2 for the between percentage of fibrosis and cell diameter (r = )y the data .30, p < .005) and percentage of fibrosis and nuclear Ignificantly area (r = .29, p < .005). This was also expected )xorubicin- because hypertrophy is a stimulus for fibrosis.23 24 The ad valvular low r value (.30) probably reflects the many factors not signifi- besides hypertrophy that promote fibrosis. Endocar- alcohol-in- dial thickness was independent of the other histologic Figure 3 markers. "ted against Cell diameter correlated weakly with cardiac func- significant- tion measures such as echocardiographic %AD (r - idiopathic -.33, p < .005), the ratio of preejection period/left he valvular ventricular ejection time (r = .23, p < .05), and mean not differ pulmonary capillary wedge pressure (r - .32, p < .01). The correlation between cell diameter with echo- cardiographic %AD improved, however, when indi- vidual groups were considered independently and [uantitative when the regression lines were forced through a nor- biopsies is mal point (figure 1). This suggests an association of nd fixation worsening heart function with increasing cellular dy was all hypertrophy. icular sep- Amount of fibrosis also correlated poorly with mea- shown that sures of myocardial function for the total population, imples are but improved for individual groups (figure 2). Percent- n addition, age of fibrosis did not correlate with mean pulmonary CIRCULATION1198 D ow nloaded from http://ahajournals.org by on June 17, 2022 PATHOPHYSIOLOGY AND NATURAL HISTORY-CONGESTIVE HEART FAILURE capillary wedge pressure. These results are consistent with those of a previous study25 that showed that fibro- sis was less important than cellular hypertrophy or myocardial adenosine triphosphate content in raising left ventricular filling pressure. Endocardial thickness did correlate weakly with ejection fraction (r = .24, p < .05) and cardiac index (r = .25, p < .005). None of the histologic measurements correlated with the duration of dyspnea or the patient age. Baan- drup et al.6 also did not find a relationship between the duration of disease and histologic change. Shirey et al. ,26 however, did find that amounts of fibrosis and cellular hypertrophy correlated with prognosis. We did not address the question of prognosis. Myocardial response to insult. One purpose of this study was to determine whether different myocardial insults resulted in certain pathologic responses. An important negative finding of this study was that, in this regard, no difference could be detected between the valvular, idiopathic, and alcohol-induced CHF groups. The doxorubicin-treated group was distin- guishable, however. For a given loss of systolic myo- cardial function (as measured by the echocardiograph- ic %/D), the alcohol-induced CHF group had the greatest hypertrophy, while the doxorubicin-treated group had the least (figure 1). Doxorubicin, an effec- tive cancer chemotherapeutic drug, intercalates DNA27 28 and interferes with protein production.29 0 Human myocardial nuclear studies have demonstrated signs of doxorubicin-induced nuclear degeneration.31 We hypothesize that the nuclear effect of doxorubicin inhibits myocardial cell growth. The mechanism for the markedly increased hypertrophy induced by alco- hol is not clear. Percentage of fibrosis in myocardium of patients also showed that doxorubicin injury was distinct from the other types (figure 2). Fibrosis was most severe in doxorubicin-treated patients and least severe in those with idiopathic CHF and alcohol-induced injuries (p < .05). Within the individual groups interesting relation- ships were observed that facilitate understanding of the myocardial response to injury. The normal group was quite interesting because of the good correlation (r = .67) between age and cell size (figure 3), but not be- tween age and percentage of fibrosis. Aging may be an insult to the human myocardium that results in hyper- trophy but not fibrosis and the previously described alterations in contractile properties in the aging heart32 may be related to cellular hypertrophy. Only 10 pa- tients were included in our normal group and further studies are needed to corroborate this data. Group 2, which included those with doxorubicin- induced cardiomyopathy, was unique because there was little to no hypertrophy observed, but amount of fibrosis was markedly increased, as previously dis- cussed. Amount of fibrosis did not correlate with the total cumulative dose of doxorubicin, suggesting a widely disparate fibrotic reaction of the individual pa- tients' hearts. This is consistent with the clinical expe- rience, which shows that some patients are less tolerant of doxorubicin because they are older,33 34 had pre- vious radiation therapy,35 or underwent other cardio- toxic chemotherapy,36" 37 while some patients can toler- ate massive doses. Group 3 patients, who had chest pain with normal coronary arteries, did not differ from the normal group in any of the histologic or hemodynamic characteris- tics. Groups 4 through 6 (familial, myotonic dystro- phy, and peripartal CHF) were small so that they could not be effectively distinguished from patients with oth- er types of CHF. However, the familial CHF patients (n = 3) had the largest cells and as a group had only modest heart failure. The peripartum CHF patients (n = 2) had the most pronounced fibrosis and only mod- erate heart failure. Valvular heart disease patients demonstrated an average amount of fibrosis and cellular hypertrophy commensurate with the degree of myocardial dysfunc- tion (figures I and 2). Alcohol-induced CHF patients, in addition to having remarkable hypertrophy, were the only patients in which a correlation between amount of fibrosis and endocardial thickness could be demonstrated (r = .56, p < .05). This relationship had been expected to hold for the total patient population but it did not. The idiopathic CHF group probably contained sev- eral ill-defined subgroups. Hypertension seemed to distinguish one subgroup with more cellular hyper- trophy. Many of the patients in group 10 were prob- ably victims of a subclinical viral myocarditis. How- ever, the light-microscopic characteristics of the postviral and the idiopathic CHF groups were not simi- lar. Further analysis and definition of this group will be necessary to understand the pathogenesis of cellular injury. We thank Dawn Orth and Pam Newton for their technical assistance and Tami Smith for the preparation of the manu- script. References 1. Noda S: Histopathology of endomyocardial biopsies from patients with idiopathic cardiomyopathy: quantitative evaluation based on multivariate statistical analysis. Jpn Circ J 44: 95, 1980 2. Baandrup U, Olsen EGJ: Critical analysis of endomyocardial biop- 1199Vol. 68, No. 6, December 1983 D ow nloaded from http://ahajournals.org by on June 17, 2022 UNVERFERTH et al. sies from patients suspected of having cardiomyopathy. I: Morpho- logical and morphometric aspects. Br Heart J 45: 475, 1981 3. Mackay EH, Littler WA, Sleight P: Critical assessment of diagnos- tic value of endomyocardial biopsy. Assessment of cardiac biopsy. Br Heart J 40: 69, 1978 4. Kawai C, Matsumori A, Kawamura K: Myocardial biopsy. Ann Rev Med 31: 139, 1980 5. Das JP, Rath B, Das S, Sarangi A: Study of endomyocardial biopsies in cardiomyopathy. Indian Heart J 33: 18, 1981 6. Baandrup U, Florio RA, Rehahn M, Richardson PJ, Olsen EGJ: Critical analysis of endomyocardial biopsies from patients suspect- ed of having cardiomyopathy. II: Comparison of histology and clinical/haemodynamic information. Br Heart J 45: 487, 1981 7. Baandrup U, Florio RA, Royers F, Olsen EGJ: Electron micro- scopic investigation of endomyocardial biopsy samples in hyper- trophy and cardiomyopathy. A semiquantitative study in 48 pa- tients. Circulation 63: 1289, 1981 8. Unverferth DV, Magorien RD, Unverferth BP, Talley FL, Balcer- zak SP, Baba N: Human myocardial morphologic and functional changes in the first 24 hours after doxorubicin administration. Cancer Treat Rep 65: 1093, 1981 9. World Health Organization/ISFC Task Force: report of the WHO- ISFC task force on the definition and classification of cardiomyop- athies. Br Heart J 44: 672, 1980 10. Sahn DJ, DeMaria A, Kisslo J, Weyman A: Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation 58: 1072, 1978 11. Lewis RP, Leighton RF, Forester WF, Weissler AM: Systolic time intervals. In Weissler AM, editor: Non-invasive cardiology. New York, 1974, Grune and Stratton, p 301 12. Unverferth DV, Lewis RP, Leier CV, Magorien RD, Fulkerson PK: Reproducibility of echocardiography in chronic congestive heart failure. J Clin Ultrasound 8: 479, 1980 13. Sandler H, Dodge HT: The use of single plane angiocardiograms for the calculation of left ventricular volume in man. Am Heart J 75: 325, 1968 14. Wade OL, Bishop JM: Symposium on the use of indicator dilution techniques in the study of the circulation. Circ Res 10: 371, 1962 15. Mason JW: Techniques for right and left ventricular endomyocar- dial biopsy. Am J Cardiol 41: 887, 1978 16. Sheehan DC, Hrapchak BB: Connective tissue and muscle fiber stains. In Theory and practice of histotechnology. St Louis, 1973, C.V. Mosby Co. 17. Weibel ER: Stereological principles for morphometry in electron microscopic cytology. Int Rev Cytol 26: 235, 1969 18. Dick MR, Unverferth DV, Baba N: The pattern of myocardial degeneration in nonischemic congestive cardiomyopathy. Hum Pathol 13: 740, 1982 19. Fenoglio JJ Jr, Ursell PC, Kellogg CF, Drusin RE, Weiss MB: Diagnosis and classification of myocarditis by endomyocardial bi- opsy. N Engl J Med 308: 12, 1983 20. Adomian GE, Laks MM, Billingham ME: The incidence and sig- nificance of contraction bands in endomyocardial biopsies from normal human hearts. Am Heart J 95: 348, 1978 21. Wikman-Coffelt J, Parmley WW, Mason DT: The cardiac hyper- trophy process. Analyses of factors determining pathological vs physiological development. Circ Res 45: 697, 1979 22. Rabinowitz M: Overview on pathogenesis of cardiac hypertrophy. Circ Res 34135: 11-3, 1974 23. Moore GW, Hutchins GM, Bulkley BH, Tseng JS, Ki PF: Con- stituents of the human ventricular myocardium: connective tissue hyperplasia accompanying muscular hypertrophy. Am Heart J 100: 610, 1980 24. Sen S, Bumpus FM: Collagen synthesis in development and rever- sal of cardiac hypertrophy in spontaneously hypertensive rats. Am J Cardiol 44: 954, 1979 25. Unverferth DV, Magorien RD, Kolibash AJ, Lewis RP, Lykens M, Altschuld R, Baba N, Leier CV: Biochemical and histologic correlates of ventricular end-diastolic pressure. Int J Cardiol 1: 133, 1981 26. Shirey EK, Proudfit WL, Hawk WA: Primary myocardial disease. Correlation with clinical findings, angiographic and biopsy diagno- sis. Follow-up of 139 patients. Am Heart J 99: 198, 1980 27. Pigram WJ, Fuller W, Hamilton LD: Stereochemistry of intercala- tion: Interaction of daunomycin with DNA. Nature New Biol 235: 17, 1972 28. Zunino F, Gambetta R, DiMarco A, Velcich A, Zaccara A, Qua- drifoglio F, Crescenzi V: The interaction of Adriamycin and its B anomer with DNA. Biochim Biophys Acta 476: 38, 1977 29. Zunino F, Gambetta R, DiMarco A: The inhibition in vitro of DNA polymerase and RNA polymerases by daunomycin and Adriamy- cin. Biochem Pharmacol 24: 309, 1978 30. Formelli F, Zedeck MS, Stemnberg SS, Philips FS: Effects of Adriamycin on DNA synthesis in mouse and rat heart. Cancer Res 38: 3286, 1978 31. Unverferth BJ, Magorien RD, Balcerzak SP, Leier CV, Unverferth DV: Early changes in human myocardial nuclei after doxorubicin, Cancer 52: 215, 1983 32. Heller LJ, Whitehorn WV: Age-associated alterations in myocardi- al contractile properties. Am J Physiol 222: 1613, 1972 33. Van Hoff DD, Layard MW, Basa P: Risk factors of doxorubicin- induced congestive heart failure. Ann Intern Med 91: 710, 1979 34. Bonadonna G, Monfardini S: Cardiac toxicity of daunorubicin. Lancet 1: 837, 1969 35. Billingham ME, Bristow MR, Glatstein E, Mason JW, Masek MA, Daniels JR: Adriamycin cardiotoxicity: endomyocardial biopsy evidence of enhancement by irradiation. Am J Surg Pathol 1: 17, 1977 36. Kushner JP, Hansen VL, Hammar SP: Cardiomyopathy after wide- ly separated courses of Adriamycin exacerbated by actinomycin-D and mithramycin. Cancer 36: 1577, 1975 37. Buzdar AU, Legha SS, Tashima CK, Hortobagyi GN, Yap HY, Krutchik AN, Luna MA, Blumenschein GR: Adriamycin and mito- mycin C: possible synergistic cardiotoxicity. Cancer Treat Rep 62: 1005, 1978 CIRCULATION1200 D ow nloaded from http://ahajournals.org by on June 17, 2022