* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Print - Circulation
Survey
Document related concepts
Heart failure wikipedia , lookup
Electrocardiography wikipedia , lookup
Remote ischemic conditioning wikipedia , lookup
Saturated fat and cardiovascular disease wikipedia , lookup
Quantium Medical Cardiac Output wikipedia , lookup
Cardiovascular disease wikipedia , lookup
Cardiac surgery wikipedia , lookup
Drug-eluting stent wikipedia , lookup
History of invasive and interventional cardiology wikipedia , lookup
Dextro-Transposition of the great arteries wikipedia , lookup
Transcript
DECEMBER 1951 NO. 6 VOLl. IV Circu latlion The Journal of the American J fert Association Anastomoses in the Human Heart, with Particular Reference to Anemia and Relative Cardiac Anoxia Interarterial Coronary Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 By PAUL M. ZOLL, M.D., STANFORD WESSLER, M.D., AND MONROE J. SCHLESINGER, M.D. In an unselected series of 1050 human hearts the coronaryv arteries were uniformly injected at necropsy with a standardized radiopaque mass. Interarterial coronary anastomoses were clearly demonstrated by this method and were studied faith respect to their incidence and pathogenesis. Anastomoses were significantly increased in hearts with coronary artery occlusion or marked narpulmonale, in cardiac hypertrophy and valvular lesions, and also in normal hearts rowing in from patients with anemia. Relative cardiac anoxia is present in all these conditions; it appears to be a common underlying stimulus for the development of interarterial coronary anastomoses. cor as bypasses or detours whereby blood reaches vessels distal to occlusions. These observations have been confirmed in the experimental animal.3 Other collateral channels, such as extracardiac anastomoses4-31 and communications of the thebesian type4-6 9, 18, 21, 29 32-76 between the cavities of the heart and coronary arteries, capillaries or veins, have not appeared to be significant in the maintenance of adequate coronary blood florw. Any regimen that would produce interarterial anastomoses so that the coronary arteries were no longer functionally end-arteries would mitigate the clinical manifestations of coronary artery disease. Knowledge of the factors influencing the development of interarterial coronary anastomoses, therefore, would be of great practical value. This paper presents the results of a study of coronary interarterial anastomoses in a series of 1271 human hearts examined by a uniform technic of injection and dissection. This unique body of data has provided a better understanding of the role of interarterial corollary anastomoses in normal and diseased states: insight into the basic mechanism of their production HE CLINICAL effects of coronary arteriosclerosis, in the form of angina pectoris, myocardial infarction and congestive failure, constitute a large portion of the serious cardiac disabilities that the physician constantly faces. Studies of atherosclerosis may, at some future date, lead to the prevention of T these lesions. In the meantime, it has become clear that narrowing or even occlusion of the coronary arteries, under certain circumstances, does not always produce clinical symptoms or myocardial damage.' This apparent inconsistency between the presence of long-standing obstructive lesions and the absence of pathologic or clinical evidence of myocardial damage has been dispelled by the demonstration in man of larger than normal interarterial coronary anastomoses.2 These anastomoses serve From the Medical Research Department of the Yamins Research Laboratories and the Pathology Department of the Beth Israel Hospital, and the Departments of Medicine and Pathology, Harvard Medical School, Boston, Mass. This investigation was aided by grants from the Life Insurance Medical Research Fund, the U. S. Public Health Service, the Patrons of Research an(1 the Sidney R. Green Heart Research Fund of the Beth Israel Hospital, Boston, Mass. 797 Circulation, Volume IV, December. 19.51 798 INT7NRARTERIAL CORONARY ANASTOMOSES IN HUMAN HEART Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 has been gained and several factors affecting their development have been identified. The subject of coronary interarterial anastomoses has been investigated in many different ways since 1669 when Lower34 stated, "The vessels which carry blood to the heart come together again, and here and there communicate by anastomoses." Simple dissections of coronary vessels in both human and animal hearts have been supplemented by injections with various materials.4-14, 16-27, 29, 32, 33, 36-41, 45, 51, 55, 64, 67, 68, 71, 72, 74-121 Water, air, suet, Max, oils, turpentine, dyes, inks, starch, bacteria, Wood's metal, mercury, iron, bismuth and lead compounds, microscopic glass spheres, radioactive erythrocytes, celloidin, gelatin, latex and liquid nylon have been used. For the most part, these studies have confirmed and led to the general acceptance of the idea that interarterial coronary anastomoses exist. Occasional disagreements 81, 82, 85, 87-89, 94, 95 have resulted largely from differences in materials or technics of injection, from anatomic variations among species, from varying pathologic conditions in the same species or, most frequently, from differences in interpretation of results. All investigators agree that, when watery solutions or suspensions of very fine particles like India ink are injected into one coronary artery, they appear promptly in the other coronary arteries and their branches.'22 Such materials, however, traverse the capillary bed and reach the coronary venous system. They cannot, therefore, be used to determine at which level (artery, arteriole or capillary) intercoronary arterial communications occur. To circumvent this difficulty it is necessary to use injection masses which do not traverse the capillary bed.7 51,109 The lead-agar mass employed in this investigation109 was regularly found in vessels as small as 40 microns in internal diameter, irregularly penetrated as far as vessels from 40 to 10 microns in diameter but was not found in vessels under 10 microns in diameter. These measurements of vessel size were made on stained microscopic sections prepared from tissue fixed in formalin and embedded in paraffin; they are therefore subject to a correction of about 50 to 100 per cent for shrinkage from the fresh unfixed state. The actual diameter of injected vessels is less important than their nature (i.e., arteriole, capillary or venule): with this method the leadagar mass did not enter the capillary bed and did not appear in the venous system. The level of penetration of injected material probably depends on many other physical properties in addition to the size of the particles. Although most of the individual particles of lead phosphate in the lead phosphate-agar mass measured about one micron in size and exhibited brownian movement in aqueous suspension, a suspension of them did not penetrate vessels many times larger. Differences in extent of penetration of the mass in different hearts, from "stumpy" injection of only the larger arterioles to very "fine" injections with filling of smallest arterioles could be produced by purposeful variations in technic. Our earlier studies2' 109, 123 on smaller numbers of hearts injected by this technic revealed interarterial coronary anastomoses in 10 to 22 per cent of human hearts that were otherwise normal at autopsy. The occurrence of anastomoses was found to be unrelated to age and sex, but it was definitely increased in the presence of valvular disease and coronary artery narrowing. Anastomoses were always found in hearts with old complete coronary artery occlusions. These observations have been confirmed by other workers."'1 117,117 METHOD AND MATERIAL An autopsy series of 1271 unselected human hearts has been studied at the Beth Israel Hospital, Boston, by a uniform injection plus dissection technic,'09 Briefly this method consisted of the following steps: (1) Injecting radiopaque lead-agar mass of different colors into the main coronary arteries at a pressure of 150 to 200 mm. Hg. (2) Unrolling the heart so that the coronar'y arterial tree lay in one plane. (3) Taking a roentgenogram of the unrolled heart. (4) Carefully dissecting and opening the coronary arteries, using the roentgenogram as a guide. With this method the demonstration of narrowings, occlusions and anastomoses in the coronary arteries was greatly facilitated. The technic of unrolling the heart reduced the difficulty in following the three-dimensional course of the vessels on a two-dimensional film. Criteria were established for the determination of anastomotic communications. It was found that the roentgenograms alone could not be relied upon P. M. ZOLL, S. WESSLER AND M. J. SCHLESINGER to prove the existence of such communications. Overlapping vessels, like the bare branches of adjacent trees,5' often deceptively gave the appearance of intercommunications. Conversely, anastomoses could be clearly demonstrated on careful dissection which were not visible on the roentgenogram. On the basis of careful dissection three types of proof of anastomoses were accepted: (1) demonstration of a continuous, stained, endothelial-lined connecting channel filled with mass between two coronary arteries; (2) presence of injection mass of any color distal to a complete occlusion and (3) visualization of a mixture of color in the injection mass. (1) Demonstration of a Continuous, EndothelialLined, Connecting Channel Filled with Mass between 799 was important in supporting the validity of the other two methods of demonstrating interarterial anastomoses. (2) Presence of Injection Mass of Any Color Distal to a Complete Occlusion. Since a complete coronary artery occlusion prevented direct filling of the peripheral vascular tree, mass of any color found beyond an occlusion must have gotten there by collateral pathways. Interarterial anastomoses could be demonstrated by finding mass of the same color both proximal and distal to an occlusion. (3) Visualization of a Mixture of Colors in the Injection Mass. Since the color of the injected mass identified its source, mixture of colors necessarily indicated a connecting pathway between two coro- Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 FIG. 1. Colored roentgenogram of injected heart (A 43-128) with multiple coronary artery occlusions and extensive interarterial coronary anastomoses. Anastomoses are demonstrated by (a) grossly dissectible channel between two coronary arteries; (b) injection mass of any color distal to complete occlusion; (c) mixture of colors of injection mass. (The use of color in this illustration was made possible by a grant from Wyeth, Incorporated, to the publication fund of the American Heart Association.) Two Coronary Arteries. Communications were proved to be unquestionable interarterial anastomoses only when a continuous, intact, vascular pathway could be demonstrated connecting two coronary arteries. Grossly dissectible anastomoses were demonstrable when the vessel was about 0.2 mm. in diameter or larger, but vessels, less than 0.5 mm. in diameter were not readily opened. Smaller anastomoses 60 to 70 microns in diameter could be traced along their entire course by laborious study under the dissecting microscope in tissue cleared by the Spalteholz technic'00 or by reconstruction of a complete series of microscopic sections. Vessels of this caliber were not visible on the roentgenogram. Indisputable demonstration of grossly dissectible pathways in this way was possible in only about one-fifth of the hearts in which anastomoses were accepted as present. Although somewhat infrequent, this method nary arteries. This was the most common method by which interarterial communications were demonstrated. The validity of accepting mixture of colors as evidence of an interarterial anastomosis rather than of an artificial extravascular connecting channel or of a common capillary bed between two arteries depended upon three lines of supporting evidence: (a) the absence of gross or microscopic evidence of extravasations of mass; (b) the absence of mass in capillaries or veins and (c) the frequent gross or microscopic demonstration of a continuous, indisputable endothelial-lined vessel connecting two coronary arteries in hearts in which mixture of colors was found. Figure 1 shows the colored roentgenogram of an injected heart in which intercoronary arterial anastomoses were demonstrated by all three methods. Although areas of even slight narrowing stood 800 INTERARTER1AL CORONARY ANASTOMOSES IN HUMAN HEART permitted classification according to the presence or absence of anastomoses, cardiac hypertrophy, valvular and congenital heart disease, coronaryl artery narrowings and occlusions and cor pulmonale. Tabulation was also made of clinical data of possibly related significance such as age, sex, arterial hypertension, pulmonary, renal and metabolic diseases, chronic wasting disease, prolonged cardiovascular collapse, surgical operations, duration of terminal illness and anemia. out on the film as irregularities in the shadow of the injected mass, the presence and extent of suspected coronary artery narrowings occlusions were determined primarily in the dissection, Mwhen the arteries were commonly opened down to small branches. The solidified mass found in them as a cast of the lumen facilitated determination of the degree of narrowing or the presence of complete occlusions. The staining of the intima by d(ye diffused from the mass also aided dissection of small or or WITH CORONARY ARTERY DISEASE NO CORONARY ARTERY DISEASE COR GROSSLY VALVULAR NORMAL DISEASE PULMONALE NARROWING slight moderate marked 80 WITHOUT CARDIAC HYPERTROPHY OCCLUSION 1 old recent 60- Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 40- E 101 91 5 5 ° 100i WITH CARDIAC 801 HYPERTROPHY No. * totl 24 19 27 17 numb or in * too few for separate analysis all coes grouped together without reference to anemIa or 33 8 * without anemIa _ m hypertrophy with anemia FIG. 2. The incidence of intercoronary arterial anastomoses in relation to anemia, cardiac hypertrophy, valvular and coronary artery disease. narrowed vessels. At sites of complete occlusion neither mass nor stained intima w-as found. After dissection of the arteries to determine anastomoses and the extent of coronary narrowing and occlusion, the mavocar(lium, valves and both cardliac surfaces were carefully examined in each heart and suitable sections taken for microscopic study. From about one-half of the total series of injected hearts a large number of individually labeled sections from exactly localized sites in the heart were made in order to permit closes correlation of the local histopathology- of the my-ocardium with changes in the coronary arteries. Of the 1271 injected hearts available, 1050 w-ere selected for the present study on intercoronaryarterial anastomoses. The qualifications for selection were an injection adequate to determine interarterial anastomosis and other (letaile(l postmortem and clinical data .* The cardiac pathologic data of simplication, !92 of the cases with all the above categories were omitted from the present series b)eca.use they exhibite(l coml)iinations of valvular heart(disease and cor* For purposes ale(uate (lata in onary artery pathology. RESULTS Table 1 presents the complete data on these 1050 hearts in all their subdivisions; some of the groups are presented individually in tables 2 to 5. Of this selected group, 480 (46 per cent) showed interarterial coronary anastomoses. Figure 2 presents an analysis of all hearts with anastomoses with respect to the relative frequency of the groups into which these 480 cases fall. The Normal Heart and Coronary Artery Anastomoses Criterion of the Normal Heart. Of the 1050 hearts in the series, only 244 (23 per cent) were considered as a normal, control group (table 1). A.ll these hearts weighed less than 350 Gm. and did not showv any valvular or congenital lesions, right ventricular hypertrophy nary artery narrowing or or coro- occlusion. Heart P. M. ZOLL, S. WESSLElR AND M. J. Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 weights under 350 Gm. w-ere considered normal for reasons which will be discussed in the section on cardiac hypertrophy. Hearts with very little or questionable coronary artery narrowing were rejected from this normal, control group. This was done in order to preserve the integrity of the normal, control group which was to serve as a standard of comparison for the entire series. It was deemed preferable thus to include some almost normal hearts in the group with slight coronary artery narrowing rather than to include them in the normal, control group. Of the 244 normal hearts rigidly selected in this way, 55 (23 per cent) showed interarterial coronary anastomoses (table 1). This figure is similar to that reported in an earlier study.123 No correlation was found between the imcidence of coronary artery anastomoses and anatomic variants such as the relative length of right and left circumflex coronary arteries'24 or the presence of a conus artery.'23 Anemia and Coronary Artery Anastomoses In an effort to disclose some clinical feature that might be related to coronary artery anastomoses, the clinical records of the entire group of patients with normal hearts were reviewed. The following factors were evaluated: (1) age, (2) sex, (3) arterial hypertension, (4) severe pulmonary disease, (5) renal disorders, (6) thyroid or other metabolic diseases, (7) chronic or debilitating disease, (8) prolonged cardiovascular collapse, (9) surgical procedures, (10) duration of terminal illness and (11) anemia. Of all these factors a definite correlation was found only between coronary artery anastomoses and anemia. Criterion of Anemia. Determination of anemia was based on hematocrit or hemoglobin levels. Hemoglobin determinations were made by the Sahli calorimetric or Evelyn photoelectric technics; 15.5 Gm. hemoglobin per cent was considered 100 per cent hemoglobin. Hematocrit determinations w-ere made by adequate centrifugation of blood in Wintrobe tubes. When necessary, hematocrit levels were equated with hemoglobin percentages by considering a hematocrit of 45 per cent equivalent to 100 per cent hemoglobin. All patients in the series were classified with SCHLESINGER80 801 respect to anemia as follows: (1) patients with hemoglobin levels of 75 per cent or more were placed in a nommanemic group; (2) those with hemoglobin levels of 70 per cent or less were placed in an anemic group; (3) the remainder with intermediate or indeterminate levels were discarded from the analysis of anemia. This last group consisted of patients with hemoglobin levels between 70 and 75 per cent and those with marked hemoglobini fluctuations. A few cases were also discarded in which hematologic data were not available or were unreliable. Finally all patients in whom aniemia had been present for five days or less before death were also discarded. Experimental studies on the production of intercoronary arterial anastomoses in the dog12m and the pig3 offer some support to this choice of a five (lay limit. The correlation demonstrated between anemia and anastomoses was not altered significantly by moving the criterion for anemia to 75 or 80 per cent hemoglobin, or by enlarging the intermediate, discarded group to include levels of 70 to 80 per cent hemoglobin. Such changes, however, redlueled the number of cases available for analysis of other factors and thereby lowered the statistical validity of the data. Effect of Anemia. Of the 244 patients with normal hearts, 101 were in the nonanemic category; 91 showed anemia as defined above. These 192 cases were used inl a correlation of anemia and anastomoses (table 2 and fig. 2). Coronary artery anastomoses were present iil only nine (9 per cent) of the nonianiemic control group and in 35 (39 per centt) of the anemic group. This difference is highly significant statistically as measured by the chi square test or standard error of difference. *126 Further analysis of the group of 91 cases with anemia was carried out to determine the relative effect of varying degrees of anemia upon anastomosis. No significant difference was found in the incidence of anastomoses in groups with (liffering degrees of anemia. For example, of 52 cases with hemoglobin levels consistently below 60 per cent, 16 (31 per cent) showed anastomoses; of 31 cases with anemia between (60 and 70 per cent * Differences were considered to be statistically significant if they were 2.5 times the standard error of the difference. 802 INTERARTERIAL CORONARY ANASTOMOSES IN HUMAN HEART TABLE 1.-Incidence of Interarterial Coronary A nastomoses inl All Groups of Total Series Normal Weight Anemia Cardiac Hypertrophy HeartswihAatmssHrs Total Number Hearts Number |Per Cent Total Number Dissectible Hearts with Anastomoses Number Per Cent Dissectible A. No Coronary or Valvular Disease Nonanemic ............ Anemic .............. 101 91 9 35 9 39 0 4 70 50 18 14 26 28 0 0 Total* ............... 244 55 23 4 153 38 25 0 B. Valvular Disease Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 Nonanemic 5 1 t 0 32 9 28 1 Anemic................ Total* .13 5 3 4 t 0 0 19 6 2 62 23 32 37 3 31 C. Slight Coronary Artery Narrowing Nonanemic .28 Anemic .19 5 2 18 11 1 0 27 17 3 6 11 39 0 0 Total* .65 11 17 2 57 13 23 0 D. Moderate Coronary Artery Narrowing Nonanemic .19 Anemic............... Total* .... 15 3 5 16 33 1 1 24 19 7 4 29 21 2 0 44 11 25 2 56 14 25 2 E. Marked Coronary Artery Narrowing Nonanemic Anemic ...... Total* . 8 3 5 2 63 t 0 0 33 8 18 7 55 88 3 2 19 12 63 1 47 26 55 5 F. Coronary Artery Occlu.ion Totalt................. 275 262 95 72 G. Cor Pulmonale Totalt............... 1.5 11 73 3 Grand Total of Entire Series .. 1050 480 46 94 * Totals include additional cases with intermediate or indeterminate hemoglobin levels. t Too few in number. t All hemoglobin levels and weight groups combined in single tabulation. hemoglobin, 15 (48 per cent) showed anastomoses. The small group of nine patients without anemia in whom the hearts were normal but nevertheless showed coronary artery anastomoses was reinvestigated in detail. Severe pul- P. M. ZOLL, S. WESSLER AND M. J. SCHLESINGER monary disease was found in six of them. The possible relation of pulmonary disease to coronary artery anastomoses is discussed below. Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 Cardiac Hypertrophy and Coronary Artery Anastomoses Criterion of Hypertrophy. A review of the literature on cardiac hypertrophy did not disclose any simple, generally accepted criterion for this factor.27-"35 The 103 hearts in the entire group which weighed 300 Gm. or more but were otherwise normal and which came from nonanemic patients were subdivided according to weight into four comparable groups of (1) 450 Gm. or more, (2) 449 to 400 Gm., (3) 399 to 350 Gm. and (4) 349 to 300 Gm. TABLE 2.-The Incidence of Interarterial Coronary Anastomoses in Relation to Anemia in Cases without Cardiac Hypertrophy, Valvular or Coronary Disease Anastomoses Total Hemoglobin Levels Absent Nonanemic .......... Anemic ............. Total 92 56 148 803 cent) of the 50 hypertrophied hearts from anemic patients (fig. 2). This incidence does not differ significantly from the 26 per cent in the group of cardiac hypertrophy without anemia or from the 39 per cent in the normal hearts with anemia. Anemia and cardiac hypertrophy individually were associated with an increased incidence of anastomoses in the human coronary artery system, but combination of the two did not show an additive effect with a still higher incidence. Valvular Heart Disease and Coronary Artery Anastomoses Criterion of Valvular Disease. Valvular heart disease without coronary artery disease was TABLE 3.-The Incidence of Interarterial Coronary Anastomoses in Relation to Cardiac Hypertrophy in Cases without Anemia, Valvular or Coronary Disease Heart Weight Total Number of Hearts Present 9 (9%) 35 (39%) 101 91 44 192 (table 3). The frequency of anastomoses was high and constant in each of these small groups above 350 Gm. and dropped sharply in the group below 350 Gm. The entire group of 70 hearts above 350 Gm. may therefore be considered homogeneous. In this way the selection of a heart weight of 350 Gm. or more was supported as the criterion for cardiac hypertrophy throughout this study. Effect of Hypertrophy. Among the 1050 hearts in the series were 153 which were normal in all pathologic respects except for hypertrophy as above defined. In the 70 hearts from this group that came from patients without anemia, the incidence of interarterial coronary anastomoses was 26 per cent (fig. 2). This incidence is significantly higher (2.9 times the standard error of the difference) than the 9 per cent in the normal, control group without hypertrophy. Effect of Anemia Plus Hypertrophy. Coronary artery anastomoses were found in 14 (28 per 450 Gm. or more 449 to 400 Gm ........ 399 to 350 Gm ........ 349 to 300 Gm ........ 25 21 24 33 Hearts with Anastomoses Number Per Cent 6 6 6 24 29 25 6 2 found in 75 hearts. The valves presented definite organic stenosis, insufficiency or a combination of defects. Only hearts with valvular disease thus established at autopsy were included in this category. Effect of Valvular Disease. Coronary artery anastomoses were present in 27 (36 per cent) of all 75 hearts with valvular disease (table 1). This group is heterogeneous in that hypertrophy or anemia or both were present together with the valvular disease in all but five cases. The largest homogeneous subgroup of 32 hearts with valvular disease plus hypertrophy from nonanemic patients (table 1 and fig. 2) showed an incidence of anastomoses of 28 per cent, which is essentially the same as the 26 per cent in hearts with hypertrophy alone. Furthermore, upon addition of anemia to valvular disease and hypertrophy, a similar incidence of anastomoses of 32 per cent in 19 cases was found (table 1 and fig. 2). The three factors of anemia, hypertrophy and valvular disease, 804 INTERARTERIAL CORONARY ANASTO1MOSES IN HUMAN HEART Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 whether taken singly or in any combination, showed similar frequencies of anastomoses, from 26 to 39 per cent, without additive effect (fig. 2). Coronary Heart Disease and Coronary A rtery Anastomoses Criterion of Coronary Artery Narrowing and Occlusion. All the hearts in the series were classified in five categories according to the condition of the coronary arteries: (1) 472 hearts without any coronary artery narrowing, already discussed under the various headings of the normal heart, anemia, hypertrophy and valvular disease; (2) 122 hearts with slight coronary artery narrowing; (3) 100 hearts with moderate coronary artery narrowing; (4) 66 hearts with marked coronary artery narrowing, but without complete occlusions; (5) 275 hearts with complete coronary artery occlusions, old or recent. In 15 additional hearts cor pulmonale was the primary pathologic condition; in them the degree of coronary artery narrowing ranged from none to moderate and was disregarded. * This classification of the degree of coronary artery disease was based on judgments by different observers at the time of dissection and subsequently on re-examination of the roentgenogram of the injected heart. Multiple, independent observations were made frequently on individual hearts by several observers with close agreement. In some of the hearts the internal diameters of the major vessels were measured by a series of graduated probes, so that the degree of reduction in diameter at sites of narrowing was quantitatively determined by comparison with adjacent zones. It was found that estimates of slight narrowing correspon(led to a reduction in diameter approximately of 25 per cent or less; marked narrowing was equivalent to reductions of 75 per cent or more. The term "occlusion" always denoted complete occlusion of a coronary artery; a complete break in the continuity of the lumen was demonstrated in the dissection by the absence of mass or colored intima in such areas. Occlusions were further classified * It Bill be noted that there are no hearts with comblinat ions of coronary anird valvular disease; all such hearts were discarded from the series to simplifthe analysis. as recent or old largely on the basis of their gross appearance. Microscopic estimation of the duration of the occlusive process auras made only in relatively few, selected instances. Effect of Coronary Artery Narrowing. All 288 cases with coronary artery narrowing were classified according to the degree of narrowing, the presence of hypertrophy and anemia (fig. 2). The incidence of anastomoses in the groups with slight and moderate narrowing did not differ significantly from their appropriate comparison groups without narrowing. Every subdivision with marked narrowing, however, showed a significantly high incidence of coronary artery anastomoses: anastomoses were found in 58 per cent of all hearts with marked coronary artery narrowing. Anastomoses were found in only 26 to 39 per cent, however, of hearts without coronary narrowing to which were added the factors of anemia, hypertrophy and valvular disease, singly or ill combination. Again, anemia and hypertrophy were not effective in increasing coronary artery anastomoses in the presence of marked coronary artery narrowing, just as combinations of them were not additive in the absence of marked coronary narrowing. Effect of Recent Coronary Artery Occlufsions. Among the 275 hearts with complete occlusions of the coronary arteries were 121 with one or more points of recent occlusion (table 4 and fig. 2). Interarterial coronary anastomoses were demonstrated in 108 (89 per cent) of these 121 hearts. Anemia and cardiac hypertrophy were not important in the development of anastomoses in this group. More than one recent occlusion was present in 27 of these 121 hearts, so that the total number of recent occlusions was 1 63 (table 4). Analysis of the local anatomic relation of the anastomoses to each of these 163 indiviidual recent occlusions throws light on the mechanism of development of interarterial anastomoses. Anastomotic filling of those branches of the coronary arteries directly peripheral to recent occlusions was demonstrated only in 110 of the 163 instances (68 per cut). Many of the recent coronary artery occlusions were clearly superimposed on the site of previous marked coronary artery narrowing. \Iarked narrowing P. M. ZOLL, S. WESSLER AND M. J. SCHLESINGER 805 Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 heart. Anastomoses were demonstrated in the coronary arteries directly distal to all but 13 of these 349 sites of old occlusion (96 per cent). Even when the instances of old occlusions found in hearts with both recent and old occlusions were added, the incidence of anastomoses directly distal to old occlusions remained high at 91 per cent (table 4). The special means of demonstrating anastomoses in hearts with coronary artery occlusions by finding injection mass of the same color on both sides of an occlusion was the sole evidence of anastomoses in only 12 (8 per cent) of these 154 hearts. It was, therefore, not a significant factor in the high incidence of anastomoses in alone without occlusion has already been shown to be associated with an incidence of 58 per cent of anastomoses. Hence it appears likely that the localized anastomoses found distal to recent occlusions were induced in most instances by previous coronary artery narrowing. Effect of Old Coronary Artery Occlusions. The group of 154 hearts with at least one old occlusion and no recent occlusions in the coronary artery tree comprised 15 per cent of the total series of 1050 hearts. Intercoronary anastomoses were present in every one of these 154 hearts. In the larger group of 236 hearts with old coronary artery occlusions (including 82 with recent occlusions as well), all but three TABLE 4.-The Incidence of Interatterial Coronary Anastomoses in Hearts with Coronary Artery Occlusions Anastomoses Distal to Occlusion Anastomoses in Heart Group Number of Hearts Number Hearts with Recent Occlusions Only 29 39 ---1 ---l Number of Occlusions Per Cent Number Dissectible 74 2 50 21 113 Number 29 Per Cent 58 ---1 Recent Occlusions 82 Hearts with Recent Plus Old Occlusions 79 96 81 72 Old Occlusions 237 I- Hearts with Old Occlusions Only _________________ - - _________________________________- Total of Hearts with Occlusions 154 - - 275 - - . per 84 154 100 49 349 336 96 262 95 72 749 644 86 1 cent) showed interarterial coronary anastomoses (table 4). In these three exceptional instances existing anastomotic pathways may have been blocked by the terminal acute occlusion. This almost invariable occurrence of anastomoses in hearts with old coronary artery occlusions is striking, whether it be compared with the incidence of anastomoses in normal hearts (9 per cent), in hypertrophied hearts (26 per cent), in hearts from patients with anemia (39 per cent), in hearts with marked coronary artery narrowing (55 per cent), or in hearts with recent coronary artery occlusions (74 per cent). In the 154 hearts with old coronary artery occlusions there were 349 separate points of old occlusion, or over two old occlusions per (99 198 __ I- hearts with old occlusions. Similarly, anemia was infrequent (13 per cent) in the 154 patients with old coronary artery occlusions and, therefore, not significant in producing all the anastomoses. Cor Pulmonale and Coronary Artery Anastomoses Criterion of Cor Pulmonale. Uncomplicated cor pulmonale was present in 15 of the 1050 hearts studied. In these 15 instances additional factors that might produce anastomoses, such as anemia, valvular disease, marked coronary artery narrowing or occlusion, were absent. Relative right ventricular hypertrophy, usually secondary to pulmonary fibrosis, was taken as the distinctive criterion of cor pulmonale, INTERARTERIAL CORONARY ANASTOMOSES IN HUMAN HEART 806 even with a normal total heart weight. All these patients had suffered from severe, chronic pulmonary disease. In these 15 hearts the right ventricle measured 5 mm. or more in thickness or was greatly enlarged and dilated. The roentgenogram of the unrolled heart showed clearly the relative size and density of the soft tissue shadows of the right and left ventricles. In this way it provided valuable confirmatory data to the observations made at the autopsy table concerning right ventricular hypertrophy. Effect of Cor Pulmonale. Anastomos3s were present in 11 (73 per cent) of these 15 hearts Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 TABLE 5.-Incidence of Etiologic Factors in 480 Hear'ts with Interarterial Coronary Anastoinoses Incidence of Total Incidence Factor as of Factor, Alone or Combined Single Condition \Iarked Coronary Artery Disease (complete occlusion or marked narrowing) Slight to Moderate Coronary Artery Narrowing Cardiac Hypertrophy Antemortem Anemia Valvular Heart Disease Cor Pulmonale Grossly Normal Hearts Number Per Cent Number Per Cent 300 63 89 19 49 10 8 2 180 38 24 6 2 18 35 1 0 4 7 0 0 9 2 114 27 11 (table 1 and fig. 2). Iespite the larg4 e standard error in this small group, there is a si tatistically significant increase in the incidence of anastomoses over the 9 per cent seen in nor mal hearts without anemia, 26 per cent in hypgertrophied hearts and 11 to 29 per cent in h(earts with slight to moderate coronary artery ]narrowing. Cor pulmonale thus is associated fith an increase in interarterial coronary anm astomoses. Relative Frequency of Factors Inducirig A nasto- lated in two ways: (1) their total incidence, whether they occurred alone or in (.oml)ina.tiollt, and (2) their individual incidence in hearts with a single, uncomplicated factor. Marked coronary artery disease, cardiac hypertrophy and aniemia were the most frequent etiologic factors. The relative frequency of these factors inl hearts with anastomoses reflects not only their etiologic effect in inducing anastomoses, but also their frequency in this particular series of 1050 hearts. The individual efficacy of these factors in inducing anastomoses has already been considered in the presentation of the percentage of hearts with anastomoses inl the various groups. Both tabulations indicate the preponderant role of marked coronary artery disease in inducing interarterial (coronoar y anastomoses and the less marked effects of cardiac hypertrophy and anemia. Grossly Dissectible AInastomoses Grossly dissectible communicating pathways were found in 94 hearts (20 per cent) of the 480 hearts with anastomoses (table 1); 72 of them occurred in the 275 hearts with coronary artery occlusions. More particularly, they were found in 70 (30 per cent) of the 236 hearts 'with old coronary artery occltusions and in only two (5 per cent) of the 39 hearts with recent occlusions alone. This latter incidence is lower than the 9 per cent found in the 66 hearts with marked coronary artery narrowing and the 20 per cent found in the 15 hearts with cor pulmonale. Grossly dissectible anastomoses were of a similar low frequency in the remainder of the series, occurring in four (4 per cent) of the 100 hearts with moderate coronary artery narrowing, in two (2 per cent) of the 122 hearts with slight coronary artery narrowing, in three (4 per cent) of the 75 hearts with valvular disease and in four (1 per cent) of the 397 remaining hearts in the series without coronary or valvular disease (table 1). moses All 480 hearts with interarterial coronary anastomoses were analyzed with resi )ect to the relative frequency of the seven grroups into which these cases fell (table 5). Siinee many hearts presented combinations of eti ologic factors, the incidence of these factors was tabu- Myocardial Histopathology Multiple, individually labeled microscopic sections from precisely localized areas of the myocardium were studied in 509 of the 1050 hearts in this series. These 509 hearts are not consecutive and do not form a representative P. M. ZOLL, S. WESSLER AND Ml. J. SCHLESINGER Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 sample of the entire group of injected hearts. Nevertheless, certain broad relations between anastomoses and histopathologic changes in the myocardium could be determined. M\icroscopic abnormalities noted included myocardial degeneration, necrosis, inflammation and fibrosis Although these lesions varied considerably, qualitatively and quantitatively, in different hearts, they were grouped together and are presented in summary form here insofar as they are pertinent to the problem of coronary artery anastomoses. As might be expected, few of the grossly normal hearts without hypertrophy, valvular or coronary disease showed histopathologic abnormalities. Nevertheless, like anastomosis, histopathology was significantly more frequent in the anemic hearts than in the nonanemic hearts of this group. In general, a parallel increasing association was found between hypertrophy, valvular and coronary lesions and the presence of histopathology and anastomosis. Of 319 hearts with histopathology, however, interarterial anastomoses were absent in 98. Conversely, it was of particular interest to find seven hearts without any microscopic myocardial pathology among the 114 with one or more complete old occlusions in which multiple microscopic sections were studied. In each of these instances there were found well-developed, extensive interarterial coronary anastomoses which may have protected the myocardium from the effects of arterial occlusion. Time Necessary for Development of Anastomoses following Sudden Occlusion of a Coronary Artery The relatively low incidence (58 per cent) of coronary artery anastomoses directly distal to recent coronary artery occlusions in comparison with the 96 per cent incidence of anastomoses directly distal to old occlusions suggests that the time interval between the occurrence of the recent vascular occlusion and death was insufficient in some instances for the development of compensatory anastomoses. Valuable information concerning the time necessary for the development of collateral circulation might be obtained by relating the age of a recent complete occlusion to the pres- 807 ence or absence of an anastomosis associated with it. In such a study it would be necessary to know the age and rate of development of the recent occlusion and how much narrowing was present in the vessel before the recent occlusion occurred. Gross observations are too crude to establish these facts. Numerous microscopic sections, preferably serial sections, through each zone of recent occlusion would provide such reliable information, but they were not available in the present series of hearts. An attempt was made to fix the time of the terminal fresh coronary artery occlusion from the clinical record. As might be expected, however, it was unsuccessful because correlations between maj or episodes of cardiac pain and coronary artery occlusion are unreliable. Such episodes of pain, even associated with myocardial infarction, may occur without concomitant fresh occlusion in the coronary artery tree and fresh coronary artery occlusions, furthermore, may occur without clinical mani- festations?136 In this series clearly defined, well-timed clinical episodes appeared in only 73 of the 121 patients dying with recent coronary artery occlusions. Furthermore, most of the 73 also had multiple, prolonged attacks of cardiac pain. Consequently, we were not able in this study to determine the rate of development of coronary interarterial anastomoses in man. DISCUSSION A definite, statistically significant relationship has been found between a number of factors and the development of human intercoronary arterial anastomoses. In the absence of these stimulating factors less than 10 per cent of human hearts presented anastomoses. It is also suggested that the incidence of anastomoses in the normal human heart would be still lower if all etiologic factors were known. Thus it is becoming more firmly established that, functionally at least, the coronary arteries in the human heart are end-arteries. It is also clear that' under a variety of stimuli the flexible coronary arterial system readily develops a functioning internal anastomotic collateral circulation to compensate for a 808 INTERARTERIAL CORONARY ANASTOMOSES IN HUMAN HEART Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 relative insufficiency of blood. Relative insufficiency may originate in the blood itself (aneemia), in the aeration of the blood (pulmonary disease), in the myocardium (hypertrophy), in the endocardium (valvular disease) or in the coronary arteries themselves (coronary atherosclerosis). The previous observation that marked narrowing and complete occlusion of the coronary arteries are most important factors in the production of interarterial coronary anastomoses has been firmly established. Their effectiveness is seen in the high incidence of anastomoses and in the frequency of large, grossly dissectible interarterial pathways under these conditions, as well as the great frequency of marked coronary artery disease in hearts with anastomoses. Slight or moderate coronary artery narrowing is much less effective. The experimental production of intercoronary anastomoses following narrowing and occlusion of the coronary arteries by surgical ligation in pigs and dogs' 125 led to similar conclusions concerning these factors. Furthermore, the functional significance of such intercoronary anastomoses was demonstrated experimentally in the prevention of death and the reduction of myocardial damage following complete coronary occlusion. The detailed histopathologic studies of human hearts with marked narrowing and complete occlusion of the coronary arteries indicate the similar functional importance of such anastomoses in the human heart. The increasing frequency of hearts with myocardial damage with increasing degrees of coronary artery obstruction was anticipated. The few hearts, however, with complete old coronary artery occlusions but without any histopathology in the multiple microscopic sections illustrate pointedly the protective effect of the anastomoses. In these hearts, the development of interarterial coronary anastomoses during gradual coronary artery narrowing must have been so adequate as to prevent organic muscle damage even with final complete occlusion of an artery. These observations find their occasional clinical corollary in the absence of cardiac pain or other symptoms after a proved acute coronary artery occlusion.136 In these instances, at least, the stimulus for interarterial coronary anastomoses did not lead also to irreversible myocardlial lamage. Besides marked coroniary artery disease, the factors of hypertrophy and of anemia have each been demonstrated to be associated individually with an increased incidence of intercoronary anastomoses. The addition of cardiac hypertrophy or anemia to other factors did not materially affect the incidence of anastomoses. The absence of a summative effect of several factors was unexpected. It is possible that some underlying fundamental mechanism such as cardiac anoxia is common to all the factors concerned. This mechanism may be activated by any single factor, but is not activated to a greater degree or does not produce a greater effect when stimulated by several factors in combination. Because of the small numbers of hearts available in the various subgroups, a small additive effect might not have been statistically apparent in the present analysis. Furthermore, in any statistical analysis based on the "null hypothesis," failure to demonstrate "significant" differences does not prove their absence."37 The data are inadequate to establish an association of valvular disease per se with increased intercoronary anastomoses, since most of the hearts with valvular disease also showed cardiac hypertrophy. Valvular disease plus cardiac hypertrophy showed the same incidence of ansastomoses as hypertrophy alone. Nevertheless, valvular disease alone may be a factor inducing interarterial coronary anastomoses; like anemia and coronary artery disease, an additive effect upon anastomoses may not be demonstrable when it is combined with cardiac hypertrophy. In half of the 27 hearts with chronic valvular disease in which coronlary artery anastomoses were present, there was also marked right ventricular hypertrophy, indistinguishable from that resulting from primary pulmonary disease. In these hearts the mitral valve was usually involved. The development of anastomoses in this group may result from: (1) valvular disease itself, if it is indeed a factor capable of stimulating anastomoses; (2) cardiac anoxia on the basis of diminished oxygenation of the blood from pulmonary insufficiency; or (3) cardiac P. M. ZOLL, S. WESSLER AND AM. J. SCHLESINGER Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 hypertrophy with increased need for blood by the enlarged myocardium. In like manner, the increased intercoronary arterial anastomoses found in cor pulmonale due to pulmonary disease may result from right ventricular hypertrophy or cardiac anoxia secondary to pulmonary insufficiency. These two mechanisms may also be involved in the six "normal" hearts with coronary artery anastomoses but without cardiac hypertrophy which came from patients with marked pulmonary disease, but without anemia. If the mechanisms were more clearly established by which primary pulmonary disease stimulated the development of coronary artery anastomoses, the number of "normal" hearts with unaccountable anastomoses could be reduced to three instead of nine in the series of 1050 hearts. Furthermore, the finding of unexplained myocardial histopathology in a few grossly normal hearts from patients without anemia suggests that the etiologic factors herein considered for the development of coronary artery anastomoses are not complete or definitive. That anemia might stimulate the development of interarterial coronary anastomoses in otherwise normal hearts might have been anticipated. In 1944, Amadeol"1 suggested that the infrequency of clinical coronary disease among the poor rural population of Puerto Rico resulted from the effect of chronic anemia upon the development of collateral anastomotic vessels of the coronary arterial system. In four hearts from patients with anemia, Scott'39 found a marked increase in interarterial coronary anastomoses, as demonstrated by injection of the coronary arteries with a barium sulfate suspension in liquid latex."21 In a study of the coronary arteries by Dock'sll method of perfusion with kerosene, we also found that the rates of flow were greatly increased in four hearts from patients with anemia.'40 In the present series the incidence of interarterial coronary anastomoses in "normal human hearts" was reduced from 22 per cent to 9 per cent when the patients with anemia were eliminated from the group of grossly normal hearts. In comparison with this low figure, the incidence of anastomoses in other groups of hearts assumed statistical significance. 809 Insight into the basic mechanism stimulating intercoronary anastomoses is afforded by the demonstration of the roles of anemia and of cor pulmonale. Relative reduction in oxygen supply to the heart (anoxia) is obviously common to these two conditions. Anoxia is associated also with all the other factors related to increased anastomoses. Accumulation of metabolites, lack of substances other than oxygen or mechanical factors such as differential pressure gradients in different parts of the coronary system consequent to anoxia, may be other mechanisms that are effective in stimulating anastomoses. The precise site at which the anoxia is the critical stimulus for the development of anastomoses is also a matter of speculation. It may be in the coronary vessels themselves. The relative insufficiency of oxygen results either from reduced blood flow in the presence of coronary artery narrowing and occlusion or from increased demand for blood by the myocardium in cardiac hypertrophy and valvular disease. In anemia and cor pulmonale there may be no reduction in the relative volume of blood flow through the coronary arteries, but there is a reduction in the oxygen content of the blood, resulting from diminished hemoglobin and oxygen capacity in anemia and from diminished oxygen saturation in cor pulmonale. Intercoronary arterial anastomoses have been found in some instances without myocardial histopathologic changes. Cardiac anoxia of various origins may thus stimulate interarterial coronary anastomoses without necessarily progressing to the stage of irreversible tissue damage. The reverse observations of hearts with myocardial damage but without anastomoses were also made in all categories. Many conditions may thus damage heart muscle irreversibly without necessarily stimulating the development of anastomoses. Nevertheless, in the control group of hearts with no intrinsic gross pathology, myocardial damage was infrequent in comparison with the histopathology found in hearts with hypertrophy, valvular disease and coronary artery disease. These normal microscopic findings add validity to the criteria that were used for the selection of the "normal, control group." 810 INTERARTERIAL CORONARY ANASTOMOSES IN HUMAN HEART In this control group there were positive correlations between myocardial damage, anemia and anastomosis. Deleterious effects of anemia on heart muscle have been reported by others."4' Our observations indicate that anemia per se may or may not produce some myocardial damage and that this potentiality is intimately related to its capacity to stimulate anastomoses. Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 Anemia differs from most of the other conditions associated with the development of intercoronary anastomoses in that it may be transient and can be voluntarily produced and controlled in man and in laboratory animals. Any desired level of anemia may be reached and maintained by repeated venesections at appropriate intervals. Furthermore, a critical degree of anemia may be found which will stimulate the development of anastomoses and yet will not result in significant myocardial lesions. Anemia might therefore be used as a readily controllable investigative tool in studies pertaining to intercoronary arterial anastomoses. The data available in the present study do not indicate clearly the duration and severity of the anemia that is necessary to stimulate anastomoses. Anemia of less than six days' duration appears to be without such effect and marked degrees of anemia seem no more effective than moderate anemia. The question also remains unanswered whether the newly developed interarterial coronary anastomoses are permanent or whether they disappear if the stimulus provoking them is withdrawn. Amadeol'38 came to similar conclusions about the degree of aniemia. He also assumed that initercoronary aniastomoses produced during the period of ineoniatal aniemia persist throughout life. Anemia may conceivably have some therapeutic application in the treatment of coronary artery disease. The presence of anemia for six months before death in five of the seven patients whose heart muscle w as entirely normal histologically despite old complete coronary artery occlusions suggests that pre-existing anastomoses produced by the anemia may have helped protect the muscle completely against the effects of the severe coronary artery disease. The old occlusions in these patients did not produce clinical manifestations, however, so that their temporal relation with the anemia could not be determined. It would theoretically be desirable to produce intercoronary anastomoses before the onset of coronary artery narrowing or occlusion; coronary atherosclerosis would be much less disastrous if the coronary arteries were not originally functional endarteries. The possibly favorable effects of anemia on the coronary circulation in experimental animals are now under investigation in this laboratory. It is a plausible assumption that the condition (anoxia) which stimulates the formation of coronary artery anastomoses also produces the microscopic myocar(lial abnormalities that are frequently found in the same hearts. On this basis, in terarterial corol1ary aiiastomosis may also be regarde(l as an organic abnormality, especially since it is so infrequent in normal hearts. The functional and organic abnormalities of anemia, chronic pulmonary disease, cardiac hypertrophy, valv ular heart disease and coronary artery disease then may be regarded as etiologic factors in the genesis of physiologic states of relative myocardial anoxia, and thereby, of the two anatomic abnormalities of anastomoses and myocardial histopathology. The latter two abnormalities tend to appear together inl the same heart but are not necessarily directly interrelated. The development of the organic abiiormality of coronary artery anastomosis is a compensatory response to these etiologic conditions. In some instances it is sufficiently protective to forestall or diminish the development of the less desirable organic abnormality of myocardial damage. These assumptions are supported by our data not only inl the large group with coronary artery disease where they might be considered axiomatic, but in every category discussed above. CONCLUSIONS In a series of 1050 human hearts studied by a method of injection plus dissection of the coronary arteries, increased initerarterial coronary anastomoses were found in coronary heart disease, cor pulmonale, antemortem aniemia, cardiac hypertrophy andlvalvular heart disease. In 101 grossly normal hearts from nonanemic P. MI. ZOLL, S. WE1SSLER AND Ml. J. SCHLESINGER patients the incidence of interarterial coronary anastomoses was 9 per cent. In 89 grossly normal hearts from anemic patients the incidence of anastomoses was 39 per cent. Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 In the remaining abnormal hearts the incidenice of interarterial coronary anastomoses was 89 to 100 per cent with coronary artery occlusion, 11 to 63 per cent with coronary artery narrowing, 73 per cent with COr pulmonale, 26 per cent with cardiac hypertrophy and 28 per cent with valvular disease. The factor of relative cardiac anoxia that is present in all these conditions appears to be a common underlying stimulus for the development of interarterial coronary anastomoses. REF1IRENCES 1BLUMGART, H. L., ZOLL, P. _M., AND WESSLER, S.: Angina pectoris, a clinical p)athologic study of 177 cases. Ti. A. Am. Physicians 63: 262, 1950. , SCHLESINGER, M. J., AND DAVIS D.: Studies of the relation of the clinical manifestations of angina pectoris, coronary thrombosis and mvocarldial infarction to the pathologic findings. Am. Heart J. 19: 1, 1940. ZYOLL, P. M., FREEDBERG, A. S., AND GILLIGAN, 1). R.: The experimental production of intercoronary arterial anastomoses and their functional significance. Circulation 1: 10, 1950. 4 LANCISI, J. M\1.: IDe motu cordis et aneurysmatibus (opus p)osthumum). Lugduni Batavorum, 1740. D1)E1 SE NAC, J. B.: Traite(de la structure du coeur, action et de ses maladies. Paris, 1749. (Ie 6 CIZUvl:ILHI1.R, J.: Anatomie dlescril)tive, vol. 1. Paris, 1834. P. 32. 7HE NLE, J.: Handbuch (let systematischen Anatomie des MIenschen. 3: 1 Abt. Geffisslehre, 86-87, 1 Aufl. Braunschweig, 1868. 8 SAPPI.Y, P. C.: Traite(d'anatomie descriptive, ed. 3. Paris, V. A. Delahaye & Cie, 1876. LANGER, L.: D)ie Foramina Thebesii im Herzen (les i\lensehen. Sitzungsb. d. k. Akad. d. Wissensch. 82: 25, III. Abt. 1880. son C. E.: Studies on the vasa vasorum. Am. J. Path. 2: 567, 1926. HUDSON, C. LJ. MORITZ, A. R., AND WEARN, IJ. T.: The extracardiac anastomoses of the c(ttOnayi arteries. J. Exper. Me(l. 56: 919, 1932. 12 ()CHRIEIN, MI.: Der Coronarkreislauf. Berlin, Julius Springer, 1932. 13 BF1K, C. S., TICHY, V. L., AND M\ORITZ, A. R.: lProduction of a collateral (irculation to the heart. Proc. Soc. Exper. Biol. & Med. 32: 759, 1935. M4MAUTZ, F. R., AND BECK, C. S.: The augmenta- W\OODRUFF, 811 tion of collateral coronary circulation by operation. J. Thoracic Surg. 7: 113, 1937. 15 THOREL, C.: Pathologie der Kreislauforgane. VI. Pathologie der Koronargefaisse. In Lubarsch and Ostertag's Ergebn. d. allg. Path. u. path. Anat. 9: 658, 1903. 16REDWITZ, E. F. VON.: Der Einfluss der Erkrankungen der Koronararterien auf (lie Herzmuskulatur mit besonderer Berucksichtigung det chronischen Aortitis. Virchows Arch. f. path. Anat. 197: 433, 1909. 17 1\LoRITZ, A. R., HUDSON, C. L., AND ORGAIN, E. S.: Augmentation of the extracardiac anastomoses of the coronary arteries through pericardial adhesions. J. Exper. M1ed. 56: 927, 1932. ROBERTSON, H. F.: The reestablishment of cardiac circulation during progressive coronary occlusion. Am. Heart. J. 10: 533, 1934. 9 BECK, C. S.: The develol)ment of a new 1)lood supplyl to the heart by operation. Ann. Surg. 102: 801, 1935. 20 , AND TICHY, V. L.: The production of a collateral circulation to the heart. I. An experimental study. Am. Heart J. 10: 849, 1935. 21 O'SHAUGHNESSY, L.: An experimental method of providing a collateral circulation to the heaIt. Brit. J. SuIg. 23: 665, 1936. 22 ARIEFF, M. J., WITUSCHINSKY, W. J., AND RADINERSON, A. B.: iTber den kollateralen Blutkireislauf im Herzen bei Verschluss der Ai.tndung det Kranzarterien (syphilitische Mesoaortitis) uncl bei adhisiver XIediastinoperikarditis. Ztschr. f. klin. Med. 130: 153, 1936. 23 LEzIUS, A.: D)ie kuinstliche Blutversorgung des Herzmuskels. Arch. f. klin. Chir. 189: 342, 1937. 24 STEPHENS, H. B., AND BENTEEN, H.: IMuscle grafts: in the surgery of the heart and lungs. California & West. MAed. 49: 366, 1938. 25 BECK, C. S.: Natural ancl artificial circulation in the heart. In Blood, Heart andl Circulation, AAAS No. 13. Lancaster, Pa., Science Press, 1940, Pp. 133-138. 26 BURCHELL, H. B.: Adjustments in coronary circulation after experimental coronaryv occlusion with pa'ticular reference to vascularization of pericardial adhesions. Arch. Int. -Med. 65: 240, 1940. 27 : Adjustments in the coronary circulation after exl)erimental coronaryI oCClusion. In Blood, Heart abnid Circulation, AAAS No. 13. Lancaster, 1Pa., Science Press, 1940. Pp. 139-144. 28 HEINBECKER, J)., AND BARTON, W. A.: Operation for development of collateral circulation to the heart. J. Thoracic Surg. 9: 431, 1940. 29 ROBERTSON, H. F.: The physiology, pathology and clinical significance of experimental coronarv sinus obstruction. Surgery 9: 1, 1941. 812 INTERARTERIAL CORONARY ANASTOMOSES IN HUMAN HEART Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 30 THOMPSON, S. A., AND RAISBECK, M. J.: Cardiopericardiopexy: The surgical treatment of coronary arterial disease by the establishment of adhesive pericarditis. Ann. Int. Med. 16: 495, 1942. 31 BECK, C. S.: Principles underlying the operative approach to the treatment of myocardial ischemia. Ann. Surg. 118: 788, 1943. 32 VIEUSSENS, R.: Nouvelles decouvertes sur le coeur. Toulouse, 1706. 33 THEBESIUS, A. C.: Disputatio medica de circulo sanguinis in corde. Lugdunim Batavorum, 1708. 34 LOWER, R.: Tractatus de corde, item de motu et colore sanguinis, et chyli in eum transitu. Amstelaedami, 1669. 35 VERHEYEN, P.: Corporis humani anatomiae. Coloniae, 1712. 36 WINSLOW, J. B.: Exposition anatomique de la structure du corps humain. Paris, 1732. P. 594. 37 HALLER, A. VON: First Lines of Physiology, vol. 1 (English translation). Edinburgh, 1786, P. 75. 38 ABERNETHY, J.: Observations on the foramina thebesii of the heart. Phil. Tr. Royal Soc. London 103, 1798. 31 BOCHDALEK, V. A.: Anatomische BeitrAge. Zur Anatomie des menschlichen Herzens. Arch. f. Anat., Physiol. u. wissensch. M~ed. 302, 1868. 40 GAD, J.: Arch. f. Anat. u. Physiol. 2: 380, 1886. 41 MICHAELIS, M.: Uber einige Ergebnisse bei Ligatur der Kranzarterien des Herzens. Ztschr. f. klin. Med. 24: 270, 1894. 42 MAGRATH, G. B., AND KENNEDY, H.: On the relation of the volume of the coronary circulation to the frequency and force of the ventricular contraction in the isolated heart of the cat. J. Exper. Med. 2: 13, 1897. 43 PRATT, F. H.: The nutrition of the heart through the vessels of Thebesius and the coronary veins. Am. J. Physiol. 1: 86, 1898. 44 PORTER, W. T.: A new method for the study of the isolated mammalian heart. Am. J. Physiol. 1: 511, 1898. 45 BAUMGARTEN, W.: Infarction in the heart. Am. J. Physiol. 2: 243, 1899. 46 MINOT, C. S.: On a hitherto unrecognized form of blood circulation without capillaries in the organs of vertebrata. Proc. Boston Soc. Nat. Hist. 29: 185, 1900. 47LEWis, F. T.: The question of sinusoids. Anat. Anz. 25: 261, 1904. 48 MORAWITZ, P., AND ZAHN, A.: tMber den Koronarkreislauf am Herzen in Situ. Zentrabl. f. Physiol. 26: 465, 1912. 4 EVANS, C. L., AND STARLING, E. H.: The part played by the lungs in the oxidative processes of the body. J. Physiol. 46: 413, 1913. 50 MARKWALDER, J., AND STARLING, E. H.: A note on some factors which determine the blood flow through the coronary circulation. J. Physiol. 47: 275, 1913. 51 CRAINICIANU, A.: Anatomische Studien fiber die Coronararterien und experimentelle Untersuchungen fiber ihre Durchgangigkeit. Virchows Arch. f. path. Anat. 238: 1, 1922. 52 PRATT, F. H.: Factors, actual and possible, in cardiac nutrition. Boston i\I. & S. J. 190: 304, 1924. 52 GRANT, R. T.: An unusual anomaly of the coronary vessels in the malformed heart of a child. Heart 13: 273, 1926. 54 DANIE'LOPOLU, D.: L'Angine de Poitrine et l'Angine Abdominale. Paris, Masson & Cie, 1927. Pp. 443. 5 WEARN, J. T.: The role of the thebesian vessels in the circulation of the heart. J. Exper. Med. 47: 293, 1928. 56 KRETZ, I.: Uber die Bedeutung der venae minimae Thebesii fur die Blutversorgung des Herzmuskels. Virchows Arch. f. path. Anat. 266: 647, 1928. 5 ANREP, G. V., BLALOCK, A., AND HAMMOUDA, M.: Distribution of blood in the coronary vessels. J. Physiol. 67: 87, 1929. 58 WIGGERS, C. J.: Alterations in the left ventricular beat produced by drugs limited to the right ventricle. Am. J. Physiol. 90: 558, 1929. 59 : The absorption of drugs from the right ventricular cavity. J. Pharmacol. & Exper. Therap. 39: 209, 1930. 60 GRANT, R. T., AND VIKO, L. E.: Observations on the anatomy of the thebesian vessels of the heart. Heart 15: 103, 1929. 61 LEARY, T., AND WEARN, J. T.: Two cases of complete occlusion of both coronary orifices. Am. Heart J. 5: 412, 1930. 62 L6HNER, L.: Uber die physiologische Bedeutung der Vasa Thebesii. Arch. f. d. ges. Physiol. 228: 457, 1931. 63 BELLET, S., AND GOULEY, B. A.: Congenital heart disease with multiple anomalies. Am. J. 1\I. SC. 183: 458, 1932. 64 STELLA, G.: The part played by the thebesian vessels in the blood supply to the heart. J. Physiol. 73: 36, 1931. 65 : Further observations on the function of the thebesian vessels in the mammalian heart. J. Physiol. 75: 181, 1932. 66 BELLET, S., GOULEY, B. A., AND MCMUILLAN, T. M.: Nourishment of the myocardium through thebesian vessels in a heart in which the large coronary arteries and veins were destroyed by tuberculous myocarditis. Arch. Int. Med. 51: 112, 1933. 67 WEARN, J. T., METTIER, S. R., KLUMPP, T. G., AND ZSCHIESCHE, L. J.: The nature of the vascular communications between the coronary arteries and the chambers of the heart. Am. Heart J. 9: 143,1 933. 68 BOHNING, A., JOCHIM, K., AND KATZ, L. N.: The thebesian vessels as a source of nourish- P. M\. ZOLL, S. WESSLER AND M. J. SCHLESINGElR8t Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 ment for the mnyocardium. Am. J. P1hysio1. 106: 183, 1933. 69 BENNETT, H. S.: The IeveloI)ment of the blood supp)ly to the heart in the embryo pig. Am. J. Anat. 60: 27, 1936. 70 KATZ, L. N., JOCHIMi, K., AND BOHNING, A.: The effect of the extravascular support of the ventricles on the flow in the coronary vessels. Am. J. Phlysiol. 122: 236, 1938. 71 EcMSTEIN, it. W., ROBERTS, J. T., GREGG, 1). L., AND \\EARN, J. T.: Observations on the riole of the thebesian veins and luminal vessels in the right ventricle. Am. J. Phvsiol. 132: 648, 1941. 72 GREGG, 1). E., SHIPLEY, R. E,, AND BIDDER, T. G.: The anterior (ar(liac veins. Their functionail iml)ortance in the venous (Irainage of the heart. Am. J. Phy-siol. 139: 732, 1943. 73 LENDRUM, B., KONDO, B., AND KATZ, L. N.: The r(le of thebesian d(rainage in the (l-nami(s of coronary flow. Am. J. Plhysiol. 143: 243, 1945. 7 ROBERTS, J. T.: The role of the small vessels and nerves of tile heart in lheart failure, coronary artery throml)osis an(l cardiac pain. M. Ann. I)istrict of Columbial 14: 483, 1945. 76 AND SPENCER, F. D.: Tile accessory miechanism for (Irainage ad(l nourishlmnett of the Ivocaro((lium )bv the thel)esian 01 arterio-luminal vessels, especially in the left ventricle. Proc. Am. Fed. Clin. Research 3: 23, 1947. 76 PRINZ\I1,TrAL, M\., SIMIKIN, 3., AND BERGMAN, H. C.: Studies on the coronary circulation. II. The collateral circulation of the normal humican hetart by (oronaryv )eifusion with radioactive eivthlov(tes and glass spheres. Am. Heait J. 33: 420, 1947. 77 RUySCH, F.: Thesauius anatomnicus quartus. Amstelale(lalnii, 1704. 78 CALDANI, L. i\I. A., AND CALDANI, F.: Icones Anatomicae. Venetiis, 1810. 79 PANUMI, P). L.: Experimentelle Beitriige zur Lehire von der Emnbolie. Virchows Arch. f. path. Anat. 25: 308, 1862. 80 LANNELONGUE, 0.: Circulation veineuse des parois auriculaires (lu coeur. These de Paris, 1867; Recherches sur Ia circulation des p)alois duI coeur. Arch. de physiol. norm. et path. 1: 22, 1868. 81 HYRTL, J.: I)ie Corriosionis-Aniatoimiie uncl ilhre 11Eigel)nisse. XVIII Tafeln. AWien, 1873. PI). 253. 82 COHNHEI11, J., AND VON SCHULTHES.S RECHI;:,RG, A.: flber (lie Folgen (ler Kranzarterienversehliessung ffir das Herz. Virchows Arch. f. path. Anat. 85: 503, 1881. 83WEIST, S.: The anastomiosis of the coronary. arteries. LaInet 1: 94.15, 1883. 84 Clses of compllete ol)literattioin of one (0oonary. artery, with remarks upon the corOnal \ 813 circulation. Ti. Path. Soc. London 35: 110, 1884. 85 NIELSEN, K. A.: Iber spontane Ruptur (les Herzen durch Verschluss (ler Koronararterie. Beitr. z. path. Anat. u. z. klin. Med. Leipzig, 1888. 86 BAUNM, H.: Die Arterienanastomosenl des Hundes und (lie Bedeutung del Collateralen fir (len thierischen Organismus. D)eutsche Zts(clhr.. f. Tliiermed. 14: 273, 1889. 87 D)RAGNEFF, S.: Recl1erclhes anIatomniquCs satl les arte'res coronaires (hu coeur clez 'lionmme. The'se, No. 26, Nancy, 1897. P1). 38. 88 ZIMIMIFIRL, U.: Richlerche tanatonmo-com)arate sui vasi car(liaci (legli axnimiiiali (lomcstici. PIrmla, 1900. B B.ANCHI, A.: \Mlorfologia (lella arteriae (co1)natiae cor(lis. Arch. ittal. (di itiat. e (di Cebl)riol. 3: 87, 1904. 90 MERKML H.: Zur Kenitnis der Kr.an.zarterlieni (les menSchlichen HeIrzen. Verb undl 1d. ( ieutsch. aIathol. Ciesellschl. 10: 127, 1906. 91 JAMIN, F., AND M\IERKEL, H.: Die Koroilnrarterien (les nienschlichen Herzens unter normiatlen undl pathologischen Verhliiltnissen. Dargestellt in steieoskopischen Rdintgenbl)il(lerin. Jena, G. Fisclher, 1907. Pp. 13, 30 Fot. 92 HIRSCH, C., AND SPALTEHOLZ, W.: C(oronaiarterien und Herzmnuskel. Anatomisehe un(l exilerimentelle Untersuchungen. )eutsclhe tne(l. Wchnschr. 1: 790, 1907. 93 MILLER, J. L., AND MATTHEWS, S. A.: lifect on the heart of experimental obstructioni of the left coronary artery. Arch. Int. MIe(l. 3: 476, 1909. 94 Oimr1nH, J.: Die Lokalisaltion dei Infarkte un(l Schwielen (let Herzmuskulatur un(l ilre Beziehiung zu (ler Geffissversorgung (les Herzens. Berl. klin. Wchns(lhr. 46: 614, 1909. 95 AMENOMIMYA, R.: tler (lie Beziehungen zmvischen Koronar'arlterien undl Papillarmuskeln imi Herzen. Viichows Arch. f. path. Anat. 199: 187, 1910. 96 PIQUAND, G.: Recherches sur l'anatomie (les vaisseaux sanguins (du coeur. J. de lanat. et physiol. 46: 310, 1910. 9 NussB.1AUrM, A.: ITber (IIaS Gefiisssystem (les Herzens. Arch. f. imikr. Anat. 80: 450, 1912. 98 SMITH, F. AI.: Concerning the Ilnatormy of the (oronary arteries. Am. J. __M. Se. 156: 706, 1918s. 9 GROSS, L.: The Blood Supply to the Heart in its Anatomical and Clinical Aspects. New York, Paul B. Hoeber, 1921. P'p. 171. 100 SPALTEHOLZ, '\V.: I)ie Arterien (ler Herzwand. Anatomische Umntersuchungen aim Menschenund Tierherzen. XV Tmafelmn. Leipzig, S. Hirzel, 1924. 1Pl). 165i. 1')1 OBE1,RtHIELMIN, H. A., AND LE COUNT, E. R.: Variations in time altilStomosis of ti e (oronary 814 INTERARTERIAI CORONARY ANASTOMOSES IN HUMAN HEART Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 arteries anCl their sequences. J.A.M.A. 82: 1321, 1924. 1( 2 KUGEL, MI. A.: Anatomical studies on the coronary arteries andl their branchlies. I. Arteria anastomotica auricular1is malgIn.a. Am. Heart J. 3: 260, 1928. 103 \\HITTFEN, i\I. B.: A review of the technical methods of (lemonstrating the (circulation of the heart. Arch. Int. Med. 42: X47, 1928. 104 CAMPBELL, J. S.: Stereoscopic radiiography of the coronary system. Quart. J. M\led. 22: 247, 1929. R. A.: The coron'arv arteries of tle clog. M0':IOORE, Am. Heart J. 6: 743, 1930. 106 LRoss, L., AND KU G L, M1. A.: The arterial hlood xvasular (Iistribution to the left and right ventricles of the human heart. Am. Heart J. 9: 165, 1933. 107 LowI.:, T. K.: Some experimental observations on the ('or'onar.lli.N circulation in man. Roy. -Melb)ourne Hospital Clin. Rep. 8: 3, 1937. 1O8 BLUM, L., SCHAUER, G., AND CALIEF, 13.: Gradual. oC(}lusion of a coronary artery, an experimental stu(ly. Am. Heart J. 16: 159, 1938. 109 SCHLESINGER, M. J.: An injection p)lus (lissection stu(ly of coronary artery o( clusions and anastomoses. Am. Heart J. 15: 528, 1938. 110 PIANNETTO, M\. B.: The coronary, .,arteries of the (log. Am. Heart J. 18: 403, 1939. 111 I)Oci WY. The capacity of the (oroiiry lbe(l in (ard(iac hypertrophy,. J. Exper. Me(l. 74: 177, 1941. 112 IOWE, T. K.: Some principles governing the 1)lood0 supply to the myoca.r(liulfl in occlusive arterial (lisease. Am. Heart J. 21: 326, 1941. 113 1PRINZMETAL, M., KAYLAND, S., MARGOLES, C., AND TRAGERMAN, L. J.: A (luantitative method for (letermining collateral coronary circulation. Preliminatry report on normal hIman hearts. J. Mt. Sinai Hosp. 8: 933, 1942. 114 HOLYOKE, J. B.: Coronary1 arteriosclerosis and myocar dial infarction as studied by aIn injection technique. Arch. Pcath. 39: 268, 1945 115 M\OKOTOFF, R., AND KATZ, L. N.: The effect of theophyllin with ethylenediamiine (amiiinol)hllin) and of papaverine hy(lrochlori(le on experimental myocardial infaretion in the (log. Am. Heart J. 30: 215, 1945. 116 SMITH, J. R., AND HENRY, M. J.: 1)emonstration of coronary- arterial system with neoprene latex. J. La11). & Clin. M\Ied. 30: 462, 1945. 117 RAVIN, A., AND GEEVER, E. F.: Coronary arterio sclerosis, coronary anastomoses an(l myocardial infarction, a clinicopathologic stul(iy ibasedl on an injection method. Arch. Int. M\Ied. 78: 125, 1946. 118 SALANS, A. H., AND TWEED, P.: A l)reliminary. study of the coronary circulation isost mortem. Am. Heart J. 33: 477, 1947. 119 1PRINZMETAL, M1., BERG.MAN, H. C., KRUGER, H. E1., SCHWARTZ, L. L., SI1IKIN, B., AND SoBIN, S. S.: Studies on the coronary circulation. III. Collateral circulation of beating lhumatn andl (log hearts with (oronary oc(lusion. Anm. Heart J. 35: 689, 1948. 120 WAGNER, A., AND POINDEXTER, C. A.: )emonstiation of tile (oronarv arteries withlon. Am. Heart J. 37: 258, 1949. 121 ScorT, it. W., YOUNG, A. F., ZIMMERMAN, II. A., AND KROH, I.: An improved nmethlo(l for visual izing the coronary arteries ait p)ost-mortem. Am. Heart J. 38: 881, 1949. 122 BLUMGART, H. I,.: Coronary arteries .as end a1teries. An. J. Med. 5: 321, 1948. 123 SCHLESINGIm,1, NI. J., ZOLL, P. A., AND S.: The conus artery: a third (oronarv artery. Am. Heart J. 38: 823, 1949. 124 Relation of anatomic l)attern to j)atliologi( conditionls of the (oronary arteries. Arch. P'ath. 30: 403, 1940. 125 BLUMIGART, H. L., FREEDBERG, A. S., ZoLL, P. NMI., I.vIs, H. I)., AND WESSLER, S.: The effect of (licumIarol on the heart in exlperiment.l acute coronatryN o(clusion. AmI. Heart J. 36: 13, 1948. 126 HILL, A. B.: Principles of Medical Statistics. London, Tile Lancet Limited, 1937. 127 AscHoFF, 1.: Lectures on Pathology. New York, Paul B3. Hoeber, 1924. P. 135. 128 SAIITH, H. I.: The relation of the weiglht of tile heart to tile weight of the b)o(ly, tuid of tlhe wveight of tile ileart to .age. Amn. Helart J. 4: 79, 1928. 129 ZEEK, P'. N.M.: Heairt weight. I. Tile weiglht of tile n1oImaii l hunmatn heart. Arch. IPtth. 34: 820, 1942. 13()BEAN, R. B.: Notes on the 1)ostnltIll growtlh of the heart, kidneys, liVem, an(l sp)leeln ill man. Ca.rnegie Institute of \a.sliington, P'ubuli(ation 272. Contributions to lmbryology 9: 268, 110. 37, 1920. 131 ROI;SSLE, R., AND ROULET, F.: \hiass uin(l Zahl in (ler IPathologie. Berlin, Julius Springer, 1932, P. 26. 132 PEACOCK, T. B.: On the weigllt and (limnmsions of the heart in healtil and (lisease. Mlonthliv J. NI. Sc. Edinburgh 19: 193, 1854. 133 ROSAHN, 1 ).: Tile weight of the normal Ileart in a(lult males. Yale J. Biol. & M\led. 14: 209, 1941. 134 THOIA, R.: Untersuchungen fiber (lie Grli.ssc n(1 (las Gewicht (ler anatomischen Bestandtheile (les menschlicher Kiirpeis. Leipzig, F. C. W. Vogel, 1882. P. 153. 135 Ross, J. A.: Post-mortem alpl)earances, e(l. 3. Lon(lon, Niorrison and Gihb Limite(l, 1937. 136 BLUNIGARTr, H. L., SCHLESINGER, NI. .J., AND ZOLL, 1P. NIM.: Angina pectoris, coroa faillhre a114(1 myocardial infarction. The r ie of coronarlx P. Al. ZOLL, S. WF2SSLER XND Occlusions andl collateral circulation. J.A.2M.A. 116: 91, 1941. Y7 XU\-IL., H. Ml. C.: P'rogress without statistics. JAM.A. 141: 195, 1949. '1( AMIADIW,(), J. A.: A suggestion for imp)roving the structure of the car(liac coronary circulatory system without surgical intervention. AIm1. Heart J. 28: 699, 1944. M. J. SCHLESINGER 815, SCOTT, R. C.: Personal communication, April 12, 1950. 140) ZOLL, P. Al., AND )1IZESDALE, 1). T.: Unpublished 139 observations. 141 T., AND BURCH, CG. 1,.: T1e electrocar(liogram an(l car(liac state in active sickle cell anemia. Am. Heart J. 29: 685, 1945. WINSOR, Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 Interarterial Coronary Anastomoses in the Human Heart, with Particular Reference to Anemia and Relative Cardiac Anoxia PAUL M. ZOLL, STANFORD WESSLER and MONROE J. SCHLESINGER Downloaded from http://circ.ahajournals.org/ by guest on June 14, 2017 Circulation. 1951;4:797-815 doi: 10.1161/01.CIR.4.6.797 Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231 Copyright © 1951 American Heart Association, Inc. All rights reserved. Print ISSN: 0009-7322. Online ISSN: 1524-4539 The online version of this article, along with updated information and services, is located on the World Wide Web at: http://circ.ahajournals.org/content/4/6/797 Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: http://www.lww.com/reprints Subscriptions: Information about subscribing to Circulation is online at: http://circ.ahajournals.org//subscriptions/